Amplification process

Abstract

There is provided a method of amplifying a signal. The method comprises: applying analogue gain to a signal, an analogue gain amount being applied at an offset amount below a set gain amount; applying digital gain to the signal, a digital gain amount applied being an amount up to the offset amount; and converting the signal between digital and analogue between the steps of applying gain. The digital gain amount is adjustable based on an amplitude of the signal. An amplifier is also provided with an analogue gain module, a digital gain module and a converter connected between the modules.

Description

[0001] AMPLIFICATION PROCESS

[0002] FIELD OF THE INVENTION

[0003] The present invention relates to processing of signal, typically a variable amplitude signal, such as a variable amplitude audio signal. The processing typically involves minimising distortion or clipping, such as by automated gain adjustments, for example by combining analogue and digital gain.

[0004] BACKGROUND

[0005] When using cameras or audio equipment, digital processing or digital capture of images or audio is now ubiquitous. Over time, this means the digital functionality, electronic capabilities and conversion of signals into a digital form have developed and advanced.

[0006] Whether it be using cameras or audio equipment, fundamentally, the input received is a waveform. When recording or capturing new audio or visual content, this waveform is generally received as an analogue waveform instead of a digital waveform. Accordingly, it is common to convert the analogue waveform into a digital waveform. Further, an analogue waveform is typically amplified so an output with an appropriate strength can be passed on to the next stage.

[0007] When amplifying a signal, gain is applied to either attenuate or amplify the signal. There are times when the gain being applied can cause the amplitude of the signal to reach its maximum capability for the equipment being used due to the amplification being applied. This can also occur for other components through which the signal is passed between an input and an output.

[0008] When such a maximum capability is reached, whichever component this occurs in, a greater amplitude signal cannot be provided. Instead, the portions of the signal that would have been at a greater amplitude are replaced with an output at the maximum capability. This provides a flat (although not necessarily noiseless or fluctuation free) portion to the output signal. This is commonly referred to as “clipping”. Clipping is generally (although not always) undesirable. As such, users seek to avoid clipping occurring. This can be addressed manually, such as by a user reducing a variable gain on an amplifier. This is difficult for a user to correctly account for, however, and usually results in overcompensating and reducing the gain by too much or by not reducing the gain enough, such as by amateur users.

[0009] To address this, a simple solution is known. This involves implementing a digital preamp instead of an analogue preamplifier (also referred to as “preamp”) and reducing the gain being applied by the digital preamp by digitally limiting the signal when the signal would otherwise clip. However, applying digital gain, or manipulating digital gain is disliked due to the increase in noise the application of digital gain causes in a signal. This reduces perceived signal quality since the digital gain alters the sound in a manner audible to a user.

[0010] In order to provide an ability to reduce clipping without manual intervention, and avoid applying digital gain, various arrangements have been attempted. One such attempt involves providing two analogue preamplifiers per channel that are operated in parallel and are controlled by firmware. A large gain offset is maintained between the two preamps during use to artificially increase the dynamic range (the ratio between the largest and smallest signal values) available. The output from the two preamps is managed by firmware. This controls the output to provide signal from one preamp or the other depending on which is classified as the appropriate one to output to avoid clipping. However, providing this ability and functionality increases component count (and thus system complexity) and increases the expense of the device operating those components considerably, which is undesirable.

[0011] As a manner of addressing this, an alternative is using an audio device with two audio input channels. This is used to record a single input through the two channels with one of the analogue to digital converters (ADC) on each channel being set at different gain levels. This results in one ADC being set at a lower gain offset, and, in post-processing software, the user is provided with the ability to replace a section of an audio stream on a primary channel where clipping has occurred with the audio from the other channel. However, implementing this approach reduces the capability of the audio device by half due to removing the ability to use the second channel for a second audio input. Further, this only allows clipping to be reduced in post-processing instead of live.

[0012] Accordingly, there remains a need for reduction of clipping without increasing system complexity or harming signal quality.

[0013] SUMMARY OF INVENTION

[0014] According to a first aspect, there is provided a method of amplifying a signal, the method comprising: applying analogue gain to a signal, an analogue gain amount being applied at an offset amount below a set gain amount; applying digital gain to the signal, a digital gain amount applied being an amount up to the offset amount, the digital gain amount being adjustable based on an amplitude of the signal; and converting the signal between digital and analogue between the steps of applying gain.

[0015] The method according to the first aspect reduces clipping of a signal while maintaining signal transparency with a minimal increase in complexity of the system that implements it thereby making the process accessible to amateur users as well as to professional users. This is because the ability to limit clipping is simplified and also maintains the full functionality of the system that implements the method. Further, the digital gain adjustment allows high-resolution adjustments, keeping the signal clean and providing imperceptible, or low- perceptibility, adjustments.

[0016] The method according to the first aspect is intended to be able to be selectively applied by a user. In other words, while the method according to the first aspect may be applied in all situations, typically, a user chooses when to apply (or turn on or off) the method according to the first aspect. This applies to any apparatus or system configured to put the method into effect, which thereby allows the user to select when to apply the method according to the first aspect.

[0017] The method according to the first aspect may be applicable to any suitable signal. This may include a signal produced by a digital camera chip. Typically, the signal is an audio signal. It is intended such an audio signal may be a live audio signal instead of a pre-recorded audio signal, but a pre-recorded signal or audio signal would also be processable using the method according to the first aspect.

[0018] The analogue gain amount that is applied may be controllable, such as digitally controllable, for example in an automated manner.

[0019] The set gain amount may be predetermined or able to be modified. It may be set automatically, such as by an autogain function, or may be set by a user manually.

[0020] The set gain amount may be received or may be pre-programmed or available by some other means.

[0021] The term "amplitude" is intended to mean or include the peak signal, also referred to as signal peak, which may correspond to the absolute value of a waveform, section or portion of a waveform. This may encompass peak amplitude, such as the size of an individual peak in the signal; peak-to-peak amplitude, such as the value of the total swing from a positive peak to the adjacent negative peak; or peak value of the signal.

[0022] It follows from the method according to the first aspect that it is intended for analogue gain to be applied to an analogue signal (i.e. the signal in analogue form) and it is intended that digital gain is applied to a digital signal (i.e. the signal in digital or digitised form). While it would be possible to apply digital gain to a signal, convert the (digital) signal to an analogue signal and then apply analogue gain (likely after having already carried out an analogue to digital conversion), typically, converting the signal includes converting an analogue signal to a digital signal after applying analogue gain to the signal. This is advantageous because it allows signal resolution to be maintained in the signal to which analogue gain is applied. This limits reductions in signal quality and increases in noise while still allowing an analogue-digital gain synthesis to be applied to the signal.

[0023] The digital gain amount may be adjustable based on an amplitude of the signal before applying gain (analogue and / or digital) or after applying digital gain. The digital gain amount may be adjustable based on an amplitude of the signal after applying analogue gain. This allows the amount of digital gain applied to take into account the amount of analogue gain applied and compensate for fluctuations in the signal following analogue gain application. This helps limit clipping since, when analogue gain is applied before the digital gain, an appropriate amount of digital gain can be applied to the signal to avoid or limit clipping of the signal in processing downstream of the application of digital gain or at the stage of applying digital gain.

[0024] Further, the digital gain amount may be adjusted based on an amplitude of the signal after applying digital gain. This may therefore be the amplitude of the signal after applying both analogue and digital gain. This is advantageous because it allows the gain of the digital amplifier to be reduced if the amplitude of amplified digital signal is too high. This helps to limit clipping of the final output signal.

[0025] In some examples, the method further comprises reducing the digital gain amount if the amplitude of the signal is greater than a threshold, and / or reducing the analogue gain amount by a second offset amount if the digital gain amount is zero. The analogue gain amount may be reduced when the digital gain amount is reduced to zero, or is sufficiently close to zero, or is less than some threshold value.

[0026] It would be possible to calculate or estimate the amplitude of the signal based on the analogue gain amount and / or digital gain amount applied. Typically, the digital gain amount is adjustable based on a measured amplitude of the signal. This allows the amplitude to be known accurately providing the ability to more precisely adjust the digital gain amount and avoids needing an estimate which is unable to account for the amplitude of the signal before gain (analogue and / or digital) is applied.

[0027] The amplitude may be measured separately from the method according to the first aspect. Typically however, the method according to the first aspect further comprises measuring the amplitude of the signal. This allows the amplitude measurement to be contemporary (i.e. made at the time the measurement is used or relevant to the process). This makes the measurement of greater relevance and accuracy to the prevailing situation instead of having been measured at a different time when the amplitude may have been different.

[0028] The digital gain amount may be adjusted based on an absolute signal amplitude or pre-set signal amplitude. Typically, the digital gain amount is adjustable based on an amplitude of the signal relative to a threshold, the threshold being less than 0 decibels relative to full scale (dBFS). This is advantageous because it allows the digital gain to be adjusted before clipping occurs, which it would when the amplitude reaches 0 dBFS. Since the analogue and digital gain are, typically, applied to a live signal or at least to a signal to which no delay has been applied, it is not possible to look ahead to predict or establish whether the signal amplitude will increase or decrease. Setting the threshold below 0 dBFS provides time for the digital gain to be adjusted further reducing the likelihood of clipping since the gain will have been adjusted before clipping occurs in the amplitude continues to increase.

[0029] The threshold may be -0.3 dBFS or -0.5 dBFS, -3 dBFS, -9 dBFS or -24 dBFS. Typically though, the threshold is -1 dBFS. We have found that this provides a suitable balance between being far enough from 0 dBFS to provide time for the digital gain to be adjusted before the amplitude potentially reaches 0 dBFS. This is achieved while the threshold is close enough to 0 dBFS to avoid a user perceiving the signal as never reaching 0 dBFS and instead only reaching the threshold value. In other words, the user is unlikely to be able to hear a difference between -1 dBFS and 0 dBFS but may be able to hear the difference when the threshold is set at a greater distance from 0 dBFS.

[0030] The adjustment of the digital gain could be an increase in the digital gain. However, typically, when the signal amplitude reaches the threshold, the digital gain is reduced to maintain the signal amplitude at or below the threshold. This is advantageous because it limits the amplitude to below a level at which the signal would clip while (only) removing gain digitally, and thus maintaining or even enhancing signal quality at the same time. While the reduction in digital gain being applied can cause the amplitude to drop below the threshold, typically, when the digital gain is reduced, the amount of reduction in digital gain maintains the signal amplitude at the threshold (such as when the amplitude of the signal is increasing while the reduction is applied). This is typically achieved by making a small change (such as a single increment) in the amount of digital gain applied. This, therefore, maintains the maximum amount of signal amplitude while minimising clipping.

[0031] When reducing the digital gain to maintain the signal amplitude at the threshold, it is intended that this is achieved by maintaining the signal amplitude within an amplitude windowwith a largest amplitude of the window being the threshold. This allows for small differences from the threshold to be achieved when, for example, the available increments of digital gain do not permit an exact match of the signal amplitude to the threshold to be achieved. As such, typically, the window is up to one increment of the digital gain range.

[0032] Once the digital gain is applied the digital gain may only be reducible. Typically however, the digital gain is able to be increased. The digital gain may be increased if the signal amplitude reduces other than due to a reduction in digital gain amount being applied. Further, or regardless of this, if the signal amplitude reduces below the threshold, the digital gain may be increased to maintain the signal amplitude at the threshold. This provides a dynamic process that is able to take account of a varying signal and allows the signal amplitude to be restored or maintained after gain was removed, which artificially reduces the signal amplitude (or keeps the signal amplitude constant). At minimum, this shifts the signal closer to the user's desired output while maintaining the minimisation of clipping.

[0033] The digital gain may be increased instantaneously. Typically however, the digital gain is increased over a release time. This smooths the associated increase in signal amplitude making the increase less perceptible or less jarring to a user.

[0034] The increase in the digital gain to be applied may be applied fully over the release time. Alternatively, the increase in the digital gain may be applied in steps or increments with one step or increment being applied over the release time before the next step or increment is applied.

[0035] The release time is able to be any suitable period of time, such as equivalent to the period taken for a full wavelength to pass or up to several seconds (e.g. up to 10 seconds). Typically, the release time is 0.1 milliseconds (ms). This provides a small buffer to avoid the rate at which the digital gain is reapplied from being too quick and potentially having to be reduced again over the same period as the signal amplitude varies independently of this (for example due to the input being received that generates the signal), while also applying it quick enough that any drop in amplitude is maintained for only a short, and (almost) imperceptible period.

[0036] The amount the digital gain is able to be reduced by may be limited to a specific amount, such as a predetermined amount or a proportion of the digital gain applied, a proportion or a total amount of the offset amount or some other amount. Typically, the digital gain is reducible up to a total amount of the digital gain applied. This gives the greatest flexibility in reducing the digital gain and allows the gain applied to the signal to be made up of an increasing portion of analogue gain as the signal amplitude increases after the digital gain is initially applied. This maintains signal quality and provides an ability to make adjustments to the total gain applied to the signal in small increments (as is possible using digital gain) instead of in larger steps.

[0037] The digital gain may be the only gain amount that is able to be adjusted. Typically though, the analogue gain is (also) adjustable. The analogue gain may be adjusted based on an amount of digital gain applied. This is advantageous because it allows the signal amplitude controlled and clipping minimised by a combination of digital gain adjustment and (when appropriate) analogue gain adjustment.

[0038] The analogue gain may be the primary gain to adjust. In other words, the analogue gain may be adjusted before the digital gain is adjusted. Alternatively, the analogue and digital gain may be adjusted proportionally to each other or according to a (predetermined) ratio. Typically however, the analogue gain is adjusted when the digital gain is reduced by a total amount of digital gain applied. This prioritises reduction in the digital gain, which maintains signal quality by reducing noise as the digital gain is reduced. Further, this allows greater flexibility in the granularity of gain adjustment that occurs and allows the digital gain to act as a buffer and provide an ability for precision and dynamic adjustment. This provides an ability for adjustment to take place at a faster rate than could be achieved by reducing analogue gain in parallel or in priority over digital gain.

[0039] The analogue gain may be reduced by the minimum amount by which the analogue gain is adjustable or by some other amount. Typically, the offset amount at which the analogue gain is applied below the set gain amount is a first offset amount, the analogue gain being adjusted by a second offset amount when reduced. This creates a step change in the amount of analogue gain being applied, which makes minimising clipping easier. This is because it moves the total gain applied a distance from the point at which the signal is likely to clip. Further, this increases the dynamic range available for minimising clipping.

[0040] The first offset amount may be greater than or less than the second offset amount. Typically though, the wherein the first offset amount and the second offset amount are the same. This simplifies processing applied in the equipment that implements the method according to the first aspect and provides consistent application and modification of the analogue gain.

[0041] The first offset amount may be 30 decibels (d B) , 1 dB, 3 dB or some other amount. Typically, the first offset amount is 6 dB. The offset amount may be determined by the noise produced when applying the analogue gain (in other words, how "clean" the analogue gain application is). We have found that 6 dB provides an acceptable balance of offset from a point at which the signal is likely to clip and minimises the amount of digital gain that may then be applied. This was found to provide better results than when applying 3 dB or 9 dB.

[0042] When the analogue gain is reduced, this may be the final stage in adjusting gain. Typically however, digital gain is re-applied when the analogue gain is reduced, the amount of digital gain re-applied may be an amount up to the second offset amount. This allows the process to be repeated. Once the analogue gain is reduced and the digital gain re-applied, the digital gain may be reduced again if the relevant criteria (such as those set out above) are applicable. This means that if the digital gain is again reduced by the total amount, the analogue gain may be reduced by an amount. This may be by the first offset amount, the second offset amount or another offset amount.

[0043] The digital gain amount that is applied may be less than the offset amount. Typically though, the digital gain amount applied is the offset amount. This directly replaces the amount the analogue gain is offset by from a set gain amount with a digital gain buffer. This makes the modification to how the gain is applied as imperceptible as possible to the user so the user is aware of as little difference as possible when the method according to the first aspect is applied.

[0044] According to a second aspect, there is provided an amplifier suitable for limiting clipping, the amplifier comprising: an analogue gain module arranged in use to apply analogue gain to a signal, an applied analogue gain amount being at an offset amount below a set gain amount; a digital gain module arranged in use to apply digital gain to the signal, an applied digital gain amount being an amount up to the offset amount, the digital gain amount being adjustable based on an amplitude of the signal; and a converter connected between the analogue gain module and the digital gain module and arranged in use to convert the signal passing between the analogue gain module and the digital gain module between digital and analogue.

[0045] The amplifier may further comprise a controller arranged in use to control the digital gain applied by the digital gain module and arranged in use to control the analogue gain applied by the analogue gain module, the controller being arranged in use to adjust the analogue gain amount based on the digital gain amount applied. Providing the controller allows for the modules and other connected components to be controlled centrally, thereby allowing control of each module or component to take account of the state of each other module or component. Typically, the controller is a digital controller and is arranged to provide digital control. This provides the analogue gain module, which may be a preamplifier, with digital control, thereby providing a digitally controlled analogue module, allowing for precise, repeatable and reliable control of the analogue gain module.

[0046] Typically, the analogue gain module is connected to an input, the converter is an analogue to digital converter and the digital gain module is connected to an output.

[0047] The digital gain module and / or controller may be implemented in software or firmware.

[0048] The amplifier may include a processor. The processor may be arranged in use to implement, run or execute a (i.e. one or more) program(s), such as a program stored in memory which may be included in the amplifier or may be an accessible external memory source. The program may provide the digital gain module and / or the controller.

[0049] BRIEF DESCRIPTION OF DRAWINGS

[0050] Example processes and apparatus are described in detail below with reference to the accompanying drawings, in which:

[0051] Figure 1 shows a block diagram of an example process;

[0052] Figure 2 shows a flow diagram of an example process; and Figure 3 shows a schematic of an example apparatus.

[0053] DETAILED DESCRIPTION

[0054] An example process in accordance with the first aspect is generally illustrated at 1 in Figure 1 . The process (and apparatus) according to an aspect described herein is generally intended to be used for processing audio (i.e. audio signals, also referred to as audio waveforms or sound waves). However, it is possible to use the process for other signal forms, such as visual or radio signals. Typically, there is a limitation to signals that are a waveform or wave of some type, but as long as analogue and digital gain can be applied to the signal, potentially after some processing or conversion, then the process and / or apparatus may be applicable to the signal.

[0055] The core elements of this process are the application of gain in the form of analogue gain and digital gain and conversion of a signal from one of these types of signal (i.e. analogue or digital) to the other between the application of gain. This is achieved in the example process shown in Figure 1 by applying digitally controlled gain 10. In various examples, this is achieved by providing digitally controlled analogue gain.

[0056] In the example shown in Figure 1 , the signal to which the digitally controlled gain is applied is received 20 from a signal input. This is optional in the sense that simply providing a signal to which gain is able to be applied allows the process according to an aspect described herein to be carried out. When the signal is received from a signal input, this is typically an input of an original signal or signal feed instead of a pre-existing or pre-recorded signal but could be either.

[0057] As set out above, the signal can be in one or many forms. However, in the example shown in Figure 1 , the signal received from the input is an analogue audio signal. In view of this, after the digitally controlled analogue gain is applied, an analogue to digital conversion 30 is applied. This is typically achieved using an ADC.

[0058] The signal then has digital gain applied 40. To try to avoid clipping, the analogue gain is applied at a lower gain amount (also referred to as a “gain value”) than the intended gain. Depending on the configuration, the intended gain may be a default gain within a system implementing this process or may be a set, pre-set, predetermined or automatically generated or user-selected gain amount. In various examples, the amount the analogue gain is set lower than the intended gain is an “offset amount”. In various examples, the amount of digital gain is less than the offset amount or more than the offset amount (although this would make clipping more likely). In this example, however, the amount of digital gain applied is equal to the offset amount. Once the digital gain is applied, the signal is passed to an output 50. In some examples, the output is a computer; a device, system or apparatus able to carry out further processing of the signal; or a speaker.

[0059] The process 1 is able to limit clipping that occurs by adjusting the overall gain that is applied to the signal over the course of the process. This is able to be achieved in some examples by (only) (further) limiting the amount of analogue gain applied or by (only) adjusting the amount of digital gain applied. In the example shown in Figure 1 , this is achieved by limiting 60 the amount of digital gain applied.

[0060] This effectively turns the digital gain into a buffer or into headroom that can be used up. The digital gain is limited based on signal amplitude or signal peak.

[0061] The signal peak is identified by monitoring (i.e. measuring). In the example shown in Figure 1 , the monitoring is carried out at the ADC 30. This means it is the signal peak of the signal once analogue gain has been applied that is monitored. In other examples, the monitoring is carried out elsewhere, such as upstream of where the analogue gain is applied, or downstream of where the digital gain is applied.

[0062] The monitoring at the ADC is used to adjust 70 the limiting that is applied. As set out in more detail below in relation to Figure 2, the monitoring includes comparing the signal peak to a threshold.

[0063] In various examples, if it is not possible to (further) limit the digital gain, then the analogue gain is stepped down 80 based on the signal peak. The process of applying gain (digital gain as well as analogue gain) is then able to be restarted in some examples.

[0064] Turning to Figure 2, this sets out an example process generally illustrated at 2 in an example order. As set out above, there are core steps, namely application of analogue gain and digital gain and conversion between digital and analogue, and the remaining steps are optional in some examples. As a first step, the process is started by applying the process 200. In some examples, the process is operated automatically, but in other examples, the process is applied at the choice of a user or based on the user’s requirements.

[0065] A gain setting is then received 202. In this example, the gain setting is a gain setting selected by a user as the amount of gain they wish to apply. In some examples, the gain setting is set or selected automatically.

[0066] A signal from an input is received 204. This is received directly from an input in some examples. In other examples, this is received via one or more other components and / or one or more other processes. While Figure 2 indicates this is received after the gain setting, this ordering is not required, and since these are two inputs to the process, these may be provided in the opposite order or in parallel to each other.

[0067] Following receipt of the gain setting, analogue gain is applied to the signal at an offset gain amount 206. In various examples, the offset gain amount is a predetermined amount of gain below the amount of gain of the gain setting. The analogue gain is applied at the lower gain amount. In some examples, the offset gain amount is 6 dB, meaning the analogue gain applies 6 dB less gain than the amount of gain of the gain setting.

[0068] Up to this stage, the signal is an analogue signal (whether it was originally an analogue signal or was converted to an analogue signal before being received). In the example shown in Figure 2, the analogue signal is then converted 208 into a digital signal.

[0069] After this, digital gain is applied 210 to the signal. In some examples, the amount of digital gain that is applied, or that is available to apply, is the offset gain amount. As such, in the example shown in Figure 2, this is 6 dB. In other examples, the amount of digital gain is different from the offset gain amount.

[0070] The signal is then output 212. However, while there continues to be signal to which the process is being applied, signal is continually output. In view of this, in some examples, monitoring takes place. A stage of the monitoring is identifying 214 if the signal peak is at or less than (or higher than) a threshold. In various examples, the threshold is -1 dBFS. The signal peak, in some examples, is monitored using a peak monitoring circuit (not shown) or is monitored in some other way using an ADC or another suitable component. It can also or instead be a signal amplitude that is monitored, which, in a number of examples can correspond to the Root Mean Square (RMS) of the signal.

[0071] Regardless of how signal peak monitoring is conducted, in some examples, this is calculated on each audio sample by taking the absolute value of a respective sample and comparing that absolute value against the threshold. As this is carried out sample by sample, the rate will be the same as the sample rate of the system and it is carried out continuously. In various examples, the maximum sample rate is 96 kilohertz (kHz). Accordingly, a number of examples include sampling the audio signal at the prevailing sample rate (whether, set, predetermined or variable), and signal peak is assessed for the sample.

[0072] If the signal peak is at (or higher than) the threshold, the digital gain is reduced 216. From this, it follows that the output signal is updated 212 to reflect the reduced gain being applied to it.

[0073] When reduced, in some examples, the digital gain is reduced in increments. In various examples, the smallest increment the digital gain is able to be reduced by is 1 / (2A24) dB (so 2A(-24) dB). The size of this increment varies from example to example depending on a configuration of any digital gain module providing the digital gain.

[0074] Depending on the example, the digital gain is reduced by a predetermined or fixed amount when reduced or is reduced by a variable amount. Any of these amounts is a multiple of the smallest increment or the smallest increment.

[0075] In some examples, the amount the digital gain is reduced by corresponds to a relationship between the current signal peak and the difference between that and the threshold. The relationship is a function of the current signal peak and the difference between that and the threshold in various examples. If at the threshold, the digital gain is reduced to maintain the signal peak at or below the threshold.

[0076] In some examples, an assessment is carried out each time the digital gain is reduced to identify 218 if the digital gain being applied is more than zero. There are other ways to carry out this assessment in other examples. For example, a check may be applied after each reduction in digital gain to identify if the next incremental reduction would cause the digital gain be 0 dB. Alternatively, the assessment may be to identify how much of the offset gain amount, or whatever the maximum available digital gain is, the digital gain has been reduced by. As long as it can be identified when the digital gain cannot be reduced further, various example assessments are able to be applied.

[0077] If the digital gain amount applied is still more than 0 dB, the monitoring 214 of the signal peak, reducing 216 the digital gain, assessment 218 of the amount of digital gain remaining continues in a loop. This is as long as the signal peak does not drop 214 below the threshold of -1 dBFS.

[0078] Should there be no digital gain applied (i.e. the amount of digital gain applied is 0 dB or the amount remaining is less than the digital gain increment), then instead of reducing the digital gain, since that cannot be carried out, the analogue gain is reduced 220. In the example shown in Figure 2, the process then returns to the step of applying 210 digital gain as set out above. As an alternative, when the analogue gain is reduced, the digital gain may be applied at a different amount from the offset gain amount in some examples, or not at all.

[0079] When the analogue gain is reduced, this is reduced 220 by the offset gain amount in the example shown in Figure 2. As noted above, this is 6 dB. In other examples, this reduction may be a different amount.

[0080] The loop referred to above is also deviated from if the monitoring 214 identifies that the signal peak is less than the threshold. Should this occur an assessment 222 of the amount of digital gain being applied is carried out. In some examples, this forms part of the same assessment 218 as to whether the digital gain is more than 0 dB, but, in other examples, the assessments are carried out independently of each other.

[0081] If the digital gain amount being applied is at its maximum amount, which, in the example shown in Figure 2 is the offset gain amount of 6 dB, then the process 2 continues the monitoring 214 of whether the signal peak has dropped below the threshold. Alternatively, in some examples, if the digital gain amount is not at its maximum amount, then the digital gain is increased 224.

[0082] When the digital gain amount is increased, as with the reduction of the digital gain set out above, in various examples, the digital gain is increased in increments. The size of this increment varies from example to example. Typically, however, the increase in the gain is intended to be sufficient to return the updated 212 signal output that follows to return to the threshold or as close as possible to the threshold.

[0083] An instantaneous increase in digital gain is carried out in some examples. In various examples, the increase in digital gain is carried out over a release time. In some examples, the release time is 0.1 ms regardless of the amount of gain that the digital gain amount is to be increased by. In other examples, the release time is set based on an algorithm, such as one included in “Digital Dynamic Range Compressor Design - A Tutorial and Analysis”, Giannoulis et al., J. Audio Eng. Soc., Vol. 60, No. 6, 2012 June. As set out above, in various examples, the digital gain is increase based on a relationship between the current signal peak and the difference between that and the threshold. This can be an increase as a function of the current signal peak and the difference between that and the threshold.

[0084] Depending on the example, the order in which the steps shown in Figure 2 are carried out is modified. In some examples, the monitoring 214 digital gain amount assessments 218, 222; digital gain reduction 216 or increase 222 and reduction 220 in the analogue gain is provided between conversion 208 of the signal to a digital signal and the application 210 of digital gain. As a further alternative, this is carried out in parallel with, during or before the conversion. By applying the process as shown in Figure 2 and as described above, clipping of the output signal is kept to a minimum.

[0085] The example processes set out above are able to be implemented using an apparatus such as the apparatus generally illustrated at 3 in Figure 3. This is an amplifier or other form of audio device or audio processor.

[0086] This includes an input 300 that is connected to an analogue gain module in the form of a preamplifier 302. The preamplifier is connected to an ADC 304, which is connected to a digital gain module 306. The digital gain module is connected to an output 308. While the apparatus 3 has other components in other examples, having these components allows the above example processes to be implemented in conjunction with a controller 310 of the apparatus, which is connected to each of the preamplifier, ADC and digital gain module.

[0087] In various examples, the controller 310 and / or digital gain module 306 are implemented in software or firmware of the apparatus 3, such as by a processor. The controller is able to execute an algorithm or program to control the application of the process by the preamplifier 302, ADC 304 and digital gain module 306. The algorithm or program may be stored in memory that is either provided in the apparatus or is held externally and accessible by the controller. Additionally, in some examples, the controller carries out the signal monitoring process to identify signal peak.

[0088] The preamplifier 302 in some examples is a digitally controlled preamplifier (for example controlled by the controller 310). In an example, this is the amplifier disclosed in WO 2021 / 074596 A1. Whatever form the preamplifier takes, the digital control aspect of the preamplifier allows for automated or electronic control of the preamplifier to be applied instead of requiring a user to manually adjust the preamplifier.

Claims

CLAIMS1 . A method of amplifying a signal, the method comprising: applying analogue gain to a signal, an analogue gain amount being applied at an offset amount below a set gain amount; applying digital gain to the signal, a digital gain amount applied being an amount up to the offset amount, the digital gain amount being adjustable based on an amplitude of the signal; and converting the signal between digital and analogue between the steps of applying gain.

2. The method of claim 1 , wherein the digital gain amount is adjusted based on an amplitude of the signal after applying digital gain.

3. The method according to claim 1 or claim 2, further comprising: reducing the digital gain amount if the amplitude of the signal is greater than a threshold; and reducing the analogue gain amount by a second offset amount if the digital gain amount is zero.

4. The method according to any one of the preceding claims, wherein converting the signal includes converting an analogue signal to a digital signal after applying analogue gain to the signal.

5. The method according to any one of the preceding claims, wherein the digital gain amount is adjustable based on an amplitude of the signal after applying analogue gain.

6. The method according to any one of the preceding claims, wherein the digital gain amount being adjustable based on a measured amplitude of the signal.

7. The method according to claim 6, further comprising measuring the amplitude of the signal.

8. The method according to any one of the preceding claims, wherein the digital gain amount is adjustable based on an amplitude of the signal relative to a threshold, the threshold being less than 0 decibels relative to full scale (dBFS).

9. The method according to claim 8, wherein the threshold is -1 dBFS.

10. The method according to claim 8 or claim 9, wherein when the signal amplitude reaches the threshold, the digital gain is reduced to maintain the signal amplitude at or below the threshold.

11. The method according to claim 10, wherein when the digital gain is reduced, the amount of reduction in digital gain maintains the signal amplitude at the threshold.

12. The method according to any one of claims 8 to 11 , wherein the digital gain is increased if the signal amplitude reduces other than due to a reduction in digital gain amount being applied.

13. The method according to claim 11 or claim 12, wherein if the signal amplitude reduces below the threshold, the digital gain is increased to maintain the signal amplitude at the threshold.

14. The method according to claim 12 or claim 13, wherein the digital gain is increased over a release time.

15. The method according to claim 14, wherein the release time is 0.1 milliseconds (ms).

16. The method according to any one of the preceding claims, wherein the digital gain is reducible up to a total amount of the digital gain applied.

17. The method according to any one of the preceding claims, wherein the analogue gain is adjusted based on an amount of digital gain applied.

18. The method according to claim 16 and claim 17, wherein the analogue gain is adjusted when the digital gain is reduced by a total amount of digital gain applied.

19. The method according to claim 18, wherein the offset amount at which the analogue gain is applied below the set gain amount is a first offset amount, the analogue gain being adjusted by a second offset amount when reduced.

20. The method according to claim 19, wherein the first offset amount and the second offset amount are the same.

21. The method according to claim 19 or claim 20, wherein the first offset amount is 6 decibels (dB).

22. The method according to any one of claims 18 to 21 , wherein digital gain is re-applied when the analogue gain is reduced, the amount of digital gain reapplied being an amount up to the second offset amount.

23. The method according to any one of the preceding claims, wherein the digital gain amount applied is the offset amount.

24. An amplifier for limiting clipping, the amplifier comprising: an analogue gain module arranged in use to apply analogue gain to a signal, an applied analogue gain amount being at an offset amount below a set gain amount; a digital gain module arranged in use to apply digital gain to the signal, an applied digital gain amount being an amount up to the offset amount, the digital gain amount being adjustable based on an amplitude of the signal; and a converter connected between the analogue gain module and the digital gain module and arranged in use to convert the signal passing between the analogue gain module and the digital gain module between digital and analogue.

25. The amplifier according to claim 24, further comprising a controller arranged in use to control the digital gain applied by the digital gain module and arranged in use to control the analogue gain applied by the analogue gain module, the controller being arranged in use to adjust the analogue gain amount based on the digital gain amount applied.

26. The amplifier according to claim 25, wherein the controller is a digital controller and is arranged to provide digital control.

27. The amplifier according to any one of claims 24 to 26, wherein the analogue gain module is connected to an input, the converter is an analogue to digital converter and the digital gain module is connected to an output.

Images