Methods and apparatus for droplet deposition

Abstract

Methods or apparatuses for correcting errors which can occur in allocating drive waveforms. The method includes provisionally assigning, based on input pixel data, each fluid chamber within said array

Description

The invention relates to methods and apparatuses for ejecting droplets, and in particular to improvements for higher frequency deposition. As may be seen from Figure 1, known droplet ejection apparatuses include a droplet ejection head 10, as well as associated actuating circuitry 100 and head controller circuitry 200. The head controller circuitry 200 is configured to receive an input set of ejection data 610. This input set of ejection data (i.e. image data) contains the information necessary for the droplet ejection head to eject droplets in the corresponding pixels in the corresponding lines of the medium. The head controller circuitry 200 uses this input set of ejection data 610 to generate sub-sets of ejection data 210(1)-210(3) corresponding to three consecutive print cycles. This means that each sub-set of ejection data may comprise the pixel information relating to a medium’s pixel line. The head controller circuitry 200 also sends the series of sub-sets of ejection data to the actuating circuitry 100, with the actuating circuitry causing ejection of droplets based on each of the sub-sets of ejection data in the series, in turn. Each sub-set of ejection data causes the ejection of droplets from particular nozzles, each sub-set of ejection data depositing a corresponding pixel line on the medium (i.e. sub-sets of ejection data 210(1), 210(2), and 210(3) result in three individual consecutive pixel lines). For each sub-set 210(1)-(3) of ejection data, the actuating circuitry 100 generates and applies drive waveforms to the actuating elements 22(1)-(n+1) of the droplet ejection head 10 such that certain nozzles 18(1)-(n+1) eject a droplet. Each set of actuation commands is then sent to the drive waveform generating circuitry. Each set of actuation commands causes a drive waveform generating circuitry to apply drive waveforms to the actuating elements 22(1)-(n+1) of the droplet ejection head 10 such that they eject a droplet from certain of the nozzles 18(1)-(n+1). The particular nozzles 18(1 )-(n+1) and the sizes of the droplets ejected therefrom are determined by the set of actuation commands, and therefore the associated sub-set 210(1)-(3) of ejection data. Figure 2 illustrates a simplified representation of the array of fluid chambers where two fluid chambers experience a decrease in their volume due to the inward movement of their actuating elements. As a consequence, pressure in those fluid chambers increases. If the voltage applied across the actuating elements is high enough, fluid is forced out of the fluid chamber through a nozzle forming a fluidic droplet. These fluid chambers are called firing fluid chambers because they eject (“fire”) a droplet. Figure 2 also shows two fluid chambers that experience no change in the volume because their actuating elements remain stationary. These fluid chambers are called non-firing fluid chambers because they do not produce a droplet. Figure 3 shows an array of fluid chambers in a droplet ejection head 10 separated by two opposed side actuating elements 22(1)-(n+1) which are formed of a piezoelectric material such as PZT. Each side of the piezoelectric material is coated internally with a metal layer that acts like an electrode, used to apply a voltage across the actuating elements. Adjacent fluid chambers are located along the +x and -x directions, forming a linear array. The z-direction is shown as being broadly the direction in which droplets are ejected (i.e. vertically in the Figure, perpendicular to the plane of the nozzle), and although not shown, for completeness, the y-direction extends into and out of the Figure. If the same voltage is applied on both sides of the actuating elements, the actuating element remains stationary, while if a different voltage is applied to the electrodes on an actuating element, the actuating element of the fluid chamber moves to the side with higher voltage by virtue of the reverse piezoelectric effect which transforms potential differences into movement. These actuating elements are called active actuating elements, while the actuating elements that remain stationary are called non-active actuating elements. In most known single cycle printing modes, the fluid chambers are divided into firing fluid chambers (which, upon the actuation of their actuating elements, the nozzle associated with said firing fluid chamber ejects a droplet) and non-firing fluid chambers (which the nozzle associated with said non-firing fluid chamber does not eject a droplet, even if one or both actuating elements move). In one example of a single cycle printing mode, the movement of two actuating elements, for example actuating elements 22(1)-22(4), in opposite directions are necessary to eject a droplet. If only one actuating element, for example actuating element 22(4), were to move, while the other remained stationary, for example actuating element 22(5), the nozzle associated with said fluid chamber 10(5) would not eject a droplet and the fluid chamber would be a non-firing fluid chamber. Likewise, if both actuating elements 22(4) and 22(5) moved in the same direction, there would be substantially no change in volume of the fluid chamber 10(5), and once more no ejection of a droplet would be seen. For example, Figures 3 and 4 shows an array of fluid chambers operating in this printing mode. Fluid chambers 10(2), 10(3) and 10(4) are firing fluid chambers while the remaining fluid chambers 10(1), 10(5), 10(6) and 10(7) are non-firing fluid chambers. Figure 4 shows a simplified schematic of the actuating elements movements, that result in three droplets to be ejected by the nozzles associated with the three firing fluid chambers. Since each actuating element is shared between two fluid chambers (i.e., the droplet ejection head has a ‘shared-wall architecture’), two of the nozzles associated with two firing fluid chambers 10(2) and 10(4) are going to release droplets during the first half of the print cycle, while the third nozzle associated with the third firing fluid chamber 10(3) is going to release a droplet on the second half of the print cycle. As shown in Figure 3 and Figure 4 the actuating elements of each firing chamber move in opposite senses; and each non-firing chamber either: has actuating elements which move in the same sense, remain stationary or one moves while the other remains stationary. Even though the ejection occurs at different times within the cycle, the droplets land on the medium substantially at the same time effectively forming a pixel line on the medium because the ejection frequency is higher than the print medium speed. This pixel line may have gaps, the size of which correspond to a number of non-firing chambers located between firing chambers. Figure 5 shows the pixel line that is printed, in a single print cycle, by moving the actuating elements according to Figure 4. At this point it becomes clear that the pixel information of the sub-set of ejection data shown in Figure 3 does not correspond to the printed pixels on the pixel line in the medium. This means that the pixel information of the sub-set of ejection data is encoded. While the printed pixels correspond to a sub-section of the input set of ejection data 610 (and, consequently, correspond to firing chambers ejecting the droplets via the corresponding nozzle), the sub-set of ejection data comprises the information related to potential applied to at least one electrode of one actuating element of the firing chamber during a particular time within the ejection period (i.e. the potential value at a particular time within a drive waveform). In the case of Figure 3, the right electrode of the left actuating element and the left electrode of the right actuating element in each fluid chamber are electrically connected, i.e. the electrode Ob from actuating element 22(0) is electrically connected to the electrode 1a from actuating element 22(1), the electrode 1b from actuating element 22(1) is electrically connected to the electrode 2a from actuating element 22(2), the electrode 2b from actuating element 22(2) is electrically connected to the electrode 3a from actuating element 22(3) and so on. Even though Figure 3 shows the left electrode and the right electrode in each fluid chamber both physically and electrically connected, this is by no means the only electrode configuration that can be used. Individually addressable electrodes on each actuator wall can achieve the same effect if the appropriate changes in the electronics that applies the drive waveforms are made. As shown in Figure 4, to eject a droplet from the nozzle associated with fluid chamber 10(2), actuating element 22(1) and actuating element 22(2) have to move inwards. To move actuating element 22(1) to the left, the potential (V) applied to the electrode 1b must be larger than the potential applied to the electrode 1a (Via <Vw). Since electrode 1b is electrically connected to the electrode 2a (Vw = XM, the pixel of sub-set of ejection data representative of the fluid chamber 10(2) is a black pixel. To move actuator wall 22(1) to the right, the voltage electrode 2a must be larger than the potential applied to the electrode 2b (V2a >V2b). Since electrode 2b is electrically connected to the electrode 3a (V2b = Vsa), the encoded pixel of the sub-set of ejection data representative of the fluid chamber 10(3) is a white pixel. This procedure can be repeated along the array of fluid chambers and 5 electrodes, resulting in the following potentials applied to the electrodes: Corresponding encoded pixel information of First half of the cycle: Second half of the cycle: the sub-set of ejection data (“traditional encoding”) vOa = vob = vla Vla <vlb Voa = vob = vla Vla >vlb ‘white’ or ‘0’ Vlb = v2a V2a >V2b Vlb = v2a v2a <v2b ‘black’ or ‘1’ V2b = V3a V3a <V3b V2b = V3a v3a >v3b ‘white’ or ‘0’ v2b = V4a Via >V4b v3b = V4a Via <V4b ‘black’ or ‘1 ’ Vlb = v5a = v5b = v6a = v6b — Vya Vlb = v5a = v5b = v6a = v6b — v7a ‘white’ or ‘0’ Since the right electrode of the left actuator wall and the left electrode of the right actuator wall in each fluid chamber are electrically connected, we can consider that each fluid chamber has a single electrode disposed therein. 10 For the sake of simplicity, in this example, having a ‘black’ encoded pixel information corresponds to the binary value T and having a ‘white’ encoded pixel information corresponds to the binary value ‘O’, although of course this is an arbitrary convention and the opposite assignment (black-0; white-1) could equally be used. A problem has been identified with this arrangement, however. Due to the shared-15 wall architecture of the droplet ejection head and high ejection frequency, the single cycle printing mode described above is susceptible to drop placement error caused by a change in the polarity of the drive waveform between the consecutive pixel lines. This causes certain droplets to have lower velocity than the remaining droplets causing the drop placement error visible in the medium. This occurs because there are effectively two drive waveforms which can be used in ejecting a droplet - a first drive waveform, referred to generally herein as “A”, in which the actuator walls first move inward then move outward (so ejecting a droplet in the initial “inward” motion phase), and a second drive waveform, referred to generally herein as “B”, in which the actuator walls first move outward then move inward (so ejecting a droplet in the later “inward” motion phase). The inventors suggest that it is the switch from A to B (or B to A) which can lead to an error in droplet placement. Figures 6 to 9 describe in detail the source of A to B and B to A transitions. For example, to form the black pixels in columns n and n+1 in pixel lines 1 to 3 of Figure 6, the right actuating element of the fluid chamber 10(n-1), the left actuating element of the fluid chamber 10(n) and right actuating element of the fluid chamber 10(n) move in opposing directions, ejecting droplets from the nozzles 18(n) and 18(n+1) associated with the pair of neighbouring fluid chambers 10(n) and 10(n+1). Chamber 10(n-2) is shown for consistency with the later inventive solutions shown below, but does not affect the discussion of this system and will therefore not be referred to again. Note that the column of n-1 is only relevant to inform that a transition happens between pixel line 1 and pixel line 2. Using the (traditional) encoding method of Table 1, the encoded pixel lines corresponding to the columns n and n+1 in pixel lines 1 to 3 are depicted in Figure 7. To move the actuator walls, potential difference is applied across each actuating element 22(n-1), 22(n) and 22(n+1). The potential difference applied across each actuating element can be represented by two drive waveforms 101 and 102, exemplified in Figure 9. Depending on the information of the input set of ejection data 610 (or depending on the information in each sub-set 210 of ejection data), the actuating circuitry applies either drive waveform 101 or drive waveform 102. The drive waveform 101 can be seen as a phase A drive waveform while the drive waveform 102 can be seen as a phase B drive waveform. The drive waveform 102 is the inverted form of the drive waveform 101 along the x axis. Even though having an inverted form along the x axis may have benefits in preparing the ejection of the neighbouring fluid chamber, this is by no means necessary and having a drive waveform 102 that is different and asymmetrical in relation to the drive waveform 101 is possible depending on the specific arrangement and design of the chambers. What is important, however, is that the drive waveforms have opposed polarity. As shown in Figure 9, drive waveform 101 has two pulses 101A and 101B and the drive waveform 102 also has two pulses 102A and 102B. When pulse 101A of drive waveform 101 and the pulse 102B of drive waveform 102 are applied across a actuator wall, that actuator wall moves in a first direction, while when pulse 101B of drive waveform 101 and the pulse 102A of drive waveform 102 are applied across a actuator wall, the actuator wall moves in second direction. In the example of Figure 8, drive waveforms 101 and 102 are applied alternately, i.e.: • In print cycle 1: o The drive waveform 101 is applied across actuator wall 22(n). The actuator wall 22(n), therefore, moves in a first direction (left) because of pulse 101A and then moves in a second direction (right) because of pulse 101B. o The drive waveform 102 is applied across actuator wall 22(n-1) and across actuator wall 22(n+1). The actuator walls 22(n-1) and 22(n+1), therefore, move in a second direction (right) because of pulse 102A and then move in a first direction (left) because of pulse 102B. o This causes the ejection of a droplet through the nozzle 18(n) of a corresponding fluid chamber 22(n) during the A phase of the cycle, while the ejection of a droplet through the nozzle 18(n+1) of a corresponding fluid chamber 22(n+1) during the B phase of the cycle. • In print cycle 2: o The drive waveform 101 is applied across actuator wall 22(n-1) and across actuator wall 22(n+1). The actuator walls 22(n-1) and 22(n+1), therefore, move in a first direction (left) because of pulse 101A and then move in a second direction (right) because of pulse 101B. o The drive waveform 102 is applied across actuator wall 22(n) The actuator wall 22(n), therefore, moves in a second direction (right) because of pulse 102Aand then moves in a first direction (left) because of pulse 102B. o This causes the ejection of a droplet through the nozzle 18(n+1) of a corresponding fluid chamber 22(n+1) during the A phase of the cycle, while the ejection of a droplet through the nozzle 18(n) of a corresponding fluid chamber 22(n) during the B phase of the cycle. • In print cycle 3: o The drive waveform 101 is applied across actuator wall 22(n). The actuator wall 22(n), therefore, moves in a first direction (left) because of pulse 101A and then moves in a second direction (right) because of pulse 101B. o the drive waveform 102 is applied across actuator wall 22(n-1) and across actuator wall 22(n+1). The actuator walls 22(n-1) and 22(n+1), therefore, move in a second direction (right) because of pulse 102A and then move in a first direction (left) because of pulse 102B. o This causes the ejection of a droplet through the nozzle 18(n) of a corresponding fluid chamber 22(n) during the A phase of the cycle, while the ejection of a droplet through the nozzle 18(n+1) of a corresponding fluid chamber 22(n+1) during the B phase of the cycle In summary, the signals applied to the to the electrodes of the fluid chambers so that they produce the pattern above in each print cycle are shown in Figure 9. Note that in the above discussion, and the illustration in Figure 9, replacing drive waveform A with drive waveform B and vice versa would result in a similar solution, and would also lead to the problems identified below. From Figure 8 and Figure 9, it is possible to observe that between print cycle 1 and print cycle 2 (between pixel line 1 and pixel line 2) there is a transition from firing A drive waveforms to firing B drive waveforms back to firing A drive waveforms applied across the actuator wall 22(n). Similarly, there is a transition from firing B drive waveforms to firing A drive waveforms back to firing B drive waveforms applied across the actuator walls 22(n-1) and 22(n+1). These transitions cause polarity of the drive waveform response between drive waveform 101 and drive waveform 102 applied across the actuator wall 22(n), resulting in at least one of the droplets having a lower velocity which, in turn causes, droplet placement error visible on the medium. Without being bound by any particular theory, the inventors consider plausible that the problem may be caused by at least two interlinked factors: • Neighbouring firing fluid chambers share actuator walls. This means that, when the actuator walls are actuated so as to eject a droplet from one nozzle at a point during the print cycle, there is an interaction with the ejection from the neighbouring nozzle that is due to eject at different point during the same print cycle. • When transitioning from applying the drive waveform 101 across a actuator wall to applying the drive waveform 102 to the same actuating element in a different print cycle, the polarity of the drive waveform applied across the actuator wall is inverted. This means that, when applying drive waveform 101 across a actuator wall a pressure wave is generated inside the fluid chamber because of the movement of the actuator wall. This pressure wave is maintained with every application of drive waveform 101 across the actuator wall. If, suddenly, a drive waveform 102 is to be applied across the same actuator wall, a different pressure wave is generated inside the fluid chamber, introducing a new state and causing ejected droplets to be slower until the fluid chamber returns to a steady state. For applications where there are multiple droplets ejected from a single nozzle to form a large drop which then lands to form a pixel in a pixel line, at least a third interlinked factor may be added to the factors above. If the large drop is composed by five droplets, for example, the drive signals applied to the actuator walls are composed by five drive waveforms 101, i.e., the application of each drive waveform 101 directly corresponds to the increase in the pressure inside of the fluid chamber. Therefore, peak pressure inside the fluid chamber occurs when applying or soon thereafter the last drive waveform 101. When there is a transition from A to B or B to A, at least the drop immediately after that transition has the peak pressure when applying or soon thereafter the penultimate drive waveform 102, causing the drop to have a lower velocity. For high frequency applications, another interlinked factor may be added. In high frequency applications, the neighbouring fluid chambers will be subjected to residual pressure waves generated by previous actuations of the actuator walls of the fluid chamber itself or previous actuations of the actuator walls of neighbouring fluid chambers. Currently, this method of ejection has been used in applications where this effect can be neglected. However, as the fields of application expand and new uses are being developed a new solution is required. The present invention aims to address some or all of the drawbacks identified above. Disclosed herein is a controller for controlling the ejection of droplets from a droplet deposition head, the droplet deposition head comprising: an array of fluid chambers separated by interspersed actuator walls extending along at least a first direction, each fluid chamber communicating with an nozzle for the release of droplets of fluid and each of said actuator walls separating two neighbouring fluid chambers; wherein each of said actuator walls has at least one electrode disposed thereon for controlling the motion of said actuator wall; the controller being configured to: receive input pixel data; wherein the input pixel data forms at least a one-dimensional array having a first extent corresponding to droplet ejections along the first direction in a single print cycle; and the input pixel data indicates intended deposition locations on a deposition medium; and provisionally assign, based on said input pixel data, each fluid chamber within said array as either a firing fluid chamber or a non-firing fluid chamber so as to produce bands of one or more contiguous firing fluid chambers separated by bands of one or more contiguous non-firing fluid chambers; provisionally assign, based on said input pixel data, each at least one electrode of each actuator wall of the firing fluid chambers at least one of a first drive waveform and a second drive waveform such that, during the single print cycle, the at least one of the first drive waveform and the second drive waveform controls the motion of the actuator walls of each firing fluid chamber; wherein the first drive waveform and the second drive waveform are different drive waveforms; the controller is further configured to: allocate, to each at least one electrode of each actuator wall of the firing fluid chambers, one of the first drive waveform or the second drive waveform; identify at least one example where, for a first firing fluid chamber of the band of firing fluid chambers, the provisionally assigned drive waveform to each at least one electrode of each actuator wall of the first firing fluid chamber of the band of firing fluid chambers does not match the allocated drive waveform for said at least one electrode of said actuator wall of the first firing fluid chamber of the band of firing fluid chambers; correct the provisional assignment of the drive waveform assigned to at least one electrode of at least one fluid chamber such that the drive waveform assigned to each at least one electrode of each actuator wall of the first firing fluid chamber of the band of firing fluid chambers of the example corresponds to the allocated first drive waveform or second drive waveform for the first firing fluid chamber of the band. As used throughout, the first direction (direction of extent of the array) is used interchangeably with the x-direction. The second direction is used interchangeably with the y-direction, and refers to the direction in which depositions in successive or preceding print cycles are (or will be) located relative to the current print cycle (for example by motion of the fluid chambers, and / or a medium onto which the droplets are deposited. The third direction is used interchangeably with the z-direction, and refers to the distance away from the ejection location (e.g., nozzle) for a given fluid chamber. These are annotated consistently throughout the Figures. As discussed in more detail below, the different drive waveforms are provided to selectively provide a desired voltage across the various actuator walls to cause or prevent actuator wall motions, and thereby to eject (or not ejects) droplets from the fluid chambers. A first drive waveform is referred to generally as the “A” drive waveform and the second drive waveform is referred to generally as the “B” drive waveform. Some examples of drive waveforms typically include a time dependent voltage over the course of an ejection cycle, including a first portion having a first polarity relative to a nominal zero voltage, and a second portion having a second, different, polarity relative to a nominal zero voltage. Other drive waveforms may include a time dependent voltage over the course of an ejection cycle, including a first portion having the second polarity relative to a nominal zero voltage, and a second portion having the first polarity relative to a nominal zero voltage. That is to say there may be a matched pair of drive waveforms in which for a first portion of an ejection cycle the drive waveforms have opposite polarity to one another (e.g. the first drive waveform is positive, while the other is negative), then in the second portion of the ejection cycle both drive waveforms switch polarity such that the drive waveform which was positive in the first portion is negative in the second portion and the drive waveform which was negative in the first portion is positive in the second portion. That is, the drive waveforms remain in opposed polarity for the ejection cycle. Note that while in some cases the waveforms may be completely symmetric (waveform A = -waveform B), this is not essential, and variations between the specific shape of the positive and negative drive waveforms in the first and / or second portions may be desirable in some cases. Indeed, the first or portion of one drive waveform may last for a different amount of time than the corresponding portion of another drive waveform. Note that other drive waveforms entirely may be used as well. A useful example is a null drive waveform having a non time-dependent voltage, for example held at the nominal zero voltage (or any other fixed DC value) for the duration of an ejection cycle. In the examples discussed herein, the first firing fluid chamber of the band of firing fluid chambers means a fluid chamber at one end of a contiguous band. That is, where the band is a single fluid chamber, that single fluid chamber, and where the band comprises two or more fluid chambers, one of the two fluid chambers which is located between a firing fluid chamber and a non-firing fluid chamber. The provisional assignment and the allocation may not match one another in some cases, and it is this mismatch which the present disclosure seeks to correct. There are various reasons why the mismatch may occur (many of which are discussed in more detail below, but a way to understand this process is via the following procedure, which consistently assigns drive waveforms to electrodes. The procedure may begin by applying a rule for consistently assigning drive waveforms to electrodes as part of the provisional assignment. This rule can take many forms, for example in a given print cycle scanning along the fluid chamber array along the first direction (in a positive or negative direction) until a firing fluid chamber is identified. The first firing fluid chamber identified is assigned the first, A, drive waveform, with subsequent contiguous firing fluid chambers in a single band being assigned second, B, drive waveforms and first drive waveforms to form an alternating string, e.g. ...NNABABN... (where N indicates a non-firing fluid chamber, and A and B the first and second drive waveforms respectively. It is clear that, depending on the offset between firing fluid chambers in subsequent print cycles, a given fluid chamber may be forced to eject A on one cycle, and B on the second (or vice versa) - this occurs most obviously when the first fluid chamber of a contiguous band in one cycle is offset from the first fluid chamber of a contiguous band in a second cycle by an odd number of fluid chambers, and where the contiguous blocks in which this first fluid chamber occurs in each print cycle overlap (i.e. include at least one fluid chamber in each print cycle which is designated as a firing fluid chamber. The simple case of an offset of one is shown below, where fluid chambers are assigned by scanning along the array from left to right: Print cycle 1 N A B A B N Print cycle 2 N N A B A B Here, each overlapping pixel necessarily changes from A to B between the first cycle and second cycle and the provisional assignment causes a conflict with an allocation in which fluid chambers are prohibited from transitioning from A to B or B to A in successive ejection cycles. A correction is therefore implemented to resolve this conflict. Another rule might be to always insist that odd-numbered firing fluid chambers in a given cycle always receive one of the drive waveforms (for simplicity, we assume A). In a similar manner to that set out above, this can cause a conflict with the allocation rule regarding transitions. In addition, it can cause induced ejection in non-firing fluid chambers adjacent to the contiguous band of firing fluid chambers or no ejection in the first firing fluid chamber. A correction is therefore implemented to resolve these sources of conflict. Whatever the source of the conflict (and indeed the exact form of the corresponding correction), an insight by the inventors in this disclosure is the process for identifying such conflicts and for applying a suitable correction thereto, to maintain high-quality (e.g. correct droplet placement), high-frequency droplet ejections. Specific corrections are set out in more detail below, but at this stage we note that the correction may be applied to the fluid chamber in which the conflict occurs, and / or in another fluid chamber. This allows for flexibility in applying various corrections so as to select a correction which best preserves ejection consistency and positional accuracy. In general, three broad classes of conflict may be identified, in which the provisional assignment leads to a situation in which: 1: a fluid chamber which should be a firing fluid chamber (because a droplet is desired to be ejected in that location in that print cycle) will in fact not eject in that print cycle, for example due to induced actuator wall motions of nearby fluid chambers. 2: a fluid chamber which should be a non-firing fluid chamber (because it is desired than no droplet be ejected in that location in that print cycle) will in fact eject in that print cycle, for example due to induced actuator wall motions of nearby fluid chambers. 3: the firing / non-firing status of a given fluid chamber remains the same, but a drive waveform assigned to at least one electrode is inappropriate. Examples of various corrections and their applicability will now be discussed. Optionally, the controller is further configured to correct the provisional assignment such that the band of firing fluid chambers containing the first firing fluid chamber are deferred or brought forward in the first direction. In other words, the whole contiguous band of firing fluid chambers is shifted in the first direction (left or right in e.g., Figure 3) by one or more firing fluid chambers. This can be used, for example, to correct a misalignment in successive print cycles which causes an A to B or B to A transition. Optionally, the controller is further configured to correct the provisional assignment such that an additional firing fluid chamber is added to the band of firing fluid chambers, for example, wherein the last non-firing fluid chamber adjacent to the first firing fluid chamber of the band of firing fluid chambers may be corrected to be a firing fluid chamber of the band of firing fluid chambers. The addition of a firing fluid chamber affects the drive waveform assignments of the fluid chambers and can thereby avoid the conflict. Optionally, the controller is further configured to correct the provisional assignment such that an additional non-firing fluid chamber is added to the band of non-firing fluid chambers adjacent to the band of firing fluid chambers, for example the first firing fluid chamber of the band of firing fluid chambers may be corrected to be the last non-firing fluid chamber of the band of non-firing fluid chambers adjacent to the band of firing fluid chambers. As above, the removal of a firing fluid chamber can affect the drive waveform assignments of the fluid chambers and avoid conflicts. Note that one or more additions and removals can be combined in superposition to form the shift correction described above. Optionally, the controller is further configured to assign one of the first drive waveform or the second drive waveform to at least one electrode of a non-firing fluid chamber preceding the first firing fluid chamber of the band of firing fluid chambers. In this example, the firing / non-firing status of a fluid chamber is retained, but the specific drive waveform applied across one or both of its actuator walls may be changed to avoid conflicts. In particular, this may involve changing the intervening actuator wall between the non-firing and firing bands, but retain the non-firing fluid chamber adjacent to a firing fluid chamber, by making that non-firing fluid chamber’s other wall (located on the other side of that non-firing fluid chamber from the contiguous band of firing fluid chambers) not move (for example, by supplying a null signal), so as to inhibit droplet ejection. Alternatively, this may include moving the other actuator wall, e.g., to match the intervening actuator wall and cancel the motion to suppress what would otherwise be a volume change leading to an ejection (i.e., both walls move in the same sense). Optionally, the input pixel data forms a two-dimensional array having a first extent corresponding to droplet ejections along the first direction in a first print cycle and a second extent corresponding to a series of droplet ejections to be enacted in second print cycle; and wherein the controller is further configured to: correct the provisional assignment of the first firing fluid chamber of the band of firing fluid chambers of the first print cycle to be the last non-firing fluid chamber of the band of non-firing fluid chambers adjacent to the band of firing fluid chambers. In this example, the status of a given fluid chamber in different print cycles is expressly considered and the correction is applied in one of the print cycles to avoid conflicts or printing errors. Note that the print cycles need not expressly be adjacent, and a correction can be applied to print cycles separated by one or more intervening print cycles. In addition, the correction can be applied to the earlier or later print cycle - that is, the correction may be applied based on what has already happened, or it may be applied in anticipation of a conflict or error which is identifiable as due to occur in a later print cycle. The correction applied may take the same general form as the corrections discussed above, that is that optionally, the controller is further configured to correct the provisional assignment of the drive waveform assigned to at least one electrode of at least one fluid chamber in the second print cycle. Optionally, the controller is further configured to correct the provisional assignment of the drive waveform assigned to at least one electrode of at least one fluid chamber in the second print cycle such that the band of firing fluid chambers are deferred or brought forward in the first direction. Optionally the controller is further configured to assign one of the first drive waveform or the second drive waveform to at least one electrode of a non-firing fluid chamber in the second print cycle. Optionally, the controller is further configured to correct the provisional assignment such that an additional firing fluid chamber is added to the band of firing fluid chambers in the second print cycle. Each of these operates in a corresponding manner as the corrections discussed above, and has the corresponding advantages. In some cases, there may be no legitimate location to apply a correction within the nearby (temporally) print cycles. In such cases, the correction may simply be dropped, i.e., no correction is made. This may occur where, for example, the nearest location in which a correction can be made occurs at a time which occurs more than a threshold number of print cycles from the source of the error. This may be more than one print cycle in some cases, but may be as many as three or even five print cycles from the source of the error in other examples. Optionally, correcting includes the controller: applying a first correction to the provisional assignment; assessing the result of applying the first correction to identify whether the drive waveform assigned to each at least one electrode of each actuator wall of a first firing fluid chamber of the band of firing fluid chambers does not match the allocated drive waveform for said at least one electrode of said actuator wall of the first firing fluid chamber of the band of firing fluid chambers; and in the event that the drive waveform assigned to each at least one electrode of each actuator wall of a first firing fluid chamber of the band of firing fluid chambers results in a conflict applying a second correction, different to the first correction, to the provisional assignment. Note that the correction may be applied to (in theory) any fluid chamber (or multiple fluid chambers), and not necessarily the first firing fluid chamber of the band of firing fluid chambers, for example. This allows for flexibility in how the correction is applied, meaning that cases where a given fluid chamber leads to a conflict may have a correction applied in a variety of ways, allowing various adaptations to be applied depending on the status (firing / non-firing, specific drive waveform, etc.) of nearby fluid chambers. This general approach allows for a hierarchical system in which a first correction is attempted (the first correction may be selected randomly; based on a statistical approach of the correction which is most often applicable; based on the specific circumstances which led to a correction being necessary, or in any other manner), and then other legitimate correction strategies attempted should the first correction not be applicable for some reason. Optionally, in the event that applying the second correction leads to at least one instance in which the drive waveform assigned to each at least one electrode of each actuator wall of a first firing fluid chamber of the band of firing fluid chambers does not match the allocated drive waveform for said at least one electrode of said actuator wall of the first firing fluid chamber of the band of firing fluid chambers, the controller is further configured to: apply a further correction, different to each of the first and second corrections, to the provisional assignment. This allows for further corrections being possible in the event that the first two tried are not applicable for some reason. Evidently, this step may be applied repeatedly to provide third, fourth, etc. corrections as needed to try to find a workable correction in the specific circumstances. The controller may form part of an apparatus for improving the consistency of droplet ejection, comprising the droplet ejection head and any of the example controllers set out above. This provides a composite system for providing the above advantages in a single unit. Optionally, each fluid chamber has a single electrode disposed on both left and right actuator walls; and wherein firing fluid chambers in each print cycle are assigned drive waveforms by: starting at a first fluid chamber, considering adjacent fluid chambers along the array in a first direction until a firing fluid chamber is identified and arranging to supply the first drive waveform thereto; where the neighbouring fluid chamber in the first direction is also a firing chamber, arranging to apply the second drive waveform thereto; and continuing to alternate between the first drive waveform and the second drive waveform in neighbouring fluid chambers in the first direction along a line of consecutive firing fluid chambers until a non-firing fluid chamber is identified. This sets out a rule for consistently assigning drive waveforms to fluid chambers, and provides a coherent basis for understanding the framework in which corrections may be desirable. Optionally firing fluid chambers in each print cycle are assigned drive waveforms by repeating the process above until all firing fluid chambers have been assigned a drive waveform. In cases where the process of scanning through the chambers of the array was not started at one end of the array, the method may include starting at a fluid chamber part way along the array, proceeding in one direction until an end is reached, and then starting again at the other end of the array. In other examples, the array may be processed in smaller blocks and the process starts again at a new chamber each time and processes that block, before moving on to process a new block, and then continuing in this manner until all chambers have been processed. Whichever procedure is used, this allows for noncontiguous firing fluid chambers (i.e., firing fluid chambers in other bands of contiguous firing fluid chambers) to also be assigned drive waveforms in a consistent manner. Optionally the controller is further configured to identify at least one example in which the input pixel data indicates that a particular fluid chamber is scheduled to transition from ejecting using the first drive waveforms in a first print cycle to ejecting using the second drive waveform in a second print cycle; and apply a correction by: deferring or bringing forward the band of firing fluid chambers containing the first firing fluid chamber in the first direction; adding an additional firing fluid chamber to the band of firing fluid chambers; optionally wherein the last non-firing fluid chamber adjacent to the first firing fluid chamber of the band of firing fluid chambers is corrected to be a firing fluid chamber of the band of firing fluid chambers; adding an additional non-firing fluid chamber to the band of non-firing fluid chambers adjacent to the band of firing fluid chambers; optionally, wherein the first firing fluid chamber of the band of firing fluid chambers is corrected to be the last non-firing fluid chamber of the band of non-firing fluid chambers adjacent to the band of firing fluid chambers; and / or assigning one of the first drive waveform or the second drive waveform to at least one electrode of a non-firing fluid chamber preceding the first firing fluid chamber of the band of firing fluid chambers. This provides a general approach for identifying and correcting conflicts which occur in the process of assigning drive waveforms to fluid chambers to enact a given deposition pattern. Also disclosed herein is a method for improving the consistency of droplet ejection in a plurality of fluid chambers, the plurality of fluid chambers comprising: an array of fluid chambers separated by interspersed actuator walls extending along at least a first direction, each fluid chamber communicating with an nozzle for the release of droplets of fluid and each of said actuator walls separating two neighbouring fluid chambers; wherein each of said actuator walls has at least one electrode disposed thereon for controlling the motion of said actuator wall; the method comprising: receiving input pixel data; wherein the input pixel data forms at least a one-dimensional array having a first extent corresponding to droplet ejections along the first direction in a single print cycle; and the input pixel data indicates intended deposition locations on a deposition medium; and provisionally assigning, based on said input pixel data, each fluid chamber as either a firing fluid chamber or a non-firing fluid chamber so as to produce bands of one or more contiguous firing fluid chambers separated by bands of one or more contiguous non-firing fluid chambers; provisionally assigning, based on said input pixel data, each at least one electrode of each actuator wall of the firing fluid chambers at least one of a first drive waveform and a second drive waveform such that, during the single print cycle, the at least one of the first and second drive waveforms controls the motion of the actuator walls of each firing fluid chamber; wherein the first drive waveform and the second drive waveform are different drive waveforms; allocating, to each at least one electrode of each actuator wall of the firing fluid chambers, one of the first drive waveform and the second drive waveform; identifying at least one example where, for a first firing fluid chamber of the band of firing fluid chambers, the provisionally assigned drive waveform to each at least one electrode of each actuator wall of the first firing fluid chamber of the band of firing fluid chambers does not match the allocated drive waveform for said at least one electrode of said actuator wall of the first firing fluid chamber of the band of firing fluid chambers; correcting the provisional assignment of the drive waveform assigned to at least one electrode of at least one fluid chamber such that the drive waveform assigned to each at least one electrode of each actuator wall of the first firing fluid chamber of the band of firing fluid chambers of the example corresponds to the allocated first drive waveform or second drive waveform for the first firing fluid chamber of the band. This method encapsulates the advantages set out above in respect of the general applicability of the ideas set out herein. Optionally, the method further comprises correcting the provisional assignment such that the band of firing fluid chambers containing the first firing fluid chamber are deferred or brought forward in the first direction. This can be used, for example, to correct a misalignment in successive print cycles which causes an A to B or B to A transition. Optionally the method further comprises correcting the provisional assignment such that an additional firing fluid chamber is added to the band of firing fluid chambers, for example the last non-firing fluid chamber adjacent to the first firing fluid chamber of the band of firing fluid chambers may be corrected to be a firing fluid chamber of the band of firing fluid chambers. The addition of a firing fluid chamber affects the drive waveform assignments of the fluid chambers and can thereby avoid the conflict. Optionally the method further comprises correcting the provisional assignment such that an additional non-firing fluid chamber is added to the band of non-firing fluid chambers adjacent to the band of firing fluid chambers, for example the first firing fluid chamber of the band of firing fluid chambers may be corrected to be the last non-firing fluid chamber of the band of non-firing fluid chambers adjacent to the band of firing fluid chambers. As above, the removal of a firing fluid chamber can affect the drive waveform assignments of the fluid chambers and avoid conflicts. Note that one or more additions and removals can be combined in superposition to form the shift correction described above. Optionally the method further includes assigning one of first drive waveform and second drive waveform to at least one electrode of a non-firing fluid chamber preceding the first firing fluid chamber of the band of firing fluid chambers. In this example, the firing / non-firing status of a fluid chamber is retained, but the specific drive waveform applied across one or both of its walls may be changed to avoid conflicts. In particular, this may involve changing the intervening wall between the non-firing and firing bands, but retain the nonfiring fluid chamber adjacent to a firing fluid chamber, by making that non-firing fluid chamber’s other wall (located on the other side of that non-firing fluid chamber from the contiguous band of firing fluid chambers) not move (for example, by supplying a null signal), so as to inhibit droplet ejection. Alternatively this may include moving the other wall, e.g. to match the intervening wall and cancel the motion to suppress what would otherwise be a volume change leading to an ejection. Optionally, the input pixel data forms a two-dimensional array having a first extent corresponding to droplet ejections along the first direction in a first print cycle and a second extent corresponding to a series of droplet ejections to be enacted in a second print cycle; and the method optionally further comprises correcting the provisional assignment of the first firing fluid chamber of the band of firing fluid chambers of the first print cycle to be the last non-firing fluid chamber of the band of non-firing fluid chambers adjacent to the band of firing fluid chambers. In this example, the status of a given fluid chamber in different print cycles is expressly considered and the correction is applied in one of the print cycles to avoid conflicts or printing errors. Note that the print cycles need not expressly be adjacent, and a correction can be applied to print cycles separated by one or more intervening print cycles. In addition, the correction can be applied to the earlier or later print cycle - that is, the correction may be applied based on what has already happened, or it may be applied in anticipation of a conflict or error which it identifiable as due to occur in a later print cycle. The correction applied may take the same general form as the corrections discussed above, that is that optionally, the method further includes assigning one of first drive waveform and second drive waveform to at least one electrode of at least one fluid chamber in the second print cycle. Optionally the method further includes correcting the provisional assignment of the drive waveform assigned to at least one electrode of at least one fluid chamber in the second print cycle such that the band of firing fluid chambers are deferred or brought forward in the first direction. Optionally the method further includes assigning one of a first drive waveform and a second drive waveform to at least one electrode of a non-firing fluid chamber in the second print cycle. Optionally, the method further includes correcting the provisional assignment such that an additional firing fluid chamber is added to the band of firing fluid chambers in the second print cycle. In some cases, there may be no legitimate location to apply a correction within the nearby (temporally) print cycles. In such cases, the correction may simply be dropped, i.e., no correction is made. This may occur where, for example, the nearest location in which a correction can be made occurs at a time which occurs more than a threshold number of print cycles from the source of the error. This may be more than one print cycle in some cases, but may be as many as three or even five print cycles from the source of the error in other examples. The method may include correcting including: applying a first correction to the provisional assignment; assessing the result of applying the first correction to identify whether the drive waveform assigned to each at least one electrode of each actuator wall of a first firing fluid chamber of the band of firing fluid chambers does not match the allocated drive waveform for said at least one electrode of said actuator wall of the first firing fluid chamber of the band of firing fluid chambers; and in the event that the drive waveform assigned to each at least one electrode of each actuator wall of a first firing fluid chamber of the band of firing fluid chambers results in a conflict, applying a second correction, different to the first correction, to the provisional assignment. Note that the correction may be applied to (in theory) any fluid chamber (or multiple fluid chambers), and not necessarily the first firing fluid chamber of the band of firing fluid chambers, for example. This allows for flexibility in how the correction is applied, meaning that cases where a given fluid chamber leads to a conflict may have a correction applied in a variety of ways, allowing various adaptations to be applied depending on the status (firing / non-firing, specific drive waveform, etc.) of nearby fluid chambers. This general approach allows for a hierarchical system in which a first correction is attempted (the first correction may be selected randomly; based on a statistical approach of the correction which is most often applicable; based on the specific circumstances which led to a correction being necessary, or in any other manner), and then other legitimate correction strategies attempted should the first correction not be applicable for some reason. The method may further include, in the event that applying the second correction leads to at least one instance in which the drive waveform assigned to each at least one electrode of each actuator wall of a first firing fluid chamber of the band of firing fluid chambers does not match the allocated drive waveform for said at least one electrode of said actuator wall of the first firing fluid chamber of the band of firing fluid chambers; the method further comprises: applying a further correction, different to each of the first and second corrections, to the provisional assignment. This allows for further corrections being possible in the event that the first two tried are not applicable for some reason. Evidently, this step may be applied repeatedly to provide third, fourth, etc. corrections as needed to try to find a workable correction in the specific circumstances. Optionally, each fluid chamber has a single electrode disposed on both actuator walls; and wherein firing fluid chambers in each print cycle are assigned drive waveforms by: starting at a first fluid chamber, considering neighbouring fluid chambers along the array in a first direction until a firing fluid chamber is identified and arranging to supply the first drive waveform thereto; where the neighbouring fluid chamber in the first direction is also a firing fluid chamber, arranging to apply a second drive waveform thereto; and continuing to alternate between the first drive waveform and the second drive waveform in neighbouring fluid chambers in the first direction along a line of consecutive firing fluid chambers until a non-firing fluid chamber is identified. This sets out a rule for consistently assigning drive waveforms to fluid chambers, and provides a coherent basis for understanding the framework in which corrections may be desirable. The method may include further firing fluid chambers in each print cycle being assigned drive waveforms by repeating the process of above until all firing fluid chambers have been assigned a drive waveform. This allows for non-contiguous firing fluid chambers (i.e., firing fluid chambers in other bands of contiguous firing fluid chambers) to also be assigned drive waveforms in a consistent manner. The method may further comprise identifying at least one example in which the input pixel data indicates that a particular fluid chamber is scheduled to transition from ejecting using a first one of the drive waveforms in a print cycle to ejecting using a second drive waveform in a second print cycle; and applying a correction by: deferring or bringing forward the band of firing fluid chambers containing the first firing fluid chamber in the first direction; adding an additional firing fluid chamber to the band of firing fluid chambers; optionally wherein the last non-firing fluid chamber adjacent to the first firing fluid chamber of the band of firing fluid chambers is corrected to be a firing fluid chamber of the band of firing fluid chambers; adding an additional non-firing fluid chamber to the band of non-firing fluid chambers adjacent to the band of firing fluid chambers; optionally, wherein the first firing fluid chamber of the band of firing fluid chambers is corrected to be the last non-firing fluid chamber of the band of non-firing fluid chambers adjacent to the band of firing fluid chambers; and / or assigning one of a first drive waveform and a second drive waveform to at least one electrode of a non-firing fluid chamber preceding the first firing fluid chamber of the band of firing fluid chambers. This provides a general approach for identifying and correcting conflicts which occur in the process of assigning drive waveforms to chambers to enact a given deposition pattern. Also disclosed herein are computer readable instructions which, when run by a controller, cause the controller to implement any of the methods set out herein. The invention will now be described, by way of example only, with reference to the Figures, in which: Figure 1 shows a schematic of the operation of a known droplet ejection apparatus; Figure 2 shows different representations of known fluid chamber actuations (actuator wall motions) and a simplified representation used herein; Figure 3 illustrates the known single cycle printing mode with corresponding sub-set of ejection data; Figure 4 illustrates a simplified representation of the known single cycle printing mode; Figure 5 shows the resulting printed pixel line from the wall movements of Figure 4; Figure 6 shows three pixel lines to be printed, in three single cycles, by two neighbouring nozzles; Figure 7 shows the encoded pixels in the known single cycle printing mode in three single cycles, by two neighbouring nozzles to achieve the printing pattern in Figure 6; Figure 8 depicts actuator wall movements that result in the ejection of each black pixel of Figure 6 caused by the encoding of Figure 7 during A phase (full lines) and during B phase (dash lines) Figure 9 shows drive waveforms applied during each of the three print cycles in the known single cycle printing mode across the walls 22(n-1), 22(n) and 22(n+1) of fluid chamber 10(n) and 10(n+1) associated with the nozzles that fire the pattern of Figure 6; Figure 10 shows a block diagram of actuating circuitry according to the present invention; Figure 11 is a block diagram of head controller circuitry according to the present invention; Figure 12A shows an example of encoded pixels in three single print cycles, by four neighbouring nozzles that print the pixels in Figure 6 according to the present invention; Figure 12B shows an example of printed pixels in three single cycles, by four neighbouring nozzles that print the pixels in Figure 6 based on the encoding in Figure 10A according to the present invention; Figure 12C shows the wall movements that result in the ejection of each black pixel of Figure 6 caused by the encoding of Figure 10A during A phase (full lines) and during B phase (dash lines); Figure 12D shows the drive waveforms applied during each of the three print cycles across the walls 22(n-2), 22(n-1), 22(n) and 22(n+1) of fluid chambers 10(n) and 10(n+1) to cause the motions shown in Figure 11; Figure 13A shows an example of encoded pixels in three single cycles, by four neighbouring nozzles according to the present invention; Figure 13B shows an example of printed pixels in three single cycles, by four neighbouring nozzles that print the encoded pixels in Figure 13A; Figure 13C shows the wall movements that result in the ejection of each black pixel of Figure 13B caused by the encoding of Figure 13A during A phase (full lines) and during B phase (dash lines); Figure 13D shows the drive waveforms applied during each of the three print cycles across the walls 22(n-2), 22(n-1), 22(n) and 22(n+1) of fluid chambers 10(n) and 10(n+1) to cause the motions shown in Figure 13C; Figure 14 illustrates the encoding process of a first line 616(1) of input set of ejection data following an ejection rule ‘1’ ‘0’ and resulting pixel line according to an example of the present invention; Figure 15 shows an example image on reproduced a medium using traditional encoding (bottom) and using the encoding of Figure 14 (top); Figure 16 illustrates an encoding process of second and third lines of input ejection data following an ejection rule ‘1’ ‘0’ and resulting pixel line according to an example of the present invention; Figure 17 illustrates an encoding process of a fourth line of input ejection data following an ejection rule T ‘0’ and resulting pixel line according to an example of the present invention; Figure 18 illustrates an encoding process of a fifth line of input ejection data following an ejection rule T ‘0’ and resulting pixel line according to an example of the present invention; and Figure 19 is a flow chart illustrating a hierarchical procedure for applying a set of different corrections. Consider now Figures 10 and 11 which show respectively actuation circuitry 100 and head controller circuitry 200 for enacting the procedures set out herein. In these blocks advantageous optional processes and features are shown in dashed lines. Note that although some aspects discussed herein may be presented as a method or as apparatus, circuitry, controllers, etc., this is presented in that format for convenience only and the principles set out herein may be implemented as a method, apparatus, controller, circuitry or otherwise without diverging from the central concepts. Indeed, the disclosure also extends to software, or computer code, which when run by a processor, is suitable for causing the processor to enact the methods set out herein. Turning to Figure 10, it can be seen that an example of the actuating circuitry 100 comprises an actuation control circuitry, which receives the sub-sets of ejection data 210(1)-(3) and stores them in memory 120 before sending consecutively or simultaneously sub-sets of ejection data 210(1)-(3) to the encoder circuitry 130. The actuating circuitry 100 may also optionally receive instructions 216 to set new rules in the memory 121, which then may be used by the encoding circuitry 130 to encode the sub-sets of ejection data 210(1)-(3), for example to provide new correction options. Additionally, or alternatively, the drive waveform generating circuitry 110 may optionally receive a clock signal (CLK) that allows synchronisation of the drive waveform generation and the application of the drive waveforms 110(1 )-(n+1) to the actuator walls 22(1)-(n+1) (not shown in this Figure). To observe the problem in the sub-sets of ejection data 210, it is necessary to observe at least two sub-sets of ejection data 210(1) and 210(2). However, the knowledge of two sub-sets of ejection data 210(1) and 210(2) is not required to encode according to the present invention. This means that the actuating circuitry 100 may receive one sub-set of ejection data 210(1), encode into the actuator ejection data 136(1), generate the drive waveforms 110(1)-110(n+1), and send those drive waveforms to be applied to the electrodes of each actuator wall 22(1)-22(n+1). This also means that in extreme and exotic cases, the encoding may be applied to each sub-set of ejection data even though, in those cases, they would not experience the problem described above. In some of the simpler examples, the original data may be manipulated in a way that any data that would cause the transitions gets removed or some additional image data is included in order to avoid these transitions. In the example of Figure 11, the head controller circuitry 200 is shown as comprising an encoding circuitry 230, which receives the input set of ejection data 610 and encodes, based on the rules stored in the memory 220, said input set of ejection data 610 into subsets of ejection data 210(1)-(3). Each sub-set of ejection data 210(1)-(3) corresponds to one pixel line on the deposition medium. The head controller circuitry 200 then sends sub-set of ejection data 210(1)-(3) consecutively or successively to actuating circuitry 100 in the droplet ejection head 10. Alternatively, the head controller circuitry 200 may comprise a buffer (or memory) 240 where one or more sub-sets of ejection data 210(1)-(3) are temporarily stored. Upon receiving deposition media speed data 310, the buffer 240 may send the sub-set of ejection data 210(1)-(3) consecutively or successively to actuating circuitry 100 in the droplet ejection head 10. Further, the head control circuitry 200 may receive a new set of rules 620 and store them in a memory 220. At this stage, a further advantage of the procedure is apparent. Specifically, that, if the droplet ejection head is operating in a grey-scale mode, then the value attributed to the problematic pixel may be distributed by the neighbouring nozzles in such a way that no ABA to BAB transitions are observed. This transfer may be made by e.g. splitting the single intended deposition of multiple droplets between nearby fluid chambers so that no electrode of each actuator wall corresponding to firing fluid chamber receives a drive waveform that is different from the allocated drive waveform for that at least one electrode of each actuator wall corresponding to firing fluid chamber. In the examples above, the input pixel data may be one or more sub-sets of ejection data 210 or input set of ejection data 610 which may have been processed by the circuitry 100, 200 in such a way as to be divided into sub-sub-sets of ejection data, with each sub-set of ejection data containing the information for each print cycle. The actuating circuitry 100, 200 of both examples uses three sub-sets of ejection data 210(1)-(3) in order to move the problematic pixel around between fluid chambers and / or print cycles 1 to 3. These sub-sets of ejection data 210(1)-(3) may be sent consecutively or simultaneously to the encoder circuitry 130 in the actuating circuitry 100. The encoder circuitry 130 may then encode each sub-set of ejection data 210(1)-(3) into actuator ejection data 136(1)-(3), based on the rules provided by the memory circuit 121, to be sent consecutively to drive waveform generating circuitry such that the drive waveforms 110(1)-110(n+1) may be applied to the actuator walls 22(1)-(n+1) of the droplet ejection head 10, causing the corresponding nozzles 18(1)-(n+1) to eject a droplet per print cycle. Each one of these examples has advantages and disadvantages for the final image quality and some might be more demanding with regards to the encoding computational power. The core point is that various options are available to choose between in any given situation. In each example, the apparatus or method is arranged to apply a correction according to one of the known acceptable corrections to improve print quality by nevertheless ensuring that ejections occur, without impacting the status of nearby fluid chambers. Although shown as simple circuitry, the invention may also include the droplet ejection head itself, that is to say the fluid chambers, electrodes, interspersed actuator walls, and so forth. There are variety of ways for avoiding the conflicts discussed above. One way to achieve a transition-free encoding may include some image data getting misplaced to a nearby pixel in either the X or Y direction (along the array, or to another print cycle). Returning to the example described in relation to Figures 6 to 9, under the presently disclosed novel encoding system, the encoding of the problematic pixel in encoded pixel line 2 is transferred from the nozzle associated with the fluid chamber 10(n-1) to the nozzle associated with the fluid chamber 10(n-2) which now is a ‘black’ pixel. This causes the assignment of the next encoded pixel (the encoded pixel associated with the fluid chamber 10(n-1)) to be the opposite (i.e. ‘white’ pixel) of the one preceding, which in turn cause, the following encoded pixels to be the opposite of the one preceding as shown in Figures 12A to 12D. Figures 12A to 12D illustrate a specific example of a conflict and a potential resolution by moving encoded pixels. While Figure 6 shows the provisional assignment which might have been made in a naive manner without knowledge of the problem set out above, it leads to print errors as set out above. Figure 12A shows the encoded pixel lines which are used to print the pixels in Figure 12B. To print without errors, the motions of the fluid chamber actuator walls must follow the motions indicated shown in Figure 12C, where full lines indicate motions which occur in the first portion of a print cycle and dashed lines indicate a subsequent motion in the second portion of the print cycle (this presentation manner is followed throughout). The drive waveforms (shown in Figure 12D in which the A drive waveform 101 and B drive waveform 102 are shown in exemplary detail) applied across actuator walls 22(n-2), 22(n-1) and 22(n) and 22(n+1) of the fluid chambers 10(n) and (10n+1) are shown in Figure 12D. In this example, in all three print cycles 1,2 and 3: • The drive waveform 101 is applied across actuator wall 22(n). The actuator wall 22(n), therefore, moves in a first direction (left in this example) because of pulse 101A and then moves in a second direction (right in this example) because of pulse 101B. • The drive waveform 102 is applied across actuator wall 22(n-1) and across actuator wall 22(n+1). The actuator walls 22(n-1) and 22(n+1), therefore, move in a second direction (right) because of pulse 102A and then move in a first direction (left) because of pulse 102B. • This causes the ejection of a droplet through the nozzle 18(n) of a corresponding fluid chamber 22(n) during the first portion of the cycle, while the ejection of a droplet through the nozzle 18(n+1) of a corresponding fluid chamber 22(n+1) during the second portion of the cycle. • Actuator wall 22(n-2) receives an A drive waveform in the second print cycle only, in order to ensure that fluid chamber 10(n-1) has both actuator walls moving in the correct manner to eject a droplet. This way no transition between AB drive waveforms to BA drive waveforms occurs and no polarity of the drive waveform response is observed. Note that the assignment of A and B is arbitrary, the core point is that each fluid chamber has either the same drive waveform in each successive print cycle, or it has a null cycle for at least one cycle interspersed in a change from A to B or B to A. Likewise, the association between motions to the left (right) and positive (negative) polarity of signal is due to the choice of actuator wall material, and may be opposite in some examples. Nevertheless, in each following example, while the above assignment of A, B, polarity, and direction of actuator wall motion is adhered to for clarity, interchanging A and B throughout that example will result in the same overall general effect, as will changing the actuator wall material to one which responds in the opposite manner to applied voltages if the polarity of the applied drive waveforms is also inverted. The above correction includes adjusting the encoded pixels, as can be seen by comparing Figures 7 and 12A. Here, fluid chambers 10(n+1) and 10(n-1) which would have been assigned drive waveforms as in Figure 7, have their encoding shifted one pixel to the left. Due to the shared wall architecture, the correct fluid chambers do indeed eject, but the change in assigned pixels means that no A to B or B to A transitions occur. In this way, the general procedures set out herein can be thought of as leveraging the proximity effects of shared wall architectures to correct errors induced by other encodings. Note that a shift of an encoding pixel can equally be thought of as deleting a pixel from one location and adding a new pixel at a different location. In this way, the shifting operation can be thought of as a composite process, which can be broken down into other operations. These are discussed below in more detail. Figures 13A to 13D show similar parts of the process in a similar context, using a different correction. Figure 13A shows encoded pixels to print a set of pixels shown in Figure 13C, and the actuator wall motions to achieve this are shown in Figure 13B. Similar drive waveforms to those shown in Figure 12D for the A drive waveform 101 and B drive waveform 102 are shown in Figure 13D. In this example, the A to B or B to A transition is avoided because the problematic pixel (pixel corresponding to fluid chamber 10(n-1) in the second pixel line, as in Figure 12C) is shifted to the next line. This clearly avoids the transition because in each of the three print cycles: • The drive waveform 101 is applied across actuator wall 22(n). The actuator wall 22(n), therefore, moves in a first direction (left) because of pulse 101A and then moves in a second direction (right) because of pulse 101B. • The drive waveform 102 is applied across actuator wall 22(n-1) and across actuator wall 22(n+1). The actuator walls 22(n-1) and 22(n+1), therefore, move in a second direction (right) because of pulse 102A and then move in a first direction (left) because of pulse 102B. • This causes the ejection of a droplet through the nozzle 18(n) of a corresponding fluid chamber 22(n) during the first portion of the cycle, while the ejection of a droplet through the nozzle 18(n+1) of a corresponding fluid chamber 22(n+1) during the second portion of the cycle. • Actuator wall 22(n-2) receives an A drive waveform in the third print cycle only, in order to ensure that chamber 10(n-1) has both actuator walls moving in the correct manner to eject a droplet. Here, the shifting of the pixel may cause a small displacement in the y direction of the droplet, but for example in high frequency applications, this shift is small, and it has been realised that in most cases the benefit in avoiding droplet ejection errors far exceeds the relatively small displacement in the y-direction. Similar comments apply elsewhere regarding misplacing pixels in the x-direction in that it may be better to eject a droplet physically close to the intended location rather than deleting the pixel entirely. Note that rather than shifting the pixel one cycle forwards in time (downward in the Figures), it could equally be brought forward and ejected sooner than it would normally have been, depending on the specific context. Additional to shifting one or more pixels to a different print cycle, other corrections may also be applied, for example shifting the printed pixel to a different fluid chamber, changing the drive waveform applied to a given electrode, and so forth. This is shown as well in Figure 13C, where the printed pixel corresponding to the fluid chamber 10(n-1) is moved to the third print cycle. In Figures 14 and 15 another example of corrections applied to the data are shown. In some of the simpler examples, the original data may be manipulated in a way that data that would cause the transitions gets removed or some additional image data is included in order to avoid these transitions. Due to the nature of a “shared wall” architecture, we know that half of the firing nozzles will eject during the first portion of the print cycle and the other half will eject during the second portion of the print cycle. This means that, if so required by input pixel data 610, the odd numbered nozzles can be set to eject during the first portion of the print cycle and the even numbered nozzles can be set to eject during the second portion of the print cycle. This is another example of a rule applied to the data to provide a provisional assignment, which may need further correction. So, if we want to fire, a rule can be formulated as follows: 1. there is a transition from a ‘white’ pixel to a ‘black’ pixel within input pixel data 610, AND 2. the actuator ejection data information associated with the previous nozzle is ‘black’. Note that in the following examples, use is made of the fact that in a shared wall architecture, a fluid chamber interspersed between two other fluid chambers which eject using a drive waveform will have actuator wall motions on each of its actuator walls because of the neighbouring fluid chambers. As an example, imagine a situation in which a given fluid chamber is supplied a null waveform, while its neighbours receive an A or B drive waveform, e.g., ...ANA... or ...BNB.... In the ANA case, the A fluid chambers will cause their actuator walls to first move inwards, then outwards, thereby ejecting a droplet. These actuator wall motions necessarily cause the N fluid chamber actuator walls to first move outwards (since the actuator walls are shared) and then inwards, in other words the motions replicate what would happen if a B drive waveform had been applied to that fluid chamber. This can be thought of as inducing a B drive waveform in the central fluid chamber of this example, albeit no B drive waveform is actually supplied. Likewise, in the BNB example, the equivalent of an A drive waveform is induced in the central fluid chamber by the same effect. This means that the provisional assignment stage need not supply two different time varying drive waveforms. It can be sufficient that contiguous bands of firing fluid chambers can be actuated by using only a single time-varying drive waveform (e.g., A or B as in Figure 9), and the other drive waveform is a null waveform. The provisional assignment then applies a consistent rule to try to replicate the intended deposition pattern. Where conflicts occur in doing so, one or more corrections of the type discussed herein is / are applied to avoid errors in droplet ejection. These may involve shifting pixels in the x or y (spatial or temporal) direction. In addition, in some cases, a third drive waveform may be supplied as part of the correction. That is, if the provisional assignment uses A and null waveforms, the third drive waveform may be a B waveform, or if the provisional assignment uses B and null waveforms, the third drive waveform may be the A drive waveform. Figure 14 shows first line 616(1) of an image (input pixel data 610) that one may wish to reproduce on a deposition medium. For simplicity’s sake, this figure also shows a nozzle map to illustrate which nozzles would correspond to the ‘white’ and ‘black’ pixels that compose the first line 616(1). Note that having a nozzle mapping tool / template on the head controller circuitry or anywhere else in the system is by no means necessary. The rule that is used in this example is: to eject a droplet, the odd nozzle has to have T actuator ejection data and the even nozzle preceding and following the black nozzle has to have a ‘0’ actuator ejection data. So, the encoding circuitry starts the encoding process from moving along the array in the ±x-direction (e.g. left to right) of the first line 616(1) of the input pixel data 610 into the first actuator ejection data 136(1) in the following manner: Nozzle 18(x) Signal information Actuator ejection data 136(1) information (encoded pixel) Result in the deposition medium 1 Default setting, ‘white’ for non-ejection, as ‘0’ No ejection. there is no pixel information that predates the pixel information corresponding to nozzle 18(1) and the first pixel of the first line 616(1) is ‘white’. White pixel in pixel line 2 The second pixel of first line 616(1) is a ‘white’ pixel, therefore the signal applied to the electrode of the fluid chamber corresponding 18(2) must be equal to the signal applied to the electrode of the fluid chamber corresponding to the nozzle 18(1). ‘0’ No ejection. White pixel in pixel line 3 The third pixel of first line 616(1) is a ‘black’ pixel, therefore the signal applied to the electrode of the fluid chamber corresponding to nozzle 18(3) must be different from the signal applied to the electrode of the fluid chamber corresponding to the nozzle 18(2). ‘1’ Ejection. Black pixel in pixel line 4 The fourth pixel of first line 616(1) is a ‘black’ pixel, therefore the signal applied to the electrode of the fluid chamber corresponding to nozzle 18(4) must be different from the signal applied to the electrode of the fluid chamber corresponding to the nozzle 18(3). ‘0’ Ejection. Black pixel in pixel line 5 The fifth pixel of first line 616(1) is a ‘white’ pixel, but because the single cycle printing mode requires two actuator walls to eject a droplet, then for the nozzle 18(4) to eject, the signal applied to the electrode of the fluid chamber corresponding to 18(4) must be different from the signal applied to the electrode of the fluid chamber corresponding to the nozzle 18(5). ‘1’ No ejection. White pixel in pixel line 6 The sixth pixel of first line 616(1) is a ‘white’ pixel, therefore the signal applied to the electrode of the fluid chamber corresponding 18(6) must be equal to the signal applied to the electrode of the fluid chamber T No ejection. White pixel in pixel line corresponding to the nozzle 18(5). 7 The seventh pixel of first line 616(1) is a ‘black’ pixel so per the rule above, it has to have a signal applied to the electrode of the fluid chamber corresponding nozzle 18(7) ‘T. However, because the signal applied to the electrode of the fluid chamber corresponding the nozzle 18(6) is T, the nozzle 18(7) will not eject T No ejection. White pixel in pixel line (Pixel lost) 8 The eighth pixel of first line 616(1) is a ‘black’ pixel, therefore the signal applied to the electrode of the fluid chamber corresponding to nozzle 18(8) must be different from the signal applied to the electrode of the fluid chamber corresponding to the nozzle 18(7). ‘0’ Ejection. Black pixel in pixel line 9 The ninth pixel of first line 616(1) is a ‘black’ pixel, therefore, the signal applied to the electrode of the fluid chamber corresponding to 18(9) must be different from the signal applied to the electrode of the fluid chamber corresponding to the nozzle 18(8). ‘1’ Ejection. Black pixel in pixel line 10 The tenth pixel of first line 616(1) is a ‘black’ pixel, therefore the signal applied to the electrode of the fluid chamber corresponding to nozzle 18(10) must be different from the signal applied to the electrode of the fluid chamber corresponding to the nozzle 18(9). ‘0’ Ejection. Black pixel in pixel line 11 The eleventh pixel of first line 616(1) is a ‘white’ pixel, but because the single cycle printing mode requires two walls to eject a droplet, then for the nozzle 18(10) to eject, the signal applied to the electrode of the fluid chamber corresponding to 18(10) must be different from the signal applied to the electrode of the fluid chamber corresponding to the nozzle 18(11). T No ejection. White pixel in pixel line 12 The twelfth pixel of first line 616(1) is a ‘white’ pixel, therefore the signal applied to the electrode of the fluid chamber corresponding to the nozzle 18(12) must be equal to the signal applied to the electrode of the fluid chamber corresponding to the nozzle 18(11). ‘1’ No ejection. White pixel in pixel line 13 The thirteenth pixel of first line 616(1) is a ‘white’ pixel, therefore the signal applied to the electrode of the fluid chamber corresponding to the nozzle 18(13) must be equal to the signal applied to the electrode of the fluid chamber corresponding to the nozzle 18(12). ‘1’ No ejection. White pixel in pixel line 14 The fourteenth pixel of first line 616(1) is a ‘white’ pixel, therefore the signal applied to the electrode of the fluid chamber corresponding to the nozzle 18(14) must be equal to the signal applied to the electrode of the fluid chamber corresponding to the nozzle 18(13). ‘1’ No ejection. White pixel in pixel line 15 The fifteenth pixel of first line 616(1) is a ‘white’ pixel, therefore the signal applied to the electrode of the fluid chamber corresponding to the nozzle 18(15) must be equal to the signal applied to the electrode of the fluid chamber corresponding to the nozzle 18(14). ‘1’ No ejection. White pixel in pixel line In summary: • the pixels corresponding to nozzles 18(1) and 18(2), 18(5) and 18(6), and 18(11) to 18(15) would not cause AB to BA transition because there is no transition from ‘black’ 5 to ‘white’ as they all are ‘white’ pixels. • the pixels corresponding to nozzles 18(3) and 18(4), would not cause a problem because the actuator ejection data information associated with the previous nozzle 18(2) is ‘0’. • the pixels corresponding to nozzles 18(8) to 18(10) would not cause AB to BA 10 transition because no transition from ‘black’ to ‘white’ as they all are ‘black’ pixels. • the only image information that can cause the AB to BA transitions is in pixel corresponding to nozzle 18(7) because there is a ‘white’ to ‘black’ pixel transition AND the actuator ejection data information associated with the previous nozzle 18(6) is T. • Because the rule requires that an ejecting odd nozzle to be T and the actuator ejection data information associated with the previous nozzle 18(6) is also ‘T, the nozzle 18(7) will not eject and a pixel is lost during the encoding process as the pixel line in the deposition medium will have three black pixels instead of four represented in the first line 616(1) of the input pixel data 610. Figure 15, shows a printed image using the traditional encoding 710 and using the encoding of Figure 14. It is clear that the loss of a pixel is unexpectedly beneficial. Even with the loss of data from the original image (i.e., the removal of problematic pixels), the image produced with the new encoding has higher quality (greater and more consistent coverage) than the image produced using traditional encoding methods. Another way of performing encoding would be to add a pixel instead of losing a pixel. Figure 16 shows a second and third lines 616(2)-616(3) of input pixel data 610 and two possible encoding methods by adding pixels to the original line of input set of ejection data. Please note that, Figure 16 shows two lines 616(2)-616(3) of input pixel data 610 to illustrate possible drawbacks from adding pixels. It is by no meaning necessary for the encoding circuitry to have the knowledge of multiple pixel lines to apply this type of correction. The rule to eject is the same as the example of Figure 15. As with the previous example, the encoding circuitry continues the encoding process from left to right of the second line 616(2) of the input pixel data 610 into the second actuator ejection data 136(2) in the following manner: Nozzle 18(x) Signal information Actuator ejection data 136(2) information (encoded pixel) Result in the deposition medium 1 Default setting, ‘white’ for non-ejection, as there is no pixel information that predates the pixel information corresponding to nozzle 18(1) and the first pixel of the second line 616(2) is ‘white’. ‘0’ No ejection White pixel in pixel line 2 The second pixel of second line 616(2) is a ‘white’ pixel, therefore the signal applied to the ‘0’ No ejection White pixel in electrode of the fluid chamber corresponding 18(2) must be equal to the signal applied to the electrode of the fluid chamber corresponding to the nozzle 18(1). pixel line 3 The third pixel of second line 616(2) is a ‘black’ pixel, therefore the signal applied to the electrode of the fluid chamber corresponding to nozzle 18(3) must be different from the signal applied to the electrode of the fluid chamber corresponding to the nozzle 18(2). T Ejection Black pixel in pixel line 4 The fourth pixel of second line 616(2) is a ‘black’ pixel, therefore the signal applied to the electrode of the fluid chamber corresponding to nozzle 18(4) must be different from the signal applied to the electrode of the fluid chamber corresponding to the nozzle 18(3). ‘0’ Ejection Black pixel in pixel line 5 The fifth pixel of second line 616(2) is a ‘white’ pixel, but because the single cycle printing mode requires the movement of two actuator walls to eject a droplet, then for the nozzle 18(4) to eject, the signal applied to the electrode of the fluid chamber corresponding to 18(4) must be different from the signal applied to the electrode of the fluid chamber corresponding to the nozzle 18(5). T Ejection Black pixel in pixel line (additional pixel)* 6 The sixth pixel of second line 616(2) is a ‘white’, so to avoid ejection the encoded pixel should be ‘T but this would cause no ejection of the nozzle 18(7) as the rule to eject requires that signal applied to the electrode of the fluid chamber corresponding nozzle 18(7) to be T. Therefore, the signal applied to the electrode of the fluid chamber corresponding to nozzle 18(6) must be different from the signal applied to the electrode of the fluid chamber corresponding to the nozzle 18(5), causing ‘0’ Ejection Black pixel in pixel line (additional pixel) ejection through the nozzle 18(6). *Note that since the signal applied to the electrode of the fluid chamber corresponding to nozzle 18(4) is different from the signal applied to the electrode of the fluid chamber corresponding to nozzle 18(5), which in turn, is different from the signal applied to the electrode of the fluid chamber corresponding to nozzle 18(6), the nozzle 18(5) will eject a second additional droplet. 7 The seventh pixel of second line 616(2) is a ‘black’ pixel, therefore the signal applied to the electrode of the fluid chamber corresponding to nozzle 18(7) must be different from the signal applied to the electrode of the fluid chamber corresponding to the nozzle 18(6) T Ejection Black pixel in pixel line 8 The eighth pixel of second line 616(2) is a ‘black’ pixel, therefore the signal applied to the electrode of the fluid chamber corresponding to nozzle 18(8) must be different from the signal applied to the electrode of the fluid chamber corresponding to the nozzle 18(7). ‘0’ Ejection Black pixel in pixel line 9 The ninth pixel of second line 616(2) is a ‘black’ pixel, therefore, the signal applied to the electrode of the fluid chamber corresponding to 18(9) must be different from the signal applied to the electrode of the fluid chamber corresponding to the nozzle 18(8). T Ejection Black pixel in pixel line 10 The tenth pixel of second line 616(2) is a ‘black’ pixel, therefore the signal applied to the electrode of the fluid chamber corresponding to nozzle 18(10) must be different from the signal applied to the electrode of the fluid chamber corresponding to the nozzle 18(9). ‘0’ Ejection Black pixel in pixel line 11 The eleventh pixel of second line 616(2) is a T No ejection ‘white’ pixel, but because EWA requires two actuator walls to eject a droplet, then for the nozzle 18(10) to eject, the signal applied to the electrode of the fluid chamber corresponding to 18(10) must be different from the signal applied to the electrode of the fluid chamber corresponding to the nozzle 18(11). White pixel in pixel line 12 The twelfth pixel of second line 616(2) is a ‘white’ pixel, therefore the signal applied to the electrode of the fluid chamber corresponding to the nozzle 18(12) must be equal to the signal applied to the electrode of the fluid chamber corresponding to the nozzle 18(11). T No ejection White pixel in pixel line 13 The thirteenth pixel of second line 616(2) is a ‘white’ pixel, therefore the signal applied to the electrode of the fluid chamber corresponding to the nozzle 18(13) must be equal to the signal applied to the electrode of the fluid chamber corresponding to the nozzle 18(12). T No ejection White pixel in pixel line 14 The fourteenth pixel of second line 616(2) is a ‘white’ pixel, therefore the signal applied to the electrode of the fluid chamber corresponding to the nozzle 18(14) must be equal to the signal applied to the electrode of the fluid chamber corresponding to the nozzle 18(13). ‘T No ejection White pixel in pixel line 15 The fifteenth pixel of second line 616(2) is a ‘white’ pixel, therefore the signal applied to the electrode of the fluid chamber corresponding to the nozzle 18(15) must be equal to the signal applied to the electrode of the fluid chamber corresponding to the nozzle 18(14). T No ejection White pixel in pixel line In summary: • the pixels corresponding to nozzles 18(1) and 18(2), and 18(11) to 18(15) would not cause AB to BA transition because there is no transition from ‘black’ to ‘white’ as they all are ‘white’ pixels. • the pixels corresponding to nozzles 18(3) and 18(4), would not cause a problem because the actuator ejection data information associated with the previous nozzle 18(2) is ‘O’. • the pixels corresponding to nozzles 18(8) to 18(10) would not cause AB to BA transition because no transition from ‘black’ to ‘white’ as they all are ‘black’ pixels. • the only image information that can cause the AB to BA transitions is in pixel corresponding to nozzle 18(7) because there is a ‘white’ to ‘black’ pixel transition AND the actuator ejection data information associated with the previous nozzle 18(6) would be T to avoid ejection of nozzle 18(6). • since the rule requires that an ejecting odd nozzle to be ‘1’ and the actuator ejection data information associated with the previous nozzle 18(6) would also be ‘1’, the nozzle 18(7) will not eject and a pixel would be lost during the encoding process. Since we wish to add a pixel instead of losing a pixel, the actuator ejection data information associated with the nozzle 18(6) should be ‘0’ to cause the ejection of both nozzles 18(6) and 18(7). • changing actuator ejection data information associated with the nozzle 18(6) to be ‘0’ not only causes ejection of nozzle 18(6) but also the ejection of nozzle 18(5). It may be sensible to apply a different rule. For example, if the preceding number of white pixels is two, the rule may be to: o remove the problematic encoded pixel, instead of adding two new pixels, while the remainder of the data would continue to be encoded with the addition of pixels; or o shift the encoded pixels as described in Figure 17, discussed below. The encoding circuitry then encodes from left to right of the third line 616(3) of the input pixel data 610 into the third actuator ejection data 136(3) in the following manner: Nozzle 18(x) Signal information Actuator ejection data 136(3) information (encoded pixel) Result in the deposition medium 1 Default setting, ‘white’ for non-ejection, as ‘0’ No ejection there is no pixel information that predates the pixel information corresponding to nozzle 18(1) and the first pixel of the third line 616(3) is ‘white’. White pixel in pixel line 2 The second pixel of third line 616(3) is a ‘white’ pixel, therefore the signal applied to the electrode of the fluid chamber corresponding 18(2) must be equal to the signal applied to the electrode of the fluid chamber corresponding to the nozzle 18(1). ‘0’ No ejection White pixel in pixel line 3 The third pixel of third line 616(3) is a ‘black’ pixel, therefore the signal applied to the electrode of the fluid chamber corresponding to nozzle 18(3) must be different from the signal applied to the electrode of the fluid chamber corresponding to the nozzle 18(2). ‘1’ Ejection Black pixel in pixel line 4 The fourth pixel of third line 616(3) is a ‘black’ pixel, therefore the signal applied to the electrode of the fluid chamber corresponding to nozzle 18(4) must be different from the signal applied to the electrode of the fluid chamber corresponding to the nozzle 18(3). ‘0’ Ejection Black pixel in pixel line 5 The fifth pixel of third line 616(3) is a ‘white’ pixel, but because single cycle printing mode requires two actuator walls to eject a droplet, then for the nozzle 18(4) to eject, the signal applied to the electrode of the fluid chamber corresponding to 18(4) must be different from the signal applied to the electrode of the fluid chamber corresponding to the nozzle 18(5). ‘1’ No ejection White pixel in pixel line 6 The sixth pixel of second line 616(3) is a ‘white’ pixel, therefore the signal applied to the electrode of the fluid chamber corresponding to the nozzle 18(6) must be equal to the signal applied to the electrode of the fluid chamber corresponding to the nozzle 18(5). ‘1’ No ejection White pixel in pixel line 7 The seventh pixel of third line 616(3) is a ‘white’ pixel, therefore the signal applied to the electrode of the fluid chamber corresponding to nozzle 18(7) must be equal to the signal applied to the electrode of the fluid chamber corresponding to the nozzle 18(6). ‘1’ No ejection White pixel in pixel line 8 The eighth pixel of third line 616(3) is a ‘white’ pixel, so to avoid ejection the encoded pixel should be ‘1’ but this would cause no ejection of the nozzle 18(9) as the rule to eject requires that signal applied to the electrode of the fluid chamber corresponding nozzle 18(9) to be T. Therefore, the signal applied to the electrode of the fluid chamber corresponding to nozzle 18(8) must be different from the signal applied to the electrode of the fluid chamber corresponding to the nozzle 18(7), causing ejection through the nozzle 18(8). ‘0’ Ejection Black pixel in pixel line (additional pixel) 9 The ninth pixel of third line 616(3) is a ‘black’ pixel, therefore, the signal applied to the electrode of the fluid chamber corresponding to 18(9) must be different from the signal applied to the electrode of the fluid chamber corresponding to the nozzle 18(8). ‘1’ Ejection Black pixel in pixel line 10 The tenth pixel of third line 616(3) is a ‘black’ pixel, therefore the signal applied to the electrode of the fluid chamber corresponding to nozzle 18(10) must be different from the signal applied to the electrode of the fluid chamber corresponding to the nozzle 18(9). ‘0’ Ejection Black pixel in pixel line 11 The eleventh pixel of third line 616(3) is a ‘black’ pixel, therefore the signal applied to the electrode of the fluid chamber corresponding to nozzle 18(11) must be different from the signal applied to the electrode of the fluid chamber corresponding to the nozzle 18(10). ‘1’ Ejection Black pixel in pixel line 12 The twelfth pixel of third line 616(3) is a ‘back’ pixel, therefore the signal applied to the electrode of the fluid chamber corresponding the nozzle 18(12) must be different from the signal applied to the electrode of the fluid chamber corresponding to the nozzle 18(11). ‘0’ Ejection Black pixel in pixel line 13 The thirteenth pixel of third line 616(3) is a ‘white’ pixel, but because the single cycle printing mode requires two actuator walls to eject a droplet, then for the nozzle 18(12) to eject, the signal applied to the electrode of the fluid chamber corresponding to nozzle 18(13) must be different from the signal applied to the electrode of the fluid chamber corresponding to the nozzle 18(12). T No ejection White pixel in pixel line 14 The fourteenth pixel of third line 616(3) is a ‘white’ pixel, therefore the signal applied to the electrode of the fluid chamber corresponding 18(14) must be equal to the signal applied to the electrode of the fluid chamber corresponding to the nozzle 18(13). T No ejection White pixel in pixel line 15 The fifteenth pixel of third line 616(3) is a ‘white’ pixel, therefore the signal applied to the electrode of the fluid chamber corresponding 18(15) must be equal to the signal applied to the electrode of the fluid chamber corresponding to the nozzle 18(14). ‘1’ No ejection White pixel in pixel line In summary: • the pixels corresponding to nozzles 18(1) and 18(2), and 18(13) to 18(15) would not cause AB to BA transition because there is no transition from ‘black’ to ‘white’ as they 5 all are ‘white’ pixels. • the pixels corresponding to nozzles 18(3) and 18(4), would not cause a problem because the actuator ejection data information associated with the previous nozzle 18(2) is ‘O’. • the pixels corresponding to nozzles 18(10) to 18(12) would not cause AB to BA transition because there is no transition from ‘black’ to ‘white’ as they all are ‘black’ pixels. • the only image information that can cause the AB to BA transitions is in pixel corresponding to nozzle 18(9) because there is a ‘white’ to ‘black’ pixel transition AND the actuator ejection data information associated with the previous nozzle 18(8) would be ‘1’ to avoid ejection of nozzle 18(8). • since the rule requires that an ejecting odd nozzle to be ‘T and the actuator ejection data information associated with the previous nozzle 18(8) would also be ‘1’, so the nozzle 18(9) wouldn’t eject and a pixel would be lost during the encoding process. Since we wish to add a pixel instead of losing a pixel, the actuator ejection data information associated with the nozzle 18(8) should be ‘0’ to cause the ejection of both nozzles 18(8) and 18(9). Another encoding method would be to shift the problematic pixel and any subsequent ‘black’ pixels in the direction of encoding as illustrated in Figure 17. Figure 17 shows a fourth line 616(4) of input pixel data 610 and third possible encoding method by shifting the problematic pixel and any subsequent ‘black’ pixels in the direction of encoding, which in this case continues to be from left to right. As with the previous examples: • the pixels corresponding to nozzles 18(1) and 18(2), 18(5) and 18(6), and 18(12) to 18(15) would not cause AB to BA transition because there is no transition from ‘black’ to ‘white’ as they all are ‘white’ pixels. • the pixels corresponding to nozzles 18(3) and 18(4), would not cause a problem because the actuator ejection data information associated with the previous nozzle 18(2) is ‘O’. • the pixels corresponding to nozzles 18(8) to 18(11) would not cause AB to BA transition because there is no transition from ‘black’ to ‘white’ as they all are ‘black’ pixels. • the only image information that can cause the AB to BA transitions is in pixel corresponding to nozzle 18(7) because there is a ‘white’ to ‘black’ pixel transition AND the actuator ejection data information associated with the previous nozzle 18(8) would be ‘1’ to avoid ejection of nozzle 18(6). • because the rule requires that an ejecting odd nozzle be ‘1’ and the actuator ejection data information associated with the previous nozzle 18(6) is also ‘1’, the nozzle 18(7) will not eject. • the pixel information corresponding to nozzle 18(7) transitions to the information corresponding to nozzle 18(8), the information corresponding to nozzle 18(8) transitions to the information corresponding to nozzle 18(9) and so on and so forth until the next pixel in the of the fourth line 616(4) of the input pixel data 610 is white, in this case, pixel corresponding to nozzle 18(12). • therefore, the pixel corresponding to nozzle 18(12) will be encoded as a ‘black’ pixel. If there are some non-firing fluid chambers preceding the firing fluid chambers, another encoding method can be applied: the actuator ejection can be encoded such that a single actuator wall of a non-firing fluid chamber moves. As explained above, the single cycle printing mode requires two actuator walls moving in opposite senses to eject, so moving one actuator wall will not cause ejection but may change the actuator ejection data information associated with the previous nozzle to be ‘ 1 ’ causing ejection on the even number nozzle. This is shown in Figure 18, which shows a fifth line 616(5) of input pixel data 610 and fourth possible encoding method actuating an actuator wall 22 of a non-firing fluid chamber preceding the firing fluid chambers in the direction of encoding, which in this case continues to be from left to right. As with the previous examples: • the pixels corresponding to nozzles 18(1) to 18(11), and 18(14) to 18(15), would not cause AB to BA transition because there is no transition from ‘black’ to ‘white’ as they all are ‘white’ pixels. • the pixel corresponding to nozzle 18(13) would not cause AB to BA transition because there is no transition from ‘black’ to ‘white’ as 18(12) and 18(13) are ‘black’ pixels. • As the rule to eject requires the odd nozzle to be ‘1’ and the even nozzle to be ‘O’, the pixel corresponding to the nozzle 18(12) would not jet as to the nozzle 18(11) would have the value ‘O’. • In this alternative encoding method, one of the actuator walls of the fluid chamber corresponding to the nozzle 18(7) moves while the other remains stationary. This means that the pixel corresponding to nozzle 18(7) would be encoded as T even though no ejection occurs. • This means that to preserve no ejection and to have the nozzle 18(11) encoded as T all the nozzles 18(7) to 18(11) would be encoded with the value ofT, allowing the nozzle 18(12) to be encoded with the value ‘0’ and eject. • This way, AB-BA transitions can be avoided. (Note that if the movement of the actuator wall of the non-firing fluid chamber occurs too close to the firing fluid chambers it may cause errors in printing due to the input of extra energy in the system. In some examples, it may be necessary, for e.g., at least four non-firing fluid chambers to exist between the non-firing fluid chamber which has one of the actuator walls moved and the firing fluid chambers.) In examples of Figures 14 to Figure 17 the encoding is performed from left to right of the input pixel data 610, but it could also be performed from right to left with similar results. These examples also show that the encoding of each individual line of the input pixel data 610 is independent, i.e., one encoding method can be applied to one part of the image and a different encoding method can be applied to another part of the image, allowing for minimal corruption of the original information in the input pixel data 610. In some cases, the encoding may be performed in the head control circuitry 200 as shown in Figure 11. Another possibility would be to choose the least disruptive encoding method for the overall input pixel data 610, allowing a single encoding method to be applied. The encoding methods could also be prioritised, for example, use pixel shifting, if not possible use pixel addition, and as the last resort use pixel removal (as that’s always possible), this procedure is illustrated in Figure 19. In Figure 19, a flow chart is shown which a procedure for trying and implementing correction methods such as those set out above may be enacted, to find an appropriate correction for the given circumstances. This flow chart begins with step S1, in which input pixel data is received, the input pixel data indicating the intended droplet deposition pattern. At step S2, a provisional assignment of fluid chambers is made, for example, according to a consistent rule for setting out which fluid chambers should have actuator wall motions and in which manner. In step S3 the provisional assignment moves on to specifically assign firing / non-firing status to fluid chambers and / or assigning a specific drive waveform to electrodes, in order to cause actuator wall motions to eject or not eject droplets in each print cycle, as needed to replicate the pattern in the input pixel data. Step S4 is performed, either in parallel or in series, and includes allocating a drive waveform to each electrode following a rule which prohibits A to B or B to A transitions. In step S5 the allocated and provisionally assigned drive waveforms for each electrode are compared. This comparison includes asking whether the provisional drive waveform is equal to the allocated drive waveform in all cases. In case the drive waveforms match, the process proceeds to step S6, in which the provisional assignment is kept and the process proceeds to print in that print cycle, using the provisionally assigned drive waveforms. However, if there is a conflict, i.e. at least one drive waveform in the provisional assignment does not match the allocated drive waveform, the method proceeds to step S7, in which a correction is applied to the provisional assignment of the electrode of a fluid chamber. The corrections which may be applied at this stage are discussed elsewhere in the document, but it suffices for this discussion to understand that several options are possible including displacing one or more pixels in the x (spatial) or y (temporal) direction, adding pixels, deleting pixels, changing assigned drive waveforms, etc., each of which may have advantages or disadvantages in the specific print circumstances (nearby ejections, recent or upcoming ejections, etc.) At step S8 a further question is asked, specifically whether the corrected assignment results in a (further) conflict. This may include situations in which the result of the correction is no better than the provisional assignment in terms of the number or severity of the conflict. If such a (further) conflict is identified, the process can include looping back to step S7 to apply a different correction. Once a different correction has been applied the method proceeds once more to step S8 and the effect of the correction is again assessed. Note that this process can continue as often as needed until either a perfect correction (which resolves the error without introducing new errors in the given print circumstances) is found, or until all possible corrections have been tried. In cases where no perfect correction is possible in the specific print circumstances, one of the already tried corrections may be used instead, or the problematic pixel is simply deleted (i.e., lost). For example, the first correction, the most recent correction, or the correction which results in the least severe error may be used. This latter example may include scoring the result of each correction in terms of how many unwanted additional pixels are printed, how many intended pixel prints are missed, how far away from the intended location a misplaced pixel is deposited, and so forth, and using the correction which achieves the best (least disrupted) outcome. Once a correction has been selected, the process proceeds to step S9 in which the assignment of firing / non-firing fluid chambers or drive waveform of a fluid chamber (e.g., a non-firing fluid chamber) may be corrected in line with the chosen correction. As with step S6, this then allows that print cycle to be enacted and droplets ejected. It will be apparent that the various corrections set out herein may be combined with one another in a given print cycle. For example, where there are multiple contiguous bands of firing fluid chambers, the procedures set out herein may be applied to each band to correct any errors introduced in the provisional assignment (i.e., conflicts between the provisional assignment and the allocation). Equally, a given conflict may best be addressed by applying more than one correction to that contiguous band of firing fluid chambers to avoid print errors. The corrections applied in step S7 above, may include such composite corrections in an attempt to find the best (or even a perfect) correction, for example.

Claims

1. A controller for controlling the droplet ejection of droplets from a droplet ejectionhead, the droplet ejection head comprising:an array of fluid chambers separated by interspersed actuator walls extending along at least a first direction, each fluid chamber communicating with an nozzle for the release of droplets of fluid and each of said actuator walls separating two neighbouring fluid chambers; wherein each of said actuator walls has at least one electrode disposed thereon for controlling the motion of said actuator wall;the controller being configured to:receive input pixel data; whereinthe input pixel data forms at least a one-dimensional array having a first extent corresponding to droplet ejections along the first direction in a single print cycle; andthe input pixel data indicates intended deposition locations on a deposition medium; andprovisionally assign, based on said input pixel data, each fluid chamber within said array as either a firing fluid chamber or a non-firing fluid chamber so as to produce bands of one or more contiguous firing fluid chambers separated by bands of one or more contiguous non-firing fluid chambers;provisionally assign, based on said input pixel data, each at least one electrode of each actuator wall of the firing fluid chambers at least one of a first drive waveform and a second drive waveform such that, during the single print cycle, the at least one of the first drive waveform and the second drive waveform controls the motion of the actuator walls of each firing fluid chamber; wherein the first drive waveform and second drive waveform are different drive waveforms;the controller is further configured to:allocate, to each at least one electrode of each actuator wall of the firing fluid chambers, one of the first drive waveform or the second drive waveform;identify at least one example where, for a first firing fluid chamber of the band of firing fluid chambers, the provisionally assigned drive waveform to each at least one electrode of each actuator wall of the first firing fluid chamber of the band of firing fluid chambers does not match the allocated drive waveform for said at least one electrode of said actuator wall of the first firing fluid chamber of the band of firing fluid chambers;correct the provisional assignment of the drive waveform assigned to at least one electrode of at least one fluid chamber such that the drive waveform assigned to each at least one electrode of each actuator wall of the first firing fluid chamber of the band of firing fluid chambers of the example corresponds to the allocated first drive waveform or second drive waveform for the first firing fluid chamber of the band.

2. The controller of claim 1, wherein the controller is further configured to correctthe provisional assignment such that the band of firing fluid chambers containing the first firing fluid chamber are deferred or brought forward in the first direction.

3. The controller of claim 1 or claim 2, wherein the controller is further configuredto correct the provisional assignment such that an additional firing fluid chamber is added to the band of firing fluid chambers.

4. The controller of claim 3, wherein the last non-firing fluid chamber adjacent tothe first firing fluid chamber of the band of firing fluid chambers is corrected to be a firing fluid chamber of the band of firing fluid chambers.

5. The controller of claim 1 or claim 2, wherein the controller is further configuredto correct the provisional assignment such that an additional non-firing fluid chamber is added to the band of non-firing fluid chambers adjacent to the band of firing fluid chambers.

6. The controller of claim 5, wherein the first firing fluid chamber of the band offiring fluid chambers is corrected to be the last non-firing fluid chamber of the band of non-firing fluid chambers adjacent to the band of firing fluid chambers.

7. The controller of any one of the preceding claims, wherein the controller isfurther configured to assign one of first drive waveform and second drive waveform to at least one electrode of a non-firing fluid chamber preceding the first firing fluid chamber of the band of firing fluid chambers.

8. The controller of claim 1, wherein:the input pixel data forms a two-dimensional array having a first extent corresponding to droplet ejections along the first direction in a first print cycle and asecond extent corresponding to a series of droplet ejections to be enacted in second print cycle; and whereinthe controller is further configured to:correct the provisional assignment of the first firing fluid chamber of the band of firing fluid chambers of the first print cycle to be the last non-firing fluid chamber of the band of non-firing fluid chambers adjacent to the band of firing fluid chambers.

9. The controller of claim 8, wherein the controller is further configured to correctthe provisional assignment of the drive waveform assigned to at least one electrode of at least one fluid chamber in the second print cycle.

10. The controller of claim 9, wherein the controller is further configured to correctthe provisional assignment of the drive waveform assigned to at least one electrode of at least one fluid chamber in the second print cycle such that the band of firing fluid chambers are deferred or brought forward in the first direction.

11. The controller of any of claims 9 or 10, wherein the controller is furtherconfigured to assign one of first drive waveform and second drive waveform to at least one electrode of a non-firing fluid chamber in the second print cycle.

12. The controller of any of claims 8 to 11, wherein the controller is furtherconfigured to correct the provisional assignment such that an additional firing fluid chamber is added to the band of firing fluid chambers in the second print cycle.

13. The controller of any one of the preceding claims, wherein correcting includesthe controller:applying a first correction to the provisional assignment;assessing the result of applying the first correction to identify whether the drive waveform assigned to each at least one electrode of each actuator wall of a first firing fluid chamber of the band of firing fluid chambers does not match the allocated drive waveform for said at least one electrode of said actuator wall of the first firing fluid chamber of the band of firing fluid chambers; andin the event that the drive waveform assigned to each at least one electrode of each actuator wall of a first firing fluid chamber of the band of firing fluid chambers results in a conflict, applying a second correction, different to the first correction, to the provisional assignment.

14. The controller of claim 13, wherein in the event that applying the secondcorrection leads to at least one instance in which the drive waveform assigned to each at least one electrode of each actuator wall of a first firing fluid chamber of the band of firing fluid chambers does not match the allocated drive waveform for said at least one electrode of said actuator wall of the first firing fluid chamber of the band of firing fluid chambers; the controller is further configured to:apply a further correction, different to each of the first and second corrections, to the provisional assignment.

15. An apparatus for improving the consistency of droplet ejection, comprising thedroplet ejection head and the controller according to any one of the preceding claims.

16. The apparatus of claim 15, wherein each fluid chamber has a single electrodedisposed on both actuator walls; and whereinfiring fluid chambers in each print cycle are assigned drive waveforms by: starting at a first fluid chamber, considering adjacent fluid chambers along the array in a first direction until a firing fluid chamber is identified and arranging to supply the first drive waveform thereto;where the neighbouring fluid chamber in the first direction is also a firing fluid chamber, arranging to apply the second drive waveform thereto; andcontinuing to alternate between first drive waveform and second drive waveform in neighbouring fluid chambers in the first direction along a line of consecutive firing fluid chambers until a non-firing fluid chamber is identified.

17. The apparatus of claim 16, wherein further firing fluid chambers in each printcycle are assigned drive waveforms by repeating the process of claim 13 until all firing fluid chambers have been assigned a drive waveform.

18. The apparatus of claim 17, wherein the controller is further configured toidentify at least one example in which the input pixel data indicates that a particular fluid chamber is scheduled to transition from ejecting using the first drive waveform in a first ejection cycle to ejecting using the second drive waveform in a second print cycle; andapply a correction by:(a) deferring or bringing forward the band of firing fluid chambers containing the first firing fluid chamber in the first direction;(b) adding an additional firing fluid chamber to the band of firing fluid chambers; optionally wherein the last non-firing fluid chamber adjacent to the first firing fluid chamber of the band of firing fluid chambers is corrected to be a firing fluid chamber of the band of firing fluid chambers;(c) adding an additional non-firing fluid chamber to the band of non-firing fluid chambers adjacent to the band of firing fluid chambers; optionally, wherein the first firing fluid chamber of the band of firing fluid chambers is corrected to be the last non-firing fluid chamber of the band of non-firing fluid chambers adjacent to the band of firing fluid chambers; and / or(d) assigning one of first drive waveform and second drive waveform to at least one electrode of a non-firing fluid chamber preceding the first firing fluid chamber of the band of firing fluid chambers.

19. A method for improving the consistency of droplet ejection in a plurality of fluidchambers, the plurality of fluid chambers comprising:an array of fluid chambers separated by interspersed actuator walls extending along at least a first direction, each fluid chamber communicating with an nozzle for the release of droplets of fluid and each of said actuator walls separating two neighbouring fluid chambers; wherein each of said actuator walls has at least one electrode disposed thereon for controlling the motion of said actuator wall;the method comprising:receiving input pixel data; whereinthe input pixel data forms at least a one-dimensional array having a first extent corresponding to droplet ejections along the first direction in a single print cycle; andthe input pixel data indicates intended deposition locations on a deposition medium; andprovisionally assigning, based on said input pixel data, each fluid chamber as either a firing fluid chamber or a non-firing fluid chamber so as to produce bands of one or more contiguous firing fluid chambers separated by bands of one or more contiguous non-firing fluid chambers;provisionally assigning, based on said input pixel data, each at least one electrode of each actuator wall of the firing fluid chambers at least one of a first drive waveform and a second drive waveform such that, during the single print cycle, the atleast one of the first and second drive waveforms controls the motion of the actuator walls of each firing fluid chamber; wherein the first drive waveform and second drive waveform are different drive waveforms;allocating, to each at least one electrode of each actuator wall of the firing fluid chambers, one of first drive waveform and second drive waveform;identifying at least one example where, for a first firing fluid chamber of the band of firing fluid chambers, the provisionally assigned drive waveform to each at least one electrode of each actuator wall of the first firing fluid chamber of the band of firing fluid chambers does not match the allocated drive waveform for said at least one electrode of said actuator wall of the first firing fluid chamber of the band of firing fluid chambers;correcting the provisional assignment of the drive waveform assigned to at least one electrode of at least one fluid chamber such that the drive waveform assigned to each at least one electrode of each actuator wall of the first firing fluid chamber of the band of firing fluid chambers of the example corresponds to the allocated first drive waveform or second drive waveform for the first firing fluid chamber of the band.

20. The method of claim 19, further comprising correcting the provisionalassignment such that the band of firing fluid chambers containing the first firing fluid chamber are deferred or brought forward in the first direction.

21. The method of claim 19 or claim 20, further comprising correcting theprovisional assignment such that an additional firing fluid chamber is added to the band of firing fluid chambers.

22. The method of claim 21, wherein the last non-firing fluid chamber adjacent tothe first firing fluid chamber of the band of firing fluid chambers is corrected to be a firing fluid chamber of the band of firing fluid chambers.

23. The method of claim 19 or claim 20, further comprising correcting theprovisional assignment such that an additional non-firing fluid chamber is added to the band of non-firing fluid chambers adjacent to the band of firing fluid chambers.

24. The method of claim 23, wherein the first firing fluid chamber of the band offiring fluid chambers is corrected to be the last non-firing fluid chamber of the band of non-firing fluid chambers adjacent to the band of firing fluid chambers.

25. The method of any one of claims 19 to 24, wherein one of first drive waveformand second drive waveform is assigned to at least one electrode of a non-firing fluid chamber preceding the first firing fluid chamber of the band of firing fluid chambers.

26. The method of claim 19, wherein:the input pixel data forms a two-dimensional array having a first extent corresponding to droplet ejections along the first direction in a first print cycle and a second extent corresponding to a series of droplet ejections to be enacted in a second print cycle; and the method further comprises:correcting the provisional assignment of the first firing fluid chamber of the band of firing fluid chambers of the first print cycle to be the last non-firing fluid chamber of the band of non-firing fluid chambers adjacent to the band of firing fluid chambers.

27. The method of claim 26, further comprising assigning one of first drivewaveform and second drive waveform to at least one electrode of at least one fluid chamber in the second print cycle.

28. The method of claim 27, further comprising correcting the provisionalassignment of the drive waveform assigned to at least one electrode of at least one fluid chamber in the second print cycle such that the band of firing fluid chambers are deferred or brought forward in the first direction.

29. The method of any of claims 27 or 28, further comprising assigning one of firstdrive waveform and second drive waveform to at least one electrode of a non-firing fluid chamber in the second print cycle.

30. The method of any of claims 26 to 29, further comprising correcting theprovisional assignment such that an additional firing fluid chamber is added to the band of firing fluid chambers in the second print cycle.

31. The method of any one of the preceding claims, wherein correcting includes:applying a first correction to the provisional assignment;assessing the result of applying the first correction to identify whether the drive waveform assigned to each at least one electrode of each actuator wall of a first firing fluid chamber of the band of firing fluid chambers does not match the allocated drive waveform for said at least one electrode of said actuator wall of the first firing fluid chamber of the band of firing fluid chambers; andin the event that the drive waveform assigned to each at least one electrode of each actuator wall of a first firing fluid chamber of the band of firing fluid chambers results in a conflict, applying a second correction, different to the first correction, to the provisional assignment.

32. The method of claim 31, wherein in the event that applying the secondcorrection leads to at least one instance in which the drive waveform assigned to each at least one electrode of each actuator wall of a first firing fluid chamber of the band of firing fluid chambers does not match the allocated drive waveform for said at least one electrode of said actuator wall of the first firing fluid chamber of the band of firing fluid chambers; the method further comprises:applying a further correction, different to each of the first and second corrections, to the provisional assignment.

33. The method of any of claims 19 to 32, wherein each fluid chamber has a singleelectrode disposed on both actuator walls; and whereinfiring fluid chambers in each print cycle are assigned drive waveforms by: starting at a first fluid chamber, considering neighbouring fluid chambers along the array in a first direction until a firing fluid chamber is identified and arranging to supply the first drive waveform thereto;where the neighbouring fluid chamber in the first direction is also a firing fluid chamber, arranging to apply a second drive waveform thereto; andcontinuing to alternate between first drive waveform and second drive waveform in neighbouring fluid chambers in the first direction along a line of consecutive firing fluid chambers until a non-firing fluid chamber is identified.

34. The method of claim 33, wherein further firing fluid chambers in each print cycleare assigned drive waveforms by repeating the process of claim 33 until all firing fluid chambers have been assigned a drive waveform.

35. The method of claim 34, further comprising identifying at least one example inwhich the input pixel data indicates that a particular fluid chamber is scheduled to transition from ejecting using a first one of the drive waveforms in a first print cycle to ejecting using a second drive waveform in a second print cycle; andapplying a correction by:(a) deferring or bringing forward the band of firing fluid chambers containing the first firing fluid chamber in the first direction;(b) adding an additional firing fluid chamber to the band of firing fluid chambers; optionally wherein the last non-firing fluid chamber adjacent to the first firing fluid chamber of the band of firing fluid chambers is corrected to be a firing fluid chamber of the band of firing fluid chambers;(c) adding an additional non-firing fluid chamber to the band of non-firing fluid chambers adjacent to the band of firing fluid chambers; optionally, wherein the first firing fluid chamber of the band of firing fluid is corrected to be the last nonfiring fluid chamber of the band of non-firing fluid chambers adjacent to the band of firing fluid chambers; and / or(d) assigning one of first drive waveform and second drive waveform to at least one electrode of a non-firing fluid chamber preceding the first firing fluid chamber of the band of firing fluid chambers.

36. Computer readable instructions which, when run by a controller, cause thecontroller to implement the method of any one of claims 19 to 35.

Images