JP2013067092A - Printer and printing method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To correct the ejection timing of ink droplets while appropriately maintaining ejection intervals of the ink droplets when printing an image by a line printer.SOLUTION: The printer includes: (A) a conveying unit to convey the medium in a conveying direction by rotating a conveying roller; (B) a head unit including a plurality of nozzles arranged in a crossing direction crossing the conveying direction, and to eject the ink from the nozzles to print the image on the medium; (C) a storage unit to store a plurality of correction values associated with each of circumferential positions of the conveying roller; (D) a control unit to control operations of the conveying unit and the head unit, and to eject the ink from the nozzles at timing corresponding to the correction value. The plurality of correction values include areas in each of which the variation of the correction value with reference to the circumferential position is kept fixed in the case of making the timing earlier than the latest timing, and in other cases, the correction value with reference to the circumferential position is represented by a periodic function.

Description

  The present invention relates to a printing apparatus and a printing method.

  2. Description of the Related Art An ink jet printer is known as a printing apparatus that prints on a medium by transporting a medium (for example, paper or cloth) in the transport direction and ejecting ink droplets from a head. In such a printing apparatus, if a conveyance error occurs when the medium is conveyed, an image or the like cannot be printed at a correct position on the medium. In particular, in an ink jet printer, if ink droplets do not land at the correct position on the medium, white stripes or black stripes may occur in the printed image, and the image quality may deteriorate.

  Therefore, a method for correcting the transport amount of the medium has been proposed. For example, Patent Document 1 proposes that a test pattern is printed, the test pattern is read, a correction value is calculated based on the read result, and the carry amount is corrected based on the correction value when an image is printed. Yes.

JP-A-5-96796

  However, in a so-called line printer type printing apparatus that discharges ink from a head fixed to a medium to be conveyed, an image is printed with a constant medium conveyance amount. difficult.

  Accordingly, a method for suppressing the deterioration of image quality by changing the ink ejection timing is conceivable. At this time, it is necessary to pay attention so that an appropriate ink droplet ejection interval cannot be maintained as a result of correcting the ejection timing. For example, if the next ejection timing comes before the operation of ejecting ink droplets from a certain nozzle is completed, an accurate ejection operation may not be performed. Therefore, it is necessary to correct the ejection timing while keeping the ink droplet ejection interval appropriate.

  An object of the present invention is to correct ejection timing while maintaining an appropriate ejection interval between ink droplets when printing an image with a line printer.

  A main invention for achieving the above object includes (A) a transport unit that transports a medium in a transport direction by rotating a transport roller, and (B) a plurality of nozzles arranged in an intersecting direction that intersects the transport direction. A head unit that prints an image on the medium by ejecting ink from the nozzle; and (C) a storage unit that stores a plurality of correction values respectively associated with the circumferential positions of the transport roller; (D) a control unit that controls operations of the transport unit and the head unit, and a control unit that discharges the ink from the nozzles at a timing according to the correction value, The plurality of correction values have a region where the amount of change in the correction value with respect to the position in the circumferential direction is constant when the timing is advanced compared to the previous timing. In this case, the correction value for the circumferential position is expressed by a periodic function, it is a printing apparatus according to claim.

  Other features of the present invention will become apparent from the description of the present specification and the accompanying drawings.

1 is a block diagram illustrating an overall configuration of a printer. FIG. 2 is a diagram illustrating an outline around a printing area of the printer. FIG. 4 is a diagram illustrating an arrangement of a plurality of short heads on the lower surface of the head 41. FIG. 4A and FIG. 4B are diagrams for explaining how ink droplets land on a medium conveyed by a conveyance roller in which eccentricity occurs. It is a figure explaining the correction value set for every position of the peripheral direction of a conveyance roller. It is a figure explaining the concept of PTS timing correction. 7A to 7C are diagrams for explaining the relationship between the magnitude of the correction value and the drive signal COM. It is a figure showing an example of the correction value used by embodiment. It is a figure showing the whole flow for calculating a correction value. It is a figure which shows an example of the test pattern printed by embodiment. It is a figure explaining the setting position to a scanner at the time of reading a reference pattern and a test pattern. It is a figure showing the flow for calculating the gravity center position of a ruled line. It is a figure explaining the method of recognizing a ruled line position about a standard pattern. It is a figure explaining the method of calculating the inclination amount of the ruled line of a reference pattern. It is a figure explaining the method of calculating the inclination amount about a test pattern. It is explanatory drawing about the margin amount X. FIG. It is a figure showing an example of the density distribution for every raster. It is a figure explaining the calculation method of a density gravity center position. It is a figure explaining the calculation method of a virtual reference pattern. It is a figure showing an example of the relationship between the density gravity center position of a virtual reference pattern, and the density gravity center position of a test pattern. It is a figure explaining the ratio of the density centroid position of a test pattern with respect to the density centroid position of a virtual reference pattern. It is a figure explaining the method of calculating the actual value of the density | concentration gravity center position of a test pattern. It is a figure explaining the outline | summary of correction value calculation. It is a figure showing the flow for calculating a correction value. It is a figure showing the relationship between the shift | offset | difference of the landing position of an ink droplet, and the time which should be corrected. It is a figure which shows the example of the ruled line used in order to calculate the corrected amount. It is a figure which shows the example of the level | step difference of a correction value. It is a figure explaining according to the case for every variation | change_quantity of a correction value about the method of level | step difference correction. It is a figure showing the mode of a change of the waveform collision margin at the time of changing a conveyance speed.

At least the following matters will become clear from the description of the present specification and the accompanying drawings.
(A) a transport unit that transports a medium in the transport direction by rotating a transport roller; and (B) a plurality of nozzles arranged in an intersecting direction that intersects the transport direction, and ejecting ink from the nozzles A head unit that prints an image on the medium; (C) a storage unit that stores a plurality of correction values associated with each circumferential position of the transport roller; and (D) the transport unit and the head unit. And a control unit that discharges the ink from the nozzles at a timing corresponding to the correction value, wherein the plurality of correction values are the immediately preceding timings. When the timing is advanced compared to the above, there is a region in which the amount of change in the correction value with respect to the circumferential position is constant. Printing apparatus correction value is represented by a periodic function, and wherein the.
According to such a printing apparatus, it is possible to correct the ejection timing while keeping the ink droplet ejection interval properly, so that even when a conveyance error occurs due to the eccentricity of the conveyance roller, a high-quality image can be obtained. Can be printed.

In this printing apparatus, the correction value is formed such that a plurality of lines along the intersecting direction are arranged in the transport direction by ejecting the ink from the nozzles to the medium transported in the transport direction. The test pattern is calculated from the image data obtained by reading the test pattern, based on the measurement result of the interval in the transport direction between the plurality of lines, and ink is ejected from the nozzle at the timing. Thereafter, when the time interval until the next ink is ejected becomes shorter than a predetermined threshold value, the correction value is preferably corrected so that the time interval becomes equal to or greater than the threshold value.
According to such a printing apparatus, it is possible to appropriately adjust the ink discharge timing while avoiding the waveform collision.

In this printing apparatus, it is desirable that the threshold value increases as the speed of transporting the medium increases.
According to such a printing apparatus, an allowable range for avoiding waveform collision can be optimized according to the conveyance speed.

In this printing apparatus, when measuring the distance in the transport direction between the lines, the amount of shift due to a transport error that occurs when the transport roller makes one rotation is corrected, and the correction between the corrected lines is performed. It is desirable that the correction value is calculated based on the result of measuring the interval in the transport direction.
According to such a printing apparatus, by removing the DC component, it is easy to suppress an extreme difference in the magnitude of the correction value before and after the rotation reference position of the transport roller.

In such a printing apparatus, the correction at the position of the front end portion is reduced so that the difference between the correction value at the position of the front end portion in the circumferential direction of the transport roller and the correction value at the position of the end portion in the circumferential direction becomes small. It is desirable that the correction value is corrected using a ratio calculated according to the difference between the value and the correction value at the position of the end portion.
According to such a printing apparatus, it is possible to suppress the occurrence of a step in the correction value when shifting from the end of the current turn to the tip of the next turn.

In this printing apparatus, the correction value at the position of the front end portion according to the inclination of the correction value at the position of the front end portion in the circumferential direction and the inclination of the correction value at the position of the end portion in the circumferential direction It is desirable to calculate a difference in correction value at the position of the end portion.
According to such a printing apparatus, the level difference can be accurately corrected according to the inclination condition of the correction value.

In this printing apparatus, the correction value is calculated in association with each position in the circumferential direction for each period in which the conveyance roller makes one rotation, and among the correction values calculated for each period of the conveyance roller The timing for ejecting ink from the nozzles is preferably corrected using a correction value obtained by averaging correction values calculated for each period other than the first period of the transport roller.
According to such a printing apparatus, the correction value for one cycle of the transport roller can be accurately calculated.

  In addition, the medium is conveyed in the conveyance direction by rotating the conveyance roller, and intersects the conveyance direction at a timing corresponding to a plurality of correction values respectively associated with the circumferential positions of the conveyance roller. Discharging the ink from a plurality of nozzles arranged in an intersecting direction, wherein the correction value is set to the position in the circumferential direction when the timing is advanced compared to the immediately preceding timing. The printing method is characterized in that there is a region where the amount of change in the correction value is constant, and in other cases, the correction value for the position in the circumferential direction is represented by a periodic function.

=== Basic configuration ===
A basic configuration of a printing apparatus for printing an image will be described. The “printing apparatus” of the present embodiment is an ink jet printer that prints an image or the like by ejecting ink droplets (ink droplets) onto a medium.

<Configuration of Printer 1>
The printer 1 will be described as an example of an ink jet printer. FIG. 1 is a block diagram illustrating the overall configuration of the printer 1. FIG. 2 is a diagram showing an outline around the print area of the printer 1.
The printer 1 includes a transport unit 20, a head unit 40, a detector group 50, and a controller 60. The printer 1 is communicably connected to the computer 110. The printer 1 and the computer 110 can also be considered as a printing apparatus (printing system).

A printer driver is installed in the computer 110. The printer driver is a program for causing a display device to display a user interface and converting image data output from an application program into print data. This printer driver is printed on a print medium (a computer-readable print medium) such as a flexible disk FD or a CD-ROM. Also, the printer driver can be downloaded to the computer 110 via the Internet. In addition, this program is comprised from the code | cord | chord for implement | achieving various functions. The computer 110 outputs print data corresponding to the image to be printed to the printer 1 in order to cause the printer 1 to print an image.
The computer 110 is installed with a correction value calculation program for calculating a correction value, which will be described later.

The transport unit 20 is a transport unit for transporting a medium (such as paper S) in a predetermined direction (hereinafter referred to as a transport direction). This conveyance unit 20 has the conveyance roller 22, the discharge roller 23, and the belt 24 (FIG. 2). A transport motor (hereinafter also referred to as PF motor) is connected to the transport roller 23 via a gear train. When the PF motor rotates, the transport roller 22 rotates, and the belt 24 and the discharge roller 23 rotate. The rotation amount of the PF motor (conveying roller 22) is monitored by a rotary encoder 51 described later.
When printing an image, a medium fed into the printer 1 by a paper feed roller (not shown) is conveyed to a printable area (an area facing the head unit 40) by the belt 24. The paper S on which an image is printed after passing through the printable area is discharged to the outside by the belt 24. The paper S being conveyed is electrostatically attracted or vacuum attracted to the belt 24.

The head unit 40 is for ejecting ink to the medium. The head unit 40 forms dots by ejecting each color ink onto the medium being transported, and prints an image on the medium. The printer 1 of the present embodiment is a line printer, the head unit 40 is provided above the belt 24 of the transport unit 20, and each head can form dots for the medium width at a time.
The arrangement of the heads constituting the head unit 40 will be described later.

The detector group 50 includes a rotary encoder 51 and an origin sensor 52. The rotary encoder 51 is provided on the rotating shaft of the PF motor, and can detect the amount of rotation of the PF motor. Thereby, the rotation amount of the transport roller 22 during the medium transport operation is detected. The rotary encoder 51 includes a disk-shaped scale and a sensor unit (both not shown), and the disk-shaped scale is provided with slits at regular intervals along the circumferential direction. In addition, a mark that defines a reference position (home position) for detecting the rotation amount of the transport roller 22 is attached to a predetermined position in the circumferential direction. The rotary encoder 51 may be provided on the rotation shaft of the transport roller 22 and directly detect the rotation amount of the transport roller 22.

  The origin sensor 52 detects the position where the above-mentioned mark is given on the disk scale. The position of the mark detected by the origin sensor 52 becomes a reference position (home position, hereinafter also referred to as HP) when detecting the rotation amount of the transport roller 22.

  The controller 60 is a control unit (control unit) for controlling the printer, and controls each unit based on print data received from the computer 110 and various detection results detected by the detector group 50. Print an image on the medium. The controller 60 includes a CPU 61, a memory 62, and a COM generation unit 63.

  The CPU 61 is an arithmetic processing device for performing overall control of the printer 1. The memory 62 is a storage unit for securing an area for storing the program of the CPU 61, a work area, and the like, and storing correction values to be described later. The memory 62 is configured by a storage element such as a RAM or an EEPROM. The CPU 61 controls the operations of the transport unit 20 and the head unit 40 in accordance with a program stored in the memory 62.

  The COM generator 63 generates a signal COM for ejecting a predetermined amount of ink from the nozzles provided in the head unit 40 and applies the signal COM to the head unit 40. COM is a voltage waveform signal. When COM is applied, a piezoelectric element (not shown) provided in the nozzle expands and contracts, and ink is ejected by pushing ink out of the nozzle. The amount of ink ejected can be adjusted by controlling the amount of expansion and contraction of the piezoelectric element by changing the shape of the voltage waveform (for example, the magnitude of the voltage).

<Configuration of head unit 40>
In the present embodiment, the head unit 40 includes a head 41 that ejects black (K) ink, a head 42 that ejects cyan (C) ink, a head 43 that ejects magenta (M) ink in order from the upstream side in the transport direction, Four heads 44 for discharging yellow (Y) ink are arranged side by side (see FIG. 2). The heads 41 to 44 have the same structure and are each composed of a plurality of short heads.

  FIG. 3 is a diagram illustrating the arrangement of a plurality of short heads on the lower surface of the head 41. FIG. 3 is a view of the head 41 viewed virtually from above. As shown in the figure, the head 41 has a plurality of short heads 41A and 41B arranged at predetermined intervals in an intersecting direction (hereinafter also referred to as a paper width direction) that is a direction intersecting the transport direction. The short heads 41A and 41B are each arranged in a staggered pattern, and the short head 41B is arranged at a position between the short head 41A and the short head 41A in the paper width direction. In each short head, a nozzle row in which a plurality of nozzles are connected in series is formed. In the nozzle row, a predetermined number (for example, 180) of nozzles are arranged at a constant pitch (for example, 360 dpi) along the paper width direction. By ejecting ink from these nozzles, dots for the paper width direction can be formed at a time.

  In the example of FIG. 2, the head unit 40 is composed of four heads (heads 41 to 44). However, when printing is performed using ink of a color other than KCMY (for example, clear or white), the head The number of may be five or more. Further, which head ejects which color of ink is not limited to the example of FIG. For example, the head 41 may eject yellow (Y) ink. It is also possible to eject ink of two or more colors from one head.

<About landing position deviation due to eccentricity of transport roller>
In a printer that transports a medium by rotating a transport roller, such as the printer 1, ink droplets that are generated when the center of rotation of the transport roller is deviated, that is, when the transport roller is eccentric. Will be described.
FIG. 4A and FIG. 4B are diagrams for explaining how ink droplets land on a medium conveyed by a conveyance roller in which eccentricity occurs. 4A and 4B, it is assumed that the transport roller is rotating at a constant speed.

When ink droplets are ejected at regular intervals on a medium that is transported at a constant speed by a transport roller that is not decentered, the intervals between the landing positions in the transport direction of the ink droplets are theoretically all. Should be evenly spaced.
However, if the transport roller is eccentric, the landing positions in the transport direction of the ink droplets are uneven even when the ink droplets are ejected from the head at regular intervals as shown in FIG. 4A. This is because the conveyance speed of the medium is not constant due to the eccentricity of the conveyance roller. In other words, a portion where the amount of transport of the medium is large and a portion where the amount of transport of the medium is small during one rotation of the transport roller, so that ink droplets ejected at equal intervals have a wide landing interval in the transport direction in a portion where the amount of transport of the medium is large. Thus, the landing interval in the transport direction is narrowed at the portion where the transport amount is small. As described above, when the landing positions of the ink droplets are deviated, the image quality of the printed image is deteriorated.
In particular, when a ruled line is printed in the medium width direction using a printing apparatus in which the heads are arranged in a staggered manner as in the printer 1 (see FIG. 3), if the ink droplet landing position in the transport direction is displaced, There may be a step in the printed ruled line. That is, the ruled line cannot be printed as a single line, and the deterioration of the image is easily noticeable.

  As a method of suppressing such landing position deviation of ink droplets, there is a method of adjusting the timing of ejecting ink droplets from the head (hereinafter also referred to as PTS timing). For example, as shown in FIG. 4B, the ink droplet ejection interval is increased in a portion where the transport amount of the medium is large during one rotation of the transport roller. On the contrary, the ink droplet ejection interval is shortened in a portion where the conveyance amount of the medium is small during one rotation of the conveyance roller. Thereby, the intervals in the transport direction of the ink droplets that land on the medium can be made uniform.

  Therefore, in this embodiment, a correction value for correcting the PTS timing is set for each position in the circumferential direction of the transport roller. By ejecting ink while adjusting the PTS timing based on the correction value, it is possible to suppress landing position deviation in the ink droplet transport direction.

FIG. 5 illustrates a correction value set for each circumferential position of the transport roller. First, the circumferential direction of the transport roller from the position detected as the home position (HP) of the transport roller to the rotation of the transport roller until HP is detected again is divided into a plurality of sections. As described with reference to FIG. 4, when the transport roller is eccentric, the transport amount of the medium per unit time varies depending on the position of the transport roller in the circumferential direction. . Therefore, the correction value is set for each of the sections divided in the circumferential direction of the transport roller, and when the medium is transported by the section, the correction value set for the section is applied to adjust the PTS timing.
In the case of FIG. 5, one round of the transport roller is divided into 720 sections # 1 to # 720. Then, 720 correction values such as the correction value c1 for the # 1 section and the correction value c2 for the # 2 section are set for each circumferential position of the transport roller. That is, a correction value for one rotation (one cycle) of the transport roller is set. Then, the PTS timing of the ink ejected while the medium is being conveyed by the # 1 section of the conveying roller is adjusted by the correction value c1.

<Discharge timing correction>
Next, a method for correcting the actual ink ejection timing (PTS timing) using the correction value will be described.

FIG. 6 is a diagram for explaining the concept of PTS timing correction. The line arranged in the horizontal direction on the upper side of FIG. 6 represents the PTS signal. The PTS signal is a signal for instructing the timing for ejecting ink from the head, and is usually arranged at equal intervals. When the printer 1 does not correct the PTS timing when printing an image, control is performed so that ink is ejected from the head after a predetermined time has elapsed after receiving this PTS signal. For example, if a predetermined time is the reference delay value (t), ink is ejected after receiving a PTS signal (t). In this case, since the PTS signal is uniform, the interval at which ink is actually ejected is also uniform. Therefore, if the transport roller is eccentric as described above, the landing position of the ink droplet is shifted.
Therefore, the timing at which ink is actually ejected is adjusted by adding the correction amount (c) defined by the above correction value to the reference delay value (t). Specifically, the ink is ejected after (t) + (c) after receiving the PTS signal. That is, if the correction amount (c) is a positive (+) value, ink is ejected later than in the case of no correction. On the contrary, if the correction amount (c) is a negative (−) value, ink is ejected earlier than the case without correction. As described above, the PTS timing is corrected by using the correction value.

  Subsequently, points to be noted about the correction amount (correction value magnitude) when correcting the PTS timing will be described. 7A to 7C are diagrams for explaining the relationship between the magnitude of the correction value and the drive signal COM. In each figure, the trapezoidal wave formed in the section between the vertical lines A and B indicates the unit of the drive signal COM ejected for each pixel arranged in the transport direction. The “pixel” is a minimum unit that constitutes an image, and an image is configured by arranging the pixels two-dimensionally.

  FIG. 7A shows an example of the drive signal COM when the PTS timing is adjusted using the same correction value in a certain pixel P (x) and the next pixel P (x + 1). First, the drive signal COM starts to be applied to the pixel P (x) at timing A, and ink is ejected. At this time, the PTS timing is adjusted by the correction value C (x). Subsequently, the drive signal COM starts to be applied to the next pixel P (x + 1) at the timing B, and the next ink is ejected. At this time, the PTS timing is adjusted by a correction value C (x + 1) equal to the correction value C (x). In the case of FIG. 7A, the drive signal COM started to be applied at the timing A is finished to be applied at the timing C in FIG. 7A. That is, at the time of applying the next drive signal (timing B), the application of the previous drive signal is completed (timing C). This means that there is a margin time (Tm) between the end of the previous ink discharge and the start of the next ink discharge. This margin time (Tm in FIG. 7A) is defined as a “waveform collision margin”.

  Next, FIG. 7B shows an example of the drive signal COM when the PTS timing at the next pixel P (x + 1) is delayed with respect to the PTS timing at a certain pixel P (x). First, the drive signal COM starts to be applied to the pixel P (x) at timing A, and ink is ejected. At this time, the PTS timing is adjusted by the correction value C (x). Then, the next drive signal COM is started to be applied to the next pixel P (x + 1) at a timing D delayed by a predetermined time from the timing B. At this time, the PTS timing is adjusted by a correction value c (x + 1) larger than the correction value C (x). In the case of FIG. 7B, at the time of applying the next drive signal (timing D), the application of the previous drive signal is completed (timing C), so the waveform collision margin (Tm) is larger than in the case of FIG. 7A. Become. Since a sufficient waveform collision margin is secured, ink droplets are accurately ejected to the pixel P (X) and the next pixel P (x + 1), respectively.

Next, FIG. 7C shows an example of the drive signal COM when the PTS timing at the next pixel P (x + 1) is advanced with respect to the PTS timing at a certain pixel P (X). First, the drive signal COM starts to be applied to the pixel P (x) at timing A, and ink is ejected. At this time, the PTS timing is adjusted by the correction value C (x). Then, the next drive signal COM starts to be applied to the next pixel P (x + 1) at the timing E which is earlier than the timing B by a predetermined time. At this time, the PTS timing is adjusted by a correction value c (x + 1) smaller than the correction value C (x1). In the case of FIG. 7C, since application of the previous drive signal is not completed (timing C) at the time of applying the next drive signal (timing E), the drive signal COM in the pixel P (x) and the next pixel There is a possibility of interference with the drive signal COM at P (x + 1). That is, a sufficiently large shape collision margin (Tm) cannot be secured, and ink ejection to the next pixel starts before ink ejection to the previous pixel is completed, so that ink dots are accurately formed. May not be possible. Hereinafter, this phenomenon is referred to as “waveform collision”. Thus, when the next PTS timing is advanced with respect to the immediately preceding PTS timing, there is a limit of the correction amount. In the above example, C (x) −C (x + 1) between the correction value C (x) for the immediately preceding pixel P (x) and the correction value C (x + 1) for the next pixel P (x + 1). When the relationship is> (Tm), a waveform collision occurs. Therefore, when the PTS timing is advanced, an image to be printed may be deteriorated unless correction is performed so as to ensure a waveform collision margin (Tm).

<Outline of this embodiment>
In the present embodiment, correction values associated with the respective circumferential positions of the transport roller are set. Then, the ink ejection timing (PTS timing) is corrected according to the correction value. By ejecting ink at a timing according to the correction value, the landing position deviation of the ink dots caused by the eccentricity of the transport roller is suppressed, and a high-quality image is printed.

  Here, the correction value for correcting the PTS timing will be briefly described. FIG. 8 is a diagram illustrating an example of correction values used in the present embodiment. The vertical axis in the figure represents the magnitude of the correction value, and the horizontal axis represents the PTS number (PTS number corresponding to the circumferential position of the transport roller). The correction value used in the present embodiment is basically represented by a periodic function (for example, a sine curve) as depicted by a broken line in FIG. That is, the correction amount fluctuates in the negative direction and the positive direction while the transport roller makes one round, and the correction amount returns to zero when one cycle is completed.

  However, as described above, when the next PTS timing is advanced with respect to the immediately preceding PTS timing, if the correction value exceeds a predetermined size, the waveform collision margin cannot be secured, and the quality of the printed image deteriorates. There is a case. Therefore, when the PTS timing corrected at a certain position (circumferential position) is earlier than the PTS timing corrected immediately before, the change amount of the correction value is prevented from being equal to or less than a predetermined threshold value. As a result, the correction value has a linear portion having a certain amount of change as shown by the solid line portion in FIG.

=== Correction Value Calculation Processing ===
A specific process for calculating the correction value will be described. FIG. 9 is a diagram showing an overall flow for calculating the correction value. The correction value is calculated by sequentially executing steps S101 to S108 by the computer 110 according to the correction value calculation program. Note that these processes are performed in the manufacturing stage (inspection process) of the printer 1. Hereinafter, each step will be described.

S101: Print Test Pattern First, the printer 1 is used to print a test pattern. The test pattern is a pattern for detecting an error in the conveyance amount that occurs when the medium is conveyed using the conveyance roller 22. FIG. 10 shows an example of a test pattern printed in the present embodiment. On the test pattern, vertical ruled lines, horizontal ruled lines, scanning marks, solid identification IDs, and the like are printed. Also, what is actually used as a test pattern is an area on the left side of the “paper cut position” indicated by a broken line in the figure, and the right area is cut off after the printing is completed.

The vertical ruled line is a straight line drawn along the transport direction, and is used when performing an inclination detection process (S401) described later. In the present embodiment, as shown in FIG. 10, one print is printed on each end in the width direction of the test pattern.
The horizontal ruled line has a plurality of straight lines drawn along the width direction of the test pattern in the transport direction, and is used when detecting a transport amount error for each circumferential position of the transport roller. In this embodiment, printing of horizontal ruled lines is started from the home position (HP) that is the rotation reference position of the conveyance roller, and the horizontal ruled lines are continuously printed while conveying the medium. Then, the printing of the horizontal ruled lines is continued until the transport roller rotates three times. In other words, horizontal ruled lines for three cycles of the transport roller are printed. Although details will be described later, in practice, three horizontal ruled lines are printed in addition to the horizontal ruled lines for three cycles (see FIG. 26).

The scanning mark is a mark that serves as a reference for setting the test pattern according to the origin of the scanner to prevent errors in the setting direction when the test pattern is set and read (scanned) on a scanner or the like. It is. In the example of FIG. 10, triangular marks are printed at the four corner positions of the cut test pattern. Of the four marks, the mark at the origin position is colored with black or the like to clarify the reference position when setting the scanner.
In the solid identification ID, various printing conditions such as a conveyance speed and a printing resolution are displayed.

  When printing a test pattern, first, a medium (paper) is set in the printer 1, and then the medium is conveyed until the HP position of the conveying roller is detected by the origin sensor 52. That is, the medium is idled without printing ruled lines up to the HP position. Then, printing of the test pattern is started from the position where HP is detected. Subsequently, printing of horizontal ruled lines is started from the position where the next HP is detected. (Actually, printing of horizontal ruled lines starts one line before the next HP.) Printing of horizontal ruled lines for three cycles is completed, and the solid identification ID and the scanning mark downstream in the transport direction are displayed. When printing is completed, the test pattern printing is terminated. After the printing is completed, the paper is cut at the paper cutting position shown in FIG. When the test pattern is set on the scanner, the most downstream side in the transport direction is aligned with the upper side of the reading surface of the scanner so as not to be affected by shakiness or position variation due to cutting.

S102: Reading Reference Pattern and Test Pattern Subsequently, the printed test pattern and reference pattern are read by the scanner. The reference pattern will be described later. Further, since the scanner is publicly known, description thereof is omitted. The reading resolution of the scanner is, for example, 720 dpi (main scanning direction) × 720 dpi (sub-scanning direction).

  FIG. 11 is a diagram for explaining the setting position on the scanner when reading the reference pattern and the test pattern. The figure shows a state when the test pattern and the reference pattern after setting are viewed from the reading surface (glass surface) side of the scanner. In FIG. 11, the upper right corner is the paper set reference position, that is, the origin of the scanner. The main scanning direction of the scanner is the X axis, and the sub scanning direction is the Y axis.

First, the reference pattern 1 is set on the left side (right side in FIG. 11) of the scanner reading surface, and the reference pattern 2 is set on the right side (left side in FIG. 11). That is, the reference patterns are set on both ends of the scanner reading surface in the main scanning direction (X-axis direction). The reference patterns 1 and 2 are the same pattern, and a large number of lines are formed at equal intervals in the sub-scanning direction (Y-axis direction). This line corresponds to the horizontal ruled line of the test pattern, and the ruled line interval is, for example, 180 dpi. In order to keep the line interval (ruled line interval) with high accuracy, the reference pattern is formed by performing laser processing on a PET film or the like.
When the reference patterns 1 and 2 are set, the reference patterns 1 and 2 are set so that the distance from the outer frame of the reading surface (glass surface) of the scanner is within 1 mm. In addition, the installation angle of the reference patterns 1 and 2 with respect to the upper reading frame outer frame of the scanner is set to be within 1 degree.

  Next, the cut test pattern is set adjacent to the reference pattern 1. That is, the test pattern is set at a position between the reference patterns 1 and 2 in the X-axis direction. At this time, by setting the colored mark of the scanning mark so as to face the paper setting reference position (see FIG. 11), the setting direction of the test pattern is not mistaken. The interval between the reference pattern 1 and the test pattern is set to be within 1 mm. The upper end of the test pattern is set to be adjacent to the upper part of the outer frame of the reading surface of the scanner, and is set so that the interval is within 1 mm. In addition, the installation angle of the test pattern with respect to the upper reading frame of the scanner is set to be within 1 degree.

  After the test pattern and the reference pattern are set at appropriate positions on the scanner reading surface, the test pattern and the reference pattern are read by the scanner. Note that the image of the test pattern in the read result is a distorted image compared to the actual test pattern. Similarly, the image of the reference pattern is also distorted compared to the actual reference pattern. Further, the test pattern image in the reading result is affected not only by the influence of the reading position error but also by the conveyance error of the printer 1. On the other hand, since the reference pattern is formed at equal intervals regardless of the transport error of the printer 1, the image of the reference pattern is affected by the error of the reading position of the scanner. It is not affected by the error.

S103: Division of read image In order to calculate the correction value, it is necessary to detect the position of the horizontal ruled line of the test pattern. However, an accurate ruled line position cannot be detected in a state where the above error is included. Therefore, it is necessary to cancel the influence of the reading position error in the test pattern image based on the reference pattern image. Therefore, the error due to the reading position is eliminated by dividing the test pattern and the reference pattern and correcting the inclination or the like separately. Details of the inclination correction (S402) will be described later.
The computer 110 divides the image read in S102 into a test pattern image, a reference pattern 1 image, and a reference pattern 2 image. Thereby, three types of pattern images are obtained.

S104: Calculation of centroid position of ruled line In order to detect the interval between ruled lines (horizontal ruled lines in FIG. 10) printed on the test pattern, the position of each ruled line in the Y-axis direction (sub-scanning direction) is determined by the gradation of each ruled line ( The density is calculated as the center of gravity position. FIG. 12 shows a flow for calculating the barycentric position of the ruled line. The calculation of the position of the vassal is performed by executing the processes of S401 to S408 by the computer 110.

  First, the amount of inclination is detected for each of the reference patterns 1 and 2 and the test pattern (S401). The inclination amount is detected using a horizontal ruled line for the reference pattern and a vertical ruled line for the test pattern.

  FIG. 13 is a diagram illustrating a method for recognizing the ruled line position for the reference pattern. FIG. 14 is a diagram for explaining a method of calculating the inclination amount of the ruled line of the reference pattern. The inclination amount of the reference pattern is detected by recognizing the position of the horizontal ruled line (ruled line aligned in the Y-axis direction). Specifically, as shown in FIG. 13, one horizontal ruled line (horizontal ruled line P in FIG. 13) is extracted from the reference pattern, and two different locations (B in FIG. 13) are extracted from the extracted horizontal ruled lines. The amount of inclination is detected by detecting and comparing the positions of point and point C).

  First, the average density is calculated for each raster in the range from KX1 to HX in the X-axis direction for pixel rows arranged in the X-axis direction in the region from the position KY to HY1 in the Y-axis direction (hereinafter also referred to as a raster). To do. Then, the raster having the highest density is extracted from the calculated average density of each raster, and the ruled line density centroid calculation range is set around the extracted raster. In the case of FIG. 13, with respect to the horizontal ruled line P, a range from the broken line P1 to the broken line P2 centering on the raster including the point A having the highest density is set as the ruled line density centroid calculation range.

  Subsequently, the raster having the highest average density in the range from KX2 to HX in the X-axis direction within the set ruled line density centroid calculation range is extracted. In FIG. 13, the raster including the point B is extracted. A first center-of-gravity calculation range in the range from KX2 to HX is set around this raster. In FIG. 13, the shaded area including the point B is set as the first centroid calculation range. Similarly, a second center-of-gravity calculation range is set for a range from KX3 to HX in the X-axis direction. In FIG. 13, the shaded area including the point C is set as the second centroid calculation range.

  The density distribution in the Y-axis direction in the first centroid calculation range and the second centroid calculation range is a distribution having a gradation value peak as shown in FIG. Using this distribution, the density barycentric position (KY2 and KY3 in FIG. 13) of each ruled line is calculated. A specific method for calculating the density centroid position from the distribution of FIG. 14 will be described later.

Using the coordinates KX2 and KX3 in the X-axis direction of the two density measurement positions (B and C) and the calculated density centroid positions (coordinates in the Y-axis direction) KY2 and KY3, the inclination θ of the horizontal ruled line P is It can be expressed by a formula.
θ = tan ⁻¹ ((KY3-KY2) / (KX3-KX2))

Next, the inclination detection of the test pattern will be described. FIG. 15 is a diagram for explaining a method of calculating an inclination amount for a test pattern. As described above, in the test pattern, the amount of inclination is calculated using the vertical ruled line drawn along the transport direction during printing (corresponding to the Y-axis direction in FIG. 15). As shown in FIG. 15, four center-of-gravity calculation ranges having a height of H and a width of W are set at positions in the ranges M1 to M2 in the Y-axis direction of the test pattern (a rectangular region surrounded by a dotted line in FIG. 15). LU, RU, LL, RL). For each of these centroid calculation ranges, the average density per raster is obtained, and the density centroid position in the X-axis direction in the raster indicating the highest density is extracted. In FIG. 15, the XLU position is extracted from the LU. Similarly, the position of XRU in RU, XLL in LL, and XRL in RL are calculated (see FIG. 15). Thereby, the inclination amounts θt1 and θt2 of the vertical ruled lines on both sides of the test pattern are expressed by the following equations.
θtl = tan ⁻¹ ((XLL−XLU) / (M2−M1))
θt2 = tan ⁻¹ ((XRL−XRU) / (M2−M1))

By averaging the inclination amounts θt1 and θt2 of the ruled lines on both sides, the inclination amount θt of the test pattern is obtained by the following equation.
θt = (θtl + θt2) / 2

In this way, after the inclination θ of the reference pattern and the inclination θt of the test pattern are detected, the inclination is corrected (S402). The test pattern image is rotationally corrected based on the inclination result θt of the test pattern image. Similarly, the reference pattern image is rotationally corrected based on the inclination result θ of the reference pattern image.
A bilinear method is used as an algorithm for image rotation processing. Since this algorithm is well known, its description is omitted.

  Next, the computer 110 calculates an offset value for the margin (S403). FIG. 16 is an explanatory diagram of the margin amount X. The solid line rectangle (outer rectangle) in the figure indicates the image after the rotation correction in S402. A dotted-line rectangle (inner oblique rectangle) in the figure indicates an image before rotation correction. In order to make the image after rotation correction into a rectangular shape, right-angled triangular margins are added to the four corners of the image after rotation when the rotation correction processing in S402 is performed.

If the inclination of the reference pattern and the inclination of the test pattern are different, the amount of added margin is different, and the position of the test pattern line relative to the reference pattern is relatively shifted before and after the rotation correction (S402). become. Therefore, the computer 110 obtains the margin amount X by the following equation.
X = (w cos θ−W ′ / 2) × tan θ
By subtracting the calculated margin amount X from the ruled line density barycentric position calculated in the following processing, a shift in the position of the test pattern line with respect to the reference pattern is prevented.

  Next, calculation of the density centroid position will be described. The method for calculating the density centroid position is the same for both the test pattern and the reference pattern. In order to calculate the density centroid position, first, an average value of density for each raster is calculated (S404). The average density value for each raster is represented by a value obtained by dividing the total density value of the pixels in the range constituting the raster (pixel column) by the number of pixels in the range constituting the raster.

  After the average density for each raster is calculated, the highest density raster is identified from the rasters (S405). FIG. 17 is a diagram illustrating an example of the density distribution for each raster. The diagram on the right side is an enlarged view of a region near a ruled line, and the diagram on the left side is a diagram showing the density distribution in that region. In this embodiment, a density threshold value for distinguishing between the density of the ruled line and the density of the printing medium is provided, and the threshold value is compared with the calculated average density. In the density distribution of FIG. 17, the average raster density is larger than the threshold value in the range indicated by the hatched portion. Among the rasters exceeding this threshold, the raster with the highest density (the raster at the position showing the peak of the density distribution) is identified as the highest density raster.

  Next, the calculation range for calculating the density centroid position is specified around the highest density raster specified in S405 (S406). FIG. 18 is a diagram for explaining a method of calculating the density centroid position. In FIG. 18, when the maximum density raster is Y_peak, for example, a range from the raster (Y_peak-N) to the raster (Y_peak + N) is specified as the calculation range.

Subsequently, the minimum density value G_min within the specified calculation range is calculated. In FIG. 18, at the point M, the minimum density value G_min (dotted line in the figure) is obtained. Further, a total density value G_sum within the specified range is calculated. If the total density of the rasters Y1 to Y2 is represented by SUM (G (Y1), G (Y2)), G_sum is represented by the following equation.
G_sum = SUM (G (Y_peak-N) -G_min, G (Y_peak + N) -G_min)
Further, an integrated value G_area of the distance from a certain raster Y_peak and the density in the raster is expressed by the following equation.
G_area = SUM ((G (Y_peak−N) −G_min) × (−N),
(G (Y_peak + N) -G_min) × N)
Here, the position of the center of gravity on the plane is calculated by the sum of (cross-sectional area × distance to the centroid position) ÷ total cross-sectional area. Therefore, the gravity center position Y_center of the density distribution of FIG. 18 is calculated by the following equation (S407).
Y_center = Y_peak + G_area / G_sum

  When reading the ruled line of the test pattern, the density of the ruled line's center of gravity is calculated inaccurately because the ruled line density of the scanned image decreases due to disturbance (vibration) or the ruled line is read twice. There is a risk that. In order to prevent such a problem from occurring, the distance between the ruled lines is determined for each of the reference pattern and the test pattern (S408). More specifically, an error in the distance between ruled lines is set as a threshold for a paper feeding error in a printer and a reading error in a scanner. If the distance between the ruled lines actually detected exceeds the threshold, an error is displayed, and the ruled line position calculation process (S104) is performed again. Thereby, the barycentric position of the ruled line can be calculated more accurately.

S105: Calculate the centroid position of the virtual ruled line of the virtual reference pattern Based on the calculated centroid position of the ruled line, the test pattern ruled line reading position (main scanning) is used to accurately detect the rule pattern spacing (horizontal ruled line position) of the test pattern. A virtual reference pattern at the position of the direction is set (S105). By comparing the ruled lines of the test pattern and the virtual reference pattern at the same position in the X-axis direction (main scanning direction), it is possible to cancel the influence of misalignment that may occur when the reference pattern 1 and the reference pattern 2 are read.

  FIG. 19 is a diagram illustrating a method for calculating a virtual reference pattern. The position of the horizontal ruled line of the virtual reference pattern is calculated using the ruled line density centroid position of the reference pattern 1 and the ruled line density centroid position of the reference pattern 2 calculated in S104.

First, between the reference pattern 1 and the reference pattern 2, a ratio T between the reading position of the test pattern in the X-axis direction and the reading position of each reference pattern in the X-axis direction is obtained. As shown in FIG. 19, the distance from the paper set reference position to the center position of the reference pattern 1 in the X-axis direction is Lr. Similarly, the distance from the paper setting reference position to the center position in the X-axis direction of the reference pattern 2 is L1, and the distance from the paper setting reference position to the center position in the X-axis direction of the test pattern is Lp. At this time, the ratio T of the reading position in the X-axis direction is expressed by the following equation.
T = (Lp−Lr) / (L1−Lr)

The nth ruled line position (density centroid position) K (n) of the virtual reference pattern is the nth ruled line position (density centroid position) K1 (n) of the reference pattern 1 and the nth ruled line of the reference pattern 2. The position (density gravity center position) K2 (n) and the ratio T are used to calculate the following equation.
K (n) = (K2 (n) −K1 (n)) × T + K1 (n)

S106: Calculate the ruled line position of the test pattern The ruled line position of the test pattern is calculated based on the density centroid position K of the virtual reference pattern calculated in S105. FIG. 20 is a diagram illustrating an example of the relationship between the density centroid position of the virtual reference pattern and the density centroid position of the test pattern. FIG. 21 is a diagram for explaining the ratio of the density centroid position of the test pattern to the density centroid position of the virtual reference pattern. FIG. 22 is a diagram for explaining a method of calculating the actual measurement value of the density centroid position of the test pattern.

  FIG. 20 is an enlarged view of a part of the horizontal ruled line group printed on the test pattern and the ruled line group of the virtual reference pattern at the corresponding position, so that the positional relationship between the groups can be easily understood. The computer 110 searches for which ruled line of the virtual reference pattern the density barycentric position of the ruled line of the test pattern is located. For example, in the case of FIG. 20, the density centroid position (373.7686667) of the top ruled line of the test pattern is the density centroid position (309.61325) of the second ruled line of the virtual reference pattern and the density centroid position of the third ruled line ( 469.430413).

After the positional relationship between the ruled line of the test pattern and the ruled line of the virtual reference pattern is clarified, the position of the ruled line of the test pattern between the two ruled lines of the virtual reference pattern is specified. As shown in FIG. 21, the density centroid position of the mth ruled line of the test pattern is S (m), and S (m) is the density centroid position K (n−1) of the n−1th ruled line of the virtual reference pattern. ) And the density barycentric position of the nth ruled line are located between K (n). Also, let L be the distance between K (n) and K (n-1), and let L (m) be the distance between S (m) and K (n-1). At this time, the ratio H of the interval L (m) to the interval L is expressed by the following equation.
H = L (m) / L = (S (m) -K (n-1)) / (K (n) -K (n-1))

  In the above example (see FIG. 20), H = ((373.7686667) − (309.61325)) / ((469.430413) − (309.61325)) ≈0.4.

By the way, since the actual reference pattern is equally spaced, the position of an arbitrary ruled line of the reference pattern can be calculated if the absolute position of the first line of the reference pattern is zero. For example, as shown in FIG. 22, the measured value of the position of the (n−1) th ruled line of the reference pattern is represented by J (n−1), and the measured value of the position of the nth ruled line of the reference pattern is J (n). n).
The mth ruled line position Sr (m) of the test pattern is calculated by the following equation.
Sr (m) = (J (n) −J (n−1)) × H + J (n−1)

  By repeating this operation from the first ruled line to the last ruled line, all ruled line positions of the test pattern are calculated.

S107: Calculation of correction value Based on the ruled line position of the test pattern calculated in S106, a correction value is calculated for each circumferential position of the transport roller. First, an outline of correction value calculation will be described. FIG. 23 is a diagram for explaining an outline of correction value calculation.

  When ink is continuously ejected to a medium transported using the printer 1 at a predetermined PTS timing, the ideal ink droplet landing positions are all equally spaced as shown in FIG. . Here, “ideal” means a state in which there is no occurrence of a medium transport error or a deviation in the flight path of the ejected ink.

  However, in reality, there is a transport error due to the eccentricity of the transport roller as described above. Therefore, as shown in FIG. Occur. That is, the landing positions are not evenly spaced. When the test pattern is printed using the printer in such a state, a ruled line is printed at a position as shown in (c). At this time, the ruled lines are printed every several dots so that the positions of the ruled lines can be easily recognized on the test pattern. For example, a line represented by a thick solid line in (c) is a ruled line that is actually printed.

  In this embodiment, a correction value for correcting the amount of deviation of the dot landing position as shown in (c) to a position where there is no deviation as shown in (a) is calculated, and the ink ejection timing is adjusted. Print high-quality images. That is, as shown in (d), the amount of correction between the detected actual ruled line position and the ideal ruled line position is the correction amount. When calculating the correction amount for one rotation of the transport roller, the correction amount of each PTS timing between ruled lines is calculated by performing linear interpolation as shown in (e). For example, when correction is performed in units of 2 PTS, one correction value is set for 2 PTS.

  Next, specific processing for calculating the correction value will be described. FIG. 24 is a diagram illustrating a flow for calculating a correction value. The correction value calculation process is performed by executing the processes of S701 to S710 by the computer 110.

First, the ruled line is turned upside down (S701). The barycentric position of the ruled line of the test pattern calculated in S106 is position information calculated from the print termination side of the test pattern (downstream in the transport direction during printing) (see FIGS. 10 and 11). On the other hand, in this embodiment, it is necessary to calculate a correction value according to the direction of conveyance by the conveyance roller. For this reason, the ruled lines in FIG. 11 are inverted so that the correction value can be calculated from the print start side of the test pattern. Assuming that the total number of ruled lines is N and the kth ruled line position before flipping up and down is LP_org (k) (0 ≦ k ≦ N), the ruled line position LP (k) after flipping up and down is expressed by the following equation. The
LP (k) = LP_org (N−1) −LP_org ((N−1) −k)

Next, the difference between the ruled line position LP (k) obtained by reversing the upper and lower sides and the ideal position of the ruled line is calculated for all ruled lines (S702). If the ideal position of the kth ruled line is IP (k), the difference ΔLP (k) for the kth ruled line is expressed by the following equation.
ΔLP (k) = LP (k) −IP (k)

  Subsequently, the DC component is removed (S703). Here, the DC component is an error that occurs when the transport roller makes one rotation. For example, it is assumed that the circumferential length of the actually manufactured transport roller differs from the design value due to a manufacturing error in the manufacturing stage of the transport roller. The amount of the medium actually transported when such a transport roller makes one rotation is different from the theoretical transport amount planned at the time of design, and an error due to a DC component occurs. If a DC component error has occurred, there is a deviation between the correction value at the rotation reference position (HP) of the transport roller and the correction value when the transport roller makes one rotation and then returns to the HP position. May occur. That is, when the correction value is set without taking the DC component error into consideration, there is a possibility that an extreme difference occurs in the magnitude of the correction value before and after the HP. Therefore, by removing the DC component, the correction amount is calculated assuming that the circumference of the transport roller is a theoretical value at the time of design.

When the number of ruled lines in one rotation of the transport roller is M and the number of test patterns is J (that is, the total number of ruled lines N = M × J), the DC component error amount ΔDC (J ) Is expressed by the following equation.
ΔDC (J) = ΔLP (J × M) −ΔLP ((J−1) × M)
The average value ΔDC_AVE of the DC component for one round is expressed by the following equation.
ΔDC_AVE = AVE (ΔDC (1), ΔDC (2),..., ΔDC (J))
Using these values, the measured value LP _DC (k) of the kth ruled line after removing the DC component is calculated by the following equation.
LP _DC (k) = LP (k) + k / M × ΔDC_AVE
Accordingly, the difference ΔLP _DC (k) between the kth ruled line position after the DC component removal and the kth ideal value is expressed by the following equation.
ΔLP _DC (k) = LP _DC (k) −IP (k)

  Even if the DC component is removed when calculating the correction value in this way, the positional deviation due to the DC component that occurs during actual printing is not corrected. As a result, for example, an image having a length of 10 m in the transport direction may actually be printed as an image of 10 m1 cm. However, when printing is performed using a line printer such as the printer 1 of the present embodiment, the positional deviation in the transport direction is different from a so-called serial printer that ejects ink while the head moves in a direction crossing the transport direction. Even if it occurs, it is inconspicuous. That is, in the line printer, the positional deviation itself due to the DC component is allowed.

  Next, based on the deviation amount of the ruled line from the ideal position, a time to be corrected for each ruled line is calculated (S704). FIG. 25 shows a relationship between the deviation of the landing position of the ink droplet and the time to be corrected. Circles represented by white in the figure represent cases where ink droplets land at ideal positions. When the ink droplet lands on the ideal position, the position of the ruled line formed by the ink droplet is also an ideal position. Therefore, three consecutive ruled lines (for example, k-1, k, k + 1th ruled lines) are formed at ideal positions in the transport direction.

On the other hand, the circles represented by black in the figure represent the case where the ink droplets land on the near side of the ideal landing position in the transport direction. That is, it represents a case where ink droplets land earlier than ideal. For example, the actual kth ruled line is formed at a position shifted by ΔLP _DC (k) to the negative side (minus side) from the ideal landing position IP (k) of the kth ruled line (FIG. 25). That is, the amount of deviation between the landing position of the ruled line formed by the ink droplets that landed earlier than ideal and the ideal landing position becomes negative (ΔLP _DC (k) <0). Here, if the time between ruled lines (for example, between the (k + 1) th ruled line and the kth ruled line) is Unit_time, the slope a of the straight line connecting the three black circles in FIG. 25 is expressed by the following equation.
a = (LP _DC (k + 1) −LP _DC (k)) / Unit_time

In order to correct the deviation amount ΔLP _DC (k) to the ideal landing position IP (k), it is necessary to delay the timing of ejecting ink droplets by ΔT (k) on the straight line with the inclination a as shown in FIG. is there. Therefore, ΔT (k) is expressed by the following equation.
ΔT (k) = − a × ΔLP _DC (k)

Also, the circles represented by the hatched portions in the figure represent the case where the ink droplets land behind the ideal landing position in the transport direction. That is, it represents a case where ink droplets land later than ideal. In this case, the deviation amount ΔLP _DC (k) between the landing position of the k-th ruled line formed by the ink droplets landed later than ideal and the ideal landing position is positive (ΔLP _DC (k)> 0)). Here, assuming that the time between ruled lines (for example, between the kth ruled line and the k-1th ruled line) is Unit_time, the slope b of the straight line connecting the three hatched circles in FIG. The
b = (ΔLP _DC (k) −ΔLP _DC (k−1)) / Unit_time
In order to correct the shift amount ΔLP _DC (k) to the ideal landing position IP (k), it is necessary to advance the timing of ejecting ink droplets by ΔT (k) on the straight line with the slope b as shown in FIG. is there. Therefore, ΔT (k) is expressed by the following equation.
ΔT (k) = − b × ΔLP _DC (k)

  When calculating the time to be corrected as shown in FIG. 25, in order to calculate the inclination with respect to the ruled line located at the end of the test pattern, plus three ruled lines are necessary in addition to the target ruled line. FIG. 26 shows an example of ruled lines used for calculating the correction amount. Consider a case in which ruled lines for three periods are printed with one period (M ruled lines) from the rotation reference position HP of the transport roller to the ruled line immediately before the next HP as shown in the figure. First, when calculating a correction value from HP1, which is the first HP in the first cycle, one ruled line is required before HP1 in order to detect the ruled line positions before and after HP1. Further, in order to calculate the correction value of the terminal portion of the third cycle, the HP of the terminal portion of the third cycle (that is, the head HP of the fourth cycle) HP4 is added. When calculating the correction value from the HP 4, one ruled line is required after the HP 4 in order to detect the ruled line positions before and after the HP 4. That is, a total of M (number of ruled lines per cycle) × J (number of laps) +3 ruled lines is printed.

  Next, using the time ΔT (k) to be corrected between ruled lines, the time to be corrected for each PTS is calculated as a temporary correction value (S705). Here, correcting the time for each PTS means that when there are a plurality of PTS timings (droplet ejection timings) between the ruled lines, the ejection timing is corrected for each of them. By adjusting the individual PTS timing, the deviation of the landing positions of the ink droplets during actual printing is suppressed.

  As a method for obtaining a provisional correction value for each PTS, specifically, using the time to be corrected at the k-th ruled line position and the time to be corrected at the (k + 1) -th ruled line position, By performing linear interpolation on the number of PTSs, the time to be corrected for each PTS is calculated.

Here, the number of PTSs between ruled lines is P, and the number of PTSs to which correction is applied is NUM_PTS. For example, if there is one correction amount in 2 PTS units, NUM_PTS = 2. Further, assuming that the number of all ruled lines is N and the PTS number to which the correction value is applied is x (0 ≦ x ≦ (N−3) × P / NUM_PTS), the time ΔPTS (x ) Is represented by the following equation.
ΔPTS (x) = (ΔT (k + 1) −ΔT (k)) × x / P / NUM_PTS
+ ΔT (k)

The calculated correction time ΔPTS (x) corresponds to a correction value indicated by a broken line in FIG. 8 described above, and the correction value is set as a temporary correction value TmpRev (x).
If this temporary correction value is applied as it is, a waveform collision as described with reference to FIG. 7C may occur, and accurate ink ejection may not be performed. Therefore, the calculated temporary correction value is checked for waveform collision, and the temporary correction value is corrected as necessary to avoid the occurrence of waveform collision (S706).

Assuming that the waveform collision margin is MARGIN, waveform collision does not occur when TmpRev (x) −TmpRev (x−1) ≧ MARGIN. In this case, there is no need to correct the temporary correction value TmpRev (x).
On the other hand, when TmpRev (x) −TmpRev (x−1) <MARGIN (in the above description, C (x) −C (x + 1)> (Tm)), a waveform collision occurs. Therefore, in order to avoid waveform collision, the x-th correction value TmpRev (x) to be corrected is corrected as follows using the temporary correction value TmpRev (x−1) immediately before.
TmpRev (x) = TmpRev (x-1) + MARGIN

  As a result, in the case where the PTS timing corrected at a certain position (circumferential position) is earlier than the PTS timing corrected immediately before, and under the condition that waveform collision is likely to occur, the broken line portion ( In the example of FIG. 8, the temporary correction value represented by a sine curve is corrected to a straight line represented by a solid line. That is, the correction value is corrected so that the amount of change is constant (for MARGIN). It should be noted that under the condition that the waveform collision is difficult to occur, the temporary correction value is not corrected, and thus the correction value represented by the broken line in FIG. 8 (sine curve in the example of FIG. 8) remains.

  Although the correction values shown in FIG. 8 are shown for three rotations of the conveyance roller, the number of rotations of the conveyance roller is not taken into account when the correction is actually performed. That is, as described with reference to FIG. 5, the correction value is set for each circumferential position of the transport roller. If the circumferential position of the transport roller is the same, the same correction value is used for the first and second rounds. It is done. For example, the same correction value is used as the correction value set at the position # 3 in FIG. 5 regardless of the number of rotations of the transport roller. Therefore, a correction value for one cycle of the transport roller is calculated from all the calculated correction values (S707).

The correction value for one period is calculated by averaging the correction values obtained for each period. At that time, the average value is calculated by excluding the correction value in the first period. For example, when correction values for three rounds as shown in FIG. 8 are obtained, the data for the first round are excluded and the correction values for the second and third rounds are averaged to obtain 1 A correction value for the period is calculated. The reason why the first round data is excluded is to consider the case where the correction value is corrected in S706. For example, the correction value at the start of the first round is always zero (see FIG. 8). On the other hand, when the correction value is corrected, the correction value at the start time of the second turn and the correction value at the start time of the third turn are both larger than zero (see the solid line portion in FIG. 8). . That is, the first round correction value may be different from the subsequent correction values. Therefore, a correction value for one cycle is calculated by taking an average of correction values for the second and subsequent rounds (second and third rounds). The average correction value REV (x) is expressed by the following equation.
REV (x) = (TmpRev (x + M × P / NUM_PTS) +
TmpRev (x + 2 × M × P / NUM_PTS)) / 2

Incidentally, the DC component is removed in the process of calculating the correction value REV (x) (S703). At this time, a step is generated between the correction value at the beginning and the correction value at the end of the correction value for one rotation of the transport roller. There is a case. FIG. 27 is a diagram illustrating an example of a correction value step. In the left diagram of FIG. 27, the correction value at the front end and the correction value at the end are different, so if the correction value is repeatedly used according to the rotation of the transport roller, the next rotation from the end of the current rotation A step of the correction value is generated when shifting to the tip of the rotation. Accordingly, the step is corrected by applying a tilt correction to the correction value as shown in the right side of FIG. 27 (S708).
Note that the gap Gap of the correction value at the beginning and the end is expressed by Rev (0) −Rev (M × P−1), and the absolute value of the gap is larger than the absolute value of MARGIN (| Gap |> | MARGIN |), a step is generated.

  The level difference must be corrected in consideration of the continuity of the correction value, but the level difference correction method differs depending on the difference (inclination) at the end of the correction value and the amount (inclination) at the beginning. Note that. FIG. 28 is a diagram for explaining the step correction method for each correction value change amount.

First, regarding the case where the slope of the correction value at the end and the slope of the correction value at the tip have different signs, FIG. 28 (1) shows an example where the end slope is positive and the top slope is negative. In this case, the correction may be performed so that the target step (Gap) to be corrected = the actual step Gap. Assuming that the inclination correction ratio is c, it is expressed by c = (Gap) / (M × P / NUM_PTS). Therefore, the correction value Rev (x) is expressed by the following expression using the correction ratio c.
Rev (x) = c × x + Rev (x)
FIG. 28 (2) shows a case where the end slope is negative and the head slope is positive. In this case as well, if correction is performed so that the target step (Gap) = actual step Gap. Good.

Next, consider the case where the slope of the correction value at the end and the slope of the correction value at the tip have the same sign. When the inclination of the end portion is represented by b1 and the inclination of the head portion is represented by b2, if both b1 and b2 are negative (in the cases of (3) and (4) in FIG. 28), the target step (Gap) to be corrected Is an amount obtained by subtracting MARGIN from the actual step Gap.
That is, (Gap) = Gap−MARGIN. Based on the target step (Gap), the inclination correction ratio c and the correction value Rev (x) are calculated in the same manner as described above.

When both b1 and b2 are positive (in the cases of FIGS. 28 (5) and (6)), the target step (Gap) to be corrected is a value obtained by subtracting the average value of the slope from the actual step Gap. . That is, (Gap) = Gap− (b1 + b2) / 2. Based on the target step (Gap), the inclination correction ratio c and the correction value Rev (x) are calculated in the same manner as described above.
By calculating the step to be corrected (Gap) in consideration of the inclination of the correction value at the front end and the end, it is possible to optimally correct the step corresponding to various cases.

After the inclination correction is completed, the correction value is offset and rounded off to make the center of the correction value zero (S709).
Finally, reconfirmation about waveform collision is performed. By performing step difference correction in S708, reconfirmation is performed so that the next correction value with respect to the previous correction value does not exceed MARGIN when the waveform collision check (S706) is performed. Suppresses the occurrence. If it is determined that a waveform collision occurs as a result of the check, that is, if it exceeds MARGIN, an error is displayed and printing is terminated.

S108: The correction value data table is stored in the memory. The correction value calculated in S107 is stored in the memory 62 of the printer 1 as a data table in association with the circumferential position of the transport roller. When printing is performed, ink is ejected while adjusting the PTS timing by applying the correction value with reference to the data table.

<Summary>
In the printing apparatus according to the present embodiment, an image or the like is printed by ejecting ink droplets from a nozzle row arranged in an intersecting direction (paper width direction) intersecting the medium transport direction to the transported medium. During printing, the medium is conveyed by rotating the conveyance roller. If the conveyance roller is eccentric, an error occurs in the medium conveyance amount for each circumferential position of the conveyance roller. In such a case, since the landing positions of the ejected ink droplets are shifted, the image quality of the printed image is deteriorated.

  Therefore, in this embodiment, a correction value is set for each circumferential position of the transport roller, and the ink droplet ejection timing is adjusted using a plurality of correction values that are entertained. At this time, the correction value set for each circumferential position of the transport roller is basically represented by a periodic function such as a sine curve. However, when the ink ejection timing is advanced compared to the immediately preceding ejection timing, the change amount of the correction value is set to be constant in a part of the region.

  According to such a printing apparatus, by correcting the ejection timing of the ink droplets, it is possible to correct the deviation of the landing positions of the ink droplets and print a high-quality image. Further, even if the current ejection timing is made earlier than the immediately preceding ejection timing, the ejection operation does not interfere. That is, ink can be accurately discharged while maintaining an appropriate discharge interval.

  In addition, a test pattern in which a plurality of lines along the crossing direction are arranged in the transport direction is printed, and the interval in the transport direction between the lines is detected from the image data obtained by reading the test pattern with a scanner. Based on the correction value, a correction value is calculated. Further, when the time interval from the ejection of the ink droplet from the nozzle to the ejection of the next ink is shorter than a predetermined threshold, the calculated correction is made so that the time interval becomes equal to or greater than the threshold. The value is corrected.

  By calculating the correction value based on the result of measuring the test pattern formed by the nozzle row used for actual printing, it is possible to obtain a correction value that can accurately correct the ejection timing.

  The above-described threshold value is increased as the medium conveyance speed is higher. In the present embodiment, the drive signal COM is applied to the head in order to eject ink. However, when the conveyance speed is high, the timing of applying the next drive signal COM is completed before the application of the drive signal COM is completed. There is a risk of coming. In this case, waveform collision occurs, and accurate ink ejection operation cannot be performed. Therefore, by adjusting the threshold according to the conveyance speed, the allowable range for avoiding waveform collision can be optimized.

  Further, in this embodiment, after removing the influence of the transport error (DC component) that occurs when the transport roller makes one rotation, the interval between the lines of the above test pattern is measured, and the correction value is calculated. Thereby, an extreme difference in the magnitude of the correction value before and after the rotation reference position of the transport roller is suppressed, and the correction value is continuously changed.

  Further, when a step is generated between the correction value at the front end portion in the circumferential direction of the transport roller and the correction value at the end portion in the circumferential direction, an inclination correction is performed on the correction value. Make the step small. Specifically, an appropriate correction ratio is calculated according to the level difference to be corrected, and tilt correction is performed according to the correction ratio. By correcting the level difference, the correction value is prevented from greatly fluctuating when shifting from the end of one cycle of the transport roller to the head of the next cycle.

  Note that the level difference must be corrected in consideration of the continuity of the correction values. If the slopes of the correction values at the front end and the end are different signs, both are the same sign and positive, and both are the same sign. In the case of failure, the size of the step to be corrected is calculated appropriately, and the step is corrected. As a result, accurate inclination correction is performed regardless of the condition of the inclination of the correction value.

  Further, the correction value is calculated for each cycle in which the roller rotates once. In the present embodiment, correction values for three periods are calculated, and among them, the correction values for the second and third periods excluding the correction values calculated for the first period are averaged, so that one period of the transport roller A correction value for the minute is calculated. Further, when the correction values for the third and subsequent cycles are calculated, that amount is also averaged. Since the beginning of the correction value calculated in the first period is always zero, there is a case where the correction value in the second period and thereafter is largely deviated. Therefore, by excluding the correction value in the first period and averaging, the correction value for one period can be accurately calculated.

=== Relationship with transport speed ===
In the above-described embodiment, the correction value is calculated in consideration of the waveform collision margin in order to avoid the waveform collision, but this waveform collision margin is largely related to the medium conveyance speed during printing.

  FIG. 29 is a diagram illustrating a change in the waveform collision margin when the conveyance speed is changed. In the figure, the vertical axis represents the voltage of the drive signal COM, and the horizontal axis represents time. The upper diagram shows the case where the conveyance speed is 300 mm / sec, and the lower diagram shows the case where the conveyance speed is 150 mm / sec.

  When the conveyance speed is high (in the case of 300 mm / sec), the waveform collision margin from the end of the application of the drive signal COM to the previous pixel to the start of the application of the drive signal to the next pixel is tm (300). It becomes. On the other hand, when the conveyance speed is low (150 mm / sec), the waveform collision margin from the end of the application of the drive signal COM to the previous pixel to the start of the application of the drive signal to the next pixel Becomes tm (150). As shown in FIG. 29, when the same drive signal COM is applied, the waveform collision margin can be increased as the medium transport speed is slower (tm (150)> tm (300)). That is, since the waveform collision margin can be changed according to the conveyance speed of the medium, the threshold value of the correction value can be appropriately changed according to the conveyance speed.

  For example, when the conveyance speed is 150 mm / sec, there is a large margin in the waveform collision margin, so the threshold can be lowered. Therefore, it is less necessary to correct the correction value in order to avoid the waveform collision, and the correction value as indicated by the broken line in FIG. 8 can be used as it is. On the other hand, when the conveyance speed is 150 mm / sec, the margin of the waveform collision margin is small, and it is necessary to set a large threshold value. Therefore, there is a high possibility that the correction value must be corrected in order to avoid waveform collision, and the correction value is corrected and used as shown by the solid line portion in FIG. As described above, since the size of the collision margin varies depending on the conveyance speed of the medium, the inclination of the correction value (inclination of the solid line portion in FIG. 8) also varies depending on the conveyance speed.

=== Other Embodiments ===
Although a printer or the like as one embodiment has been described, the above embodiment is for facilitating the understanding of the present invention, and is not intended to limit the present invention. The present invention can be changed and improved without departing from the gist thereof, and it is needless to say that the present invention includes equivalents thereof. In particular, the embodiments described below are also included in the present invention.

<Ink used>
In the above-described embodiment, an example of printing using four colors of CMYK inks has been described. However, the present invention is not limited to this. For example, printing may be performed using inks of colors other than CMYK, such as light cyan, light magenta, white, and clear.

<Regarding the arrangement of nozzle rows>
The nozzle rows of the head portion are arranged in the order of KCMY along the transport direction, but the present invention is not limited to this. For example, the order of the nozzle rows may be changed, or the number of nozzle rows for K ink may be greater than the number of nozzle rows for other inks.

<About the printer driver>
The printer driver processing may be performed on the printer side. In this case, a printing apparatus is configured by the printer and the computer on which the driver is installed.

1 Printer 20 transport unit, 22 transport roller, 23 discharge roller, 24 belt,
40 head units, 41 to 44 heads,
50 detector groups, 51 rotary encoder, 52 origin sensor 60 controller, 61 CPU, 62 memory, 63 COM generator,
110 computer

Claims (8)

(A) a transport unit that transports the medium in the transport direction by rotating the transport roller;
(B) a head unit that has a plurality of nozzles arranged in an intersecting direction intersecting the transport direction, and prints an image on the medium by ejecting ink from the nozzles;
(C) a storage unit that stores a plurality of correction values associated with each circumferential position of the transport roller;
(D) a control unit that controls the operation of the transport unit and the head unit, the control unit discharging the ink from the nozzle at a timing according to the correction value;
A printing apparatus comprising:
The plurality of correction values are:
When the timing is advanced compared to the immediately preceding timing, the amount of change in the correction value with respect to the circumferential position is constant,
Otherwise, the correction value for the circumferential position is represented by a periodic function.
A printing apparatus characterized by that. The printing apparatus according to claim 1,
The correction value is
The test pattern was read for a test pattern in which a plurality of lines along the intersecting direction were formed in the transport direction by ejecting the ink from the nozzles to the medium transported in the transport direction. From the image data, calculated based on the result of measuring the spacing in the transport direction between the plurality of lines,
When the time interval between the time when ink is ejected from the nozzle and the time when the next ink is ejected is shorter than a predetermined threshold at the timing, the correction is performed so that the time interval becomes equal to or greater than the threshold. A printing device, characterized in that the value is modified. The printing apparatus according to claim 2,
The faster the speed of transporting the medium,
The printing apparatus, wherein the threshold value is increased. The printing apparatus according to claim 2 or 3,
When measuring the interval in the conveyance direction between the lines, correct the deviation due to the conveyance error that occurs when the conveyance roller makes one rotation,
The printing apparatus, wherein the correction value is calculated based on a result of measuring an interval in the transport direction between the corrected lines. The printing apparatus according to any one of claims 1 to 4,
The difference between the correction value at the front end portion in the circumferential direction of the transport roller and the correction value at the end position in the circumferential direction is reduced.
The printing apparatus, wherein the correction value is corrected using a ratio calculated in accordance with a difference between a correction value at the position of the tip portion and a correction value at the position of the end portion. The printing apparatus according to claim 5,
The correction value at the front end position and the correction value at the end position according to the inclination of the correction value at the front end position in the circumferential direction and the inclination of the correction value at the end position in the circumferential direction. A printing apparatus characterized in that a difference between the two is calculated. The printing apparatus according to any one of claims 1 to 6,
For each period in which the transport roller makes one rotation, the correction value is calculated in association with each position in the circumferential direction,
The timing at which ink is ejected from the nozzles using a correction value obtained by averaging correction values calculated for each cycle other than the first cycle of the transport roller among the correction values calculated for each cycle of the transport roller A printing apparatus, wherein: is corrected. Conveying the medium in the conveying direction by rotating the conveying roller;
Discharging ink from a plurality of nozzles arranged in an intersecting direction intersecting the transport direction at a timing corresponding to a plurality of correction values respectively associated with positions in the circumferential direction of the transport roller;
A printing method comprising:
The correction value is
When the timing is advanced compared to the immediately preceding timing, the amount of change in the correction value with respect to the circumferential position is constant,
In other cases, the correction value for the position in the circumferential direction is expressed by a periodic function.