JPH0852899A - High-resolution printing method and space light modulator array - Google Patents

Abstract

PURPOSE: To achieve a gray scale within the restriction of a line time and to remove the stepwise artificial effect in a print image by resetting a cell in a space light modulator when the cell is loaded with data and using a method of either one of timing delay, horizontal offset and pixels different in dimension. CONSTITUTION: By at least one method among (1) a method for temporally delaying image data so as to finely control a print image in a vertical direction, (2) a method for offsetting a pixel image in a horizontal direction so as to finely control the print image in the horizontal direction and (3) a method using a pixel set to a partial dimension so as to finely control the solid coating of the print image, data is processed and analyzed so as to change a final image. Therefore, printing of high image quality is enabled by using a DMD space light modulator consisting of torsion hinges 5a, 5b, posts 2a, 2b, a high reflectivity mirror 1 and address designating electrodes 8a, 8b.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to printing technology,
More particularly, it relates to printing techniques with improved image quality and gray scale.

[0002]

Digital copiers and printers have inherent problems in reproducing high resolution images and gray scales. Desktop printers are generally 300 dpi
It has only (dots per inch) resolution, and is printed with slightly oversized pixels to allow overlap to determine the actual printer resolution. These constraints appear as artifacts due to jagged edges on the text and aliasing on the graphic because at the nominal resolution, individual pixels cannot accurately reproduce curved lines and diagonal lines. .

The alternate (alternate) dot operating method is
Commercially it has drawbacks and cannot reproduce grayscale levels. Instead, this method changes the shape, size or position of the digital black pixels by controlling the laser power and timing. By shortening the dot exposure time and delaying the exposure time, the spot becomes smaller and elliptical, which allows the dot to be positioned within the boundaries of a standard pixel cell. Further, the spot can be moved across the width of the pixel cell in the direction of the raster scan motion. Reducing the power level in the raster allows the spots to be flattened in the print process direction, which is orthogonal to the scan motion direction. Generally, the scan direction is horizontal on the page and the print process direction is vertical.
Smoothing the edges can be achieved by the proper placement of appropriately distorted pixels, a method commonly referred to as resolution enhancement.

Laser printers can easily achieve resolutions higher than 300 dpi by using more complex controllers and printer process subsystems. For example, the printer page description language can represent a document or image on a digital printer at its full resolution. However, the burden on the printer's memory, microprocessor and printer device and optical scanner's ability to support higher resolutions generally increases with the square of the linear resolution. The system used for applications requiring higher resolution is 1200 to 2500 dp
i can be operated by i, but a general desktop type 300d
It is more expensive in proportion to its resolution than a pi printer. Increasing the linear resolution has the advantage that the binary printer can simulate a grayscale image by a process called halftone.

Xerographic processes, such as those used in printers, copiers and facsimiles, are binary in nature, making it difficult to obtain varying shades of gray. The development process, which deposits charged toner particles on the exposed latent image on the photoreceptor, operates by nature as if it were digital (ie, the method is a very high contrast analog process). Therefore, there is a need to use a higher resolution binary xerographic system that allows the gray scale to be simulated in a so-called halftone manner. Smaller pixels are increasingly populated to form larger pixels or halftone cells.
This allows the number and placement of black and white elements to be varied, resulting in a visually gray halftone cell.

The accuracy and computational power to generate such cells is higher than that for binary desktop laser printers and is a limiting factor in laser printer capabilities in achieving gray scale. Due to the inherent complexity of the halftone process, gray scale has recently become necessary in printers that are designed to reproduce scene images from a photography source or computer display outside the perceived higher resolution. It has become a characteristic.

In the latter case, the computer display can take advantage of the integral response of the human eye to change the gray scale or intensity of the image over a wide dynamic range, and the pixel is of a predetermined time within the display time of the frame that the eye integrates. It can be turned off during a period of time so that continuous tone intensity gradations can be recognized. Computer displays frequently produce images for printing, so the image produced on the display is incompatible with the ability of digital binary printers to print this image. This is also true for digitally scanned continuous tone photography images.

[0008]

In summary, printing with binary digital printers is limited. There are problems reproducing characters at lower resolutions without distortion, loss of detail, sampling artifacts or position errors. As a result of grayscale simulations using higher resolution printers, it is acceptable to eliminate grayscale and visual artifacts.

[0009]

SUMMARY OF THE INVENTION The invention disclosed herein enables both resolution enhancement and multi-grayscale levels by manipulating pixel size and fitting within standard print parameters. It consists of a printing system. These capabilities result from the ability to perform sub-pixel modulation using a spatial light modulator that includes a range of user selectable pixel element dimensions.

The advantage of this system is that it combines more gray shades and higher resolution than conventional binary printers using the halftone method. Another advantage of the system is that it requires less memory at the same resolution and grayscale than the higher resolution halftone printers currently available and uses no dither algorithm to simulate grayscale. Is also good. Both factors have led to improved system efficiency, reduced cost and improved print image quality.

For a more complete understanding of the present invention and its advantages, the following detailed description is set forth with reference to the accompanying drawings.

[0012]

DETAILED DESCRIPTION Digital copiers and printers for all types of media have inherent problems in acceptably reproducing both high resolution images and gray scales. The performance of laser printers has improved the print quality by the pixel size compared to dot matrix printers, and has replaced the office typewriter as the standard device for print quality. However, typical desktop laser printers are still found to be limited in both resolution and grayscale capability. If you look carefully at the letters, the edges will be jagged and the graphic image will often look rough, because at the nominal resolution, individual pixels cannot reproduce curves and diagonals well.

Standard desktop laser printers and many resolution-enhanced systems have 300 dpi (1
Dots per inch) resolution and slightly larger (overlap) to determine the printer's basic resolution
It has only pixels. Similar to this, display systems usually have larger dots than the addressing grid to help in image overlap and smoothing. However, while displays can modulate the intensity of individual displays, digital xerographic printers cannot.

However, for both the printer and the display, the human eye's ability to resolve minimal size features and white-to-black transients exceeds the resolution capability of display systems. The main difference between a display and a digital printer is that the display can vary the gray level or intensity over a wide dynamic range within a given pixel element, whereas the digital printer only produces black and white (binary) spots. Computer displays often produce images for printing, so the image produced on the display and the ability of a digital (binary) printer to reproduce it are inherently incompatible. This also applies to continuous tone (contone) images produced by photography. The printer must map the image to the higher resolution system and then use the halftone method to simulate gray scale.

Laser printers can easily achieve resolutions higher than 300 dpi at the cost of increased controller and printer engine subsystem complexity.
You can reduce the appearance of hardcopy text and graphics to almost invisible elements. As a standard, a suitable analog (bright lens) electrophotographic copying system exceeds the eye's ability to detect distortion6.
Features corresponding to a resolution of 00 dpi can be decomposed and reproduced. The printer page description language, which is device independent, can display a sentence or image to a digital printer at the printer's own limiting resolution. However, for example at 600 dpi, the image contains four times as many pixels to manipulate, rasterize, and image. To support such high resolutions, the burden on the printer's memory and microprocessor and the required capabilities of the printer device and optical scanner also generally increase with the square of the linear resolution. Device reliability and consumables (eg toner and paper) quality are also major limiting factors.

The system used for typesetting, plate cylinder manufacturing, film processing, and applications requiring high resolution is 120.
It operates at a resolution of 0 to 2500 dpi. Therefore, these devices are usually more expensive than typical 300 dpi printers.
Larger, slower, and more expensive to buy and maintain. Increasing the linear resolution has the advantage that the digital printer can simulate a grayscale or photographic image by a method called the halftone method, at the cost of grayscale generation in the image.

Grayscale images where the image area is not solid black or solid white require the ability to accurately simulate the desired image with various shadings, which is a fundamental problem in laser xerographic printers. ing. Part of the problem stems from the nature of the process itself. In xerographic printers, copiers, plain paper fax machines, etc., the latent image is tonered, ie, developed, by exposing the photosensitive medium (from the original by lens or electronic means) to light and then developing the toner. Printing is performed by transferring onto paper by electric means. In this case, the charged toner moves to an exposed portion of the photoreceptor on which a latent image is formed or to an unexposed portion thereof, which depends on whether the development is positive or negative.

The typical exposure-development process is digital rather than analog in operation and, as a result, is extremely high-contrast in attracting a variable portion of toner to the photoreceptor at each exposed pixel position. Have various problems. Toner particle size is also a factor in the process of obtaining a low noise grayscale image. This problem also exists with similar settings for various types of xerographic printers and fax machines in copiers. Controlling the exposure process, the sensitivity of the photoreceptor and the development process so that the pixels can be reproduced in smooth and accurate gray levels is a very demanding task.

As a result, high resolution binary xerographic systems rely on a so-called halftone method to simulate gray scale for reproduction of continuous tone images. A macropixel consisting of an array of minimal printer pixels is generated by a variable number of black and white addressable elements to simulate a desired level of gray. The disadvantage is that the linear resolution, the available levels of gray scale and the computational requirements for generating halftone cells during printing must be compromised. Since cells can be filled in many different ways, artifacts and pixel-to-pixel interactions also occur.

For example, a macropixel with m subpixel elements can generally generate (m + 1) gray levels, so a 2x2 pixel of 4 elements will have a black and white level and three intermediate levels. A total of 5 levels can be generated. Adjacent cells can interact to produce unwanted artifacts, such as edge effects or roughness, depending on the solid fill sequence. Other solid patterns can produce symmetrical screen patterns that are visible to the human eye.

As a direct result of forming such a halftone cell, the linear resolution is lost by a factor of two, and for only 5 levels to produce a valid gray scale for a 2x2 cell, the overall Inappropriate. Eye calculations suggest that there is a certain relationship between linear resolution, so-called screen size, and gray scale for good quality prints.
Screen with a minimum of 64 gray levels and 100 lines, preferably 128 gray levels and 200
A screen of lines is desired. For a 1200 dpi laser printer, the compromise yields about 100 gray levels on a 120 line screen, which gives good performance only for low end applications. In commercial typesetting systems, to obtain a gray scale for photo reproduction,
It uses a resolution of at least 2500 dpi.

To improve text and line graphics, varying the size or intensity of individual pixels is also effective, but typical laser printers lack the capability. Another solution that has recently emerged in the laser printer industry is a simulation process called resolution enhancement RET. This process also does not produce gray level pixels. This process changes the shape, size and position of digital pixels by controlling the laser power and timing,
By shortening the dot exposure time and delaying it,
It allows the positioning of smaller, elliptical spots on full-size pixel boundaries. The spot can be moved across the width of the pixel in the direction of the laser scanning movement, this process effectively increasing the resolution in the scanning direction. The spot can be flattened by lowering the laser power level. That is, the size can be reduced in the print processing direction orthogonal to the scan axis. Moving around the bits of a pixel enhances the appearance of the character and smooths jagged features.

The problems of common printing methods can be summarized as follows. Characters must be faithfully reproduced in the human eye without distortion, blurring or loss of detail, sampling artifacts or position errors.
The grayscale construction must produce an acceptable artifact-free combination of gray levels and linear resolution from the formation of halftone cells. Furthermore, the system approach must be reliable and able to reproduce consistent results day by day.

In addition to the widely adopted laser polygon scan printer systems, xerographic printers and film exposure systems that use various forms of spatial light modulators (SLMs) and light emitting arrays such as light emitting diode (LED) arrays. Is being developed. SLM
Although desirable in terms of cost, size and manufacturing, it is generally a digital system with one or more limitations in terms of the printing process as described above. Examples of such SLMs are liquid crystal devices (LCDs), electro-optical crystals, magneto-optical cells and digital micromirror devices (DMDs), also known as deformable mirror devices.

Many of these modulators consist of an array of cells that can be addressed to turn on and off by changing the transmission state. Reflective modulators, such as DMDs,
It is generally considered to be in the on state when the cell deflects light towards the imager lens and then towards the imaging surface, either the view screen or the photosensitive medium. Some of the other arrays absorb incident light in the off state and transmit light in the on state. Also, some have a limited spectral bandwidth. In some cases, the architecture has crosstalk issues between pixels. While any hybrid technology, such as electro-optical technology, generally has manufacturing problems, DMDs do not have any of the above problems.

DMD SLM Devices Digital micromirror devices (DMDs) are manufactured using a monolithic silicon metal oxide semiconductor (MOS) process. The substrate includes addressing circuits and logic circuits that capture digital data input signals and route these input signals to a memory cell array manufactured in the shape of the desired SLM integrated circuit (IC). An array of fine (17 micron x 17 micron) metal mirror elements is fabricated on top of the addressing circuitry, which is rotated about an axis in the plane of the array to allow the addressing circuitry below Can respond (Fig. 1
a). The typical rotation angle is ± 10 degrees, and the response time is 10 to 19 μsec. The array can be square in nature, as in displays, or elongated for printing. An example of the former is 640x480 pixels and an example of the latter is 2500x16 pixels.

In operation, the DMD must be illuminated by an external light source which is a passive reflective SLM. See US Pat. No. 5,101, March 31, 1992 for placement of optics to form an actual SLM system for printing.
No. 236, "Light Energy Control System and Method of Operation". The same optical principle is used for DM in the form of surface array.
It can also be applied to a display system using D.

The modulator cells for the display are normally on and are repeatedly turned off for a whole on-time which is a fraction of the frame time and a fraction of the frame time is on. It remains on and is then switched off. By any of the three methods of reducing the overall light level by the pulse width modulation (PWM) method,
The display can simulate gray scale by taking advantage of the integral characteristics of the human eye. Of course, gray scale is a general condition for intensity that can also be applied to color displays, and the gray scale technique by PWM is the least significant bit (LSB) to the most significant bit (M
SB) and is typically displayed in fractions of a binary scale (eg, 1/2, 1/4, etc.) 8
By dividing each color field into individual binary segments, the shade of color can be changed.
The LSB has the shortest on-time and the MSB has the longest on-time. With the proper combination of eight binary segments, 256 gray levels can be generated for each color, giving a total of 16 million colors.

A planar array display reimages each DMD pixel at a fixed point in the display plane (60
Since the time for a (Hz) display frame is 5.7 ms for each field of color, grayscale of color can be very easily achieved using 8-bit PWM. A problem for printing is the relatively short raster line time available to perform PWM. In order for a 300 dpi printer to be able to produce reasonable output in pages per minute, all lines must be able to print in less than a millisecond. Achieving the above gray scales for displays with frame times as long as 16 ms is impractical. This limits the number of times the raster line can be effectively turned on and off at print speeds faster than 15 pages per minute. Therefore, only a limited gray scale range is possible.

FIG. 1a shows a pixel of a DMD spatial light modulator.
An example of an element is shown. Bistable DMDs are generally
On the posts 2a and 2b by the flanges 5a and 5b.
Provided on both sides of the rotation axis of the supported highly reflective mirror 1.
Two addressed electrodes 8a and 8b.
In addition, the DMD cell is designed to prevent the risk of contact welding.
Landing electrode 3 held at the same voltage as the Raler element 1
And 4. The addressing electrodes are
0 volt in response to the input signal from the specified logic circuit
Alternating between 5 volts, so the mirror is electrostatic
Rotates in response to suction force. "Spatial light modulator and its
U.S. Pat.
These devices are described in more detail. Actuation
In the middle, the individual elements are responsive to either electrode 8a or 8b.
In response, rotate around the hinge, and landing gear and mirror
-Deflect by ± 10 degrees until the edges touch. electrode
8a and 8b are for rotating the mirror by a predetermined angle.
And these electrodes are Φ_AAnd Φ _BOften referred to as an electrode
Yes. The mirror element is on the edge (0.0003 mm2)
The nominal thickness is 17 microns, but its size and shape
The pitch and pitch vary depending on the design, such as the landing angle θ.
Can be obtained.

FIG. 1b details the timing and voltages required to operate the DMD element. This figure shows the control function bias 10 and the address bias 14 in the rotating state of the mirror line 18 when the cross-hatched surface shows the mirror on state (or the off state after t ₃ ).
16 and 16 show the action of the reset pulse train (t _R ).

In the sequence of FIG. 1b, the address voltages 14 and 16 (+5 volts) of the address electrodes corresponding to 8a and 8b are turned on at t ₀ (on by way of the respective memory cells in the silicon substrate 6 in FIG. 1a). To the off state). The mirror remains latched in its previous state due to the pulling action of the bias voltage 10 (-10 volts) applied in parallel to all mirror elements. At t ₀ , there is a negative bias, but no reset sequence 12. After the address voltage goes true, the reset sequence 12 begins at t ₁ and continues for several cycles until t ₂ . During this period, the bias voltage V _bias 10 is off. The reset pulse train is tuned to the plate resonance of the particular DMD mirror architecture, electrically pumping mechanical energy into the mirror pixel between t ₁ and t ₂ (2 μsec). This is typically a train of 5 pulses of -24 volts.

At t ₂ , both the bias voltage and the reset voltage are zero and the reset mirror is free to rotate from the deflected state 18 to the flat state (slightly before t ₃ ) to the new address voltages Φ _B and t ₃ . By the cooperative pulling force of the bias voltage 10 (−10 V) re-applied by
At t _{4, the} reverse deflection state 19 is set. The absolute value of the negative bias and the positive address voltage combine to 15 volts, which is sufficient to fully rotate the mirror element to the opposite landing (-10 degrees) condition. The bias must be applied exactly after the mirror has physically completed the angular offset as a result of the new address voltage condition, ie at t ₃ . The flight time t ₂ -t ₃ of the mirror is about 8-10 μsec and is described in US Pat.
The reset sequence is described in more detail in No. 6,279.

One of the problems with this sequence is the time it takes to reset the cell. When the reset cycle is complete, the DMD generally returns to its undeflected state and then waits for the application of a bias voltage that allows it to move to the new state. In the above example, the data causes a state change, but if the same reset and return to the neutral state occur, then the state change is unnecessary, so that
In addition, the off time is low in efficiency.

FIG. 1c actually illustrates the above pulse width modulation (PWM) method for achieving gray scale with DMD. Corresponding to 4 bits of gray scale in the line time, that is, frame time in the case of display,
In this example, the DMD cycle is turned on and off four times. The reason for this is that each on-segment is half the duration of the previous segment. Therefore, the on time 18a is the most significant bit, and the on time 18d is the least significant bit (LSB). The combination of these binary pulse widths simulates 2 4 or 16 different gray levels, corresponding to each portion of the total line (ie frame) time that a given pixel element is on. it can. Note that each reset time t _{R on} line 10 is subtracted from the allowable on-time on line 18 of the mirror. The reset cycle t _R has been expanded to show details, but is actually L
It is on the order of SB.

In the case of a display, each pixel image is spatially fixed on the screen, the binary PW of FIG.
M integrates the light intensity as a fixed spot as perceived by the observer's eyes. However, for printing methods where the photosensitive medium or organic photoreceptor (OPC) moves relative to the spatially fixed image of the exposure DMD, unwanted artifacts can occur. As a result, the OPC movement is P
The spatial separation of the components of the WM signal results in an effective mapping of the pulses from the time domain to the spatial domain according to the relation:

X (position) = V (OPC speed) * t (seconds)
Where t is the worst case time difference between MSB and LSB, for example, and X defines the combined physical separation of the imaged spots.

The resulting binary PWM pattern is 1
Pair sizes of 5 to 10 lines per mm give particularly unpleasant artifacts to the human eye at a given spatial frequency. This spacing is very close to the method of a 300 dpi printer, so the spatial separation from PWM is readily apparent. Two very even gray levels, eg MS
B and MSB-1 will be very spatially different because all of the low order bits must be combined to generate level MSB-1. Obscure changes in gray level can be eliminated due to the substantial spatial difference between the two pixel patterns. Therefore, it is preferable to print using gray scale by PWM using some combination of patterns that are not seen in this effect rather than binary. Simple splitting of the pixels into evenly spaced on-off bars in a linear pattern can improve gray level smoothness, but is too complex to perform beyond a few levels of gray.

In particular, subdividing a relatively slow dot line of 1 millisecond so that 16 gray levels occur would require 16 pixel changes per line, but 4 bit 16 level In the binary example, only 16 is needed. Each line pair corresponds to a total acceptable time of only 62 μs. The reset address time is 20 μs,
Since two address times are required for each line pair,
Almost all the time is used to address the DMD, and the exposure efficiency drops to only 30%. In either case, a line time of 1 millisecond and only 16 gray levels for printing are of limited interest. Therefore, there is a need for an improved technique for producing 120 gray levels in a line time corresponding to a printing process of 50 pages per second, ie 300 μs, without artifacts at the pixel level.
One possibility arises if a method of selectively readdressing other DMD elements while latching one DMD element can be implemented.

Operation of DMD Deflection of the DMD is caused by an electrostatic attraction force between the mirror element and the lower substrate, especially the address electrode. The attractive force is proportional to the square of the magnitude of the potential difference and inversely proportional to the square of the air gap between the mirror and the substrate. The recovery force due to the torsion hinge is linearly proportional to the spring constant k and the torsion angle θ, and at a certain percentage of the maximum (landing) deflection angle, the secondary force of the suction force becomes larger than the linear recovery torque, Simultaneously drops into the electrostatic potential well until the tip contacts the landing electrode, which is producing an equilibrium reaction force. The pixel stops at a precise angle determined by the air gap and the size of the pixel from the tip to the axis of rotation.

The potential difference consists of two components. The first component is a positive address voltage of nominally zero or 5 volts,
The second component is the negative bias voltage applied directly to the mirror structure. In the case of an ideal mirror structure without addressing, the symmetry of the attractive forces on each side of the torsion hinge prevents the mirror from rotating solely by the application of a bias voltage. Actually address electrode Φ _A or Φ _B
Applying a +5 volt address signal on either phase of the mirror causes the mirror to tilt in that direction by a small percentage of the maximum allowable deflection. Subsequent application of a negative bias voltage acts to increase the amount of deflection until collapse to full rotation occurs. For the device geometries described in the cited patents, a 5 volt address combined with a -10 volt bias to generate a total of 15 volts produces sufficient suction to make a full angular displacement. The average voltage that causes full rotation, called the collapse voltage, is nominally 12 volts, so that deflection can in principle be achieved with an address of +2 volts and a bias voltage of -10 volts. It is necessary to increase the address margin by an additional 3 volts to ensure full deflection of all pixels over a range of operating conditions and to accommodate device changes over time.

If the pixel state must be updated, the first step is to turn off the bias voltage. Since the address voltage is insufficient to hold the pixel in a fully rotated state against the torsion spring, it is ideal to release all pixels and return to a near flat position.
In practice, a controlled -24 volt preset pulse is used to electrically "pull" a pixel to store mechanical energy in the pixel at resonance, helping release and return to the flat state. Has become. The mirror is flat (θ =
When equilibrium is reached on either side of the 0) state, the bias is reapplied, thus causing the pixel to rotate again to ± θ _max .

The ability to finely control the relationship between the generation of bias, reset, address, and hold voltages and their amplitudes allows the selective reset and rotation of mirrors to change data. The mirror whose data does not change remains fixed throughout the reset process. A fully rotated DMD pixel can latch and hold data even when a reset signal is generated.

It is preferable to avoid the delays associated with returning to the flat state (θ = 0) and then re-rotating to the initial state. One such method alters the current reset process that returns V _bias to zero during and after reset. The residual reset or bias voltage is subtracted directly from the margin of the address voltage, the addressing of 5 volts is required V _bias = 0, it is necessary voltage related thereto for 3 volt address margin . Since the pixel cannot rotate by θ = 0 and therefore cannot respond to alternating address conditions, reliable operation with residual bias voltage is not possible during the period when the pixel returns to θ = 0. . Leaving 5 volt addressing on the address electrodes only
It is noted that it is not enough to latch the pixel after reset.

5 according to the current CMOS address structure design
It is possible to have addressing at voltages between 10 and 10 volts. If Vaddress were increased to 6.5 volts, for example, _Vbias could be reduced to -8.5 volts for a constant value of the address voltage plus the bias voltage of 15 volts. More importantly, V _bias can be maintained at -1.5 volts over the reset process, allowing for +3 volts of true address voltage margin. It has a residual bias voltage (-1.5 volts) and a minimum threshold address voltage (+2.0 volts).
Volt) is subtracted to reach the desired address voltage margin (3 volts).

The address voltage is increased and each increased voltage at the address can be balanced with a corresponding increase in the residual V _bias (ie, the bias voltage that occurs immediately after the reset cycle during the flight time of θ _max to θ = 0 of the pixel). As a result, the magnitude of the holding potential increases by 2 volts for every volt at the address. When the combined voltage reaches the holding voltage of about 10 volts, the pixel is latched down and does not return to θ = 0 after reset. FIG.
d illustrates this situation corresponding to FIG. 1b.

The reset 12a allows the consecutively addressed pixels to be momentarily free, but the applied voltage (1
0 volt) 15a before returning the pixel to θ _max ,
angle or twice certain of theta _{ma x} = 10 degrees, only rotates in the opposite direction. There is no loss of light efficiency or exposure time as indicated by a slight drop in light output 11. In order to switch to another rotation state, Vaddress is applied to the counter electrode 16, and the pixel has θ =
Only 0 is recovered, and regardless of the residual holding voltage 15b,
You can return to the newly addressed state. The magnitude of arrows 15a and 15b does not correspond to the attraction potential of 15 and 2.5 volts, respectively. The reduced bias voltage setting, ie -2.5 volts in Figure 1d, is reset to full V to ensure latching of the selected pixel.
It is only maintained for 10 or 15 microseconds before applying _bias (which causes the reset pixel to rotate completely). Similarly, the increased address voltage is
In order to prevent stress from being applied to the S circuit, it is only necessary to apply it instantaneously.

The flexibility of the DMD hinge provides another variation that is believed to constitute a reliable mode of operation of the type described above. The reason is that more flexible hinges require lower bias voltages and lower latching potentials.

This method is used for individual DM for rewrite.
The selection of the D cell requires another feature of the randomly addressable CMOS structure. DMD CMO
S addressing arrays are typically SRAM or DRA
Since it is an M memory array, the configuration is not difficult. Word and bit record and addressing features for accomplishing this are well known to those skilled in the art. Since hardcopy DMD devices are typically configured as very long arrays with fairly shallow columns, xy addressing (eg, 16-128 bits) is not a complication.

This configuration has the advantage of simplifying the addressing bandwidth problem and shortening the DMD operating cycle when only a small portion of the data array is deflected. For printing, it has the additional advantage of overcoming the artifacts associated with binary PWM, the resolution limit of linear PWM.

As an example of the many ways to utilize these latching and DMD reset characteristics, one of the four pixels used to emulate a gray scale
Consider groups. Consider four adjacent pixels showing four different exposure sequences, as shown on line 18 in FIG. 1b. At the start of the line all four pixels are on. The first pixel shows the most significant bit of group 18a, which means that it remains above the longest one. In the first pulse on reset line 12, the electrode state for the three higher order bits 18a, 18b and 18c has not changed, but the lower pixel 18d is switched off. The next bit is switched off with the second pulse and the third bit is switched off with the third pulse.
Finally, the last bit, i.e. the pixel in question, receives new data corresponding to the next printed line along with updates of other pixels, and the process begins again.

Thus, the time overhead of achieving gray scale with pulse width modulation is significantly improved. This allows the line time of the competing page rate per minute to be maintained and further control of the number of grayscale levels achieved. Tuning the voltage also allows for easy manipulation of the data to complete the proper image.

Manipulating the data with respect to timing can also be applied in improving the resolution of the image. One of the biggest problems in improving the resolution of spatial light modulator printers arises from the transfer of the image of the cells to the paper. Even on high resolution page printers, the printed image has a staircase effect on the curved edges. This phenomenon occurs because the cell is almost square and has a step shape to fill the curve.

FIG. 2a shows a typical layout of a spatial light modulator array. The pixel 20a is the center of the horizontal line 24 and the vertical line 22a. As it moves to the right, the pixels on the line are centered on the lines 22a, 22b and the like one after another. FIG. 2b shows the print image produced from column 22a with the data loaded in the standard way. The image of the first third pixel from the array of Figure 2a is transferred to area 26 on Figure 2b.

FIG. 2c shows one way to improve the staircase effect at the edges of the printed image 29. The top pixel 20 from the array receives its data at a later time, the drum continues to rotate and the pixel image appears at offset position 30. As the next pixel image is transferred, the bottom of the pixel is at the next standard timing pixel 29.
Become part of the statue. This allows the vertical width of the pixels to be manipulated so that detailed edges of print features can be represented in finer steps. With a delay of 8 steps, for example, the curve results as a series of pixels of the size of the previous pixel minus one eighth of its height. This is shown in Figure 2d. Therefore, selective re-addressing of pixels is used to move print-symmetric edges on the address grid that are finer than the resolution of individual elements in the process direction.

The limitation in FIG. 2 is the positioning of the center of the pixel. All the different sized pixels remain centered on the same xy grid as the standard sized pixels. There are many ways to improve the resolution along the direction of the SLM array in order to complement the above resolution enhancement methods in the process direction. One of these methods is to maintain the current 300 dpi standard when the decomposed features can be easily printed, and then to the higher dpi horizontal when printing curves or fine feature symmetry. It is a method of switching to the mode. FIG. 3 shows an array of spatial light modulators that do this. By selecting the proper columns, the timing can be varied as shown in FIG. 2 to incrementally offset the edges of the image horizontally as much as they are offset vertically or in the process direction.

The standard pixel 40 of FIG. 3 remains the same size. Pixel 4 starting after pixel 40
Column 01 moves to the right an increased distance. Eg 9
If 00 dpi is desired, then the next column of pixels should be offset by one third pixel from the first column, as shown as position 42. The column starting with pixel 402 moves one third further from the previous column, so the distance 4
4 equals two-thirds the width of a standard pixel. There is no limit as to how the columns can be offset from each other. This system requires first generating an indented column, then a standard column, and then another indented column. Further, the designer may think that two or more columns must be offset by a predetermined amount.

The addressing circuit for these pixels is
It is the same as the circuit for a normal array (see Figure 2a).
The printer controller determines when the delayed edge is needed to more accurately display the image features and waits for the line, such as line 401 or 402, to align with the latent image position on the OPC. Delay the data.

The print lines resulting from this offset technique are shown in FIG. The features of the first two prints are indicated by pixel lines 50. The feature-to-feature gap 52 results from a nominal two-pixel-wide gap during print exposure, but features 54
The third aspect of is printed by the pixels in column 401 (from FIG. 3) offset by a distance 42 equal to one third of the width of a standard pixel. Another gap in the line occurs only two-thirds after the offset gap 44 exposed by the DMD column 402 (from FIG. 3). At pixel column 45, the processor determines if it needs to return to the non-offset pixel or column 400 that occurred in the non-integer gap 47. Line 56 shows the composite image transferred to the paper. l
The line labeled (El) shows the unadjusted pixel resolution grid corresponding to column 400 (FIG. 3).

A more preferred method of manipulating the feature size of the printed image is to physically make the pixels in the DMD array shorter, narrower, or even smaller than the resolution grid configured for the printer. To change. This is shown in FIG. The standard pixel 20 is referred to as dimension 1. In this example, the quarter size reduction is reduced, but the method is not limited to such a specific setting. Pixel 34 has a linear dimension of 0.75 and an area of pixel 20 of 0.5. Similarly, pixel 36 has a dimension of 0.25,
Pixel 38 has an area of 0.0625 (1/16). These small size pixels can be considered for modulators with rows and columns of each size limited only by the application in which they are used. For example, the designer can decide to have three columns of pixels of each size.

Clearly combining the advantages of the above methods,
Along with gray scale imaging, a more powerful combination of features, both horizontal resolution enhancement and vertical resolution, is achieved. FIG. 6 shows an array capable of timing delay (vertical offset), variable pixel size, shape and grid position, and horizontal offset.

For reference, a normal xy grid (see FIG. 2)
A standard size pixel 20 is shown centered on lines 22 and 24) of a. Pixels 58a and 58
b shows a half-sized pixel, offset to align with the edge of pixel 20. These pixels 58a and 58b can operate separately or in association to simulate a square pixel element at the expense of increased complexity and addressing circuitry. Purely square pixels have optical advantages because movement in the xerographic process blurs the pixel image in the process direction. Pixels 60a and 60b show square sized pixels and timing delays, and pixel 60b is the pixel image with a delay of half a dot line. Pixel pairs 62a-b, 64a-b show that smaller pixels can be shifted left and right to match the left and right edges of a standard pixel. Although the pairs show three quarters and one quarter pixel, the pairs can be sized to fit the width of the standard pixel 20. Designers generally do not randomly arrange these pixels, but this figure illustrates various combinations of the above methods.

Next, questions concerning the manufacture of such a device will be described. The manufacture of such arrays for most modulators is obviously difficult. While this is true, it is believed that any of the above modulators can accommodate the results in such an array. A particularly suitable modulator for this type of array is a digital micromirror device (DMD), especially with a hidden hinge architecture.

For the manufacture of DMD, 1991, 10
US Pat. No. 5,061,049, issued Jan. 29, the manufacture of a double level DMD is Jan. 2, 1992.
It is described in US Pat. No. 5,083,857 issued on the 8th. Both US patents are incorporated herein by reference. The silicon substrate has digital addressing circuits manufactured by conventional CMOS methods, after which electrodes are deposited on this substrate and connected to the CMOS. The electrodes are then coated with an organic material, which can be removed by plasma etching. The vias are cut in the organic material and a first layer of metal is deposited thereon, filling the vias and then patterned to form support posts. Then a second layer of metal is deposited and patterned to form mirror elements. The entire structure is then etched to remove the organic material, leaving the mirror suspended on the electrodes and supported by the posts. There are many variations to this process, including variations where the mirror is supported by the rest of the organic material and not the posts.

The hidden hinge architecture is another adaptation of this basic process. Hidden hinge DMD in Figure 7a
FIG. Substrate 66 has an original layer of electrodes, such as 61, on and within the substrate. Organic material is present in the layer originally shown as gap 73.
The metal layers designated as 68a and 68b remain separated from the electrodes by the spacer layer. At this point, the original DMD is completed. Hidden hinges are an adaptation of the process of adding images. In the hidden hinge embodiment, the electrodes 70a and 70b are provided above the substrate by a gap 73 and are now mirrors 75 within the original single height structure connected to these electrodes. I have things. Another layer of metal is formed on the raised electrode layer to form a second layer of organic material, to form a bias for metal layer 58b, and to form post 72 and second level mirror 74. Stack. The second level mirror is typically formed in the central region of the original mirror 75. The structure thus formed is equivalent to the electrode / mirror 68a ...
a mirror element 74 suspended above an air gap 76 after removing the first and second layers of organic material from above and below b.
Becomes The electrodes / mirrors 68a-b are then separated from the address circuit by an air gap 73.

According to another adaptation, which provides another advantage, the first layer 68 of metal is separated from the electrode 71 without using the layer 68 as an electrode as described above. A conventional DMD structure is fabricated in metal layer 68 and completed with hinges, support posts and electrodes that stay on level 71. In this method, the addressing pulse actuates the first level mirror for addressing, causing air gap 7
4 uniquely determines the deflection angle. This separates the electronically active layer from the optically active layer, which is the second level mirror 74. Another advantage of this structure is
It is possible to fabricate the deformable mirror with a second uniform spacer that produces a uniform air gap 73 and an air gap 76 with the hammering angle of rotation operable. This structure is shown in Figure 7b.

The main advantage of this embodiment of the digital micromirror is that a second level mirror can be installed. This second level mirror is not restricted to being addressable, so it can move relative to the center post position,
Further, as shown in FIGS. 5 and 6, different sizes are possible.
FIG. 7b shows a schematic side view of this embodiment. Addressing electrodes 70a-b address a conventional first level mirror 75. This first level mirror 75 has posts 72 thereon corresponding to the spacer layer to ensure an air gap between the electrical layer element 75 and the optical layer elements 74 and 77. The entire DMD array has electrodes 70a ...
b and in response to control signals on the mirror 75, it also consists of a uniform-sized electrical element 75 that operates at angles ± θ. An optical element 74 and a flat special metal surface 77 around it are supported on posts 72 and 78. The optical element 74 mounts up to a precise deflection angle ± θ along the control element 75 regardless of size, position or shape.
The inert metal emphasis structure 77 prevents light from entering the lower mirror element 75 and entering the optical system.

The first level bias portion is filled with the planarizing material 80, and the second level portion is immediately above the first level bias portion.
It is possible to manufacture the level bias section 78. In another configuration shown in FIG. 7c, the post supporting the non-modulating metal light shield 77 may be located at 81, spaced apart from the active control element 75 and not directly above the biasing portion 72. it can. FIG. 7c is a plan view of two samples with alternating dimensions and locations 82, 83 of this structure and photopixel element. Hinge 8 suspending an octagonal control element 75 on a control electrode (not shown)
4 is a biasing portion 72 due to the planarized filler 80.
Are shown to be attached to.

Shielding the control level pixel structure from the optics requires a non-working optical level metal cross-hatched sample 77. This sample has a bias section 81 or 78 depending on design conditions.
(Fig. 7b) supported above.

Pixel elements 82 in the first two columns
Is the square (shaded) optically active portion of the array attached to the lower element 75 by the biasing portion 72. In the second embodiment, half the size of the pixel (shaded) is arranged along the four columns of actuating elements 75 to match the cell boundary edges, as shown in FIGS. In addition, it is meandering in a state of being offset from the center. The combination of the four elements 83 when electronically re-interlaced can be aligned to form a line image at twice the resolution of the basic control elements. This combination is constrained to meet the electrical operating requirements. The element 83 is 64 in FIG.
It also corresponds to the concepts shown in a and 64b.

Although FIG. 7c shows a constraining hinge array for simplicity, the method is also compatible with 45 degree hinges or other angular arrays.

Many other combinations such as those mentioned above can also be implemented on this structure. The electrostatic force constraint on the deflection is handled by the first level mirror, so the second level mirror provides more optical imaging with little, if any, constraint. Have the potential. The separation of optical and addressing and electrical operating conditions is the main advantage. Fabricating arrays that enable all of these possibilities allows for high resolution printing in gray scale. FIG. 8a is the result of running a standard 300 dpi printer with print image features. The desired print image characteristics are indicated by the dot lines. As can be seen from this figure, there are many resolution-related drawbacks that reduce the resolution of the image. Number 78
Artificial effects marked with make up dents that cannot be decomposed at present. The protrusions 76 exhibit artifacts with the opposite problem when they are too large to fill the gap in the image without extending beyond the limit.

In FIG. 8b these artifacts have been resolved. Pixels labeled 80 and 82 with cross-hatching are 0.75 times the standard pixel size,
Pixels with different sizes, either 0.5 or 0.25. The dotted pixel 86 is
Since they are delayed timing pixels, they occur vertically shifted. Pixels 84 filled with the cross pattern are pixels that are horizontally offset. Further combinations such as those mentioned above are possible. For example, pixel 80 is not only a 0.5 times larger pixel, but is also horizontally offset to fill its corners. Pixel 82 is a 0.25 times larger pixel, which can be further offset in the horizontal direction by a timing delay. Gray scale is not shown, but it is possible to gray scale with this array, as described above. Thus, gray scale can be used to further enhance the appearance and outline of the printed features. This combination simulates a true grayscale image by using pulse width controlled density modulation or area modulation without the loss of linear resolution obtained by using the smaller subpixels shown in FIG. Another advantage is that

The ASIC reformatter chip, which can contain a DMD on the board, provides the necessary datapath control switching and the appropriate delay to activate this array. The processor must determine which type of columns, or differently sized pixels, or which of the horizontally offset or standard pixels will receive a particular portion of the data stream. . In addition, the processor has 300 dp image features.
i or 900 dpi, or which effective offset dpi, depending on the final selection of the array.
You have to decide what to get. The DMD decodes the appropriately encoded data subject to timing, position or pixel size to obtain the desired printed image characteristics. The processor is also required to monitor the reset process and obtain the proper level of gray.

This allows the system to take advantage of all the possibilities for higher quality printing: gray scale, timing delay data, horizontally offset pixels and pixels of different sizes.

Although specific embodiments of methods of producing higher quality prints have been described above, such specific descriptions are deemed to limit the scope of the invention, except as set forth in the following claims. Should not

With respect to the above description, the following items will be further disclosed. (1) i. A method of time delaying the image data so as to finely control the printed image in the vertical direction; ii. A method of horizontally offsetting the pixel image so as to finely control the printed image in the horizontal direction; iii. A method of using partially dimensioned pixels for finely controlling the solid fill of the printed image; a. A high-resolution printing method in which data is processed and analyzed so as to change the final image by at least one of the three methods.

(2) The method according to item 1, further comprising a method for achieving gray scale. (3) a. Loading data into the addressing circuitry for each cell of the spatial light modulator to activate a predetermined number of said cells, b. Energizing the predetermined number of cells, c. Selectively deenergizing selected cells of the predetermined number of cells to reduce the exposure of the photosensitive surface and achieve gray scale; d. A grayscale printing method comprising repeating the deenergization until a desired shade of gray is obtained. (4) The method according to item 3, wherein the spatial light modulator is a digital micromirror device. (5) The method according to claim 3, wherein the spatial light modulator is a double-level digital micromirror device.

(6) Delay the transmitted data so as to energize the cells of the spatial light modulator and deactivate selected ones of said cells, pixels on the photosensitive medium in the vertical direction of the image. A method for improving resolution, which comprises finely controlling the size of an image. (7) Pixels in the array are obtained by offsetting the pixels in the array by a partial horizontal increment to obtain some multiple of the resolution of the original number of dots per inch. The method of claim 4 including horizontally offsetting and a method of achieving gray scale. (8) Pixels in the array are offset by offsetting the pixels in the array by a partial horizontal increment to obtain multiples of the original dot-per-inch resolution. A resolution enhancement method comprising horizontally offsetting.

(9) Use a pixel of a partial size of the standard pixel to fill the area of the printed image where the standard pixel leaves undesired artifacts, and grayscale The method of claim 8 including a method to achieve. (10) A method for improving resolution, which comprises using a pixel having a partial size of a standard pixel so as to fill the area of a printed image where the standard pixel leaves an undesired artifact. (11) Including a method for achieving gray scale,
The method according to item 10. (12) a. a cell of standard size spatial light modulator centered on an xy grid; b. a spatial light modulator cell of a partial dimension of the standard dimension about the xy grid, c. A spatial light modulator cell of a partial dimension of the standard dimension that is horizontally offset from the cell about the xy grid, and d. A standard dimension spatial light modulator cell horizontally offset from the cell about the xy grid;
And a spatial light modulator array having optically active and electrically active levels.

(13) The modulator according to item 12, wherein the modulator is a digital micromirror device. (14) The modulator according to item 12, wherein the modulator is a double-level micromirror device. (15) The modulator according to Item 12, wherein the optically active level and the electrically active level are different levels. (16) The modulator according to item 12, wherein each of the cells can be uniformly deflected to substantially the same deflection angle regardless of size.

(17) In the spatial light modulator printer,
It is difficult to perform high quality printing. There are two main problems: achieving gray scale within line time constraints and eliminating staircase artifacts in the printed image (81). This problem involves resetting the cell in the spatial light modulator, timing delay (86), horizontal offset (84) and another size pixel (80, 82) when loading data into the cell. It can be improved by using different methods.

[Brief description of drawings]

FIG. 1a is a diagram of a digital micromirror.

FIG. 1b shows a method of resetting data on a digital micromirror.

FIG. 1c shows a method of resetting data on a digital micromirror.

FIG. 1d shows a method of resetting data on a digital micromirror.

FIG. 2 illustrates a standard layout of a spatial light modulator and how data timing can be used to better control aliasing effects.

FIG. 3 shows horizontally offset pixels.

FIG. 4 illustrates a method of using horizontally offset pixels and the resulting printed image of the method.

FIG. 5 shows a layout of partial sized pixels centered on an xy brid.

FIG. 6 is a diagram showing an example in which various pixels are combined so as to improve resolution.

FIG. 7 is a diagram of a double level digital micromirror.

FIG. 8 shows print features using a standard spatial light modulator array and the same features using an adapted array.

[Explanation of symbols]

 1 Mirror 2a, 2b Post 3, 4 Landing electrode 5a, 5b Torsional hinge

Claims (2)

[Claims] 1. i. A method of time delaying the image data so as to finely control the printed image in the vertical direction; ii. A method of horizontally offsetting the pixel image so as to finely control the printed image in the horizontal direction; iii. A method of using partially dimensioned pixels for finely controlling the solid fill of the printed image; a. A high-resolution printing method in which data is processed and analyzed so as to change the final image by at least one of the three methods. 2. A. a cell of standard size spatial light modulator centered on an xy grid; b. a spatial light modulator cell of a partial dimension of the standard dimension about the xy grid, c. A spatial light modulator cell of a partial dimension of the standard dimension that is horizontally offset from the cell about the xy grid, and d. A standard dimension spatial light modulator cell horizontally offset from the cell about the xy grid;
And a spatial light modulator array having an optically active level and an electrically active level.