JPH03110513A - Method and device for multi-beam exposure - Google Patents

Abstract

PURPOSE:To record a pattern border line extremely smoothly by making a specific-pitch subscan with a multi-beam consisting of a specific number of recording beams and modulating respective recording beams in following cycles between recording beams of a 1st-cycle main scan with image data in scanning line sequence. CONSTITUTION:The subscan is made with the multi-beam MB consisting of the number of recording beams B1 - By (y) shown by an equation I for one main scanning period at pitch (l) shown by an equation II for each rotation of a rotary cylinder when, for example, a rotary scan type recording device is used. In the equations I and II, (y) is the number of recording beams constituting the multi-beam, (n) is the number of overlaps and an integer larger than 3, and (x) is an integer >= 1. In a main scan of (n) cycles of (n)-multiple exposure, respective recording beams in following cycles between recording beams of the 1st-cycle main scan are modulated with the image data of scanning lines which are changed in order to perform >=3-multiple exposure. Consequently, a drawing pattern which has an extremely smooth border line can be recorded relatively easily.

Description

[Detailed Description of the Invention] <Industrial Application Field> The present invention relates to an exposure method commonly used in image scanning recording devices such as laser plotters for manufacturing printed wiring boards, color scanners for plate making, or laser printers. ,
In particular, the present invention relates to a multi-beam exposure method and apparatus for scanning an image recording surface in parallel with a plurality of recording beams (multi-beams).

<Prior art> When performing simultaneous recording using multi-beams, if the beam spots BS are arranged and scanned so that they are adjacent to each other as shown in FIG. 21, then the adjacent portions of each beam spot are exposed. If the amount is insufficient, the density in that area may become low, resulting in so-called scanning line cracks.
As shown in FIG. 2, it is known that jitter occurs at the boundary line in the sub-scanning direction of a recorded image, resulting in a decrease in image quality.

In order to eliminate this inconvenience, a method can be considered in which a portion of the beam sponsons BS arranged in series are overlapped with each other for exposure as shown in FIG. 23, but for example, it is possible to When used, since laser light is coherent light, if adjacent beam spots are simply superimposed, light interference will occur in that area, and exposure and recording will not be performed properly. Therefore, the present applicant has proposed various methods for avoiding such interference of laser recording beams and realizing multi-beam overprinting.

The first method is to overlap the beam spots by making adjacent laser recording beams form a pair of orthogonal polarizations (Japanese Patent Application Laid-open No. 1-110959). A method of superimposing beam spots by providing the above optical path length difference (Japanese Patent Application Laid-Open No. 64-2736
No. 2) and third, as shown in Fig. 24, a plurality of beam spores) BS are arranged in a so-called staggered pattern and scanned.
There is a method (Japanese Unexamined Patent Publication No. 6427360) in which the beam spots BS are partially overlapped in the later state of the burn-in.

Furthermore, similar to the method described above, the applicant proposed the following image method as a method that allows exposure without making any special changes to the optical system when exposing parts of the beam spots BS by overlapping each other. A laser exposure method for a scanning recording device was previously proposed (Japanese Patent Application No. 63-152).
No. 901). In this laser exposure method, in successive main scans of multi-beams that are scanned in parallel, the latter part of the beam spot row that is scanned later in the main scan passes between the beam spots in the first half of the beam spot row that is scanned first. By sub-scanning the multi-beam to fill in the second
Double exposure as shown in FIG. 5 is performed.

According to the exposure method of overprinting as described above, as shown in FIG. 25, scanning line cracks do not occur in the recorded image, and relatively smooth border lines can be obtained.

<Problems to be Solved by the Invention> By the way, some drawing patterns place particular importance on the smoothness of the border lines, and for such patterns, the double exposure proposed earlier can be used. It is also possible that this is not sufficient.

Furthermore, as described above, when a light source that generates incoherent light such as an LED is used instead of a laser light source that generates coherent laser light as the light source for generating the recording beam, , since the beam spot cannot be narrowed down sufficiently, the border line will not be sufficiently smooth even with double exposure.

The present invention was made in view of the above circumstances, and avoids complication of the optical system and delicate adjustment of data output timing, and uses a relatively simple configuration to achieve triple or more multi-beam multi-beam processing. It is an object of the present invention to provide a multi-beam exposure method and apparatus that can make pattern boundaries extremely smooth by performing overprinting.

<Means and operations for solving the R problem> The first invention will be explained with reference to FIG.

The figure shows the positional relationship in the sub-scanning direction of each multi-beam at the same timing (same phase) in each main scanning, but to make it easier to understand, the positions of each are shifted in the main scanning direction. There is. In this respect, Figures 2, 3, and 8 to 1, which show the positional relationship of the multi-beams described later,
The same applies to FIG.

Now, the multi-beam MB composed of y recording beams B, ~B, is transmitted at a pitch l (however , l is the number of beam spots, ie, the number of dots, with spot pitch as a unit). Here, the sub-scan feeding may be continuous feeding or intermittent feeding.

Then, by performing n main scans, the multi-beam MB is sub-scanned by a distance represented by "nxI!".

The distance between each beam of the multi-beam MB in the first main scan is 2
At equal intervals by the beam of each main scan from the th time to the nth time,
In order to uniformly perform n-fold overprinting so as to fill in the gaps, the first beam B1 of the multi-beams MB in the rn+1''th main scan must be the same as the first beam B1 of the multi-beams MB in the first main scan. It must be in contact with the last beam B, that is, rnXl'' is expressed by the following equation (1)
need to be met.

nXj! -y・・・・・・(1)
In other words, when performing n-fold overprinting, the relationship between the sub-scanning feed pitch l and the number of beams y must be set so as to satisfy equation (1).

Suppose that a condition that does not satisfy equation (1), for example, Fig. 2 (a)
As shown in (n)l!, when a multi-beam consisting of three recording beams (y=3) is sent in the sub-scanning direction at a pitch l of r l + 1/3 J dots, (n)l! f-y n
is an integer). Then, since there is no multibeam with leading beam B, t that is in contact with the last beam B of multibeam MB of the first main scan in the subsequent main scan, each of the multibeams of each main scan Because overprinting is not performed so that the spaces between the beams are filled in at equal intervals by each of the other main scanning beams, the boundaries of the pattern drawn using this exposure method are as shown in Figure 2 [
As shown in [Yes]), there is uneven wobbling, and the effect of multiple overprinting cannot be fully exhibited.

On the other hand, in the case of triple overprinting (n = 3) and the sub-scanning feed pitch l is rl + 1/3'' dots, if the number of beams is 4 (y-4), then the above formula (+1 is satisfied). In this case, as shown in FIG. 3(a), the last beam B of the multi-beam MB of the first main scan is added to the Q beam ahead of the multi-beam M B of the fourth main scan. 4B, are in contact with each other, and each beam of the multi-beam of each main scan is uniformly overprinted by each beam of the other main scan, so the boundary line of the recorded pattern is as shown in Figure 3(b). becomes smooth.

In the above formula CI), if we set "1-x+1/m [x and m are integers]", ' [x+1/m)XnJ must be an integer, so m=n.

Therefore, sub-scan feed pinch! is the following formula (2)% formula%
(2) The above equation (1) can be rewritten as the following equation (3).

y = [x + −) X n ・
...(3) However, y is the number of recording beams constituting the multi-beam, n is the number of overlapping, and is an integer of 3 or more. X is an integer of 1 or more. Therefore, for example, uniform triple overprinting ( The number y of multi-beams that can perform 5-fold overburning (n-3) and 5-fold overburning (n=ε) is calculated by adding 1 to X in the above equation (3).
, 2, 3.4. The results are shown in a zero-order table that can be easily obtained by sequentially substituting .

Table 1 Note that when x-0, the number of beams in the multi-beam is one, which is outside the spirit of the present invention, and is therefore excluded.

8 to 1O are examples of drawing by the multi-beam exposure method according to the first invention, in which the number of overlaps in FIG. 8 is 3 and the number of beams is 4, and in FIG. 9 the number of overlaps is 3. heavy, the number of beams is 10, the number of overlaps is 5 in the first theta diagram, and the number of beams is 1
It shows the positional relationship of the multi-beams in each main scan in the case of one beam. Note that in each figure, the numbers shown within the beam spots indicate the numbers of the scanning lines corresponding to each recording beam. In addition, the beams do not overlap with each other at the beginning of exposure in the sub-site area indicated by the broken line in the figure, so
It represents a beam that is set in advance to a non-exposure state.

The output order of image data in the first invention will be described below with reference to FIG.

Figure 8(b) shows that the positional relationship of the multi-beams shown in Figure 8(a) has been rearranged into three consecutive main scans (1 to 3 tile main scans) that are triple-overprinted. This is an indication. As is clear from the figure, the main scanning of the first cycle (i.e.,
■, ■, ■, ...) between the beams, l
Main scan of the second cycle (i.e., main scan of ■, ■, ■, ...) and main scan of the third cycle (i.e., main scan of ■, ■, ■)
, .
The scanning lines are arranged sequentially in the order of 4→5→6. This also applies to other examples such as those shown in FIGS. 9 and 10.

Therefore, when performing n-fold overprinting using the multi-beam exposure method according to the first aspect of the invention, between the recording beams of the main scan of the first cycle, each recording beam of l cycles is used to print images sequentially in the scanning line. If you modulate it with data, the image will be reproduced correctly.

Next, the second invention will be explained.

As understood from the above formula (2), in the first invention, n
When performing heavy overprinting, the sub-scan feed pitch l was set so that the multi-beams in each successive main scan were shifted by 1/n dots. However, the sub-scanning feed pitch l does not necessarily have to be set as described above, and can also be set to be shifted by 2/n or 3/n, for example.

In this case, the sub-scanning feed pitch l can be expressed by the following equation (4).

However, n is the number of stacks and is an integer of 3 or more n' is an integer of 2 or more smaller than n X is an integer greater than or equal to 0.

When the multi-beams are sent in the sub-scanning direction at such a pitch l, the number y of the multi-beams that can perform uniform overprinting can be expressed by the following equation (5).

The following table shows the number of multi-beams that can perform triple overprinting (n-3) and quintuple overprinting (n=5).

(Hereinafter, blank spaces) Table 2 In Table 2 above, the numbers in parentheses indicate the number of beams of the multi-beam obtained by the first invention.

11 to 14 are examples of WJ images obtained by the multi-beam exposure point method according to the second invention, in which the number of overlaps in FIG. 11 is 3 and the number of beams is 5, and FIG. 12 shows the number of overlaps. is triple,
The number of beams is 2, the number of stacked beams in Fig. 13 is 3, the number of beams is 8, and the number of stacked beams in Fig. 14 is b (5 stacked).

It shows the positional relationship of the multi-beams in each main scan when the number of beams is nine.

The output order of image data in the second invention will be described below with reference to FIG.

The 11th @(b) is the 11th rj! in the same way as explained in FIG. The positional relationship of the multi-beams shown in J(a) is rearranged and shown in three consecutive main scans (1 to 3 tile main scans) in which triple printing is performed.

As is clear from the figure, between the beams of the first cycle of main scanning (i.e. main scanning of ■, ■, ■, ...),
Main scanning of the second pile of l (i.e. ■, ■.

(main scanning of ■, ...) and the main scanning of the third cycle (i.e., main scanning of ■, ■, ■, ...)
→3→2”, “4→6→5”, etc., and are arranged in an order different from the scanning line order. This is Figure 12~
The same applies to other examples as shown in FIG.

Therefore, when performing n-fold overprinting using the multi-beam exposure method according to the second aspect of the invention, between the recording beams of the main scan of the first cycle, the image of the subsequent cycle that fills the space between the recording beams of each of the l cycles. If you modulate it with data, the image will be reproduced correctly.

By the way, in an apparatus that performs exposure recording by processing human data in real time, the number of multi-beams to be used depends on the data processing time, that is, the complexity of the drawing pattern. If the drawing pattern is relatively simple, the data processing time is short, so the recording speed can be increased by increasing the number of beams; conversely, if the pattern is complex, the number of beams must be reduced. Therefore, in this type of multiple exposure apparatus, it is preferable to be able to finely switch the number of multiple beams.

However, as can be understood from Table 2 above, for example, in the case of triple exposure, according to the method according to the first invention, the number of beams is "4" → "7" → "10", etc. According to the second invention, all the values are discrete, such as "2" → "5" → "8"...

Incidentally, if both are used, the number of beams can be switched continuously and a practical device can be realized. Taking these points into consideration, the multi-beam exposure apparatus according to the third invention automatically selects the first or second multi-beam exposure method that is compatible with the number of overlapping beams n and the number of beams y. This allows exposure recording to be performed by selecting the The specific configuration will be explained in detail in the examples described later.

<Example> Embodiments of the present invention will be described below with reference to the drawings.

FIG. 4(a) is a block diagram showing a schematic configuration of a laser plotter which is an embodiment of the present invention.

This laser plotter has an image data creation section 1 that creates contour side vector data from CAD data of a drawing pattern.
0, a data conversion section 20 that converts the vector data into dot data for multi-beam control, and an image recording section 30 that performs exposure recording by controlling the multi-beam ON/OFF based on the dot data.

The image data creation section 10 is configured as follows.

When a desired drawing pattern is designated from the keyboard 11, the CAD data of that pattern is imported from a storage medium such as a magnetic tape 12 or a magnetic disk 13 into a minicomputer 14 serving as an image data creation means, and is converted into outline vector data. is created.

The keyboard 11 also has a function as a drawing condition setting means, and arbitrary drawing conditions including the number n of overlapping beams and the number y of beams are input and set from here. mini computer 14
also has a function as a sub-scanning feed pitch calculation means, and calculates the sub-scanning feed pitch 2 based on the set number of overlapping beams n and the number of beams y. Note that the CRT 15 is for displaying input drawing conditions and the like.

The data converter 20 is configured as follows.

The vector data created by the image data creation section 10 is
It is given to the intersection point calculation unit 21 as an intersection point calculation means. The intersection calculation unit 21 calculates the intersection position between the drawing pattern represented by vector data and each scanning line, and converts the intersection information into data (intersection address data) in a format corresponding to the address of the buffer memory 24, which will be described later. Convert and output. This intersection address data is given to a data output order switching section 22 as an output order switching means.

The data output order switching unit 22 is for switching the output order of the intersection address data according to the preset number n of overlays and the number y of beams.
As shown in the figure, a number of first-in, first-out memories (FIFOs) equal to the maximum number n of overprints by the laser recording beam
) 1 to n, and a data write control unit 22A and a data read control unit 22B that individually control writing and reading to these FIFO I to n.

As shown in FIG. 6(a), the output order control data storage section 23 as a control data storage means stores a plurality of types of control data for controlling the output order of intersection address data according to drawing conditions. It is composed of a ROM 23A and a latch circuit 23B that latches read control data. The read control data is given to the data read control section 22B.

As shown in FIG. 5, the intersection address data read from FIFO I-n is given to selectors 28A and 28B as write addresses.

The selectors 28A and 28B are connected to the buffer memory control section 25.
Either one is selected based on the switching control signal SW from , and the intersection address data or the read address given from the read counter 29 is output to the buffer memory 24 .

The buffer memory 24 as an intersection information storage means is composed of two buffer memories 24A and 24B controlled by a buffer memory control section 25, and each buffer memory 24A and 24B stores a number of lines equal to the maximum settable number of beams m. It is equipped with memories (1) to (m). Each line memory [1] to (m) has the same number of bit areas as the number of pixels included in the - scanning line, and the selectors 28A and 2
"1" or "1" input from the input data switching circuits 43A, 43B to the address given via 8B.
0'' data (intersection data) is written.

Dot data creation unit 26 as dot data creation means
is a circuit for converting the intersection point data read out from the buffer memory 24 into dot data, and the line memory (
1) - AND game corresponding to (m)) G 1 - G m
and the flip gates connected to each AND gate 01~Gm.
It is composed of flops FF1=FFm.

The modulator control unit 27 individually turns ON/OFF, for example, a multi-channel acousto-optic modulator provided in an image recording unit 30, which will be described later, based on the dot data of the required line given from the dot data creation unit 26. Outputs a signal for @il.

The image recording section 30 as an image recording means is configured as follows.

As shown in FIG. 4(a), the rotary cylinder 31 on which the photosensitive material (recording medium) F is mounted is rotated at a constant speed by a drive mechanism (not shown), and the rotation speed is detected by the rotary encoder 32. Ru. rotary encoder 3
The second detection signal is given to the timing generation circuit 42.

The timing generation circuit 42 creates an output timing signal based on this detection signal and outputs it to the buffer memory control section 25.

As shown in FIG. 4(b), the exposure head 33 includes a light source 132 for emitting laser light, an optical means 133 for dividing the laser light emitted from the light source into a plurality of laser recording beams, and each of the lasers. The recording beam is turned on individually based on the signal given from the modulator control section 27.
an optical modulator 134 such as an acousto-optic modulator that transmits 1 m, an optical system 136 that reduces and focuses a laser recording beam controlled by the optical modulator 134 onto the photosensitive material F;
including.

The exposure head 33 is screwed into a screw ring 34, and this screw ring 3
4 is rotationally driven by a motor 35, the exposure rod 33 is sent along the axis of the rotary cylinder 31 in a sub-scanning direction. The position of the exposure head 33 in the sub-scanning direction is detected by a linear encoder (not shown).

A motor control circuit 41 serving as a sub-scan feed control means drives and controls the motor 35 based on the sub-scan feed pitch l given from the minicomputer 14 of the image data creation unit IO.

Next, the operation of the above-described device will be explained.

When drawing by this apparatus, first, the number n of overlapping laser recording beams and the number y of beams are set as drawing conditions using the keyboard 11 of the image data creation unit 10.

The mini-computer 14 calculates the sub-scanning feed pitch l by substituting the set number of overlapping beams n and the number of beams y into the following equation (6).

j! = y/n (6
) The calculated sub-scan feed pitch l is output to the motor control circuit 41 for sub-scan feed control, so that the exposure head 33 moves the multi-beam by l dots per revolution of the rotary cylinder 31. - Sub-scanning is carried out at a constant speed.

On the other hand, the vector data created by the minicomputer 14 is given to the intersection calculation section 21, and intersection address data is calculated. This will be explained below with reference to FIG.

As shown in FIG. 7(a), the intersection calculation unit 21 uses vector data of a drawing pattern (indicated by a diagonal area in the figure) given from the image data creation unit 10, a scanning line (+),
(2) Sequentially calculate intersection address data with . In order to simplify the explanation, the drawing patterns in FIG. 3 +
D l g is transferred to the data output order switching section 22 . Here, if the drawing patterns overlap each other,
The data writing controller 22A performs sorting for one line and overlap processing of intersection points.Next, the data writing controller 22A performs the FIF
A write control signal Write is output to F
The intersection address data is sequentially written to the IFOI, and an acknowledge signal (Ack) is sent each time writing of each data is completed.
) is output to the intersection calculation unit 21 to inform that the next intersection data can be accepted.

When all the intersection address data of the scanning line (1) is output, the intersection calculation unit 21 outputs the last intersection address data D14.
Following, outputs line feed data. This line feed data is F
The data is written to the IFOI and is also provided to the data write control unit 22A. Thereby, the data write control unit 22A switches the write target to FIFO2.

As a result, the intersection address data D2++ D22D of the scanning line (2), which is transferred next from the intersection calculation unit 21, is
, D, 4 and the subsequent line feed data are written to FIFO2, and then the writing target is switched to FIFO3. Thereafter, by the same writing process, in the case of n-fold overprinting, each intersection address data of scanning lines (1) to ([scanning lines) is sequentially written to FIFO I to n. Fig. 7 (■)
indicates the intersection address data written to FIFO l-n in this manner.

Note that if a full signal is output to the n section 22A during data writing from the FIFO that is the target of writing, the new intersection address data is It is determined that there is no write area, and writing of the intersection address data is waited until a free area is created.

Each intersection address data of scanning lines (1) to (n) is FIF
When written to OI~n, the intersection address data of the next scanning line (1+n) is transferred to the FIF as shown in FIG. 7(C).
Written to OI. Similarly, the intersection address data of each scanning line after scanning line (2+n) is sequentially written into FIFOs 2 to n. As a result, the FTFOI has scan line (1),
(1+n).

The intersection address data of (1+2n), ... is FI
FO2 has scanning lines (2), (2+n), (2+2n).

The intersection address data of scanning lines (n), (2n), . . . are written in PIFOn, respectively.

Next, a read operation using intersection address data from PIFOI~n will be described.

As mentioned above, there is a case where the exposure method according to the first invention (the first drawing method) is used, and a case where the exposure method according to the second invention (the second drawing method) is used.
Since the reading order of the intersection address data is different depending on the case where the drawing method is used, the output order control data storage unit 2
Data output order control data corresponding to various drawing conditions is stored in advance in the ROM 23A of 3, and the control data matching the drawing conditions given is read out from the minicomputer 14 of the image data creation section 10, and the intersection address is set. I am trying to control the output order of data.

As shown in FIG.
", rFIFo number F", number address A".

The "number of spaces S" is data for setting the number of laser recording beams to be maintained in a non-exposed state during a required main scan at the beginning of exposure.

"Number of output lines L" indicates the number of laser recording beams used for exposure recording in the main scanning.

"rFIFo number F" specifies the number of the FIFO to be read out in the main scan among FIFO I-n.

"Next address A" is data specifying the lower bit of the address where control data for the next main scan is stored, and is fed back to the ROM 23A via the latch circuit 23B.

The data output operation of the first drawing method will be described below with reference to FIGS. 15 and 16. FIG. 15 is a flowchart of data output processing 1. FIG. 16 schematically shows part of the control data stored in the ROM 23A. Here, a case is taken as an example in which the drawing conditions (the drawing conditions shown in FIG. 8) are set such that the number of beams is "4" and the number of overlaps is "3".

First, the drawing conditions of the number of beams "4" and the number of overlaps "3" are given to the ROM 23A from the image data creation unit 10 as upper address data. Further, at the start of exposure, a clear signal is given from the data readout control section 22B to the latch circuit 23B (step S in FIG. 15).
1”), and “0” as the lower address A of ROM23A.
" is input.

As a result, the ROM address "rynA" is set to "430", and the first output order control data "r2211 as 5LFAJ" is read out (step S2, see FIG. 16).
Based on the latch signal from B, it is latched by the latch circuit 23B (step S3).

Data read control unit 22 that read this control data
B determines whether the number of spaces S in the control data is "rO" (step S4).

In this example, since the number of spaces is S-2, the data read control unit 22B outputs a line feed request signal NReq to the buffer memory control unit 25 (step S5).

When writing the intersection data of the first main scan, the buffer memory 1liIJ control unit 25 selects the buffer memory 24A and the input data switching circuit 43A, so the buffer memory control unit 25 that receives the line feed request signal NReq:
A line feed signal is output to the input data switching circuit 43A. As a result, the input data switching circuit 43A writes data to the line memory of the buffer memory 24A.
Move from [1] to line memory [2]. In the initial state, each bit of each line memory of the buffer memory 24 is "0", so all bits of the line memory (1) of the buffer memory 24A are maintained as rQ.

When a line feed is made from line memory [1] to [2], data readout control unit 22B from buffer memory control unit 25
A line feed end signal NAck is issued. As a result, the data read control unit 22B subtracts 1 from the number of spaces S in the control data (step S5), and the result is "0".
It is determined whether or not (step S4).

In this example, since it is S-2-1-1, the process goes to step S5 again, and the same line feed process as the first is performed, so that the data write target moves from line memory [2] of the buffer memory 24A to line memory [ 3].

When s-o is reached through the above two line feed processes, the process proceeds from step S4 to step s6, where the data read control unit 22B determines whether the number of output rices in the control data is r□. Here, the number of output lines is "
2'', the process advances to step S7.

In step S7, intersection address data is read from the FIFO indicated by the FIFO number in the control data.Here, since the FIFO number F is "1", the FIFO
By outputting a read control signal Read to the FIFOI, the intersection address data of the FIFOI is read. Then, it is determined whether the read data is a line feed bit (step S8).If we assume that the intersection address data as shown in FIG. Since the intersection address data D, which is displayed is not a line feed bit, the intersection address data D++ is transferred from the FIFOl to the selector 28A.
(Step S)
9).

When the intersection address data is transferred to the buffer memory 24A, the data read control section 22B sends a data write request signal DReq to the buffer memory control section 25. This causes the input data switching circuit 43A to
1" is set. Then, the buffer memory control section 2
Based on the write signal W from 5, the intersection data ``1'' is written to the address transferred from the selector 28A on the line memory [3] designated as the write target by the above-mentioned line feed process.

When writing of the intersection point data is completed, a data write end signal DAck is output from the buffer memory control section 25 to the data read control section 22B. As a result, the data read control unit 22B reads the next data from the FIFOI (step S7), and the same operation as described above is performed, and the second intersection address data D18 of the scanning line (+) is read out. This intersection address data t)
+z is transferred to the buffer memory 24A via the selector 28A, thereby writing the intersection data "1" into the corresponding address of the line memory [3].

Thereafter, similarly, the intersection address data is sequentially read from the FIFOI, and the intersection data rl of the scanning line (1) is written to the corresponding address on the line memory [3].

If the data read from the FTFOI is a line feed bit, the process proceeds from step S8 to step SIO, where the buffer memory control unit 25 is instructed to perform a line feed in the line memory. As a result, the intersection data is transferred from line memory [3] to line memory [4] of the buffer memory 24A. Then, 1 is subtracted from the number of output lines in the control data (step 5IO), and the process returns to step S6 to determine whether the subtraction result is "0". In this example, since it is L-2-1-1, the process proceeds to step S7, and the subsequent data written in the FIFOI, that is, the intersection address data of the fourth scanning line (4) is read out, and the above-mentioned By the same operation as above, the line memory [4
] Intersection data “1” is written to the corresponding address above.

FIG. 17 is a schematic diagram showing a state in which intersection data is written to the buffer memory 24, and FIG. 17(a) shows a state in which writing to the buffer memory 24A is completed during the first main scan described above. Note that (1) in Figure 17,
(2), . . . indicate a group of intersection data for each scanning line written in each line memory.

When two lines of intersection data are written to the buffer memory 24A, the number of output lines becomes L=O, so the process returns from step S6 to step S2 and the next control data necessary for the second main scan (■) is read out. It will be done. Then, the buffer memory control unit 25 selects the buffer memory 24B and the input data switching circuit 43B, and also selects the selector 28
Switch to output the intersection address data for B.

At this time, the lower bits of the address of ROM23A are
The next address "l" in the previous control data latched into the latch circuit 23B is input.

Therefore, control n data r1322J is read from ROM address 31J (see FIG. 16).

Based on this control data, a write process similar to that described above is performed. That is, since the number of spaces is "lJ", all bits of the line memory [1] of the buffer memory 24B are "0".
” will be maintained. Also, the number of output lines is r3J, FIF
Since the O number is "2", the intersection address data written in FIFO2 is read out in order and the line memories (2), (3), (
4) Scan to the corresponding address M(2), (5), (8
) are written respectively. Figure 17 (
b) shows a state in which writing to the buffer memory 24B has been completed.

While writing to the buffer memory 24B is being performed,
Intersection data is read from the buffer memory 24A to which writing has been completed first. Below, Figures 18 and 19
This will be explained with reference to the figures.

Note that FIG. 18 is a schematic diagram showing a state in which the intersection data of the first main scan is written into the buffer memory 24A, and FIG. 19 is a timing chart for reading the intersection data.

The buffer memory control unit 25 includes a timing generation circuit 42
An output timing signal as shown in FIG. 19(a) is input from. One cycle of this output timing signal corresponds to a pixel pitch. The buffer memory control unit 25 selects the buffer memory 24A, and also selects the selector 28
A is switched to the read counter 29 side, and a pulse signal synchronized with the output timing signal is output to the read counter 29. The read counter 29 outputs the count value of this pulse signal to the selector 28A, so that the selector 28A outputs a read address as shown in FIG. 19(b).

The buffer memory control unit 25 outputs an R/W signal as shown in FIG. 19(C) to the buffer memory 24A. When this signal is at "H" level, selector 28
The intersection data within the read address given through A is read out in sequence (see FIG. 19(d)).The reading of the intersection data is performed in parallel for each line memory of the buffer memory 24A.

Each read intersection data is stored in the dot data creation section 26.
(AND game) G (G1 to Gn). AND gate G is the buffer memory controller 25
As a result of opening the flip-flop F in accordance with the timing pulse given from
Intersection data as shown in FIG. 19(f) is input to F (FFI to FFn).

As a result, the output Q of the flip-flop FF is the 19th
As shown in the figure (comb), the data will be reversed at each point of change in the intersection data.

That is, the first intersection data "l" shown in FIG. 19(d)
is the change point where the drawing pattern changes from white to black,
Assuming that the next intersection data "1" is a change point from black to white, the [HJ level research area of the Q output of the flip-flop FF shown in FIG. Corresponds to the ON state (exposure state B).

The output Q of each flip-flop FF obtained in this way is given as dot data to the exposure radar 33 of the image recording section 30 via the modulator control section 27 at the next stage, as shown in FIG. As shown, in the first main scan (■), the laser recording beam B1°B is maintained in a non-exposed state, and beam B is the dot data of scanning line (1), and beam B is the dot data of scanning line (4). The dot data is modulated 0N/OFF respectively, and a scanning line (1),
Pixels along (4) are recorded.

Note that while the intersection data is being read from each line memory of the buffer memory 24A, the buffer memory control unit 2
5, the input data switching circuit 43A stores data such as shown in FIG. 19 (b) in each line memory.
0". As a result, after the intersection data is read out, the content of each address from which the intersection data was read becomes "0" during the "L" level period of the R/W signal shown in FIG. 19(C). can be rewritten as

When the first main scan ■ is completed, the 17th V! As shown in J(C), by outputting the intersection data written in the buffer memory 24B, the laser recording beam B of the second main scan (■) is maintained in a non-exposed state, and the beams B*, Bs and B4 are scanning lines (2) and (
5), (8147) Intersection data "i?0N/OFF modulated, pixels along scanning lines (2), (5), (8) are recorded. Then, the intersection data from the buffer memory 24B is read. While the data is being output, the control data "o433" at the given address r432.
”, the intersection address data of the third main scan (■) is read out from P I FO3 and stored in the buffer memory 24A.
The intersection data of scanning lines (3), (6), (9), and 02) are written in .

Hereinafter, according to the "next address" of the control data, ROM address address 33J → r434J → "435" → r433
”→..., and writing/reading of each intersection point data to and from the buffer memories 24A and 24B is performed alternately through cyclic addressing, thereby creating a multi-beam composed of four laser recording beams. Triple overprint exposure is performed in real time.

Note that if an empty signal E+wp is output to the data read control unit 22B from the FIFO to be read, it is determined that writing of the intersection address data to that FIFO has not been completed, and the empty signal is output. Reading of the intersection address data is waited until Emp is released.

Next, as shown in FIG. 11, the data output operation of the second drawing method is performed when the number of beams is "5" and the number of overlaps is r3. In this example as well, which will be explained by taking as an example the case where the drawing conditions are set, data is read from the FIFO and written to/read from the buffer memory 24 according to the procedure explained in FIG.

From the image data creation unit 10 to the ROM 23A, the number of beams is "5", the number of overlaps is r3. The drawing conditions of A are given as the upper address data, and the lower address A at the start of exposure is given as the upper address data.
Since is "0", ROM address 30J is specified. As a result, the output order control data "3211" stored at the address is read out (see FIG. 16). Based on the number of spaces S “3” in this control data,
As shown in FIG. 20(a), each bit of the line memory (N, (2), (3)) of the buffer memory 24A is maintained at "0".
, the scan line (1
).

The intersection address data (4) is read and written to the corresponding addresses in the line memories (4) and (5) of the buffer memory 24A.

After writing the intersection data for the required output lines,
Address r5 specified by the next address of the previous control data
From 31J, second main scan ■ control data r2332J
is read out. The number of spaces S for this control π data is “
2", so as shown in FIG. Therefore, in the second main scan (■), the intersection address data of scanning lines (3), (61, (91) stored in FIFO3 is read out, and the intersection address data of the line memory (3),
(4).

Intersection data "1" is written to the address corresponding to [5].

The reason why the intersection address data of FIFO 3 is read out first by jumping over FIFO 2 is that under these drawing conditions, as shown in FIG. 11, the drawing order is reversed between the second cycle main scan and the third main scan. Therefore, it is necessary to change the output order of the intersection address data.

In parallel with the writing to the buffer memory 24B, the intersection data is read from the buffer memory 24A, and each pixel along the scanning lines (1) and (4) is recorded.

In the third main scan (2), the ROM address "532J control data r0523" is read out.

With this control n data, the intersection address data of FIFO 2 is read out, and as shown in FIG. 20(C), the scanning line (2 ), (5), (8),
(11).

04 intersection data are written.

Below, according to the "next address" of the control data, ROM7
Fres r533"-+r534"-+r535"→r5
33J→..., intersection point data is written/read out alternately to and from the buffer memories 24A, 24B, and as a result, a multi-beam consisting of five laser recording beams is used. Triple overlaying is done in real time.

Note that the present invention can also be modified as follows.

(+) In the embodiments, explanations have been given taking as an example the case where triple printing is performed using a multi-beam consisting of four or five laser recording beams.
It goes without saying that the method can be applied to triple or more overprinting using a multi-beam consisting of more than one laser recording beam.

(2) In addition, although the embodiments have been explained using a rotary scanning type laser plotter as an example, the present invention can also be applied to a flat scanning type laser plotter or other image scanning recording devices such as a color scanner for plate making or a laser printer. can also be applied.

(3) In the embodiment, the case where the recording beam is a laser beam has been described as an example, but the present invention can also be applied to other light sources such as LEDs.

(4) In the above-mentioned embodiments, only the case where recording is performed by superimposing the beams on each other has been described, but depending on the required image quality, it is possible to use conventional recording without superimposition or the superposition of the present invention. It is also possible to selectively switch between the records to be subjected to waging.

<Effects of the Invention> As is clear from the above description, according to the multi-beam exposure methods according to the first and second inventions, special optical (In addition, there is no need to make delicate adjustments to the data output timing such as when overprinting is performed using a multi-beam with multiple beam spots arranged in a staggered manner.) By the above-described overprinting, a drawing pattern with extremely smooth border lines can be recorded relatively easily, and is particularly effective when the beam diameter is large.

Further, according to the multi-beam exposure apparatus according to the third aspect of the invention, the number of overlapping beams can be arbitrarily specified according to the degree of smoothness of the required boundary line, and the number of overlapping beams can be arbitrarily specified, and It is possible to realize a practical multi-beam exposure apparatus that can finely adjust the number of beams depending on the complexity of the drawing pattern.

[Brief explanation of drawings]

FIG. 1 is an explanatory diagram of a multi-beam exposure method according to the present invention,
Figure 2 is an explanatory diagram of an example of undesirable multiple overprinting;
The figure is an explanatory diagram of a multi-beam exposure method according to the first invention. 4 to 20 relate to embodiments of the device of the present invention, FIG. 4(a) is a block diagram showing a schematic configuration of the device, and FIG.
b) is a perspective view showing an overview of the exposure head, FIG. 5 is a block diagram showing details of the data conversion section, FIG. 6 is a block diagram showing details of the output order control data storage section, and FIG. An explanatory diagram of the operation of writing intersection address data to FIFO, FIGS. 8 to 10 are explanatory diagrams of drawing examples by the multi-beam exposure method according to the first invention, and FIGS. 11 to 14 are diagrams according to the second invention. FIG. 15 is a flowchart showing a procedure for reading out intersection address data; FIG. 16 is an explanatory diagram of control data stored in an output order control data storage unit; 1st
FIG. 7 is an explanatory diagram of the write/output operation of the buffer memory according to the drawing example shown in FIG. 8, FIG. 18 is an explanatory diagram of the intersection data written to the Hahanofa memory, and FIG. 19 is a timing diagram of the read operation of the intersection data. , Figure 20 is the 11th
FIG. 6 is an explanatory diagram of write/read operations of the buffer memory according to the illustrated drawing example. Figures 21 to 25 are explanatory diagrams of conventional examples. Figure 21 is an explanatory diagram of a general non-overlapping multi-beam, and Figure 22 is an explanatory diagram of a recorded image exposed by a non-overlapping multi-beam. Figure 23 is an explanatory diagram of the multi-beam with overlap, the second
Figure 4 is an explanatory diagram of multi-beams arranged in a staggered pattern.
FIG. 5 is an explanatory diagram of a recorded image obtained by double overprinting. 10... Image data creation unit 11... Keyboard (drawing condition setting means) 14...
Mini computer (image data creation means, sub-scanning feed pitch calculation means) 20...Data conversion section 21...Intersection point calculation section (intersection point calculation means) 22...Data output order switching section (output order switching means) 23 ...Output order control data storage unit (control data storage means) 24...Buffer memory (intersection information storage means) 25.
...Buffer memory control unit 26...Dot data creation unit (dot data creation means)

Claims (3)

[Claims] (1) A multi-beam consisting of a plurality of independently controlled recording beams is irradiated onto the image recording surface to form a series of beam spots, and the multi-beam is aligned in the spot arrangement direction with respect to the image recording surface. Relative scan in the cross direction (
In a multi-beam exposure method in which exposure recording is performed on a recording medium by relative scanning (main scanning) and relative scanning (sub-scanning) in the spot arrangement direction, the multi-beam is )]×n However, n is the number of overlapping beam spots, and x is an integer of 3 or more. /n=x+(1/n) In the n-cycle main scan in which sub-scan feeding is performed at a pitch l (unit: number of dots) and n-fold overprinting is performed, the recording of the first main scan cycle is A multi-beam exposure method characterized by performing overprinting three times or more by modulating each recording beam in subsequent cycles that fills in the gaps between the beams with image data in sequential order of scanning lines. (2) A multi-beam consisting of a plurality of independently controlled recording beams is irradiated onto the image recording surface to form a series of beam spots, and the multi-beam is aligned in the spot arrangement direction with respect to the image recording surface. Relative scan in the cross direction (
In a multi-beam exposure method in which exposure recording is performed on a recording medium by relative scanning (main scanning) and relative scanning (sub-scanning) in the spot arrangement direction, the multi-beam is n)]×n where n is the number of overlapping beam spots, n' is an integer of 3 or more, and x is an integer of 2 or more smaller than n, which is composed of y recording beams given by an integer of 0 or more. This multi-beam is sent in the sub-scanning direction at a pitch l [unit: number of dots] given by the following formula l=y/n=x+(n'/n), and n-fold overprinting is performed n. In the main scanning cycle, overprinting is performed three times or more by modulating each recording beam of the subsequent cycle, which fills in the space between the recording beams of the main scanning of the l-th cycle, with the image data of the scanning line whose order has been changed. A multi-beam exposure method characterized by: (3) A multi-beam consisting of a plurality of independently controlled recording beams is irradiated onto the image recording surface to form a series of beam spots, and the multi-beam is aligned in the spot arrangement direction with respect to the image recording surface. Relative scan in the cross direction (
A multi-beam exposure device that performs three or more overprints on a recording medium by performing relative scanning (main scanning) and relative scanning (sub-scanning) in the spot arrangement direction, and is optional, including the number of overlapping beams n and the number of beams y. a drawing condition setting means for setting drawing conditions; and a sub-scanning feed pitch l [unit: number of dots] given by the following formula l=y/n based on the set drawing conditions.
sub-scanning feed pitch calculation means for calculating sub-scanning feed pitch, sub-scanning feed control means for performing sub-scanning feed at the calculated pitch l, and scanning output order of intersection point information of image data according to the set drawing condition. control data storage means storing control data for controlling each line; output order switching means for switching the output order of the intersection information based on the control data given from the control data storage means; and the output order The intersection information output from the switching means is
An intersection point information storage means that is stored in association with each recording beam making up the multi-beam, and dot data for ON/OFF modulation of each recording beam is created based on the intersection information stored in the intersection information storage means. A multi-beam exposure apparatus comprising: dot data creation means for performing exposure recording by ON/OFF modulating each recording beam constituting the multi-beam based on the dot data.