Analog video transport to a display panel and source driver integration with display panel
By transmitting video signals as analog signals directly to the display panel, the SAVT technique addresses performance bottlenecks and complexity in display source drivers, enhancing bandwidth and reducing costs through integrated source driver technology.
Patent Information
- Application Number
- KR1020267037598
- Authority / Receiving Office
- KR · KR
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2024-01-26
- Filing Date
- 2024-02-15
- Publication Date
- 2026-11-17
AI Technical Summary
Existing video transmission technologies face performance scaling bottlenecks and high costs due to the need for digital-to-analog conversion in display source drivers, which become increasingly difficult as display resolution and frame rates increase, leading to inefficiencies and complexity.
The implementation of sampled analog video transmission (SAVT) technique that transmits video signals directly as analog signals, eliminating the need for digital-to-analog conversion in the display source driver, thereby integrating the source driver with the display panel's glass substrate and reducing complexity and cost.
This approach increases bandwidth, reduces production costs, and simplifies display connectivity by allowing the source driver to be integrated with the display panel, enabling flexible and efficient video transmission with lower power consumption.
Smart Images

Figure PAT00011_ABST
Abstract
Description
Technology Field Cross-reference of related applications This application claims priority to U.S. preliminary application No. 63 / 500,341 (document No. HYFYP0015P2), filed on May 5, 2023, under the title “Analog video transmission to a display panel and source driver integration with a display panel,” and to U.S. preliminary application No. 63 / 447,241 (document No. HYFYP0015P), filed on February 21, 2023, under the title “Analog video transmission to a display panel,” the two patents of which are incorporated herein by reference. This application claims priority to U.S. preliminary application No. 63 / 611,274 (document No. HYFYP0017P2), filed December 18, 2023, and 63 / 516,220 (document No. HYFYP0017P), filed July 28, 2023, under the title “Video transmission within a mobile device,” both of which are incorporated herein by reference. This application claims priority to U.S. Preliminary Application No. 63 / 625,473 (Document No. HYFYP0018P) filed on January 26, 2024, titled “In-vehicle signal transmission”, which is incorporated herein by reference. This application refers to U.S. Patent Application No. 17 / 900,570 (HYFYP009), filed August 31, 2022; U.S. Patent Application No. 18 / 098,612 (HYFYP013); U.S. Patent Application No. 18 / 117,288 (Docket No. HYFYP014), filed January 18, 2023, currently U.S. Patent No. 11,769,468; and U.S. Patent Application No. 18 / 117,288 (Docket No. HYFYP014), filed March 3, 2023, currently U.S. Patent No. 11,842,671. This application incorporates, by reference, U.S. Patent Application No. 18 / 442,447 (HYFYP017) filed herein on the same date. Background Technology The present invention generally relates to video transmission. More specifically, the present invention relates to transmitting analog video samples, for example, within or to a display unit. Image sensors, display panels, and video processors are constantly competing to achieve larger formats, higher color depth, higher frame rates, and higher resolutions. Local-site video transmission involves performance scaling bottlenecks that limit throughput and degrade performance while consuming more cost and power. Eliminating these bottlenecks can yield benefits. For example, as display resolution increases, the data rate of video information transmitted from a video source to the display screen is increasing exponentially, from 3Gbps for Full HD 10 years ago to 160Gbps for new 8K screens. Generally, a 4K display requires a bandwidth of approximately 18Gbps at 60Hz and 36Gbps at 120Hz (divided by P physical channels). An 8K display, on the other hand, requires 72Gbps at 60Hz and 144Gbps at 120Hz. Until now, data has been transmitted digitally using a variation of Low Voltage Differential Signaling (LVDS) data transmission that uses a bit rate of 16 Gbps per signal pair and parallelizes the pairs to achieve the required total bit rate. In a digital connection with a wiring delay of 5 ns / m, the wavelength of every bit is 12 mm, which is close to the limits of this type of connection and requires extensive data synchronization to obtain usable data. This digital information must then be converted into analog pixel information on the fly using ultra-fast digital-to-analog (D-to-A) conversion or ultra-parallel slow conversion in the display's source driver. Currently, D-to-A converters use 8 bits, but 10 or 12 bits may soon be required. If this happens, it will become very difficult to convert accurately at sufficiently fast data speeds. Therefore, displays must perform D-to-A conversion within a very short time, and as the time available for conversion is becoming increasingly shorter, stabilizing D-to-A conversion is also becoming a problem. Therefore, a new device and technology that eliminates the need for D-to-A conversion in the display source driver, increases bandwidth, utilizes analog video signals within the display device, and can transmit video signals from other locations is desirable. To achieve the aforementioned objectives, and in accordance with the purpose of the present invention, a sampled analog video transmission (SAVT) technique that resolves the aforementioned defects of the prior art is disclosed. This technique may also be referred to as "clock analog video transmission" or CAVT. It can be seen that the requirements for bit-perfect communication between computing devices (e.g., text, spreadsheets) are very different from those for communicating video content for human viewing. Fundamentally, since a video signal is a list of brightness values, accurately maintaining fixed bit-width (i.e., digital) brightness values is inefficient for video transmission; furthermore, because there is no need to accurately reproduce these brightness values in bit units, analog voltage provides higher resolution. The unnecessary requirements for bit-perfect video transmission impose a cost burden known as "digital overhead." Therefore, this invention proposes transmitting video signals as analog signals rather than digital signals. While conventional digital transmission uses expensive mixed-signal processes for high-speed digital circuits, embodiments of the present invention utilize fully depreciated analog processes to increase flexibility and lower production costs. Additionally, using analog signals for data transmission between a display controller (e.g.) and a source driver of a display panel reduces complexity compared to conventional transmission between a signal source (via an LVDS or Vx1 transmitter) and a source driver receiver equipped with a D-to-A converter. In one embodiment, a transmitter is disclosed for processing a received digital video sample, converting it into an analog and transmitting it to a display panel, and a source driver of a display panel is also disclosed for receiving an analog sample and driving it to a display panel. The analog signal is used to transmit digital video data received from a video source (or storage device) to a video sink for display. Since the analog signal can be generated by a transmitter of a computer (or other processor) and transmitted to a source driver of a display unit for display on a display panel, the analog signal can be generated outside the display unit or by a transmitter within the display unit itself. In another embodiment, part or all of the source driver may be integrated with the glass substrate of the display panel, taking into account the required analog speed and accuracy. Source drivers of the prior art were mounted at the edge of the display panel but were not integrated due to the complexity of high-speed digital circuits and the large area required for D-to-A conversion. The present invention allows the source driver to be integrated with the glass itself, as the source driver does not require a D-to-A converter or a decoder, and due to the low-frequency sampling of the SAVT signal (e.g., the SAVT video signal reaches the source driver at a frequency of 1 / 10 the data transfer rate of a 3 GHz digital video signal). The present invention can be used on any active matrix display substrate. Highly mobile substrates (e.g., low-temperature polysilicon (LTPS) or oxide (IGZO) TFTs) are most suitable. As a result, the display panel can be connected to the GPU with only signal cables of any length and a power supply when the entire source driver is integrated. Since there is no need to connect additional electronics to the glass, the edge width can be further reduced and the module made thinner. The present invention is particularly applicable to displays used in computer systems, televisions, monitors, game displays, home theater displays, retail signage, outdoor signage, etc. Embodiments of the present invention are also applicable to in-vehicle video transmission, such as in automobiles, trains, airplanes, ships, etc., and apply to video transmission from a transmitter to a display or monitor in the vehicle, as well as video transmission within such display or monitor. The present invention is also applicable to video transmission to or within a mobile device, such as a telephone. In certain embodiments, the present invention is useful within a display device used to transmit and receive video signals. For example, the transmitter of the present invention may be used to implement the transmitter described in U.S. Patent No. 11,769,468 (HYFYP013), and the receiver of the present invention may be used to implement the receiver described in U.S. Application No. 17 / 900,570 (HYFYP009). Brief explanation of the drawing The present invention, along with its additional advantages, is best understood by referring to the following description together with the accompanying drawings: FIG. 1 is a diagram illustrating the transmission of an electromagnetic (EM) analog signal to a display panel of a display unit using conversion within the display unit; Figure 2 shows the architecture of the transmitter within the display unit; FIG. 3 shows an example of a distributor in which other predetermined permutations are used; FIG. 4 is a block diagram of an alternative embodiment for each DAC following an image processor; Figure 5 shows the architecture of the source driver of the display panel; Figure 6 shows another architecture of the source driver of the display panel; Figure 7 is a block diagram showing the integrated transmitter and timing controller located immediately behind the SoC of the display unit; Figure 8 shows the integrated transmitter and timing controller in more detail; FIG. 9 is a block diagram showing the integrated transmitter, timing controller, and SoC within the display device; FIG. 10 shows a video transmission system within a display unit; FIG. 11 shows a specific subpixel transmission order in which subpixels are grouped by color to minimize transmission bandwidth; FIG. 12A illustrates a specific subpixel transmission sequence in which subpixels are still grouped by color and control signals all reach a single dedicated amplifier channel to minimize transmission bandwidth; FIG. 12B illustrates a specific subpixel transmission sequence in which subpixels are grouped by color and transition bands are placed between them; Figure 13 shows an SAVT receiver integrated with a source driver where each distribution amplifier drives an adjacent column and all control signals are processed by a single amplifier; FIG. 14 illustrates a source driver input of a source driver (820) for interleaving multiple input amplifiers capable of satisfying speed requirements; FIG. 15 is a diagram summarizing the pixel transmission sequence showing how pixel and control signals are transmitted from the transmitter to the source driver of FIG. 13 and the amplifier assigned to each; FIG. 16 is a block diagram of the input vector of a transmitter having a predetermined permutation that provides the subpixel transmission order required by FIG. 15; FIG. 17A is a diagram illustrating the source driver of FIG. 13, which shows the control channel in more detail, and three comparators used to extract phase alignment information from the control signal; FIG. 17B illustrates an alternative source driver 820'' for the source driver of FIG. 17A, showing the details and control signals added from the first amplifier; FIG. 17C is a drawing illustrating a preferred source driver 820''' for the source driver of FIG. 17B; FIG. 17D summarizes the subpixel order collected by the input amplifier of FIG. 17C; FIG. 18 illustrates a technique for introducing a timing reference into a sequence of control signals; Fig. 19A shows the control sequence commands and parameters; Fig. 19B shows another technique for transmitting commands and parameters using MFM; FIG. 20A illustrates a technique for performing phase alignment; FIG. 20B illustrates a sampling phase adjustment circuit; FIG. 20C shows a special synchronization video pattern to facilitate locking; Fig. 20D shows an example of MFM flag loss due to PLL phase wrap-around; FIG. 20E shows an example where the MFM flag is lost when the phase extends past the end of the control bit; Fig. 20F is a flowchart explaining how phase alignment occurs with MFM flag search; Figure 21 shows the analog data path of one channel of the example source driver; Fig. 22 is a block diagram showing the transmission of analog video samples within a mobile phone; FIG. 23 illustrates the implementation of source driver function integration in various embodiments; FIG. 24A shows the placement of a gate driver and a source driver on the glass of a display panel; FIG. 24B shows a source driver fully implemented in glass; FIG. 25 shows another embodiment of the placement of EM signals; FIG. 26 shows an analog architecture for commanding and receiving feedback; Figure 27 shows latching values in a column amplifier; FIG. 28 shows a digital architecture for commanding and receiving feedback; Figure 29 shows latching values in a column amplifier; FIG. 30 illustrates an SAVT transmitter arranged to transmit various video samples from various sources; and FIG. 31 is a drawing illustrating a SAVT receiver located in an SoC, processor, legacy display, or other location. Specific details for implementing the invention Since the wiring weaving of the display unit closely matches the design values, it is realized that the resilience provided by the use of spreading codes (video sample encoding and decoding for transmission within the display unit as described in U.S. Patent No. 10,158,396) can be greater than the circuit overhead of decoding in the source driver. In particular, the use of spreading codes provides some degree of resilience against thermal noise from the transmitter's DAC and the source driver's sample and hold amplifiers. Nevertheless, since this thermal noise is probabilistic, it must be at an undetectable level. Therefore, in some applications, encoding and decoding in the source driver are not required, so code spreading is not strictly necessary. Accordingly, it has been proposed to transmit video data from the transmitter to the source driver of the display panel as an analog signal. It is known that the digitization of video signals typically occurs at the system's signal source (often a GPU), after which the digital signal is transmitted to a display panel source driver—usually using a combination of high-performance wiring systems—where it is converted back into an analog signal and loaded onto the display pixels. Therefore, the sole purpose of digitization is to transmit data from the video source to the display pixels. Consequently, it is known to be more advantageous to avoid digitization whenever possible and transmit analog data directly from the video source (or from an appropriate transmitter) to the display source driver. Since these analog signals are highly accurate (though this may vary depending on circuit imperfections) and represent continuous values, the resolution of the possible values is always higher than what can be expressed by an arbitrarily long digital representation. In other words, the sampling rate is at least 10 times lower than in the case of digital transmission, allowing for greater bandwidth expansion. Additionally, it may be easier to perform D-to-A conversion at a point requiring less power than the end point where the display panel is driven. Therefore, instead of transmitting digital signals from the video source (or SoC or timing controller) to the location where the analog signal needs to be generated, the signal can be converted to analog near the SoC or timing controller within the transmitter and then transmitted to the display panel at a much lower sample rate than is typically used for digitization. In other words, instead of transmitting gigabits per second over multiple lines, analog signals transmit only hundreds of megasamples per second, reducing the bandwidth of the channels that need to be used. This speed is about one-tenth the level of the digital speed required for the same number of physical communication paths. Furthermore, while conventional digital transmission occupies only about 1.25 cm (considering that cable propagation is about 0.2 m / ns, 16 Gbps is 1 / 16 ns / bit, so 1 bit is 0.2 / 16 meter), transmitting analog data increases available time tenfold, allowing for additional bandwidth to be utilized. Furthermore, the bits of digital data must be well-defined. This definition is highly sensitive to errors and noise, and must be able to detect highs and lows with great accuracy. The present invention can be applied particularly to high-resolution, high-dynamic-range display devices used in computer systems, televisions, monitors, machine vision, automotive displays, aviation displays, virtual or augmented reality displays, mobile phones, billboards, scoreboards, etc. Analog signal transmission to display panel FIG. 1 illustrates the transmission of an electromagnetic (EM) analog signal to a display panel (150) of a display unit (100) using conversion within the display unit (100). In this embodiment, the conversion of a digital video signal to an analog signal is performed within the display unit (100) itself, thereby improving display connectivity. The illustrated video signal (110) represents a video signal (110) transmitted to a display unit using an HDMI interface (such as LVDS, HDBaseT, MIPI, IP Video, etc. interfaces may also be used). Typically, a system-on-chip (SoC) (120) and a timing controller (TCON) (130) are illustrated to transmit digital video samples from the video signal to a transmitter (140). The SoC (120) performs functions such as display control, decompression, and specific digital signal processing, and outputs the video signal to the TCON. Typically, LVDS or V-by-One is used to transmit digital video data (122) from the SoC to the TCON. For example, when using LVDS pairs, the number of pairs depends on the implementation and the data transfer rate per pair, panel resolution, frame rate, bandwidth, etc. Additionally, as is known to those skilled in the art, various physical layers, including serial-deserializer or SerDes layers, may be used to transmit video data from the SoC (120) to the TCON (130), and when the transmitter (140) is integrated with the TCON (130), this physical layer transmits video data from the SoC (120) to the integrated TCON and transmitter as illustrated in FIG. 7. For example, up to 48 SerDes channels or more may be used to transmit this video data. Some or all digital or image processing may be performed on the SoC, in which case no image processing is performed after the line buffer and before the DAC in FIG. 2. Preferably, image processing includes some form of gamma correction and dimura correction, and may include image enhancement or modification (e.g., motion correction or correction for adjustment between the bottom and top of the panel). Image processing may be performed in a serial format (e.g., processors 250-259) or using sequential pixel conversion to a serial format, but it is easier to perform in a parallel format. Various embodiments are possible: a discrete implementation in which the transmitter (140) is embedded in a mixed-signal integrated circuit and the TCON and SoC are separate components; a mixed implementation in which the transmitter (140) is integrated with the TCON in a single IC and the SoC is separate; and a fully integrated implementation in which as many functions as possible are integrated into a custom mixed-signal integrated circuit in which the transmitter is integrated with the TCON and SoC. In this example of FIG. 1, the display panel (150) is located within the panel frame (151) of the display unit. As illustrated, both the transmitter (140) and the panel frame (151) are located within the display unit (100). The display panel (150) may be a display panel of any size, such as a monitor, a large-screen television, an advertising board, or a scoreboard; it may be a display within a VR headset; or it may be a head-up display (HUD) where the display is projected onto a windshield, a screen of a visor, etc. For the purposes of this disclosure, "display panel" refers to the internal part of a display unit (also referred to as "glass") that implements pixels that generate light for viewing, and "display unit" refers to an entire (generally) rectangular enclosure comprising a display panel, a panel assembly, a frame, a driver, cables, and associated electronics for producing video images. A mass-producible display panel containing approximately N^2 pixels is controlled by approximately N voltages that are updated N times per display interval (the inverse of the frame rate). Even when the signal input to a display device is a digital video signal, using an analog signal for transmission within the display device offers significant advantages. Conventional display devices have full-bitrate digital data that must be decompressed from the HDMI signal and then transmitted from the receiving point of the display device to all source drivers within the device. For 65-inch or 80-inch displays, these connections can be quite long, requiring the transmission of digital data from one location within the device with the input to another location (perhaps on the opposite side) where the final source driver is located. Therefore, converting the digital signal internally into an analog signal and then transmitting that analog signal to the source driver offers advantages similar to those of using a low-frequency signal. Also illustrated in FIG. 1 is a transmitter (140) that generates an analog EM signal (192) for a source driver (186). It includes individual flexible PCBs (184) each having a rigid PCB (182) and a source driver (186) that generates a source voltage for a display panel. As will be described in more detail below, the signal (180) optionally provides information about the display panel back to the transmitter (140) to support the processing of video samples. The generation of the gate driver control signal (190) may be performed by a timing controller (or other specific hardware) as is known to those skilled in the art, and may be based on the synchronization information of the source driver. Typically, the transmitter (140) and the receiver (in this case, the source driver (186)) are connected by a transmission medium. In various embodiments, the transmission medium may be a cable (e.g., HDMI, flat cable, fiber optic cable, metal cable, non-metallic carbon track flex cable, metal trace, etc.) or wireless. The transmission medium may have multiple EM paths, and there may be one path per EM signal 192. The transmitter includes a distributor that distributes received video samples to the EM paths. The number of paths may range from one to more than one. In this example, the transmission medium will be a combination of cables, traces on a PCB, internal connections within an IC, and other media used by those skilled in the art. During operation, a digital video sample stream (110) arranged in chronological order containing color values and pixel-related information is received from a video source of a display unit (100) and transmitted to a transmitter (140) via an SoC and a TCON. The number and content of the input video samples received from the video source depend on the color space in which the source is operating (the samples may be black and white). Regardless of the color space used, each video sample represents the amount of light detected or measured in the specified color space. It is a signal from an SoC (typically an LVDS digital signal, but other signals may be used) in which pixel values are listed in row-major order across consecutive video frames. One or more pixel values (e.g., 2, 4, etc.) may arrive at a time, and it is serial in the sense that groups of pixels are transmitted progressively from one side of the line to the other. For example, a processing unit, such as an unpacker in a timing controller, can be used to unpack (or expose) these serial pixel values into parallel RGB values. Additionally, it should be understood that the color information exposed for each set of samples can be arbitrary color information (e.g., Y, C, Cr, Cb, etc.) and is not limited to RGB. To use color information other than RGB subpixels, additional processing may be required before the source driver drives the columns (essentially subpixel intensity values). The number of output sample values S for each set of pixel samples is determined by the color space applied by the video source. For RGB, S=3, and for YCbCr 4:2:2, S=2. In other situations, the sample value S of each sample set may be one or three or more. The unpacker may also unpack digital signal framing information in the form of framing flags provided along with pixel values. Framing flags indicate the position of a pixel in a specific video frame and, as is known to those skilled in the art, indicate the start of a line, the end of a line, an active video section, horizontal and vertical blanking sections, etc. Framing flags are used to inform the gate driver of the line currently being transmitted to the display panel and also control the timing of the gate driver's operation. Framing flags may be included within the gate driver control signal (190), as is known to those skilled in the art. Generally, symbol and sampling synchronization occur before extracting framing information such as H-sink and V-sink (and other line control information). The TCON (130) provides a reference clock (170) to each source driver (186), that is, each source driver chip (e.g., Hi-Fi HY1002 chip) has a clock input provided by the TCON (whether it is an FPGA or an IC). Although the clock (170) is shown as an input only for the first source driver for clarity, each source driver receives a reference clock. This reference clock can be a relatively low frequency, such as, for example, about 10.5 MHz. Details regarding the reference clock are shown in Figure 17C. transmitter FIG. 2 shows the architecture of a transmitter (140) within a display unit. A splitter (240) containing two line buffers (241 and 242), a splitter controller (230), multiple P image processors (250-259), digital-to-analog converters (260-269) following each image processor, and analog EM signals (270-279) output from each DAC is shown. In this example, there are 24 source drivers, i.e., 24 EM paths (P=24), and there may be a single EM path or multiple EM paths. Depending on implementation and design decisions, multiple outputs can improve performance but may require more paths. Controller 630 stores a line of pixels for the display in one of the line buffers, and when the line is completed, that line is output (to the DAC or another line buffer as described below). Typically, pixels for a line on the display panel arrive sequentially from the SoC, but since the gate driver can display pixels for a line simultaneously, the source driver must prepare pixels for the entire line at the same time. Therefore, each line buffer provides storage space for pixels for a line. Additionally, since there are cases where only half of a line of pixels is activated on the display panel by the gate driver, one line is stored in the line buffer, and then half is extracted and transmitted while a new line is being stored. Generally, when a stream of input digital video samples is received within the transmitter (140) in row-priority order, the input digital video samples are (1) repeatedly distributed to one of the EM paths according to a predetermined permutation (in this example, row-priority order, i.e., identity permutation), (2) converted to analog, and (3) transmitted as an analog EM signal through a transmission medium, with one EM signal transmitted per EM path. The incoming analog EM signal from each source driver (186) is received at the input terminal, and each analog sample is sequentially distributed to the storage cells of a specific column driver using the inverse of the predetermined permutation used in the transmitter through a sampling circuit. Once all samples from the corresponding source driver are placed in place, they are driven to the display panel. Consequently, the original stream of video samples, arranged in chronological order and containing color and pixel-related information, is transmitted from the video source to the video sink. The inverse permutation effectively stores the incoming samples in a single row in the storage array (for display on the panel) in the same order as the sample row received from the distributor. Samples may arrive sequentially, such as R, G, and B, or in parallel as three separate signals, i.e., in parallel, such as RGB. The order of the samples can be changed as needed using a divider (240). In one embodiment, four control signals are inserted into the sample stream of the splitter for every 60 video samples and transmitted to the source driver. As illustrated, each input vector (280) of the line buffer contains a total of 1024 values, including four control signals per 60 video samples. The control signals can be inserted at various locations in the input vectors; for example, "samples" 960–1023 of the input vectors (280–288) may actually be control signals. Any number of control signals may be used in each input vector. Additionally, the number of control signals may be arbitrary but finite. The more control signals transmitted, the higher the required data transmission rate. Ideally, it is recommended to limit the number of control signals so that the number of control signals corresponds to the blanking cycle and the indicated lines (thus reducing the required storage capacity or the amount of additional resynchronization). Additionally, the control signals may be inserted into the sample stream at the splitter, or the insertion of the control signals may be performed at other locations. A splitter (240) is positioned to receive pixel color information (e.g., R, G, and B values) exposed to an input set of samples. The splitter (240) takes the exposed color information and writes multiple input vectors (280–288) to the first line buffer (241) according to a predefined permutation (one input vector per EM path). When the line buffer (241) is full, each input vector (280–288) is read to the corresponding DAC or optionally to the corresponding image processor (250 259) through the corresponding output ports (281–289). While these input vectors are being read from the line buffer (241) (or when the line buffer (241) is full), the RGB input samples of the next line are written to the input vectors (290–298) of the second line buffer (242). Therefore, when the second line buffer (242) is full (and the DAC or image processor has finished reading the input vector from the first line buffer (241), the DAC or image processor begins reading samples from the second line buffer (242) through the output ports (291-299). Writing and reading to the first and second line buffers continue in this "ping-pong" manner as long as input samples arrive at the transmitter. Output ports 281-289 and 291-299 may be bit serial communication, but are more likely sequential word-whole samples or parallel word-whole samples. In a preferred embodiment of writing to and reading from a line buffer, when a sample arrives at the transmitter (140), it is written only to one of the line buffers, e.g., buffer (241). When that buffer is full, all samples are written in parallel from buffer (241) to line buffer (242). Then, the samples are output only from buffer 242 to the DAC (or image processor). The process proceeds sequentially such that when buffer 242 outputs samples, buffer 241 is filled, and when buffer 242 is depleted, all samples from buffer 241 are written to buffer 242. Samples can be written from buffer (241) to buffer (242) during a horizontal blanking period. The number of line buffers required depends on the relative time required to load and then unload the buffers. A continuous data stream is received at the RGB input. If it takes time T to load all samples into the buffer and the same time T to unload them, two buffers are used (so that one can be unloaded while the other buffer is being loaded). If the time required for unloading becomes shorter or longer, the buffer length can be adjusted (i.e., by adjusting the number of input vectors or N of each input vector) so that the number of line buffers required is always two. Nevertheless, if desired, more than two buffers may be used and any of the aforementioned embodiments may be used for writing to and reading from the buffers. The distributor controller (230) controls the operation and timing of the line buffer. In particular, the controller is responsible for defining the number of samples (N) and the permutation used when constructing the four input vectors. In this example, N = 1024. The controller (230) may also include a permutation controller that controls the distribution of RGB samples to the positions of the input vectors. The controller (230) may also include a permutation controller that controls the distribution of samples for the location of the input vector. The controller is also responsible for adjusting the clock domain that crosses from the first clock frequency to the second clock frequency. In a specific embodiment, samples are clocked in at the frequency of F pixels, and samples are clocked out sequentially at each input vector at Fsavt, which is the sampled analog video transmission (SAVT) frequency. It is also possible to clock in two samples at a time rather than one at a time, or to clock in three samples at a time. The analog samples are transmitted to the SAVT receiver as analog EM signals (270-279) along the electromagnetic path of the transmission medium. In one embodiment, each line buffer (241 or 242) has three input ports for incoming RGB samples, and the samples are clocked in at the frequency of F pixels; each line buffer also has 24 output ports, e.g., 281 or 291 (if there are 24 EM signals, each is transmitted to one of the 24 source drivers), and the samples are clocked out at the Analog Video Transmission (SAVT) frequency Fsavt sampled from each input vector. It is also possible to clock in two R, two G, and two B samples at a time rather than one each, or to clock in three samples at a time. In one embodiment, Fsavt = 663.552 MHz for 24 channels. For illustrative purposes, one possible permutation is one in which each input vector contains N samples of color information and control signals. In this example, the RGB samples of the exposed sample set are assigned to the input vectors from left to right. That is, the "R", "G", and "B" values of the first sample set, the "R", "G", and "B" values of the next sample set, etc., are assigned to the input vector (280) in order (e.g., RGBRGB, etc.). Once N samples and control signals are assigned to the input vector (280), the above process is repeated sequentially for other input vectors until each input vector has N values. The number of N values per input vector can vary greatly. As in this example, this predetermined permutation follows the row-order of the incoming samples; that is, the first input vector (280) contains samples 0 through 1023 of the first row in that order, and subsequent input vectors continue to maintain that permutation (including control signals). Accordingly, the distributor controller (230) performs permutations by assigning incoming samples to specific addresses within the line buffer. Additionally, the distributor (230) may use any permutation scheme, and it should be understood that whatever permutation scheme is used by the transmitter, the inverse is used by the control logic of each source driver to distribute incoming samples to the column driver. In a situation where only one electromagnetic path is used and video samples are received from the SAVT transmitter, the distributor writes to one input vector in each line buffer. Although the image processor (250-259) is shown after the line buffer and before the DAC, it is preferable to place the image processor (or processor) before the line buffer to reduce the number required, that is, it is desirable to reduce the number of image processors because image processing is performed when RGB samples arrive and the samples are distributed to the line buffer. It shows that pixels arrive one at a time, such that one image processor is used when pixels arrive one at a time, and two are used when two arrive at a time. Even though the image processor is located before the line buffer, certain processing, such as gain management, can be performed after the line buffer. Generally, image processing is performed as follows: a) applying gamma correction to each sample, b) level-shifting each gamma-corrected sample by mapping the range (0 . . . 255) to (-128 . . 127) to remove the DC component from the signal, c) applying path-by-path amplifier dispersion correction to each gamma-corrected and level-shifted sample, performing gain correction for each sample, performing offset adjustment for each sample, performing Demura correction for each sample, etc. Other corrections and adjustments may also be performed depending on the target display panel. Individual image processors (250-259) may process each output stream of the sample (e.g., 281 and 291), or a single monolithic image processor may process all outputs (e.g., 281 and 291, 285 and 295, etc.) at once. To avoid performing image processing on control signals within a line buffer, since the control signal timing and position within the buffer are known, the logic may determine that image processing of the control signal should not be performed. As mentioned above, image processing does not need to occur within the transmitter (140) and can occur at other locations such as the SoC (120), TCON, or receiver. For example, gamma correction is traditionally performed at the receiver (source driver), but dimura and more complex image processing are not performed at the source driver. The processed digital samples of each input vector are input sequentially into one of DACs 260-269 (image processing occurs either before or after the line buffer), and each DAC converts these modified digital samples at the frequency of Fsavt and transmits them along the electromagnetic path of the transmission medium as analog EM signals 270-279 to the source driver of the display unit. Each DAC converts the samples received in the digital domain into a single analog level, which can be transmitted as a pair of differential voltage signals of magnitude proportional to the received digital value, and the analog levels are transmitted serially when output from each DAC. The output of the DAC can vary from a maximum voltage to a minimum voltage, with a range of approximately 1V to 4V, Vpp (peak-to-peak), and approximately 2V Vpp works well. In certain embodiments, a signal with a dynamic range of + / -500mV or 1V (actually, the dynamic range of the input is about 30% higher or about 1.3V) is represented. Two line buffers are shown within the splitter (240) (which is preferred), but a single line buffer may also be used, and when a sample of a specific input vector is read to the image processor (or its DAC), the splitter refills the input vector with the received sample so that the continuous delivery of the sample from the line buffer to the DAC or image processor is not interrupted. Additionally, although less preferred, it is also possible to perform image processing on analog samples by placing each DAC (or multiple DACs per EM path) after the splitter and before the image processor (if any). FIG. 3 illustrates an embodiment of a splitter (240') in which a different predetermined permutation is used. This predetermined permutation can be used to reduce the wiring complexity of each source driver. The splitter (240') arranges the samples and control signals of each input vector (380-388) in this order: 0, 64, 128, . . . 960, 1, 65, 129, . . . , 961, and 63, 127, 255, ..., 1023 in order. (Indices corresponding to the indices shown in FIG. 6, i.e., each amplifier processes 64 samples and control signals.) Thus, each input vector inputs 1,024 values for a specific EM path at a time (assuming 64 control signals are added to each of the 960 actual samples of the TCON), and the controller distributes the samples to each input vector as shown using the predetermined permutation. Although not shown in the figure, 64 control signals are added to the end of each input vector, so that 4 control signals can be distributed to each amplifier as shown in the source driver of Figure 6. The numbers 64 and 960 vary depending on the implementation, and different numbers of control signals and different numbers of columns can be used per vector. Then, samples of the input vectors (380-388) are output from the line buffer (245) to the image processor (250-259) through the output ports (381-389). As in the divider of FIG. 2, the divider alternately writes samples to the line buffer (245) or line buffer (246) and reads samples from the line buffer (246). The line buffer (246) outputs samples to the image processor through the output ports (391-399). Thus, by rearranging the samples of the transmitter, adjacent columns can be driven while each interleaved input sampling amplifier of the source driver (e.g., shown in FIG. 6) operates in rotation. Output ports 381-389 and 391-399 may be bit serial communication, but are more likely sequential word-whole samples or parallel word-whole samples. In another embodiment (not illustrated), a specific permutation used by the splitter (240) sorts samples by color for each input vector, namely, all 320 red subpixels, all 320 green subpixels, all 320 blue subpixels, and all 64 subband signals. Thus, taking the first input vector as an example, sample positions 0-319 contain red subpixels, sample positions 320-639 contain green subpixels, sample positions 640-959 contain blue subpixels, and positions 960-1023 contain 64 subband signals for a specific row. The samples are then sent to an image processor (all red, all green, all blue, all subband). Other input vectors use the same permutation to group samples by color. Of course, the color groups of subpixels in the input vector can be in any order (not necessarily red, green, blue), and the 64 subband signals can be inserted anywhere within the group. The reason for specifying this order is to reduce potential electromagnetic interference signals generated by the system when grouping samples in this way, by leveraging a heuristic in natural images that individual color components tend not to exhibit high spatial frequencies. In fact, significant EMI is reduced when almost all subpixels of a specific color are grouped together. Additionally, using this order allows for slower operation of the source driver's S / H amplifier and reduces the bandwidth required for the transmitter and receiver communication channels. The control logic of each source driver then uses the inverse of this permutation to send incoming samples to the correct column driver. FIG. 4 is a block diagram of an alternative embodiment for each DAC following the image processor. As illustrated, instead of a single DAC following the image processor (250), there are four interleaved DACs (361-364) that process modified samples output from the image processor. The number of DACs can be any number, but DACs that are multiples of 2 (e.g., 2, 4, 8, 16) are common. Latches 351-354 are used to sample specific sample outputs from the image processor and then pass them to the corresponding DACs. Thus, for example, when using four DACs, the first four samples (Sample0, Sample1, Sample2, Sample3) are passed to DACs 361-364 in order, and the next four samples (Sample4, Sample5, Sample6, Sample7) are also passed to DACs 361-364 in order. The multiplexer (340) is used to multiplex the analog samples output from each DAC so that they are aligned to the EM signal (270) in the same order as those output from the image processor (250). The multiplexer (340) may be implemented as a selector, and the control of the selector may be a rotation enable bit of a shift register or a counter plus decoder. Although not illustrated, each of the other image processors will have a similar interleaved DAC. In this embodiment, a slower speed DAC may be used. The image processor (250) may also follow the DAC, in this case, an analog signal processor. Source driver architecture FIG. 5 shows an example of the architecture of a source driver (400) of a display panel. Typically, a display panel may have 24 such source drivers, i.e., 24 P (for an 8K display, fewer for a 4K display). An analog-to-digital converter (ADC) is not required in the source drivers to convert samples for the display into analog. Table 1 shows the parameters, values, and units of the source drivers to be used for an 8K144 display panel. Thus, each of the 24 source drivers drives 960 columns to provide subpixels for one row of the display (23,040 subpixels per row). 8K144 example value Parameters Hpix Vpix Screen Refresh Rx Chips H subpix subpixel rate subpixel overhead sample rate rate per chip Value 7680450014424230401492992000001.06715925248000663552000 Unit Pixel Pixel Hz Chip / System Subpixel / Line Samples / Sec = 64 / 60 (Synchronization and Control Overhead) Samples / Sec Samples / Sec / Chip One of the EM signals from the transmitter (140) is input from the input terminal (410) to the source driver (400). In this example, terminal 410 receives 1,024 analog values in succession at a time and then stores them in rows of storage array A or B 434 or 436 via S / H amplifiers 420-429. Analog video samples arrive in a natural order according to a predetermined permutation shown in the example of FIG. 2, but more complex wiring is required in the source driver as shown in the figure. Other permutations may also be used. The source driver (400) consists of 16 interleaved sample / hold input amplifiers (420-429) that sample the input (410) at Fsavt / 16. There are 16 blocks (430 being the first block) consisting of 60 video and 4 control signals. Each S / H amplifier (420-429) sequentially samples a specific analog sample and stores it in one of 64 storage arrays (434 or 436), 60 of which are fed directly to the column driver (440). Since the input amplifiers (420-429) are interleaved, they can run 16 times slower than the input signal, and each amplifier is phase-shifted by one SAVT interval. As shown in the figure, S / H amplifier 0, 420, drives columns 0, 16, 32, etc., and S / H amplifier 1, 421, drives columns 1, 17, 33, etc. Thus, the output of each S / H amplifier spans the range of all 960 columns (an amplifier output every 16 columns). In one embodiment, each of these storage elements of storage array A or B is a storage cell, such as a switch capacitor. Other terms for a storage cell include "sampler capacitor" or "analog latch". Additionally, there are 960 high-voltage column drivers (440), each of which drives a column (450) through an output pin to provide the necessary voltage to the display panel (480). As shown in the figure, there are 16 blocks consisting of 60 videos and 4 synchronization signals 460 (e.g., Hsync, Vsync, CTRL) per block. Once these 960 samples are stored (e.g., in row A), they are driven to each output column (450) via a column driver (440), while the next set of 960 analog samples is stored in another row (e.g., B). Thus, while one set of incoming 960 samples is driven to a column in either row A or B, the next set of 960 samples is stored in the other row. In a specific implementation, each analog level is a differential signal arriving at an S / H amplifier (420) that swings between approximately +0.5V or -0.5V and has a maximum swing of approximately 18V around the mid-range voltage of each single-ended column driver, thus requiring amplification. In addition to amplification, at some point the differential signal (- full scale is dark and + full scale is bright) is converted to a single-ended signal and drive polarity is applied so that, depending on the polarity setting, the dark subpixel becomes, for example, 9V, and the bright subpixel becomes full scale plus (e.g., 18V) or minimum voltage (e.g., 1V). Figure 21 shows an example of a source driver in more detail. The control logic (470) implements the inverse of a predetermined permutation used in the transmitter (140), controls the timing of the S / H amplifier (when each specific received analog sample is sampled) via the control line (471), controls the timing of the storage elements of column A and column B (when each column latches a specific analog sample) via the control line (472), and controls the driving timing of each column driver (440) via the control line (473) (when each column latches each input analog sample). For example, if the transmitter uses the predetermined permutation described above in which specific subpixel colors are grouped together in an input vector and then transmits these groups within an EM signal, the control logic (470) can use the inverse of this predetermined permutation to route the incoming sample to the correct column driver. At least the control logic (470) knows how the splitter controller (242) has positioned the incoming samples in the input vector, and uses this prior knowledge to send the incoming samples to the correct column driver so that the samples are displayed as they appear in the original source image. Note that the number of S / H amplifiers, 420–429, represents a trade-off between number and quality. As more amplifiers are added, the execution speed slows down, noise increases, and the load driven by each amplifier decreases. However, the load delivered to the input terminal (410) increases with the number of S / H amplifiers, which affects the transmission quality. Therefore, the number of input amplifiers to use must be determined in the design. To address RFI / EMI emission issues, the interval of each clock cycle can be slightly modified. The input of the SHA amplifier generates only a + / -250mV swing in the common mode voltage (positive and negative inputs, respectively), leading to a + / -500mV signal (1v dynamic range). This is a voltage swing similar to conventional digital signals such as CEDS or LVDS. Clock modulation can be performed in both cases to reduce RFI / EMI emissions, but this modulation is not preferred as it occupies the sampling window. Additionally, to optimize the performance of the source driver (to respond to process changes in the implemented S / H amplifier), a low-frequency feedback network can be added off-chip to characterize the gain and offset of all amplifiers in the source driver, but this technique is not preferred due to area and performance constraints. Another method to optimize the performance of the source driver output is to utilize the existing calibration technology of the display device itself. Modern OLED (and Micro LED) manufacturing technology characterizes the response of every subpixel in the array and pre-calibrates individual offsets in a manufacturing data table stored in the TCON and used during sample generation. Therefore, depending on the physics of the entire display device (transmitter, amplifier, source driver, each pixel, etc.), each subpixel may have a different characteristic response, such as being too bright or too dark. This table contains the individual offsets for each characteristic response. Note that in the source driver architecture (400 or 500), a predetermined interleaved sampling amplifier (420, 429, or 520-529) stores the pixel voltage in a switch capacitor and then amplifies it to a given column. Thus, all columns are driven through the same amplifier in each row. During manufacturing, the amplifier's linear error (e.g., gain error) is overlaid in a consistent pattern on other errors measured for individual subpixels of the column through conventional compensation techniques. Therefore, these conventional OLED error correction techniques can also correct all linear errors of the proposed source driver amplifier. This observation suggests that low-cost implementation can be enabled by relaxing design requirements (e.g., regarding gain accuracy). In one preferred embodiment, there are three amplifier stages, and the amplifier includes a common-mode feedback amplifier. FIG. 6 illustrates another embodiment of the architecture of a source driver (500) of a display panel. As in FIG. 5, an input terminal (510), input S / H amplifiers (520-529), rows of storage elements of arrays (534 and 536), a column driver (540), an output (550), subband signals (560 (e.g., synchronization signals, e.g., H-sink, V-sink) and CTRL), control logic (570), and a display panel (580) are illustrated. In this embodiment, each S / H amplifier (520-529) drives 64 adjacent locations (60 columns and 4 subband values), thereby reducing the wiring complexity of the source driver, reducing the physical distance that one of the amplifiers (520 529) must drive, and also making Demura compensation easier. For example, among the 60 video and 4 subband signals, the first block 530 is driven by amplifier 520, block 531 by amplifier 521, and block 539 by amplifier 529. This configuration also causes the gain error of a given input sampling amplifier to appear in 60 adjacent columns, thereby facilitating existing high MTF Demura compensation solutions so that the Demura system can easily detect it. To implement this embodiment, as shown in FIG. 3, the distributor (240') transmits samples in the following order: 0, 64, 128, . . . 960, 1, 65, 129, . . . , 961, and 63, 127, 255, ..., 1023, the last value 64 is a control signal. Thus, by rearranging the samples in the transmitter, each interleaved sampling amplifier can drive adjacent columns while operating in rotation. The control logic (570) implements the inverse of a predetermined permutation used in the divider (240') of FIG. 3, controls the timing of the S / H amplifier (when each sample a specific received analog sample) via the control line (571), controls the timing of the storage elements of rows A and B (when each row latches a specific analog sample) via the control line (572), and controls the timing when each column driver (540) drives each column via the control line (573). In this embodiment, each sampling amplifier drives 60 columns and 4 subband signals. Source driver architecture implementation example Figures 5 and 6 above reveal the architecture of the source driver. Depending on the specific implementation, other architectures may also be used, such as manipulating the transmission order of subpixels at the transmitter so that control signals are distributed among amplifiers for various reasons, transmitting control signals to only one amplifier through one channel, or grouping subpixels by color. FIG. 11 illustrates a specific subpixel transmission order in which subpixels are grouped by color to minimize transmission bandwidth. Table 800 shows the RGB subpixel indices of the subpixels collected by each input amplifier of the source driver. As illustrated, each amplifier (802) receives a series of subpixels (804) after first receiving a control signal, in this example, receiving them first in the order of red, green, and blue. The transmission order of subpixels from the transmitter is received at the specific amplifiers from top to bottom and left to right, as shown in the figure. Since color information tends to change more slowly than luminance information in a given video image, grouping subpixels by color reduces transitions and reduces bandwidth. While this is a desirable architecture for certain applications, it can be difficult to extract control information when control signals are distributed across all S / H amplifiers (as is the case in Figures 5 and 6), and in this architecture (Figure 5), the load on each amplifier channel increases because each amplifier must drive all 960 columns of the display. FIG. 12A illustrates a specific subpixel transmission sequence in which subpixels are still grouped by color and control signals all reach a single dedicated amplifier channel to minimize transmission bandwidth. Table 810 shows the RGB subpixel indices of the subpixels collected by each input amplifier of the source driver. As illustrated, amplifiers (812) are numbered 0 through 15, and each amplifier 0 through 14 receives a series of subpixels (804), receiving them first in the order of red, green, and blue in this example. Amplifier 15 exclusively receives the control track 816 of the control signal. As described above, this has the advantage of fewer transitions, reduced bandwidth, and simpler extraction of control information. Additionally, each amplifier spans only 320 columns, rather than the entire 960 columns.
[0071] While grouping by color offers several advantages and is a desirable architecture for certain applications, 960 subpixels, 15 amplifiers, and 3 colors are not integers, so padding must be applied to the video data. The additional overhead for padding means that 66 samples are transmitted per line instead of 64 per amplifier. In other words, the transmission frequency must be increased by 64 or 66 times, partially neutralizing the purpose of reducing transmission bandwidth by grouping colors. Furthermore, transmitting across 320 columns is not as desirable as transmitting only 64 columns. FIG. 12B shows a specific embodiment in which subpixels are grouped by color and there is a transition band between groups. This embodiment can be used in the implementation of the HiPi integrated circuit chip "1002". Since color has a lower frequency content than luminance, the change in color can be abrupt, so the change is performed more slowly. For example, if the bandwidth of the transmitter channel is not sufficiently wide, it may take time to move from pixel index 958 to 2 (from green to blue). To slow down the speed of this change and minimize bandwidth, a blank transition band (817) is added. This 1002 chip uses a permutation to minimize SAVT bandwidth requirements, where all subpixels of each color are transmitted as a group and a blanked transition band (i.e., blanking transition signal band) is used between groups to lower the required bandwidth between groups. SHA amplifier 0 is a control channel (818) that transmits control signals 0 to 64, i.e., 65 samples per line. As shown in the figure, the red subpixel indices extend from 0 to 957, green from 1 to 958, and blue from 2 to 959. Sample 815 is the red blanking transition signal (TR0 . . . TR4), sample 816 is the green blanking transition signal (TG0 . . . TG4), and sample 817 is the blue blanking transition signal (TB0 . . . TB4). These bands (815, 816, 817) provide blanking transitions between colors. As discussed above, a different architecture is proposed considering that in certain applications, bandwidth limitations may not be critical, control information is virtually random, padding may be undesirable, and grouping control signals into a single channel is advantageous. FIG. 13 shows another architecture of a source driver where each distributor amplifier drives an adjacent column and all control signals are processed by a single amplifier. When bandwidth is adequate, FIG. 13 provides an architecture that minimizes routing. An input terminal (822) is shown that demultiplexes incoming pixel data and control signals from the transmitter and distributes them to an S / H amplifier (pixel data inputs numbered 0 to 14) 824 and an amplifier 826 that receives the control signals. Pixel data from amplifier (824) is transmitted to storage array A (828) or storage array B (830) as described above, and control signals are processed by component (836) and output from 838. Then, pixel data from either storage array is input to a column driver (832) and output to a column (834) as described above. Control logic for controlling the timing of the input amplifier, storage array, and column driver is not shown, but such control logic is as described above with reference to FIG. 5 and 6. Pixel data is received sequentially in a single channel (per chip) and stored sequentially in an A / B collector (at intervals of one Fsavt cycle), but it is also possible to store 15 subpixels in parallel in an array from 15 SHA amplifiers. Therefore, 15 interleaved S / H amplifiers receive incoming pixel data and each drive 64 adjacent columns—that is, 64 video tracks—so the span of the columns driven by each amplifier can be minimized. This architecture provides 15 blocks consisting of 64 video samples and one 64-bit subband channel (control signal) per display line (per source driver). For example, Amplifier 0 drives columns 0 through 63, Amplifier 2 drives columns 64 through 127, Amplifier 15 drives columns 896 through 959, and Amplifier 826 drives the control signal. Having all control signals on one channel means there are no differences in amplitude, delay, etc., between the signals (unlike when they are on different channels). Additionally, the control signal may arrive at Channel 0 (i.e., Amplifier 0) instead of Amplifier 15, which is advantageous in that the control information arrives before the pixel data. Another advantage of this architecture is that synchronization is simplified because control signal extraction is not distributed across all amplifiers, but only one deinterleaving amplifier output needs to be checked. This diagram contains 15 video amplifiers, each driving 64 subpixels = 960 subpixels / chip. There is one dedicated control channel, which delivers 64 symbols per line (per source driver). If MFM is used for timing synchronization (see description below), 64 symbols are transition-encoded, and considering flag and command bits, 24 or 25 control bits remain per line. As shown in the figure, the control channel receives a control signal from the amplifier (826), and this signal is input to a comparator (836) operating at the 16th of Fsavt, or approximately 41.5 MHz, where the reference voltage is 0 V. Assuming the control signal is in the range of -0.5 V to +0.5 V, the comparator detects whether the control signal is greater than 0 V (meaning digital 1) or less than 0 V (meaning digital 0). This digital data is output to 838 to provide a single control bit every 16 samples. The control signal provides synchronization and phase alignment as described below. This particular embodiment is for an 8K144 display, and examples of parameter values are shown in Table 1 above. Those skilled in the art will find it simple to modify the architecture to suit different display sizes and speeds. By rearranging the sample of the transmitter, each interleaved S / H amplifier can drive adjacent columns while rotating and operating as described below. FIG. 14 shows the source driver input of a source driver (820) for interleaving multiple input amplifiers to meet speed requirements. (A single amplifier may be used, but the transmission speed is reduced.) As illustrated, an input terminal (822), distribution amplifiers 0 through 14 (824) and amplifier (826), and an associated switch (842) that rotates to effectively connect one amplifier at a time to receive either an incoming subpixel or a control signal are shown. Thus, the inputs are interleaved in 16 ways, and the outputs of the switches are demultiplexed into 16 channels running at 1 / 16 of the data transmission rate. The 960 subpixels in one line are conveniently grouped into 15 groups, each consisting of 64 subpixels, and one channel is used exclusively for the detection and processing of the control signal. FIG. 15 summarizes the pixel transmission order (300), showing how pixels 0-959 and control signal 0-63 are transmitted from the transmitter to the source driver of FIG. 13 and the amplifiers assigned to each. What is shown is the natural order of subpixels transmitted from the TCON to the source driver, for example, the subpixels are read from left to right and then arrive from top to bottom. Since the input data in the source driver is interleaved in 16 directions, the preferred method of transmitting subpixels to the source driver is from the top left to top to bottom and from left to right, that is, the indices of the transmitted subpixels (and control signals) start with 0, 64, 128, etc. What is shown is an example of the index of the S / H amplifier (302), the subpixel index (304) of the 16th amplifier, and the control track (306). In this permutation, 15 of the amplifiers (0–14) each drive 64 adjacent columns with subpixel values, and amplifier 15 processes all 64 control signals. This variation minimizes the hardware of the source driver and the wiring load of the input amplifiers. Additionally, since this method does not require padding in the data sequence, it supports the slowest possible SAVT (Sampled Analog Video Transmission) transmission rate (64 x 16 sampling per line). To best display text and other abrupt transitions in intensity, it is desirable for the sampling amplifier to be able to set a new value every 1 / Fsavt, i.e., every 1.5ns per sample. To implement this architecture, the order of subpixel indices to be transmitted from the transmitter is 0, 64, 128, .. . 832, 896; 1, 65, . . . 897; . . . ; 63, 127, 191, . . . 895, 959. FIG. 16 is a block diagram of an input vector (320) of a transmitter having a predetermined permutation that provides the sequence of subpixel transmissions required by FIG. 15. As previously described, when subpixels arrive at the distributor from the timing controller, they are distributed into the input vector (320) in the order shown. When the input vector is full, samples of the input vector are serially output to an image processor via output port 321 as described above, converted, and then transmitted to a source driver having the architecture described in FIG. 13. Other input vectors in the line buffer are not shown, each input vector has a similar permutation, and the other source driver corresponding to each input vector has the same architecture as shown in FIG. 13. Control samples at specific locations are also shown. Synchronization signals, subbands, control tracks, and digital data of FIGS. 5, 6, 11, 12A, 12B, 13, and 16, i.e., signals transmitted for control rather than actual video samples to be displayed, such signals may be distributed across a number of S / H amplifiers as illustrated or transmitted on a dedicated channel or track, for example, track 15 of FIG. 12A or track 0 of FIG. 12B. The architecture of the source driver (820) of Figure 13, along with the transmission order above, provides the above advantages while maintaining the slowest possible SAVT clock speed. Accurate sampling of each subpixel within the available time is provided through synchronization as described below. Transmitter integrated with timing controller As mentioned above, in other embodiments, the transmitter is integrated with a timing controller rather than the individual implementation shown in Figure 1. FIG. 7 is a block diagram showing an integrated transmitter and timing controller (640) located immediately after the SoC (120) of the display unit (600). FIG. 7 illustrates inputting a digital video signal (110) via an HDMI connector (or LVDS, HDBaseT, MIPI, IP video, etc.) to a system-on-chip (120) that performs functions such as display controller, inverse compression, brightness, contrast, overlay, etc. Then, the modified digital video signal (664) is transmitted to the integrated transmitter and timing controller (640) using LVDS, V-by-one, etc. In this embodiment, the timing controller is integrated with the transmitter, and both are implemented within a circuit, preferably an integrated circuit of a semiconductor chip. Note that since the transmitter and timing controller chip (640) is located immediately after the SoC chip (120), the transmission of the digital signal (at that point) is easier. Preferably, the chip (640) is located about 10 cm or less from the SoC chip, about 5 cm or less in another embodiment, and about 2 cm or less in another embodiment. Due to the physical characteristics of LVDS, the maximum communication distance between chips is limited to about a few inches. The integrated transmitter / timing controller (640) receives the digital video signal (664), distributes it to line buffers, performs image processing, converts the digital samples into analog samples, and transmits the EM signal to the source driver as described above. Typically, the EM signal (192) is transmitted to the source driver (186) using a differential wire pair (or metal trace) (e.g., one pair per source driver). The gate driver control signal (190) controls the gate driver (160) to synchronize with the source driver so that the correct line of the display is activated. Since each source driver chip performs self-synchronization, a single reference clock (170) of the transmitter and timing controller (640) can be fanned out to all source drivers, but the actual reality of driving intensity may mean that it may be desirable to distribute multiple clocks. In any case, frequency locking between source driver chips is maintained. FIG. 8 shows the integrated transmitter and timing controller (640) in more detail using multiple identical blocks of FIG. 2. The signal (664) is typically an LVDS digital signal as described above. The unpacker (620) unpacks (or exposes) these serial pixel values into parallel RGB values and outputs a frame flag (627) to the splitter controller (630) and the gate driver controller (650). The splitter (642) is positioned to receive the exposed color information (e.g., RGB) from the unpacker (620) and fill line buffers (241 and 242) according to a predetermined permutation. As described above, the controller (630) coordinates the storage and retrieval of pixel values within and from the line buffer. As previously mentioned, the framing flag (627) is provided from the unpacker (620) and input to the distributor controller (630), which uses this flag to determine the position of a pixel within a line, stores it in the correct input vector, and then places it. After the frame flag is output from the controller (630) (typically delayed) and input to the gate driver controller (650), a number of gate driver control signals (671) are generated to control the timing of the gate driver. These signals (671) will include at least one clock signal, at least one frame strobe signal, and at least one line strobe signal. When a pixel value is pushed to the source driver for a specific line, the line-strobe signal is used for the specific line activated by the panel gate driver controller. Thus, the line strobe signal drives the selected line at the appropriate timing. Timing control of the gate driver can be performed as known to those skilled in the art. Additionally, bidirectional communication (637) between the controller (630) and the gate driver controller (650) is also shown, and this communication is used for timing management between the source and the gate driver. The operation of the two line buffers, image processors (250-259), and DACs (260-269) may occur as described above. Preferably, image processing occurs after the unpacker (620) and before the line buffers, in which case the image processing blocks (250-259) are removed and replaced by a single image processing block between 620 and 241, 242. Additionally, as mentioned above, image processing does not need to occur within the transmitter (640) but may occur in the SoC (120) or at another location. Transmitter integrated with timing controller and system-on-chip As mentioned above, in other embodiments, the transmitter is integrated with the timing controller and SoC rather than the individual implementation shown in Figure 1. FIG. 9 is a block diagram showing a transmitter, a timing controller, and an SoC (684) integrated within a display unit (680). In this embodiment, the conversion of a digital video signal (110) into an analog signal (192) is performed within a single chip (684) that integrates the transmitter, the timing controller, and the SoC. As illustrated, a digital video signal (110) is input to a display unit (680) via an HDMI connector (or via LVDS, HDBaseT, MIPI, IP video, etc.) and is internally transmitted to an integrated SoC (111). The SoC performs traditional functions such as display controller, inverse compression, brightness, contrast, overlay, etc. After the SoC performs its traditional functions, the modified digital video signal (not illustrated) is internally transmitted to an integrated transmitter and timing controller using an appropriate protocol such as LVDS, V-by-one, etc. In this embodiment, the timing controller and the transmitter are both integrated with the SoC, and all three are implemented within a single circuit, preferably an integrated circuit of a semiconductor chip. The transmitter in the circuit (684) converts the modified digital video signal into an analog EM signal (192), which is transmitted to the display panel (690). Preferably, the signal (192) is transmitted to the source driver (186) using a differential wire pair (or metal trace) (e.g., one pair per source driver). The gate driver control signal (190) controls the gate driver (160) to synchronize with the source driver so that the correct line of the display is activated. Typically, the distance between the chip (684) and the source driver (186) ranges from about 5 cm to about 1.5 m depending on the panel size. A single reference clock (170) of the transmitter, timing controller, and SoC (684) can be fanned out to all source drivers. The integrated chip (684) can be implemented as described in this specification, i.e., as illustrated in FIGS. 14 and FIGS. 7 and 8, with the understanding that the functions of the transmitter, timing controller, and SoC are all integrated into the same chip or circuit. This embodiment of FIG. 9 has the same advantages as listed above in relation to FIG. 7. Additionally, by integrating the transmitter and timing controller into the SoC chip, additional benefits such as a reduction in the number of chips, reduced complexity, reduced required area, and reduced power consumption can be obtained. Furthermore, a DAC (digital-to-analog converter) for video signal conversion is not required within the display panel or source driver. An example of a video transmission system, FIG. 10, illustrates a video transmission system (700) within a display unit. The drawing illustrates a timing controller (702) that outputs a set of color samples as described above, such as subpixel values in a digital format representing the brightness values of an image or video to be displayed on a display panel (710). The samples are input into a transmitter (704), converted into analog, and transmitted to a source driver array (708) via a low-voltage wiring harness (706) to be displayed on the display panel (710). The transmitter's splitter includes a line buffer (720), an arbitrary number of input vectors (or banks) (722, 726), and a splitter controller (728). RGB samples (or black and white or other color spaces) are received sequentially in the splitter and distributed into input vectors according to a predetermined permutation controlled by the splitter controller (728). In this example, a row-first permutation is used, such that the first part of a row is stored in the input vector (722) and the last part of a row is stored in the input vector (726) from left to right of the received video frame (or image). Thus, when the line buffer (720) is full, it contains all pixel information of the first row of the video frame, which is transmitted and displayed as the first row of the video frame on the display panel (710). Each input vector is read sequentially by the corresponding DAC (732-736), and each sample is converted to analog for transmission. When samples arrive continuously from the timing controller (702), the samples are distributed, converted, and transmitted, and finally displayed as video on the display panel (710). As shown and described in Figures 2, 3 and 8, there may be two or more line buffers. Connecting the transmitter (704) to the source driver array (708) is a low-voltage wiring harness (706) consisting of differential wire pairs (742-746), each wire pair transmitting a continuous stream of analog samples (electromagnetic or EM signals) from one of the DACs (732-736). Each differential wire pair is terminated at input 760 of one of the source drivers 752-756. Other transmission media (e.g., wireless, optical) may also be used instead of the wiring harness. Each source driver in a source driver array, such as the source driver (752), includes an input terminal (760), a collector (762), and a plurality of column drivers (764) (corresponding to the number of samples in each input vector in this example). Samples are received continuously at the input terminal (760) and then collected by the collector (762), which can be implemented as a one-dimensional storage array or an array having a length equal to the size of the input vector. Each collector can be implemented using an A / B sampler (storage array) as shown in Figure 5 or Figure 6. When each collector is full, all collected samples are output in parallel to all column drivers (764) of all source drivers, amplified to the appropriate high voltage required for the display panel, and output to the columns (766) using a single-ended format. As samples continuously arrive through the wiring harness, each collector continues to collect samples and output them to the display panel, thereby affecting the video representation. Synchronization of sampled analog video transmission Synchronization can be used to provide horizontal synchronization (start of a display line), vertical synchronization (first display line of a frame), and sample phase alignment (when sampling incoming subpixel samples). That is, a receiver, such as a source driver receiving a video pixel stream, requires information from a transmitter that tells it the starting position of the frame, the starting position of the line, and the point at which to sample data representing a specific subpixel. For example, each switch (842) must know when the input samples are valid and stable so that it can sample accurate values, and determines the sampling time through a process called sample phase alignment. Sample phase alignment accounts for various delays from the TCON to the geographically distributed source drivers of the display and optimizes the phase of the locally generated clock (171) (derived from the reference clock (170)) for locally transmitted samples as described in detail below. Synchronization is useful for several reasons, but it can also make things difficult. First, a constant stream of video subpixels inherently lacks information indicating the start of a frame, the start of a line, or phase alignment data. Additionally, cable delay from the transmitter to the receiver is potentially variable and generally unknown. Cable attenuation (which can vary depending on the path to different source driver chips) can also be an issue. Finally, the waveform of the received subpixel values may be unknown or variable due to ringing, overshoot, filtering, the rate of change of input values, or other signal distortions. Recognizing that synchronization is important but can be difficult due to the factors mentioned above, the techniques described here provide commands for synchronizing and phase aligning analog samples. Here, a timing criterion called a "flag" (i.e., a special timing violation that does not occur during normal data transmission) is used to inform the receiver, i.e., the source driver, that the clock can be reset, followed by a known command or data for synchronization. For example, when the flag is received, a command is received to start phase alignment, and then the optimal sampling phase for sampling the incoming subpixel is determined. First, the sample transmission frequency (i.e., the rate at which samples arrive at the input terminal of the source driver) is known; in this example, Fsavt is approximately 664 MHz (673.92 MHz for the HY1002). Since transmitting such a high-frequency clock may be impractical, in one embodiment, a slower clock, Fsavt / 64 = 10.375 MHz (10.53 MHz for the HY1002), is transmitted to each source driver, and each source driver uses a phase-locked loop to increment that frequency up to the high-frequency clock. Also, it can be seen that 15 subpixel data streams arrive at Fsavt / 16 from each input amplifier, each input amplifier delivers 64 samples, there is a control stream (analog or digital control signal) arriving at the 16th amplifier, and there are 64 control signals per line. FIG. 17A shows the source driver of FIG. 13, which shows the control channel in more detail, and three comparators used to extract phase alignment information from the control signal. In this particular embodiment, the control channel processed by the amplifier (826) includes three comparators (835-837). Comparator (836) outputs a logic "1" or a logic "0" depending on the value of the control signal as described above. Comparator 835 has a high reference voltage used to detect the upper threshold, and comparator 837 has a low reference voltage used to detect the lower threshold, and these references can be set by the DAC. Comparators 835-837 operate at Fsavt = 650 MSPS. The central comparator is reliable as it acts as a zero-crossing detector, and generally, the data extracted from this central comparator matches the data extracted from the high and low channels, assuming everything is functioning correctly (i.e., if the control signal is +.4V, both the central comparator 836 and the high comparator 835 detect a logic "1"). However, if sampling occurs at an incorrect time, the central comparator will provide the correct bit, but the high and low values of the other two comparators are highly likely to be lost. The concept here is that high and low data can only receive correct values when the input is nearly stable, whereas the zero-crossing detector can receive correct values even if the input samples are still fluctuating. It is also possible to use only one of the high or low comparators in conjunction with the central comparator. The use of this information is explained below in relation to phase alignment. FIG. 17B shows an alternative source driver 820'' for the source driver of FIG. 17A, showing more detailed information and control signals from the first amplifier. A timing generator 821 (described below), an input terminal 822, a DAC 838 for adjustable thresholds, and a sampler block 839 are shown. Comparators (835-837) are clock comparators for extracting control signals. Amplifier 832 is an amplifier stage including a preamplifier, level conversion, and high-voltage driver. Sixteen interleaved sampling amplifiers (SHA amplifiers and offset control) with offset removal functions, including amplifier (826), are shown. Preferably, in this embodiment, amplifier 0 (826) is used for the control signal (instead of amplifier 15 as in FIG. 17A) so that the control information arrives with a time margin before the display line time ends. This allows some time to decode the control channel and set the signal to be used within the next line time. FIG. 17C shows a source driver 820''' preferred for the source driver of FIG. 17B. A timing generator (821), an input terminal (822), and a sampling block (839) are illustrated. A comparator (836) is a comparator for extracting control signals. Amplifier (832) is an amplifier stage including a preamplifier, a level converter, and a high-voltage driver. Sixteen interleaved sampling amplifiers (SHA amplifiers and offset control) with offset removal functions, including amplifier (826), are illustrated. Preferably, in this embodiment, amplifier 0 (826) is used for the control signal (instead of amplifier 15 as in FIG. 17A) so that the control information arrives with a time margin before the display line time ends. This allows some time to decode the control channel and set the signal to be used within the next line time. In this embodiment, it can be seen that synchronization requires only a single comparator 836 (zero-crossing detector) for a single SHA channel and that a DAC is not required to set the comparison threshold. The algorithm for synchronization is executed in the digital domain (zero-crossing detector output) and can perform both clock-level synchronization (alignment of SHA outputs so that the side channel is visible at one specific SHA output) and phase-level synchronization (selection of the optimal sampling phase within a clock cycle). Input terminal 822 features a single analog input differential with matching termination and ESD protection functions. This is driven by a source impedance of 50R per side via a 50R transmission line. Consequently, the receive voltage is reduced by 50% relative to the transmit voltage. The PLL of the 821 increments 11 selectable phases (e.g., approximately 675 MHz for the HY1002) per clock cycle of a relatively slow reference clock 170 (e.g., TCON) up to a maximum speed Fsavt clock 171 (e.g., approximately 675 MHz) (e.g., selectable per clock cycle). Additionally, high-speed timing generation allows for the creation of a sampling strobe, a reset signal, and an output transmit strobe for SHA amplifiers 0-15. The 16-way deinterleaver 840 is constructed using SHA amplifiers as shown in Figure 14, and the ON switch rotates so that only one is actually turned on at a time. (In the HY1002, two amplifiers are always turned on simultaneously and overlap for one SAVT cycle.) Therefore, 16 consecutive samples from 16 amplifiers are sequentially deinterleaved, allowing each amplifier more time to stabilize. As shown in the figure, 15 SHAs drive the display columns by driving 64 adjacent subpixel columns, each consisting of a preamplifier, a level converter (differential to single-ended), and a high-voltage driver. One of the SHAs drives the control samples (each control sample has a 16-sample interval). The control samples represent digital values to make the system robust using a type of switching coding (e.g., MFM) to provide timing and control information. A bandgap voltage reference circuit (840) provides current and voltage references for various input amplifiers. Figure 17D summarizes the order of subpixels collected by the input amplifiers of Figure 17C. This summary shows how pixels 0-959 and control signals 0-63 are transmitted to the source driver of Figure 17C and the amplifier assigned to each subpixel. Due to the 16-way interleaving of the source driver, the preferred method of transmitting subpixels to the source driver is to start from the top left, go from top to bottom, and proceed from left to right, so that the indices of the transmitted subpixels (and control signals) are transmitted sequentially to the 16 amplifiers, such as 0, 64, 128, etc. The figure shows the index of the S / H amplifier (842), an example of the subpixel index (844), and the control signal (306) of the 0th amplifier (826). This subpixel order minimizes the hardware of the source driver and the wiring load of the input amplifier. To best display text and other abrupt transitions in intensity, it is recommended that the sampling amplifier be set to 1 / Fsavt, i.e., a new value every approximately 1.5ns per sample. As shown, SHA 0 carries control and timing, while SHA 1-15 carry video data, allowing each SHA to drive 64 adjacent columns of the display. Since the SHAs are sampled sequentially, the transmission order is as follows: CTL[0], V[0], V
[64] , . . .V
[0896] , CTL[1], V[1], V
[65] , . . .V
[0897] , . . ., CTL
[63] , V
[63] , V
[0127] , . . . V
[0959] . This order provides 64 control bits per line and 960 video samples per line, and a total of 1,024 samples are transmitted per line (per source driver). FIG. 18 illustrates a technique for introducing timing references into a control signal sequence to provide synchronization and other commands to the control channel of a source driver. Since control signals appear as a continuous stream of bits, timing references are introduced into the control sequence to interpret the sequence. In radio transmission, Modulated Frequency Modulation (MFM) is typically used to provide timing references. While MFM is known in the art, it has never been used for encoding control sequences for video transmission. It is known that, in the context of the present invention, timing references can be introduced and used to transmit commands and data. Furthermore, MFM has not previously been used to transmit these types of commands in the context of video transmission. For example, by using horizontal and vertical synchronization commands together with specific parameters, all parameters supported by CEDS can be transmitted. Additionally, phase alignment commands are a new feature, as are the specific phase alignment techniques described below. The timing reference indicates the point in time following the command and data in the control sequence, and the timing reference is an MFM flag, which is an intentional timing violation. It is assumed that the wire length from Tx to Rx exceeds one SAVT cycle and that the wire length can be a variable. Since digital data is extracted from the control channel, it is robust even if the analog samples are not perfect. Additionally, the level value is not important, only the transition value is important, and the true and flat values have the same meaning. The timing signal (852) is Fsavt / 16, which corresponds to the output timing of the input amplifier, i.e., the rate at which each amplifier outputs data. The control bit cell (854) represents a 64-bit sequence received as a control signal from the output (838) of one of the source drivers. The MFM cell (856) represents a bit encoded in one MFM cell for every two control bits, and the payload (858) is the command and data. The control sequence (860) is a sequence of control bits received, for example, from one of the source drivers to the control channel and output from the control (838) of the source driver (820'''). As described, the control sequence includes an MFM flag (862), which is a sequence that does not typically occur in a stream of control bits. Flag 862 consists of a sequence in which control bits are switched at 4-3-4 intervals, with the end of the fifth switch marking the end of the flag (a timing violation). Then there is a trailing 0 864, which is an MFM-encoded 0 that is ignored by the data receiver before the actual payload begins. The payload then typically transmits the LSB first, but the MSB may also be transmitted first. In the control sequence, the LSB 866, located at the 0 position of the MFM bit cell 856, has a value of "0". A total of 25 MFM-encoded cells are transmitted, and the payload 867 is shown to the right of the control sequence. The second control sequence shown at 868 differs from sequence 860 but has the same MFM flag and trailing 0; however, since the payload being transmitted is different, it reflects the various commands and data that can be transmitted over a single control channel. Another example of a control sequence is found at 869. The payload being transmitted may represent commands, parameters, or be reserved for later use. Synchronization is complex because the deinterleaving amplifier receives data across 16 channels. The source driver cannot know which channel holds the control sequence until synchronization is achieved. One proposed method is to use a flag sequence that appears on only one channel output. Identifying the MFM flag instructs the source driver chip to resynchronize and prepare commands or data, such as horizontal or vertical synchronization commands or phase alignment modes. Upon power-on (or after a power outage), the transmitter sends the MFM flag and control sequence on all channels, and the correct control channel (channel 16 in this example) recognizes the flag, resynchronizes the timing, and recognizes the commands and parameters. Once resynchronization is complete, the control sequence only needs to be transmitted on the control channel, and video data can be transmitted on the remaining 15 channels. Another more desirable method is to initially transmit the MFM flag on a single channel rather than all 16 channels. The receiver looks for the flag on one channel (which may be the wrong channel before synchronization is complete). If the flag is not detected after one line time (~1.5 µs), the clock sleeps (skips) for one cycle, effectively rotating the amplifier usage. Therefore, it may take up to 16 display line times to synchronize with the clock cycle. Because the control stream is continuous, commands and parameters can be extended across multiple display lines if necessary. Furthermore, while conventional CEDS transmits approximately 28 to 32 control bits per display line, it is known that a low transmission frequency is sufficient, as some of these bits do not need to be transmitted per line (e.g., cold mode, power control, etc., and some may be frame parameters). Additionally, the control channel disclosed herein is robust enough that a CRC is not required because it extracts digital data and the same channel transmits analog video samples with 10-bit accuracy. Nevertheless, a CRC may be added. Another advantage of using MFM in the context of video transmission is that it provides an immediate and accurate timing reference when a flag occurs, so there is no need to wait for correlation as is the case with other technologies such as Kronecker. Moreover, this control channel is suitable for transmitting commands such as frame synchronization, line synchronization, parameter data, and phase alignment information. Commands are distributed (transmitted sequentially, one symbol at a time, across a single line), and the received control information is applied to the next display line. Generally, the control sequence does not end. One control packet is transmitted per line cycle. Information such as the polarity and driver strength of the column driver is transmitted per line. FIG. 19A shows an embodiment of transmitting commands and parameters via MFM encoding. As described above, the timing clock Fsavt / 16 at the bottom and the control bit field (879) used to encode MFM data are shown. FIG. 2 shows a control sequence (870) containing an initial MFM flag and 25 MFM cells as described above. Control sequence (871) shows how to use 3 MFM cells to hold 3-cell commands, leaving 21 cells for parameters. Depending on the implementation, more or fewer cells may be used for commands. Sequence 872 shows the location of the MFM flag associated with the MFM cells. Sequence 873 shows 21 cells available for 3-cell horizontal synchronization commands and parameters. Sequence 874 shows 21 cells available for 3-cell vertical synchronization commands and parameters. Sequence 875 describes parameters that may include a phase alignment mode setting command and a comparator threshold. Sequence 876 shows the phase alignment mode exit command. This command is not strictly necessary, as phase alignment mode exits upon receiving another command. Sequence 877 shows the possible values for the command and the space for its parameters reserved for future use. The synchronization stream (878) is a stream continuously transmitted on the control channel during phase alignment mode. To determine the threshold of the upper comparator (835) (where the phase alignment of FIG. 20 is used), this stream is considered to be modulated to include two levels, namely 1V for every different pulse and 1.2V for every different pulse, but these voltages may vary. Phase alignment is typically performed before turning on the display and during frame blanking (since the phase best suited for sampling may change due to heat or other effects). Once phase alignment is established, the phase alignment mode is terminated and the video is displayed normally. Figure 19B shows another embodiment of transmitting commands and parameters via MFM encoding. While Figure 19A shows different command lines where each command is identified by a different 3-cell command, Figure 19B shows a single command line containing parameters for various commands. An MFM cell 856 is shown, containing the command / parameter 858 and control bit 854, which includes flag 862 and trailing 0 864. As indicated, two samples are used to represent each MFM cell, with one MFM packet per line, and an MFM flag is prepended to each MFM packet with a (4, 3, 4, 2) transition. Thus, 25 MFM cells are generated per line. Upon receiving the flag, all data reception is terminated, and the 25 command / parameter cells are not valid until all 25 are received; if a new flag is received before then, it is ignored. Depending on the last state of the previous message, a normal message (1) or (2) may be used. During initial synchronization, up to 4 flags per line can be sent using fast flag messages to increase the synchronization speed and provide control signals to channel 0 faster. The point is that parameter and control information can be transmitted line by line along with sampled analog video data. When VSYNC 881 is asserted, it indicates that the currently receiving line is in a vertical blanking period, so no data is displayed and the video controller state machine is re-initialized. Polarity control bit 882 determines the polarity with which column pairs are driven based on the dark level. Each column pair is driven in a complementary manner (one column output is driven more positively than the dark level, and an adjacent column is driven more negatively), and the direction is determined by the polarity control of the column pair. The four polarity controls independently control the four column pairs in an 8-column group, and the pattern repeats every 8 columns for all sub-pixel columns in a line. Polarity controls can be updated for each line. In practice, however, to reduce power consumption, it is likely that the polarity control bits are changed at most once per line. The short-circuit control bit 883 contains "short_gena"; if this bit is asserted, adjacent columns are short-circuited only when the corresponding polarity control bit is changed, and short_all is not asserted either. If the assertion is released, no shorting occurs at all. "Short_all" enables shorting of all column pairs regardless of the polarity control change state, but is effective only if short_gena is also asserted. The drive time control 884 specifies the number of Fsavt / 16 cycles from the start of the high-voltage driver drive period until the driver reaches the 3-state or charge-sharing state (depending on SHORTCTL). The high-voltage driver sampling phase 885 is a chopper clock that swaps both the input and output of the main amplifier to cancel the offset. The SHA calibration control signal 886 contains two SHA calibration control signals, sha_video(sha_cal[0]) and sha_meas(sha_cal[1]), which can be directly controlled from the side channel control bits. This signal controls the Calibration_Phase1 and Calibration_Phase2 signals of the SHA, respectively. Phase Alignment Techniques As mentioned above, it is desirable to determine the optimal sampling phase of the received samples for a specific source driver due to cable delay and attenuation, the presence of ringing in input samples, and inter-symbol interference. This phase alignment can be performed at regular intervals, such as when powering on or during frame synchronization or frame blanking, and the phase alignment mode can be set by executing the "Phase Alignment Mode Setting" described above. Two phase alignment techniques are proposed below. FIG. 20A illustrates a useful technique for performing phase alignment when there are three comparators in the source driver of FIG. 17A and 17B. The synchronization stream (878) is shown in more detail (890), and for convenience of explanation, a square wave is shown. Each sample or pulse of the stream (890) is 16 SAVT sample lengths (sampled output of a single channel of the deinterleaving amplifier), and 16 identical values can be seen when compared to the threshold at the SAVT rate. For example, it is desirable to modulate the stream using a triangular wave or a trapezoidal wave to generate two different levels 891 and 892 (and corresponding negative levels 897 and 898). For the sake of convenience of explanation, levels of approximately 1.0V and 1.2V (and corresponding negative levels of -1.0V and 1.2V) are shown, but in reality, it is considered that the positive level is approximately 0.5V and the negative level is approximately -0.5V. As shown in the figure, the synchronization stream continuously repeats two positive levels and two negative levels until the phase alignment mode ends. When entering phase alignment mode, a synchronization stream (878) is transmitted along the control channel of the source driver (e.g., the synchronization stream arrives at the input amplifier (826), and this stream is detected by the center comparator (836), the upper threshold comparator (835), and the lower threshold comparator (837). This synchronization stream is preferably a valid MFM data stream (i.e., MFM 0 and MFM value) having a regular 50% duty pattern of positive and negative values of known amplitudes. Since this is a valid MFM data stream, the "exit phase alignment mode" command can be executed at any time. Comparators 835 and 837 must be sufficiently fast, but absolute accuracy is not required if the offset is smaller than the difference between the last two amplitude levels. Basically, the central comparator (836) provides zero-crossing detection and indicates whether the detected input is positive or negative. If the sample is positive and the sampling step is correct, the upper threshold comparator (835) must also generate a positive value; otherwise, it means that sampling occurred too early or too late, that is, before or after the transition to a positive value. The lower threshold comparator (837) provides similar information if the sample is negative. If the upper or lower comparator does not match the central comparator, the sampling phase is adjusted. A detailed technique for adjusting the phase to accurately sample positive inputs is described below, and those skilled in the art will be able to apply this technique to negative inputs. For example, the upper comparator must report the same value as detected by the zero-crossing detector. As the sampling phase rotates (by advancing the phase in the PLL), it eventually reaches the point where the next transition begins to occur. This transition causes the upper comparator to provide a result that does not match the zero-crossing detector. Then, recognizing that a transition has been detected, the sampling step is set back one step (or two steps for safety) to allow sampling in the latter part of the symbol cycle, after the sample has stabilized and before transitioning to the next sample. Note that if the transition is very fast, the zero-crossing detector may also flip when the symbol transition occurs; in this case, since there is no need to look at the upper comparator, the "transition" can be determined by the OR of these two events. As mentioned, the first step is to enter phase alignment mode and transmit a synchronization stream (890) along the control channel. Preferably, the amplitudes (891 and 892) are set far enough apart to provide a window (in this example, about 0.2 V) that can set an upper threshold (893) so that pulse (892) triggers without pulse (891) triggering the upper comparator when the sampling phase is approximately accurate, and that the pulse (891) does not trigger the upper comparator when the sampling phase is approximately accurate. For example, if the expected amplitudes of the control signal are about 1.5 V and about -1.5 V, the amplitudes of pulses (891, 892) are set below the expected amplitudes as shown in the figure. The corresponding amplitudes of pulses 897 and 898 can also be set in the same way. One goal in selecting the initial voltage for these two pulse amplitudes is to set the modulation level so that the transition can be detected. In one embodiment, 50% amplitude and 75% amplitude (positive pulse) are used for these two amplitudes of the synchronization stream. This allows the DAC threshold between the two amplitudes to be easily set (allowing for noise, etc.) and to clearly identify when the transition occurs (when the 75% amplitude pulse falls below the upper DAC threshold). The initial sampling phase can be selected randomly, a reset value (e.g., phase 0), or another phase. Since the zero-crossing detector is used to determine the expected signal level, it is highly unlikely that the sampling phase will be selected during a symbol transition where the zero-crossing output appears randomly (a probability of ~1 / 16, but possible). However, if this occurs and the flag is not displayed after all 16 clock skips have finished, the phase is advanced to a position where the zero-crossing detector can operate. The implementation has 11 phases, and it is expected that moving the phase two or three positions will be sufficient. Other implementations may differ. When the synchronization stream arrives, the source driver's logic and circuitry (not shown) can determine the optimal voltage by adjusting the upper threshold up or down. For example, if the upper threshold is too low, the upper comparator is triggered on both pulses 891 and 892; if the upper threshold is too high, it is not triggered on pulse 892; and if the upper threshold is correctly set, it is not triggered on pulse 891 but is triggered on pulse 892. The source driver cannot know the amplitudes of pulses 891 and 892 due to attenuation, etc., but if the amplitudes are sufficiently spaced apart to accurately set the upper threshold, the source driver does not need to know the exact values. This adjustment process uses a sampling step that is approximately accurate but not optimal. If the upper threshold is correctly set, it is not triggered by the ringing of pulse 891 but by pulse 892. The upper threshold can be adjusted using a DAC. Once the upper threshold is set, the source driver's logic and circuitry begin rotating the sampling phase around 11 different phase positions to determine the phase best suited for the sample. For example, the initial sampling phase might occur approximately in the middle of pulse 892, but this may not be the optimal point for sampling as the pulse is not stable at this point and could yield an incorrect value. Generally, the best time to sample is just before the transition to the next pulse, when the signal is most stable—that is, before the trailing edge of the pulse. When the sampling phase rotates to point 894, both the center comparator and the top comparator are triggered to signal that a positive value has been received, and when it rotates to point 895, both are triggered again. However, when the sampling phase rotates to point 896, the top comparator is suddenly not triggered, indicating that the transition has just passed. There is no correspondence between the top comparator and the center comparator. (Even if a perfect square wave is displayed, the center comparator still signals a positive value because the transition drops gradually rather than abruptly to -1.2V). Therefore, by returning the phase tap one or two steps earlier, namely to point 895 or 894, the optimal sampling phase can be found. A different number of phase positions can be used, and 11 were determined by the process (the number of inversion steps that fit within a 1.5ns clock cycle). Depending on the VCO structure, odd positions may also work well. By using two pulses (891, 892) to set an upper threshold, one can be assured that the upper threshold is sufficiently high to search for the optimal sampling phase as described below. Providing an upper threshold lower than the amplitude of pulse (892) ensures that there is no correspondence between the upper comparator and the center comparator when sampling occurs after the pulse (892) transition, making it easier to select the optimal sampling phase. Although a single sampling phase detected once can be used, it is preferable to average multiple measurements to handle noise and overshoot. The above describes a technique for determining the optimal sampling phase of a positive pulse. The same technique can be applied to a negative pulse, and the results can be averaged. In one embodiment, the lower threshold comparator (837) is not required, and only comparators (835 and 836) are used to determine the optimal sampling phase using the positive pulse as described above. In another embodiment, the upper threshold comparator is not used, and only the lower threshold comparator and the center comparator are used in relation to the negative pulse to determine the optimal sampling phase. In yet another embodiment, the upper threshold comparator is used only for positive pulses that all have the same amplitude in the synchronization stream, and the upper threshold is set lower than the amplitude of these positive pulses, and the sampling phase is rotated back and forth depending on when the trigger of this upper comparator is interrupted. The upper threshold comparator may also be used exclusively only when the synchronization stream contains alternating positive pulses of different amplitudes, as illustrated in Fig. 20A. Similarly, in another embodiment, the lower threshold comparator is used exclusively only when there are only negative pulses having the same amplitude in the synchronization stream, the lower threshold is set lower than the amplitude of these negative pulses, and the sampling phase is rotated back and forth depending on when the trigger of this lower comparator is interrupted. In this embodiment, the negative pulses may be as shown in Figure 20A. Figure 20A above illustrates one technique for phase alignment. While phase alignment commands included in the video stream can be used, it is preferable to perform phase alignment by a state machine that monitors flag timing in the MFM sequence, in which case the commands in the stream are not used and the payload dynamically changes the parameters. Below, we describe a second preferred technique in which phase alignment is combined with retrieving MFM flags (or other flags) from the control sequence. Figure 20B shows a sampling phase adjustment circuit useful for the source driver of Figures 17C and 20C-20F. The input controls are clock cycle and phase adjustment controls, and the sample_phase output (sampling clock) is sent to each SHA amplifier. Phase rotation (and clock skipping) affects all sampling. This occurs first in the PLL as shown in Figure 20B. Skipping the clock cycle allows the SHA channel where control information is displayed to be easily changed. (Since the sample-and-hold amplifier / deinterleaver samples to a different channel every cycle, skipping the clock cycle rotates the information delivered to each channel by one.) Phase adjustment is performed by rotating (selecting) one of the 11 phases of the PLL (only one phase step is allowed at a time to prevent glitch clocking). The VCO high-frequency clock is the clock whose phase is adjusted to generate the sampling clock output. The input reference clock (170) is kept constant. The inputs "skip", "adv", and "ret" come from the synchronization state machine in the chip's digital logic as described in Figure 20F.
[0139] Figure 20C illustrates a special synchronization video pattern designed to facilitate locking. During synchronization, this special video pattern is used where all video samples in a line (per source driver) are negative constant values (zero-crossing detector output is 0), except for the last video sample 944, which is transmitted as a positive constant value (zero-crossing detector output becomes 1). The figure shows control data 940 for SHA Channel 0 and SH amplifier 1 15 (Video 0 - Video 14). Therefore, using this knowledge, one can determine the direction in which the phase must be adjusted to sample the control bit at the optimal time when SHA Channel 0 (the control channel) attempts to extract control data by sampling the video stream. For example, sampling at data 940 and shifting the phase backward results in a "1," while shifting the phase forward results in a "0." Since all SHA channels share the same (relative) timing with respect to the sampling step, this also implies that the video samples are sampled optimally (at the end of the sample's settling time). Figure 20D shows an example where the MFM flag is lost due to a wrap-around of the PLL phase. In this situation, a clock skip is required. The 11th sampling phase step 950 is shown, and the optimal step is 952. As shown in the figure, the PLL phase rotates from step 10 to step 0, shifting the sampling back to the previously transmitted sample 949. If the currently selected sampling phase enters the CTRL field, the flag is detected (Y). If it falls outside the CTRL field, the flag is not displayed (N: illegal MFM in the video channel), indicating that the sampling has shifted too far back, and the sampling step can be shifted forward by 3 steps. Figure 20E shows an example where the MFM flag is lost when the phase extends past the end of the control bit. Here, sampling at steps 9 and 10 is carried over to the next sample, 942. In this situation, no clock skip is required, and if MFM flag detection is lost at step 9, the phase selection is shifted two steps back to the optimal step 7. Figure 20F is a flow diagram of the synchronization state machine implemented in the digital logic of the source driver chip, illustrating how phase alignment occurs through MFM flag detection. In this diagram, EOL signifies the end of receiving the display line (64 control samples received for a single source driver), V=1 signifies a zero-crossing detector output of 1, indicating that the last video sample of the previous line is being viewed instead of the control bit when the MFM flag disappears, V=0 signifies a zero-crossing detector output of 0, indicating that the first video sample of the same line is being viewed instead of the control bit when the MFM flag disappears, and "!" signifies negation (e.g., "! flag not detected" means the flag is detected). If the flag is not detected at step 974 and the detector output is 1, the process moves to step 972, as sampling too far forward to sample 949 of the previous line as shown in Figure 20D. Therefore, step 973 shifts the phase forward by 3 steps. If the flag is not detected at step 974 and the detector output is 0, the process moves to step 976, as shown in Figure 20E, as sampling too far forward to the next sample 942. Therefore, at step 976, the phase shifts backward by 2 steps to the optimal phase, and synchronization is achieved at step 977. Once the optimal phase is determined, the output of the sampling phase adjustment circuit is transmitted as the sampling clock to the SHA amplifier and implemented within the source driver. It is desirable for all amplifiers in all source drivers to work together; that is, there exists only one sampling phase alignment circuit controlling all SHA amplifiers and one clock cycle alignment. Within each source driver, these input SHA amplifiers are time-interleaved to generate sample outputs that are temporally distorted by 1 Fsavt cycle between adjacent channels. The SHA amplifier then transmits these samples to the collector (A / B sampler), but after the collector has gathered all samples from a line, the preamplifier transmits all samples simultaneously to the next stage (level converter). This effectively 'wast' 16 Fsavt cycles of transmission time at the preamplifier output, but there is sufficient time because the sampling rate is reduced. Other synchronization techniques can also be used. For example, a low-frequency clock can be transmitted to provide horizontal and vertical synchronization. Another technique to adjust the phase for the sampling position is to transmit a known black / white reference in the subband and tune the receiver's PLL until the blackest and whitest values are found. In another synchronization technology, the reference clock is not merely a reference clock but also includes data (e.g., parameters), although its frequency is lower. The clock and its parameters are transmitted via a separate wire from the SAVT samples (which already exists). Since there is no need to mix side-channel data with video data, the SAVT speed is reduced, and because only 60 * 16 = 960 samples need to be transmitted per line, lower bandwidth is required for communication. Using subpixel color grouping further reduces bandwidth requirements. Additionally, because the video stream does not contain side-channel bits, there is no problem with side-channel bits bleeding into the video bits. Example source driver Figure 21 shows the analog data path of one channel of an example source driver. In this particular example, the source driver drives an LCD display, and some of the components shown are specialized for that type of display. However, this invention is applicable to other types of displays as well. Input terminal 902 and SHA[0] 824 (indicated on 904), one of the 16 input distribution amplifiers, are shown. Switch 842 on the input terminal and the other 15 distribution amplifiers driving the video samples are not shown. Input sampling is indicated by the switch sampling to the capacitor. The input sampling switch is symbolically controlled by signals b and t. There are 15 other identical amplifiers (with time delays applied) transmitting the video, and SHA[0] transmits additional channel information. The SHA channel transmitting the additional channel can be selected arbitrarily, but the advantage of using SHA[0] is that control information arrives before, rather than after, the video samples, allowing for setup time before the control information is needed. Each amplifier 904 has a nominal gain of 1, but it can vary. Each SHA channel drives 64 columns 920 through the sampling block / collector 908, preamplifier 910, level converter 912, HV driver 914, and column short switch 918, according to the array notation [63:0] used in the component specifiers of the figure. The level converter 912 is also called a differential-single-ended converter. The preamplifier 910 provides the necessary gain for the signal coming from the transmission medium. Figure 21 shows one of the 64 A / B sampling capacitor blocks 908 connected to the channel's amplifier 904 and associated preamplifier 910. In a specific implementation example, timing skew caused by the deinterleaving process of the input distribution amplifier can be corrected through the source driver's preamplifier timing. For example, when all 64 samples arrive at either the A or B sampling capacitor block, preamplifier 910 waits 16 clock cycles after the first sample arrives (or waits until the last sample arrives) and then begins sending all 64 samples to the level converter 912. Thus, all 960 preamplifiers of the source driver send samples simultaneously. 912, one of the channel's 64 level converters, converts the differential signal into a single-ended signal, adds an offset, changes the signal polarity, and provides amplification. Output out_p 913 = vmax + 0.5*(Vinp-Vinn) if pol=0. Output out_p 913 = vmin - 0.5*(Vinp-Vinn) if pol=1. The high-voltage driver 914 is one of the channel's 64 drivers, amplifying the input signal to provide the voltage (plus or minus) required by the display. The column short switch 918 provides a short circuit to the LCD display, as is known in technology. Finally, the expected voltage is output to the column via 920. The preamplifier 910, level converter 912, and HV driver 914 can be considered as amplification stages preceding each column, in which case they may be called pipeline amplifiers or simply “amplifiers.” Switch 842, the source driver's input amplifier 904, and the A / B sampling block 908 in Figure 14 may also be referred to as collector 915. This is because the collector receives the input serial analog samples and stores (or “collects”) them in the sampling block. The stored analog samples are then output in parallel to the column driver 916 to be displayed as a single line on the display panel. Mobile phone specific implementation example Figure 22 is a block diagram illustrating the transmission of analog video samples within a mobile phone. U.S. Patent Application No. 18 / 442,447, titled “Video Transmission within a Mobile Device” (Patent Attorney Registration No. HYFYP017), discloses details regarding various technologies. Existing displays in conventional OLED DDIC devices (e.g., mobile phones) require improvement due to the high refresh rates of 4K smartphone displays, MIPI receivers, SRAM, digital image processing, and the extensive use of analog signals requiring approximately 1,000 digital-to-analog converters (DACs). The split OLED DDIC architecture shown in Figure 22 has the following advantages: it enables optimized splitting of DDIC-TCON and DDIC-SD; provides short-distance MIPI transmission from the SoC; optimizes the digital DDIC-TCON for SRAM and image processing; provides a simplified DDIC composed entirely of analogs; and requires only a few digital-to-analog converters integrated with the transmitter in the DDIC-TCON. The mobile phone (or smartphone) 980 shown in the figure may be a similar portable mobile device for communication and image or video display. Device 980 includes a display panel 982, a traditional mobile SoC 984, an integrated DDIC-TCON (Display Driver IC - Timing Controller) and transmitter module 988, an integrated analog DDIC-SD (DDIC - Source Driver) and receiver 992. Although the mobile SoC 984 and module 988 are shown outside the mobile phone for the sake of illustrative convenience, they are actually internal components of the phone.
[0156] The mobile SoC 984 is a standard SoC used in mobile devices and transmits digital video samples to Module 988 via the MIPI DSI 986 (Mobile Industrial Processor Interface Display Serial Interface). This operates in a manner similar to the Vx1 input signal discussed above. Module 988 includes a DDIC-TCON integrated with the transmitter, as described above. For example, it is similar to the transmitter shown in Figures 1-4 and 7-10. By reading this disclosure and referring to the previous drawings, a technician in the field can understand how to implement the transmitter to output an arbitrary number of analog EM signals 990. In this example, the transmitter outputs 12 pairs of analog EM signals at a rate of 380 Msps. The gate driver control signals transmitted from Module 988 to the gate driver of the display panel 982 are not shown. Typically, in mobile phones, the DDIC is located at the bottom narrow edge of the phone, while the SoC is located in the middle part of the device. Therefore, the integrated DDIC-TCON / transmitter is located near the SoC, within approximately 10 cm or even within approximately 1-2 cm. Since digital data transmission takes place at ultra-high frequencies, it is advantageous to keep the conductor length as short as possible. For tablet computers, the distance is within approximately 25-30 cm. This analog signal 990 is received by the integrated analog DDIC-SD and receiver 992. The DDIC-SD receiver 992 receives multiple pairs of analog signals and generates a voltage to drive the display panel 982, which can be implemented, for example, as shown in Figures 5, 6, 10, 13, or 17. Advantageously, only a single source driver may be required to drive the display panel 982, and the module 992 does not require a digital-to-analog converter. The Analog DDIC-SD Receiver 992 can be a single integrated circuit containing 12 source drivers internally (each processing a single pair) or can be composed of 12 independent integrated circuits, each circuit functioning as a source driver and processing one of the 12 signal pairs. Naturally, if the number of signal pairs is small, the number of source drivers can be reduced accordingly. Integration of analog video transmission source driver and display panel As discussed above, analog video transmission within the display unit is used to deliver video information to the source drivers of the display panel. Current large display architectures consist of a large area of active matrix display pixels. Initially, display drivers (source and gate) were mounted on the edges of the glass but not on the glass itself, providing the source and gate driving circuits. Due to the complexity of high-speed digital circuits and the large area required for D-to-A conversion, the additional integration of driving electronics onto the glass surface stalled. For example, the digital signals transmitted to the source driving circuit operate at a frequency of approximately 3 GHz, which is too high a frequency to allow for integration with the glass. Furthermore, it was recognized that driving a full-resolution LCD or OLED screen requires attaching a large number of display drivers to the edges of the display. A typical driver has about 1,000 outputs, and since a typical 4K display requires 4,000 × RGB = 12,000 connections, 12 source drivers are needed. Increasing the panel resolution to 8K increases this number to 24 source drivers. Data speed, synchronization difficulties, and combined logistics make it difficult to continue in this direction. Display panels (e.g., LCD panels) are fabricated from glass substrates on which thin-film transistors (TFTs) are formed. That is, field-effect transistors fabricated via thin-film deposition technology. These TFTs are used to implement the pixels of the display. Therefore, these TFTs (along with appropriate capacitors, resistors, and other suitable analog components) can also be used to create a logic circuit that implements the elements of the novel source driver described in the present invention, which can be integrated with the glass. These elements are integrated into the edge portion of the glass, outside the pixel display area but inside the periphery seal of the glass. Thus, the source driver disclosed herein can be integrated with the glass plate using these transistors, capacitors, resistors, and other necessary analog components, and can be implemented in this way in the embodiments described below. Accordingly, the source driver (or its elements), which were previously located on the outside and edges of the display panel glass plate, are now moved to the display panel glass plate itself. Additionally, the gate driver function of the gate driver can also be moved onto the display panel glass. SAVT video signals can be transmitted along the edge of the display glass with relatively simple wiring and are less sensitive to interference than the existing Vx1 interface. The low sampling rate allows the necessary analog electronic circuits of the source driver (with a simpler structure) to be designed on the edge of the TFT panel itself within the display panel glass. By constructing the source driver circuit on the glass edge, the following source driver elements can perform their general functions in conjunction with integration with the glass: input terminals and switches (receiving analog samples via the SAVT signal and distributing them to the collector); the collector (receiving analog samples via the input amplifier and collecting them into a storage array or line buffer); a level converter (providing single-ended conversion, voltage inversion, and voltage offset); and amplifiers such as high-voltage drivers (providing the amplified voltage and current required to charge the display source line capacitance). Figure 23 illustrates the integration of source driver functions in various implementation methods. Depending on the quality of the transistors used in the glass plate, various elements of the source driver can be integrated with the glass plate. As is known in the field, TFTs (e.g., amorphous silicon) can handle frequencies ranging from low to high. There are three main technologies used in TFT manufacturing: amorphous silicon (a-Si); oxides (indium-gallium-zinc oxide (“”) or similar materials), which can routinely handle frequencies from about 50 kHz to 100 kHz and up to about 200 kHz depending on the voltage used (oxides can handle frequencies of about 1 MHz at 50 volts in specific components); and low-temperature polysilicon (LTPS), which can handle frequencies of about 5 MHz depending on the voltage used and can exceed up to about 10 MHz. Additionally, crystalline silicon TFTs implemented using CMOS technology can handle higher frequencies. Other types of TFTs may also be used. Using high-quality, high-speed TFT transistors allows the high-frequency portion of the source driver to be integrated into a glass plate. Additionally, a smaller device size increases the switching speed of the transistors, enabling the elements using such devices to be implemented on the glass plate. For example, the channel length of the TFT affects its size. For oxide TFTs, a channel length of approximately less than 0.2 μm is preferred, while for LTPS TFTs, it is preferred to be approximately less than 0.5 μm. Reducing the channel length by 50% increases the speed fourfold. Furthermore, implementation may vary depending on the display type. Displays with 2K, 1K, and smaller resolutions may use elements that do not require the high frequencies found in 4K and 8K displays. Generally, amorphous silicon transistors are not used because they exhibit a threshold shift tendency and are unstable. The source driver disclosed herein does not require a digital-to-analog converter to convert video samples or a decoder to decode input video samples. In the first implementation example 102, the level converter 620 and amplifier 621 are integrated into the glass plate. This is because the level converter requires only a relatively low-frequency clock. Since the level converter switches once per line, a switching frequency of approximately 50 kHz is required for a 2K display and approximately 100 kHz for a 4K display. Therefore, the integration in the first implementation method can use TFTs capable of operating at a minimum clock frequency of approximately 50 kHz, assuming a 2K panel (100 kHz for a 4K panel, etc.). Thus, IGZO or LTPS TFTs can be used in the first implementation method. In the second implementation method 104, faster transistors are used, and the level converter 620, amplifier 621, and collector 786 can also be integrated with the glass plate, allowing for the integration of the entire source driver. Since each collector manipulates the pixel sequence, the collector 786 requires a higher frequency clock; a switching frequency of approximately 50 MHz is required for a 2K display, and approximately 100 MHz for a 4K display. Therefore, for the integration in the second implementation method, TFTs capable of operating at a minimum clock frequency of approximately 50 MHz can be used, assuming a 2K panel. Consequently, LTPS TFTs can be used for 2K panels in the second implementation method. FIG. 24A shows the arrangement of a gate driver and a source driver on a display panel glass plate. Typically, a display panel is implemented with two glass plates, consisting of an upper (common) glass plate and a lower (active) glass plate, with the lower glass plate being smaller than the upper glass plate. The TFT is implemented on the lower glass plate, and the following description refers to this lower glass plate, with only the lower glass plate shown in the drawing. What is shown is the display panel glass plate 150 itself (for clarity, the panel frame or surrounding display unit is not shown), and there are two rectangular areas 130 and 132 with a width of a few millimeters on both sides of the display panel glass plate (in this example, an LCD panel). In this implementation, the gate driver is integrated on the glass using a TFT device as a switching element. Since the gate driver is typically implemented as a simple shift register, this shift register may be located in area 130 or 132. Additionally, a rectangular area 140 located on the glass itself is indicated, in which elements of the source driver may be located. The source driver function can be partially or completely integrated with the glass by utilizing TFT switches on the glass surface in this area 140. In the first implementation, the amplifier and level converter are integrated on the glass surface (these are formed in area 140), and in the second implementation, the amplifier, level converter, and collector are integrated (these are also formed in area 140). The source driver disclosed herein does not receive digital signals and does not include a D-to-A converter and related circuitry, or a decoder, for processing digital video samples. Therefore, the low processing frequency and small size of such a driver allow it to be integrated into a glass plate. For example, the pixel width of a typical 64-inch 4K TV panel is 80 µm (40 µm for an 8K display), and since the size of the output amplifier is expected to fit within this space, there is sufficient width to integrate the driver directly into the glass plate. Depending on the pixel width of a specific implementation, a specific TFT can be selected. The interconnect printed circuit board 182 receives the EM signal 602 via the flexible circuit board 184 and transmits it to the source driver located in integrated circuit 186a and to the source driver partially integrated with the glass portion of TFT 186b. Transmitting the EM signal in this manner is implemented such that in Implementation Method 1, a portion of each source driver (minimum collector) is still located within the flexible PCB 184 of IC 186a, while the level converter and amplifier are located on the glass of TFT 186b. As shown in the figure, each integrated circuit 186a transmits the analog signal 187 to the corresponding circuit. The nature of this analog signal depends on Implementation Method 1 or Implementation Method 2. An example of Implementation Method 2 is shown below. Gate clocks 190 and 192 are transmitted to the gate driver via circuit board 182 and the flexible PCB 184. As is known, PCB 184 is attached to the panel glass 150. Fig. 24B shows a source driver 186 fully implemented on glass. This second implementation method implies the complete integration of the source driver function with the glass. As shown in the figure, the flexible PCB 184 contains only the EM signal 602 and does not include source driver functions such as the collector, level converter, and amplifier. All source driver functions are implemented as 186 on the TFT (and other analog components) on the glass surface. Although not shown, each different source driver can have its own PCB 184 and EM signal 602 (from the corresponding transmitter). In a specific implementation example, there are 24 such source drivers. Figure 25 shows another implementation example for the placement of the EM signal 602 when Implementation Example 2 is applied. As previously mentioned, in Implementation Method 2, since all functions of the source driver are integrated into the glass plate, there is no need to transmit the EM signal 602 through a large circuit board 182 the length of the display and then through multiple flexible PCBs 184 as shown in Figures 7A and 7B. Therefore, a much smaller printed circuit board 183 and a single flexible PCB 185 are attached to one location on the display panel glass 150, and the EM signal 602 is transmitted to the glass through 183 and 185, then travels along the glass surface to be delivered to each source driver 186 located on the glass within area 140. Additionally, the integrated circuit 186 on the flexible PCB is not required because all functions of each source driver are now implemented on the glass plate. As shown in the figure, the EM signal 602 is transmitted in parallel to each source driver. Returning to the example source driver in Figure 21, it can be seen that parts of the source driver can be implemented on the glass plate in many different implementations. For example, the high-voltage driver 914 (and optionally the thermal short 918) can be implemented in glass area 186b, and the remaining indicated elements are implemented in integrated circuit 186a outside the glass edge. Alternatively, the driver 914 and level converter 912 can be implemented on the glass, and the remaining upstream elements can be implemented outside the glass edge. Alternatively, the driver 914, level converter 912, and preamplifier 910 can be implemented on the glass, and the remaining upstream elements can be implemented outside the glass edge. Alternatively, the driver 914, level converter 912, preamplifier 910, and A / B sampling block 908 can be implemented on the glass, and the remaining upstream elements can be implemented outside the glass edge. Alternatively, the driver 914, level converter 912, preamplifier 910, A / B sampling block 908, and input distribution amplifier 904 are implemented on the glass, and the remaining upstream elements are implemented outside the edge of the glass. Alternatively, all indicated elements (including input terminal 902 and its switch 842) are implemented on the glass, and the elements of the source driver are not implemented on the flexible PCB 184 as shown in Figure 24B. Alternatively, all column drivers 916 of the source driver are implemented on the glass, and the remaining upstream elements (i.e., collector 915) are implemented outside the edge of the glass. Or, all column drivers 916 and collector 915 of the source driver are implemented on the glass, and the elements of the source driver are not implemented on the flexible PCB 184 as shown in Figure 24B. In a specific implementation example, the SHA input amplifier operates at an input rate of approximately 664 MHz, the A / B sampling block operates at 1 / 16 of the input rate, and the preamplifier and downstream components operate at 1 / 1024 of the input rate (1 / 64 of 1 / 16 of the input rate). Of course, the input rates may vary, and the ratio of the input rate at which downstream components operate may depend on the implementation method, number of columns, interleaving technology, etc. In other implementation examples, only the HV driver is implemented on glass because the output of the Level Converter 912 is single-ended (making implementation easier). Alternatively, the preamplifier, Level Converter, and HV driver may be implemented on glass because they require lower frequencies than the SHA amplifier and A / B block. Additionally, since the Amplifier 904 operates at the highest frequency, only the SHA Amplifier 904 may be implemented on the source driver chip, while the remaining downstream components are implemented on glass. Video image improvement through thermal amplifier feedback in the display Figure 1 shows an example of a display panel typically equipped with 24 new source drivers 186. Figure 13 is an example implementation of an SAVT receiver integrated with the source drivers, and Figure 21 shows in detail how the 64 columns 920 of the corresponding source drivers are driven starting with a sample input at 902. Unfortunately, due to manufacturing variations between the display panel and its source drivers, 23,000 (example) column amplifiers (and the analog circuits connected to them) produce different output levels when driven by the same input. We disclose a technology that compensates for such fluctuations by transmitting appropriate feedback to the display unit's timing controller (TCON). This invention allows 24 (or more) demultiplexing / source driver chips to feed back the levels of a single column amplifier to the TCON. After collecting performance information from all 23,000 column amplifiers, the TCON can equalize performance between columns by pre-scaling the values assigned to specific columns. Advantageously, there are no high-speed performance requirements: this invention can be used in the pre-sales phase for screen calibration purposes. That is, this technology can be used in the production test phase where all columns are available and drive characteristics can be measured without the additional area burden caused by column-by-column sampling per chip. This pre-scaling based on individual column feedback is in addition to the pre-scaling performed by the TCON to equalize the performance of different rows of the display. Rows far from the column driver may require additional current to achieve the same optical output. There are two main implementation methods: analog feedback and digital feedback. Both can utilize interfaces such as JTAG, I2C, or SPI, which allow the TCON to issue commands to specific column drivers to return values from their column amplifiers. These commands may also be issued using MFM commands as described above. In the analog version, the corresponding value is sampled onto an analog bus via an analog switch—a single analog connector shared by all source driver chips returns to the TCON. The TCON then performs analog-to-digital conversion (ADC) and digital processing of the result. Since the source driver output is high voltage, high-voltage transistors are required for multiplexing, or the column voltage may be generated by converting it to a low voltage before multiplexing (via resistor or capacitor dividers). In either case, area overhead per column occurs. Figure 26 shows an analog architecture 1010 that issues commands and receives feedback. The new pre-scaling controller 1020 of the TCON 130 uses JTAG, I2C, or SPI 1022 (or similar protocols and communication lines) to send commands to a specific column driver (e.g., driver 916) in one of the source drivers 186 to sample the analog values of the amplifier output from the HV driver 914 of that driver. The sampled values are returned to the ADC 1032 within the TCON via a single analog bus 1024. The converted digital values collect all values from all column amplifiers in a similar manner, process all values, determine a pre-scaling method to equalize each value, and then output the results to a pre-scaling value store 1034 that stores the actual values, offset values, percentages, ratios, etc., for each column. The values in 1034 are used to modify the samples transmitted from the TCON to the display during display operation (along with other pre-scaling). Figure 27 illustrates the process of sampling values in a thermal amplifier. The HV driver 914 for a specific column (in this case, column 63 of SHA[0] in Figure 21) is shown; this driver provides the output of the internal amplifier 1040 to an analog switch 1042, and the output of the switch is connected to an analog bus 1024 to feed back the column output to TCON. Commands arrive via a control signal 1041 (e.g., JTAG, I2C, SPI, etc.), and this signal samples the value. Switch 1042 is an example, and other types of analog switches (e.g., using FETs) or similar sampling circuits can also be used. In the digital version, each source driver chip is equipped with its own A-to-D converter to perform thermal amplifier sampling locally (thus avoiding the load effect of a long analog return path), and returns the digital value to the TCON via the same JTAG, I2C, or SPI path as the command. Figure 28 shows the digital architecture 1050 for receiving commands and feedback. The new pre-scaling control unit 1020 of the TCON 130 uses protocols and communication lines such as JTAG, I2C, or SPI 1052 (or similar) to send commands to a specific column driver (e.g., driver 916) in one of the source drivers 186, causing the output analog value of that amplifier to be sampled by the HV driver 914. The sampled and converted digital value is returned to the TCON's control unit 1020 via line 1052. The control unit 1020 collects all values from all column amplifiers in a similar manner, processes all values, determines a pre-scaling method to equalize each value, and outputs the result to the pre-scaling value store 1034. This store stores the actual value, offset, percentage, ratio, etc., for each column. Value 1034 is used to modify the samples transmitted from the TCON to the display during display operation (along with other pre-scaling). This is a known technique. Figure 29 illustrates the process of sampling values from the thermal amplifier. The HV driver 914 of a specific column (in this case, column 63 of SHA[0] in Figure 21) provides the output of the internal amplifier 1040 to the analog switch 1042, which can be implemented and controlled as described above. An ADC 1060 is included per source driver to convert the analog value of each thermal amplifier into a digital value and transmit it to the TCON. An ADC per HV driver or more ADCs may also be used (provided the thermal voltage drops sufficiently low to avoid damaging the ADC input circuit). The result is a simple screen measurement. The display can be recalibrated after manufacturing (pre-sale during production testing) or self-calibrated at another time upon user command. For post-sale recalibration, it is advantageous to incorporate additional hardware within the chip instead of calibration during production testing. In other implementation methods, the performance of a pixel is modeled in relation to neighboring pixels (e.g., within the same column). This model is used to pre-scale subpixel inputs before they are transmitted to the source driver chip to generate the desired brightness. Typically, it can be difficult to determine the light output of each subpixel, and such measurements require complex test setups. Instead, the current passing through each subpixel at a specific input value is measured, and that measured current is used as a proxy for light emission to model the performance of the pixel (or subpixel). In the modified case, we model the performance of the display using actual screen measurement techniques, specifically analyzing the impact of adjacent subpixel values on brightness (such as slew rate issues). Utilizing this model, we generate the desired brightness by pre-scaling the subpixel input signal before transmitting it to the source driver chip. Video signal transmission in different environments The above description is an example of an implementation for transmitting video signals into a display panel and a display unit. The present invention also includes examples of implementations for transmitting video signals using SAVT in other environments, such as receiving SAVT signals directly from a camera or other image sensor, from an SoC or other processor, or from an SAVT receiver not integrated with the display panel (unlike the one described above). For example, SAVT signals may be received from an SoC, a computer processor, or an SAVT receiver not integrated inside a legacy display panel. U.S. Patent Application Nos. 63 / 611,274 and 63 / 625,473 (HYFYP017P2 and HYFYP018P), cited by reference above, disclose examples for other environments, such as a mobile device and inside a vehicle, respectively. FIG. 30 shows a SAVT transmitter 1240 configured to transmit various video samples from various sources. A splitter 1241 includes two line buffers 1242 and 1243 having input vectors, a splitter controller 1230, optional digital-to-analog converters 1260-1269, and analog EM signals 1270-1279 output from each input vector, and can be implemented as follows. It can be implemented as shown in Figure 2. Outputs 1281 and 1291 represent serial outputs, but the outputs can be implemented as described in Figure 2, and all samples can be output in parallel from the first buffer to the second buffer. As in Figure 2, each DAC (if present) converts the received samples from the digital domain to a single analog level (if no DAC is present, the analog level is output from the line buffer). This analog level can be transmitted as a pair of differential voltage signals of magnitude proportional to the input digital value, and is transmitted serially when output from each DAC. Although not shown in the figure, image processors 250-259 may be optionally present depending on the implementation and can be placed after the buffer or, if preferred, before the buffer. In this example, multiple EM paths exist. There may be a single EM path or multiple EM paths. Depending on implementation and design decisions, multiple outputs can improve performance but require more paths. To use as few cables as possible from the transmitter 1240, a method of transmitting a single EM signal 1270 through a single path may be used. The SAVT transmitter 1240 can be implemented similarly to the transmitter in Figure 2, but the inputs may differ, an image processor may not be essential (it can be implemented downstream in an SoC or other processor), and a DAC is optional. Additionally, the number of input vectors per line buffer and the number of samples N per input vector can vary significantly depending on the implementation method, the type of input signal, the desired bandwidth, whether the transmitter is implemented in a camera, an SoC, or another processor, etc. Furthermore, to use as few cables as possible from the transmitter 1240, only a method of transmitting a single EM signal 270 through a single path may be used. Depending on the implementation method discussed below, analog RGGB video sample 1239a may be input, analog or digital RGB sample 1239b may be input, digital G sample 1239c may be input, or analog BGBG . . RGRG sample 1239d may be input. If the samples are digital, DACs 1260-1269 are used. Generally, the transmitter can accept analog or digital video samples regardless of the color space used, and they do not necessarily have to be RGB. Samples may arrive sequentially (e.g., R followed by G followed by B) or in parallel (i.e., RGB in parallel as three separate signals). Splitter 1241 can be used to reorder the samples as needed. As previously mentioned, the input may vary depending on the implementation method. Input 1239d can be generated in the following manner. The image sensor can output raw analog samples that have not undergone “demosizing” via interpolation without using an ADC. Therefore, the image sensor output is a continuous serial stream of analog video samples arranged in chronological order, where each sample represents each pixel in row-first order (e.g., left to right) and continues for every frame while the image sensor detects. Of course, other orders may also be used. If Bayer filtering is used, the samples are BGBG row... followed by RGRG row..., which is often referred to as the RGGB format because each 2x2 pattern contains one RGGB each. This analog video sample row 1239d is input to SAVT transmitter 1240 and transmitted as EM signals 1270-1279, then transmitted to the SAVT receiver in Figure 31, and subsequently output as 1360 from the SAVT receiver. Since this sample is still raw data from the image sensor (i.e., the sensor's Bayer filter output), a downstream ADC is used as needed, and the ISP performs “demosicing” using CFA interpolation to interpolate “missing” color values at each location, generating RGB samples per pixel. Input 1239c can be generated as follows: Raw analog samples from the image sensor are converted to digital by an ADC, and then “demosizing” is performed within an Image Signal Processor (ISP) to generate digital RGB samples per pixel. From the set of RGB samples per pixel, only the green channel (i.e., one G sample per element of the array) is selected and sent to input 1239c. Rows of these green digital video samples 1239c are input to SAVT transmitter 1240 and transmitted as EM signals 1270–1279, which are then sent to the SAVT receiver in Figure 31 and output sequentially from the corresponding SAVT receiver. This significantly reduces image latency on the display, providing immediate feedback to viewers in applications using proximity displays, such as virtual reality and augmented reality. Alternatively, since only the green channel is transmitted, G samples can be obtained by performing interpolation only on the sensor's R and B components. Interpolation is simpler and faster because it is not required on the G component, and R and B samples are not needed on that component. Since the green channel corresponds to the luminance (or “luma”) channel, there is no perceived loss of resolution, but the downstream display shows a monochrome image. Input 1239a can be generated as follows: It modifies the read output of an image sensor and reads two or more rows simultaneously. For example, it reads the first two bottom rows of the image sensor simultaneously to output a stream of continuous values such as BGRGBGRG... or GBGRGBGR.... The reading order is as follows: the blue value of the first row, the green and red values of the second row, the green value of the first row, and so on, resulting in a continuous output of BGRGBGRG. Alternatively, an alternative reading order is used: the green value of the second row is read first, followed by the blue and green values of the first row, and then the red value of the second row, resulting in a continuous output of GBGRGBGR. Other reading orders that mix the color values of two adjacent rows may also be used, and the order of pixel values may vary depending on whether a specific row starts with a red, green, or blue value. Because two rows are read simultaneously, the four values from those two rows (e.g., BG at the beginning of the first row and GR at the beginning of the second row, i.e., two Gs, one R, and one B) can be output sequentially, creating a continuous pattern such as BGRG . . . or GBGR . . . After the first two rows are read, the next two rows are read, and a similar output is possible. Each group of four values may contain two green values, one red value, and one blue value. The image sensor can be read starting from a specific corner, from top to bottom or bottom to top, row by row or column by column, or in other similar ways. Therefore, the output of the video source is a series of values such as BGRGBGRG . . . or GBGRGBGR or similar values. Using this series of values, “demosizing” can occur in the analog domain of the SoC, and there is no need to convert these values to digital or use digital processing. The order of these color values facilitates interpolation in the analog domain. Other color spaces may be used, in which case reading two or more rows simultaneously and mixing the color values of different rows in the continuous output also facilitates color interpolation in the analog domain. A video source including an image sensor outputs a pattern such as BGRGBGRG . . . or GBGRGBGR, which is indicated and referenced as “. .”1239a in Figure 30. This RGGB video sample is input sequentially to transmitter 1240, converted into EM signals 1270-1279, transmitted to the SAVT receiver in Figure 31, and then output sequentially from the SAVT receiver. Input 1239b can occur as follows. Digital RGB samples can be input as shown in Figure 2, and analog RGB samples 1239b can also be input as shown in Figure 30. These analog RGB samples can be generated within a processor that performs demosicing in the analog domain on the raw image sensor output, or within a processor that receives digital RGB samples and converts them into analog RGB samples, or they can be generated in other ways. These analog RGB video samples are input to transmitter 1240, converted into EM signals 1270-1279, transmitted to the SAVT receiver in Figure 31, and sequentially output from the SAVT receiver. Figure 31 shows a SAVT receiver 1300 located on an SoC, processor, legacy display, or other location. This receiver receives EM signals 1270–1279 at random and inputs them into a collector 1320 equipped with two line buffers 1301 and 1302. Similar to the SAVT transmitter splitter in Figure 30, each line buffer has output vectors 1304, 1306, 1308 (or 1314, 1316, 1318) containing video samples (e.g., N=1024) corresponding to the input vector. When operating, each output vector 1304-1308 of the first line buffer 1301 is filled with samples from the corresponding EM signals 1270-1279, and while buffer 1301 outputs samples (via outputs 1305, 1307, 1309) and transmits them to receiver output 1360, the second line buffer 1302 is filled with the corresponding EM signals 1270-1279. When the first line buffer is empty, refilling begins, and during this process, the second line buffer is output to receiver output 1360. As with the SAVT transmitter in Figure 30, two line buffers are preferred, but more buffers may be used as needed, and the buffer lengths can be adjusted as mentioned. The output of the collector 1320 is serial, but the output of each buffer can be parallel (i.e., N samples at a time from each output vector), and the output time per sample may take longer than the input sampling time. Therefore, if 100 samples are output at a time, they may be transmitted to the output 100 times slower than the input sampling rate (assuming input sampling occurs one at a time). Collector controller 1330 specifies the order in which samples are loaded from inputs 1270 . . 1279 and controls the timing for unloading samples for further processing. Because the input stream is continuous, while the collector controller loads samples into one line buffer, samples from another line buffer are sent to the output for further processing. An example of a receiver implementation compatible with the implementation method discussed in Figure 30 is shown. In this implementation, rows of the image sensor are output sequentially and transmitted to collector 1320 via SAVT transmitter 1240 (input 1239d, i.e., BG . . . RG . . .), where they are output sequentially from the collector in the same format as the input (i.e., BG . . . RG . . .). The output samples are transmitted for further processing or display. As previously mentioned, the output from the collector can be output in a parallel format if parallel input is required by the processor. Whatever permutation was used by the SAVT transmitter to distribute input samples to the line buffer, the reverse permutation is used by the SAVT receiver, so receiver output 1360 outputs samples in the order received from the SAVT transmitter (i.e., the order received from 1239d). The SAVT receiver 1300 displays only the BG . . . RG . . . output, but it can also receive and output input samples from other implementations described in Figure 30. Thus, when input 1239a is input to the SAVT transmitter, output 1360 becomes RGGB . . . samples, and when input 1239c is input to the SAVT transmitter, output 1360 becomes G . . . samples. Also, when input 1239b is input to the SAVT transmitter, output 1360 becomes RGB . . . samples. Since output 1360 is an analog sample, the sample can be converted to digital using ADC 1362 or an ADC as needed. Thus, each output vector can provide a continuous digital sample stream 1364 by outputting samples one by one through the analog-to-digital converter (ADC) 1362. Additional Examples The present invention includes the following additional embodiments. C1. Display Driver Integrated Circuit (DDIC) - A device integrating a Timing Controller (TCON) and a transmitter, which includes the following components: A distributor that receives multiple digital video sample streams generated from a system-on-chip (SoC) of a mobile phone and distributes the corresponding digital video samples into multiple input vectors according to a predetermined permutation; A plurality of digital-to-analog converters (DACs), each DAC configured to receive a digital video sample from an input vector, convert the said digital video sample into a series of analog video samples, and output the said series of analog video samples to a display panel of a mobile phone via an electromagnetic path; and Gate driver control signal output to the gate driver of the display panel. C2. In the transmitter described in Clause C1, the distributor further comprises the following components: As the first line buffer, storing the corresponding multiple input vectors; and It includes a second line buffer storing multiple second input vectors, and the splitter is further configured to alternately distribute a line of digital video samples between the input vector of the first line buffer and the second input vector of the second line buffer, and the image processor performs a read from the first line buffer when the splitter performs a write to the second line buffer, and performs a read from the second line buffer when the splitter performs a write to the first line buffer. C3. A transmitter described in Clause C1, wherein digital video samples distributed to an input vector constitute a line of an image. C4. A transmitter described in Clause C1, wherein digital video samples are distributed to an input vector at a first frequency, and said digital video samples are sequentially output from each input vector at a second frequency different from the first frequency. C8. An integrated transmitter and timing controller as described in Clause C1, wherein the device is located within approximately 2 cm of the system-on-chip. C9. A device described in Clause C1, wherein the device is integrated within a single integrated circuit of a mobile phone. C10. A device in which the device is also integrated with the system-on-chip of the mobile phone in the transmitter described in Clause C9. C11. Additionally included in the device described in Clause C1: A plurality of image processors, each image processor configured to sequentially read digital video samples from one of the corresponding input vectors and perform at least gamma correction on the digital video samples of the corresponding input vector. D1. An analog DDIC-SD (Display Driver Integrated Circuit - Source Driver) for mobile phones, comprising the following components: An input terminal configured to receive an analog electromagnetic signal containing a continuous series of continuous analog video samples through an electromagnetic path; A plurality of sampling amplifiers, each configured to exclusively sample a portion of the corresponding analog video sample and write the sample to a location in a storage array assigned to the corresponding sampling amplifier; and It consists of multiple column drivers, each column driver configured to read one of the corresponding analog video samples from the corresponding location in the storage array, to amplify the corresponding analog video sample, and to transmit the amplified analog video sample to the column of the display panel of the corresponding mobile phone. D2. An analog DDIC-SD described in claim D1, additionally comprising a second storage array having a position assigned to each sampling amplifier, wherein the sampling amplifier is configured to alternately write each portion of the analog video sample to the storage array or the second storage array, wherein the sampling amplifier writes to the second storage array while the column driver reads from the storage array, and the sampling amplifier writes to the storage array while the second storage array reads. D3. The analog DDIC-SD described in Section D2 includes the following additional components: A control logic circuit enables each sampling amplifier to sample a portion of the corresponding analog video sample, enables the sampling amplifier to write to a storage array or a second storage array, and enables a column driver to read from the storage array or the second storage array. D3. In the analog DDIC-SD described in claim D1, a portion of the analog video samples is used for synchronization purposes and is not transmitted to the columns of the display panel. D4. In the analog DDIC-SD described in paragraph D1, the analog DDIC-SD does not include a digital-to-analog converter (DAC) used to convert video samples. D5. In the analog DDIC-SD described in paragraph D2, the column driver is additionally configured to read in parallel from the storage array when the storage array is full, or to read in parallel from the second storage array when the second storage array is full. D6. In the analog DDIC-SD described in Clause D1, the series of analog video samples arrives in a predetermined order and each sampling amplifier outputs the analog video sample of the corresponding portion to a contiguous storage location of the storage array. D7. In the analog DDIC-SD described in Clause D1, the said electromagnetic signal includes a control signal for synchronization and is not transmitted to a column of the said display panel, and the said source driver further includes the following: A dedicated sampling amplifier for sampling the corresponding control signal. F1. As a video transmission device, Includes a transmitter A splitter that receives a digital video sample stream from a system-on-chip (SoC) of a mobile phone and distributes the digital video samples into multiple input vectors in a line buffer according to a predetermined permutation, and a digital-to-analog converter (DAC) for each input vector, wherein each DAC is configured to sequentially receive the corresponding digital video samples from the corresponding input vector and convert the corresponding digital video samples into a series of analog video samples; A plurality of electromagnetic paths, each path configured to transmit one of the series of corresponding analog video samples to the display panel of the corresponding mobile phone; and, A source driver array including a source driver corresponding to each DAC, each source driver is: A collector that receives the corresponding analog video sample series from each DAC and stores the analog video samples of the corresponding input vector; and It includes multiple column drivers that receive analog video samples stored in parallel from a collector and amplify the stored analog video samples into columns of a display panel. F2. In the device described in paragraph F1, the said transmitter is integrated with DDIC-TCON, and said device includes the following additional components: Gate driver control signal output to the gate driver of the display panel. F3. In the device described in paragraph F2, the transmitter is located within approximately 2 cm of the system-on-chip. F4. A device described in claim F2 in which the transmitter is integrated within a single integrated circuit of the mobile phone. F5. A device described in paragraph F2 in which the transmitter is also integrated with the system-on-chip (SoC) of the mobile phone. F6. A device described in paragraph F1, wherein each source driver is an analog DDIC-SD (Display Driver Integrated Circuit - Source Driver) of a mobile phone. F7. In the device described in paragraph F6, the analog DDIC-SD does not include a digital-to-analog converter (DAC) used to convert video samples. I1. The source driver of the display unit includes the following components: Input terminal for receiving analog sample values representing a video stream; A collector that receives a corresponding analog sample value from an input terminal, stores it in a storage array, and outputs it in parallel, wherein the input terminal and the collector are implemented outside the edge of the display panel of the corresponding display device; and A plurality of thermal drivers that input the corresponding analog sample values in parallel and output voltage to the columns of the corresponding display panel, wherein the plurality of thermal drivers are implemented using transistors on the glass substrate of the corresponding display panel. I2. In the source driver described in I1, the analog sample values are an ordered sequence consisting of analog levels of continuous amplitude. I3. In the source driver described in paragraph 1, the plurality of thermal drivers are located on the glass substrate of the display panel between the pixel display area and the periphery of the glass substrate. I4. In the source driver described in paragraph 1, said source driver does not include a digital-to-analog converter for converting video samples. I5. In the source driver described in paragraph 1, the transistor can operate at the clock frequency required by the thermal driver. I6. In the source driver described in paragraph I5, the transistor is a thin-film transistor (TFT), and is either a low-temperature polysilicon (LTPS) transistor or an indium-gallium-zinc oxide (IGZO) transistor. I7. In the source driver described in claim I1, the pixels of the corresponding display panel are implemented using the same type of transistor used to implement the corresponding column driver. I8. In the source driver described in claim I1, the display unit is the display of a mobile phone. I9. In the source driver described in claim I1, each column driver includes a high-voltage driver. I10. In the source driver described in claim I9, each column driver further comprises a level converter, and each level converter operates to convert the output voltage of the corresponding column driver. I11. In the source driver described in claim I1, the collector is a two-stage collector. J1. The source driver of the display unit includes the following components: An input terminal receiving an analog sample value representing a video stream; a collector receiving the analog sample value from the input terminal, storing the analog sample value in a storage array, and outputting the analog sample value in parallel; and A plurality of thermal drivers that input the corresponding analog sample values in parallel and output voltage to the columns of the corresponding display panel, wherein the corresponding collector and the plurality of thermal drivers are implemented using transistors on the glass substrate of the corresponding display panel. J2. In the source driver described in claim J1, the analog sample values are a sequence of analog levels of continuous amplitude. J3. A source driver as described in claim J1, wherein the collector and the plurality of column drivers are positioned between a pixel display area and a periphery of the glass substrate on the glass substrate of the display panel. J4. In the source driver described in claim J1, the source driver does not include a digital-to-analog converter for converting video samples. J5. In the source driver described in claim J1, the transistor can operate at a first clock frequency required by the thermal driver and a second clock frequency required by the collector. J6. In the source driver described in paragraph J5, the transistor is a thin-film transistor (TFT), and is either a low-temperature polysilicon (LTPS) transistor or an indium-gallium-zinc oxide (IGZO) transistor. J7. In the source driver described in paragraph J1, the pixels of the corresponding display panel are implemented using the same type of transistor used to implement the corresponding collector and column drivers. J8. In the source driver described in paragraph J1, the corresponding display unit is the display of a mobile phone. J9. In the source driver described in claim J1, each column driver includes a high-voltage driver. J10. In the source driver described in claim J9, each column driver further comprises a level converter, and each level converter converts the output voltage of the corresponding column driver. J11. In the source driver described in claim J1, the input terminal is implemented outside the edge of the display panel of the display unit. J12. In the source driver described in claim J1, the input terminal is implemented on the glass substrate of the display panel. J13. A source driver described in paragraph J1, wherein the collector is a two-stage collector. K1. The display unit is configured as follows: A display panel having a glass substrate; and With multiple source drivers, each source driver is A collector configured to receive analog sample values representing a video stream, store the analog sample values in a storage array, and output the analog sample values in parallel; and It includes multiple column drivers that receive the corresponding analog sample values in parallel and output voltage to the columns of the corresponding display panel. K2. A display device described in paragraph K1, wherein the collector of each source driver is implemented outside the edge of the display panel of the display device, and the thermal driver is implemented using a transistor on the glass substrate of the display panel, and the transistor is capable of operating at the clock frequency required by the thermal driver. K3. A display device described in paragraph K1, wherein the collector and thermal driver of each source driver are implemented using at least a transistor on a glass substrate of the display panel of the display device, and the transistor is capable of operating at the clock frequency required by the collector and thermal driver. K4. In the display device described in paragraph K1, the analog sample value is an ordered sequence consisting of analog levels of continuous amplitude. K5. In the display device described in paragraph K2, a plurality of thermal drivers of each source driver are located between the pixel display area and the periphery of the glass substrate on the glass substrate of the corresponding display panel. K6. In the display device described in paragraph K1, each source driver does not include a digital-to-analog converter for converting video samples. K7. In the display device described in paragraph K2, the transistor is a thin-film transistor (TFT), and is a low-temperature polysilicon (LTPS) transistor or an indium-gallium-zinc oxide (IGZO) transistor. K8. In the display device described in paragraph K3, the transistor is a thin-film transistor (TFT), and is a low-temperature polysilicon (LTPS) transistor or an indium-gallium-zinc oxide (IGZO) transistor. K9. A display device as described in claim K1, wherein the display device is a display device of a mobile phone. K10. A source driver as described in claim K1, wherein each collector of each source driver is a two-stage collector. P1. A feedback device for a source driver of a display unit, wherein the feedback device includes the following components: The display unit includes an analog switch having an input terminal connected to the amplifier output of a thermal driver of a display panel, a control input terminal configured to receive a control signal from a timing controller of the display unit, and an output terminal connected to an analog line connected to the timing controller, wherein the analog switch is configured to latch the voltage value of the amplifier and transmit the voltage value to the timing controller when receiving the control signal. P2. In the feedback device described in paragraph P1, the corresponding control signal commands the corresponding analog switch to latch the corresponding voltage value when the corresponding column driver drives the column of the corresponding display. P3. Additionally included in the feedback device described in Paragraph P1: An analog-to-digital converter located between the output of the analog switch and the timing controller is configured to convert the corresponding voltage value from the output into a digital voltage value and transmit it to the timing controller. P4. In the feedback device described in claim P1, the voltage value transmitted to the corresponding timing controller is an analog voltage value. P5. The source driver of the display unit is configured as follows: Each column of the display panel includes the feedback device described in claim P1. Q1. A method for providing feedback from the source driver of a display unit to a timing controller is configured as follows: A step of transmitting a command from the timing controller to the column driver of the source driver to latch the voltage value output by the amplifier of the corresponding column driver to the display panel of the display unit; A step of latching the corresponding voltage value at the output of the corresponding amplifier; and A step of receiving the corresponding voltage value from the corresponding timing controller and storing the corresponding voltage value. Q2. In the method described in paragraph Q1, if the received voltage value is an analog voltage value, the method further includes the following steps: A step of converting the analog voltage value into a digital voltage value within the timing controller and storing the digital voltage value. Q3. In the method described in paragraph Q1, if the voltage value at the output of the amplifier is an analog voltage value, the method further comprises the following steps: Converting the above analog voltage value into a digital voltage value within the source driver; A step of receiving the digital voltage value from the timing controller. Q4. According to the method described in paragraph Q3, the method additionally includes the following steps: A step of transmitting the corresponding command to the corresponding column driver through a digital control interface; and A step of receiving the corresponding digital voltage value from the corresponding timing controller through a digital control interface. Q5. According to the method described in paragraph Q2, the method additionally includes the following steps: A step of transmitting the corresponding command to the corresponding column driver through a digital control interface; and A step of receiving the corresponding analog voltage value through an analog line. Q6. According to the method described in claim Q1, the method further comprises the following steps: A step of performing the corresponding transmission, corresponding locking, and corresponding receiving steps for a plurality of column drivers; and A step of pre-scaling the output voltage value of the timing controller assigned to the column of the display panel using the stored voltage value. R1. Transmitter Components: A distributor that receives a video sample stream generated from an image sensor, distributes a first selection set of the said digital video samples into a plurality of first input vectors in a first line buffer according to a predetermined permutation, and distributes a subsequent selection set of said video samples into a plurality of second input vectors in a second line buffer according to a predetermined permutation; said distributor is further configured to alternately distribute the selection of said video samples between the first line buffer and the second line buffer; and Multiple output ports, each output port configured to alternately read from one of the first input vectors of the first line buffer while the splitter performs a write to the second line buffer, alternately read from one of the second input vectors of the second line buffer while the splitter performs a write to the first line buffer, and output an analog level series corresponding to a video sample of one of the first or second input vectors as an electromagnetic signal. R2. A transmitter described in claim R1, wherein the video sample is an analog RGB video sample. R3. A transmitter described in claim R1, wherein the video sample is a digital RGB video sample, and the transmitter further comprises the following components: It includes multiple digital-to-analog converters (DACs), each DAC being configured to receive a corresponding digital RGB video sample from one of the first or second input vectors, convert the corresponding digital RGB video sample into an analog video sample, and output the corresponding analog video sample to one of the output ports. R4. A transmitter described in claim R1, wherein the video sample is a raw analog BGRG video sample. R5. A transmitter described in claim R1, wherein the video sample is a raw analog RGGB video sample output from an image sensor in a modified form. R6. A transmitter described in claim R1, wherein the video sample is a digital G video sample, and the transmitter further comprises the following components: It includes multiple digital-to-analog converters (DACs), each DAC being configured to receive a G digital video sample from one of the first or second input vectors, convert the digital video sample into an analog video sample, and output the analog video sample to one of the output ports. S1. As a receiver: It is composed of a plurality of input terminals, each input terminal receives an electromagnetic signal through an electromagnetic path, and each electromagnetic signal contains a sequence of analog levels representing analog video samples; A collector configured to collect analog video samples of each electromagnetic signal into one of the first output vectors of the first line buffer and one of the second output vectors of the second line buffer, wherein the collector is further configured to alternately collect a selection of analog video samples between the first line buffer and the second line buffer, and wherein the collector outputs the corresponding analog video samples in a predetermined order from the first output vector of the first line buffer while simultaneously filling the second line buffer, and outputs the corresponding analog video samples in a predetermined order from the second output vector of the second line buffer while simultaneously filling the first line buffer, thereby causing the receiver to continuously output a continuous stream of the corresponding analog video samples. S2. The source driver described in claim S1 includes the following additional components: The collector controller is configured to collect input analog levels from electromagnetic signals into first and second line buffers, and to output analog levels from the first and second line buffers in a predetermined order. S3. In the receiver described in paragraph S1, when the first line buffer is full, an analog video sample is output from the first output vector. S4. In the receiver described in paragraph S1, the receiver does not include a digital-to-analog converter (DAC) used to convert video samples. *S5. In the receiver described in paragraph S1, the receiver includes at least one analog-to-digital converter (ADC) to convert an analog video sample stream into a digital video sample. G1. Method for sending commands to the source driver of the display unit: A step of transmitting a control sequence from a transmitter within a display unit to a source driver of the display unit, wherein the control sequence is an MFM (modified frequency modulation) encoded sequence; A step of introducing an MFM flag into the control sequence to indicate to the source driver that at least one command follows the MFM flag; and A step of sending at least one video synchronization command to the source driver via the control sequence after the MFM flag. G2. A step additionally included in the method described in claim G1: A step of performing vertical or horizontal synchronization of an image displayed on a display panel of the above-mentioned display device using the above-mentioned video synchronization command. G3. A step additionally included in the method described in claim G1: A step of inserting a control sequence into an analog video sample stream transmitted from a transmitter to a source driver so that the control sequence is received by a single input amplifier of the source driver. G4. A step additionally included in the method described in claim G1: A step of introducing an MFM flag into the control sequence when the power of the display unit is turned on. G5. Methods additionally included in the methods described in paragraph G1: Transmit the corresponding control sequence (including inserted MFM flags) through all input channels of the source driver until resynchronization occurs; After resynchronization, the step of transmitting only the corresponding control sequence through a single input channel of the source driver. G6. Methods additionally included in the methods described in paragraph G1: A step of simultaneously transmitting the corresponding control sequence and the introduced MFM flag to all source drivers of the display unit. H1. How to receive commands from the source driver of the display unit: A step of receiving a control sequence from a transmitter within a display unit to a source driver of the display unit, wherein the control sequence is an MFM (modified frequency modulation) encoded sequence; Receiving and recognizing the MFM flag of the control sequence and indicating to the source driver that at least one command follows the MFM flag; and Interpreting the next MFM cell of the control sequence after the corresponding MFM flag as a video synchronization command in the corresponding source driver. H2. A method additionally included in the method described in claim H1: A step of performing vertical or horizontal synchronization of an image displayed on the display panel of the corresponding display unit using the corresponding video synchronization command. H3. A method in which, in the method described in paragraph H1, the control sequence is received at a single input amplifier of the source driver. H4. Steps additionally included in the method described in paragraph H1: A step of receiving the MFM flag of the corresponding control sequence when the power of the corresponding display unit is turned on. H5. In the method described in claim H1, the source driver receives the control sequence (including the introduced MFM flag) on all input channels of the source driver until resynchronization occurs; After resynchronization, the step of receiving the corresponding control sequence only on a single input channel of the corresponding source driver. H6. A method comprising the following steps in addition to the method described in paragraph H1: A step of simultaneously receiving a control sequence and an introduced MFM flag from all source drivers of the display unit. L1. A method for performing phase alignment to determine the source driver sampling phase, wherein the method comprises the following steps: A step of receiving a command to enter phase alignment mode from the source driver of the display unit; A step of receiving a positive pulse stream having a first amplitude; A step of setting an upper threshold of an upper comparator receiving the corresponding synchronization stream so that the upper comparator is triggered when it receives a positive pulse; A step of rotating the sampling phase of a positive pulse toward the rear edge of the corresponding positive pulse until the upper comparator is not triggered at a specific sampling phase; and If it is determined that the above-mentioned upper comparator is not triggered, rotate at least one step backward from the above-mentioned specific sampling phase, and set the rotated sampling phase as the sampling phase of the source driver to sample the analog video sample input from the source driver. L2. In the method described in claim L1, the synchronization stream comprises a positive pulse that intersects a second positive pulse having a second amplitude smaller than the first amplitude, and the method further comprises the following steps: A step of adjusting the upper threshold of the upper comparator to set the upper comparator so that it is triggered on a positive pulse but not on a second positive pulse. L3. A method as described in claim L1, comprising the following steps: A step in which a source driver samples at least one of the input analog video samples using a source driver sampling step, and displays the input analog video sample on a display panel of a display device. L4. In the method described in paragraph L1, the source driver includes a central comparator that indicates whether the pulse of the synchronization stream is positive or negative, and the method further includes the following steps: A step of rotating the sampling step backward in that specific sampling step only if the result of the central comparator does not match the result of the upper comparator. L5. A method described in paragraph L1, wherein the command is an MFM (modified frequency modulation) encoded command and the synchronization stream is an MFM encoded stream. L6. A method described in paragraph L1 in which the command and the synchronization stream are both received through a single channel of the source driver. L7. A method described in paragraph L1, wherein the synchronization stream comprises a negative pulse that intersects a positive pulse, and the negative pulse has a negative amplitude, wherein the method further comprises the following steps: A step of setting a lower threshold of a lower comparator receiving the synchronization stream so that the lower comparator is triggered when it receives the negative pulse; A step of rotating the second sampling phase of the corresponding negative pulse toward the rear edge of the corresponding negative pulse so that the corresponding lower comparator is not triggered at a specific second sampling phase; If it is determined that the corresponding lower comparator is not triggered, a step of rotating at least one step backward from the corresponding specific second sampling phase and setting the rotated second sampling phase as the second sampling phase of the source driver; and A source driver sampling step for sampling analog video samples input from a source driver, and a step for determining the average of the corresponding rotated second sampling step and the corresponding rotated second sampling step. M1. A sampling step adjustment circuit included in the source driver of a display unit, comprising the following components: Input reference clock; The above phase adjustment control; A phase selector that moves the sampling phase of the reference clock forward when the phase adjustment control indicates that sampling occurred before the control sequence of the video sample received from the source driver, and moves the sampling phase of the reference clock backward when the phase adjustment control indicates that sampling occurred after the control sequence of the video sample received from the source driver; and As an output sampling clock having a sampling phase output by the above-mentioned phase selector, the output sampling clock is transmitted to an amplifier that receives video samples of the source driver. M2. Additionally included in the sampling phase adjustment circuit described in claim M1: Input clock cycle skip control that controls the corresponding sampling phase adjustment circuit to skip one cycle of the reference clock. N1. A method for transmitting a command to a source driver of a display unit, wherein the method is: A step of receiving a sample stream from the source driver of the display unit; A step of interleaving the sample into multiple channels; A step of searching for an MFM flag in one of the channels; and A step of searching the next channel if it is determined that the MFM flag is not found in the current channel, and determining that the MFM flag exists in the next channel. N2. A method additionally included in the method described in paragraph N1: Step to implement clock cycle skip to retrieve the MFM flag in the next corresponding channel. N3. Methods additionally included in the methods described in paragraph N1: A step of transmitting the deinterleaved sample to a plurality of sample-and-hold amplifiers (SH amplifiers), each SH amplifier implements one of the corresponding channels. N4. Steps additionally included in the method described in paragraph N1: A step of confirming that an MFM flag exists in the next channel, and then identifying a command for the display panel of the display unit following the MFM flag of the next channel. N5. A step additionally included in the method described in claim N1: A step of confirming that an MFM flag exists in the next channel, and then performing sample phase alignment using the sample received from that next channel. O1. A method for performing a phase alignment method for sampling a video signal within a display unit, wherein the method comprises the following steps: A source driver of the corresponding display unit receives multiple video samples and control samples, and the video sample that arrives immediately before the control sample has a different analog level than the video sample that arrives immediately after the control sample; Sample the corresponding sample by changing the sampling phase for one of the control samples to a phase step; If it is determined that sampling occurred in a sample prior to the control sample, skip a clock cycle to sample at the control sample; and If it is determined that sampling has occurred in a sample after the control sample, a step of shifting the sampling phase backward by at least one phase step; O2. Steps additionally included in the method described in paragraph O1: A step of skipping a clock cycle to perform sampling in a control sample, and then moving the sampling phase forward by at least one phase step. O3. Methods additionally included in the methods described in O1: A step of determining whether sampling occurred in a previous sample or in a sample after a control sample using a single comparator. O4. Methods additionally included in the methods described in paragraph O1: A step of skipping the corresponding clock cycle or moving the corresponding sampling step backward, and then transmitting the corresponding sampling step to each of a plurality of amplifiers that sample video samples and control samples.
Claims
Claim 1 As a source driver for a display unit, An input terminal that receives an analog electromagnetic signal through an electromagnetic path containing a continuous series of analog video samples; A source driver comprising: a plurality of sampling amplifiers each configured to exclusively sample only a portion of the analog video sample and write the corresponding portion of the analog video sample to a location in a storage array designated to each sampling amplifier; and a plurality of column drivers configured to read one of the analog video samples from one of the locations in the storage array, amplify one of the analog video samples, and apply one of the amplified analog video samples to a column of a display panel. Claim 2 A source driver in claim 1, further comprising a second storage array having a designated position for each sampling amplifier, wherein the sampling amplifier is further configured to alternately write each corresponding portion of the analog video sample to the storage array or the second storage array, and while the sampling amplifier writes to the storage array, the column driver alternately reads from the storage array, and while the sampling amplifier writes to the second storage array, the column driver is configured to alternately read from the second storage array. Claim 3 A source driver in claim 2, further comprising a control logic circuit that enables the sampling amplifier to sample a corresponding portion of the analog video sample, enables the sampling amplifier to write to the storage array or the second storage array, and enables the column driver to read from the storage array or the second storage array. Claim 4 In claim 1, the electromagnetic signal includes a control signal for synchronization and is not applied to a column of the display panel, and the source driver further includes a dedicated sampling amplifier for sampling the control signal. Claim 5 In claim 1, the source driver does not include a digital-to-analog converter (DAC) used to convert video samples. Claim 6 In claim 2, the source driver is further configured such that the column driver can read in parallel from the storage array when the storage array is full, or read in parallel from the second storage array when the second storage array is full. Claim 7 A source driver according to claim 1, wherein the series of analog video samples arrive in a predetermined order such that each sampling amplifier can output a corresponding portion of the analog video sample to an adjacent storage location within the storage array. Claim 8 In claim 7, the source driver, wherein the predetermined order indicates that one of the sampling amplifiers exclusively samples only the control signal.
