WO2004015679A1

WO2004015679A1 - Methods for measuring column nonuniformity in a projection display system

Info

Publication number: WO2004015679A1
Application number: PCT/US2003/024969
Authority: WO
Inventors: Philip Odom; Steven H. Linn; David L. Keith; Howard V. Goetz
Original assignee: Iljin Diamond Co., Ltd.
Priority date: 2002-08-09
Filing date: 2003-08-08
Publication date: 2004-02-19
Also published as: AU2003264026A1; TW200410196A

Abstract

Column nonuniformities (751, 752 and 753) are measured using a gradient image signal and a dataramp (720) configured to produce sets of contiguous pixels associated with different pixel brightnesses. A boundary between the sets is measured or a number of pixels (160, 161) in one of the sets in measured and a column correction value is based on the measurement. Alternatively, alternating dataramp signals (720)are used to produce display flicker, and measured value of flicker can be associated with column corrections.

Description

METHODS FOR MEASURING COLUMN NONUNIFORMITY IN A PROJECTION DISPLAY SYSTEM

Cross Reference to Related Applications This application claims the benefit of U.S. Provisional Patent Application

No. 60/402,194, filed August 9, 2002, and U.S. Provisional Patent Application No. 60/403,595, filed August 13, 2002, both of which are incorporated herein by reference.

Field of Invention

The invention pertains to display systems (e.g., liquid crystal projection displays) and, more particularly, to methods for measuring nonumformity in column brightness in such systems

Background

Column nonumformity is a defect found in many types of liquid crystal display systems. For example, column nonuniformity is exhibited in liquid crystal projection systems that use liquid crystal light valves to project images. To correct for nonuniformity, an offset correction value can be added to or subtracted from the data signal associated with a given column. In order to determine the appropriate correction values, however, the nonuniformity needs to be accurately quantified and measured. The liquid crystal material used for the liquid-crystal light valves has a distinctive, nonlinear EO response curve, increasing the difficulty with which the mid-level gray values can be found. Thus, the correction process creates a relatively large amount of data, and can be time intensive due to the settling rate of liquid crystal materials and the low frame-rate used by conventional computer displays.

Accordingly, there is a need for improved methods and apparatus for measuring column nonuniformity in hquid crystal light valves and for deterrnining the appropriate offset correction values. Summary

Methods of measuring column nonuniformity in a projection display comprise delivering a gradient image signal to a column driver of the projection display, wherein the gradient image signal comprises at least two gray-level values. The gradient image signal is compared with a ramp signal and a column of the display is driven using a dataramp signal based on the comparison, wherein the dataramp signal is configured to produce a first pixel brightness in a first set of contiguous pixels and a second pixel brightness in a second set of contiguous pixels. The first set of pixels and the second set of pixels define a boundary between the first set and the second set. A location of the boundary is determined, and an offset correction value is established based on the location. In representative examples, the first pixel brightness is associated with a substantially white pixel and the second pixel brightness is associated with a substantially black pixel. In other examples, the dataramp signal comprises a square wave. In additional examples, the data ramp signal is configured so that the first set of pixels and the second set of pixels include substantially all pixels in the column. In other examples, the gradient image signal is configured to produce at least two contiguous sets of pixels having the first pixel brightness and at least two contiguous sets of pixels having the second pixel brightness. Methods of measuring column nonuniformity in a projection display comprise delivering a gradient image signal to a column driver of the projection display, wherein the gradient image signal comprising more than one gray-level values. The gradient image signal is compared with a ramp signal and a column of the projection display is driven using a dataramp signal based on the comparison. The dataramp signal is configured to produce a first pixel brightness in a first set of pixels and a second pixel brightness in a second set of contiguous pixels. An offset correction value is established based on a number of pixels in the first set of pixels or the second set of pixels, hi representative examples, at least one of the first set of pixels and the second set of pixels includes at least two pixels. In other examples, a brightness associated with the column is measured, and the number of pixels in the first set of pixels or the second set of pixels is based on the measured brightness. In other examples, the first pixel brightness is associated with a substantially white pixel and the second pixel brightness is associated with a substantially black pixel, and the dataramp signal comprises a square wave. In additional examples, the gradient image signal is configured to produce at least two contiguous sets of pixels having the first pixel brightness and at least two contiguous sets of pixels having the second pixel brightness.

Methods of measuring column nonuniformity in a projection display comprise driving a column of the projection display using a first dataramp signal, the first dataramp signal producing a first pixel brightness to one or more pixels of the column. The column of the projection display is driven using a second dataramp signal, wherein the second dataramp signal produces a second pixel brightness in one or more pixels of the column, and the second dataramp signal is an inverse of the first dataramp signal. A difference between the first column brightness and the second column brightness is determined and an offset correction value is selected based on the difference. In representative examples, the first dataramp periodically ramps from a substantially zero value to a substantially maximum value, and the second dataramp periodically ramps from a substantially maximum value to a substantially zero value. In other examples, the substantially zero value is associated with a white pixel, and the substantially maximum value is associated with a black pixel. In still other examples, the first dataramp is an opposing-in dataramp, and the second dataramp is an opposing-out dataramp.

Methods of measuring column nonuniformity in a projection display comprise driving a column of the projection display using a first dataramp signal. The first dataramp signal produces a first pixel brightness in one or more pixels of the column. A second dataramp signal drives the column to produce a second pixel brightness in one or more pixels of the column, wherein the second dataramp signal is an inverse of the first dataramp signal. The first dataramp periodically ramps from a substantially zero value to a substantial value, and the second dataramp periodically ramps from the substantial value to the substantially zero value. In other examples, the first dataramp and the second dataramp have alternating positive and negative portions that produce corresponding positive pixel voltages and negative pixel voltages. Methods of measuring column nonuniformity in a projection display comprise driving a column of the projection display using a dataramp signal. The dataramp signal comprises a first period in which the dataramp signal ramps from a first nonzero value to a substantially zero value having a first polarity, and a second period in which the dataramp signal ramps from the substantially zero value to a second nonzero value. The second nonzero value has a second polarity opposite the first polarity, and the first period and the second period are alternately applied to the column in consecutive display frames. A degree of flicker is measured, and an offset correction value is obtained based on the measurement. In representative examples, the column is driven based on a substantially constant video signal and the degree of flicker is measured by measuring a brightness variation of the column.

Methods of measuring column nonuniformity in a projection display comprise determining a voltage-to-transmission response curve for a selected column and measuring transmission values of a column-under-test at selected gray- level values. The measured transmission values are mapped to the determined response curve in order to determine corresponding gray-level values for the column-under-test. An offset correction value is determined based on a difference between the selected gray-level values and the corresponding gray-level values of the selected column. In other examples, the selected gray-level values comprise a minimum gray-level value, a maximum gray-level value, and mid-gray-level value between the minimum gray-level value and the maximum gray-level value. In additional examples, the mid-gray-level value corresponds to a 50% transmission value from the voltage-to-transmission response curve and the selected column is from substantially the middle of the display. ' Methods of measuring column nonuniformity in a projection display comprise measuring a brightness of one or more baseline columns and a column- under-test. An offset correction value associated with the column-under-test is adjusted until the brightness of the column-under-test is substantially equal to the brightness of the baseline columns. In other examples, these steps are repeated for substantially all columns of the display.

Display systems comprise a gradient image signal source configured to provide a gradient image signal to a column driver of a projection display and a comparator configured to compare the gradient image signal with a ramp signal. A dataramp signal generator is configured to produce a first pixel brightness in a first set of contiguous pixels and a second pixel brightness in a second set of contiguous pixels based on the comparison. These and other features are described below with reference to the accompanying drawings.

Brief Description of the Drawings

FIG. 1 is a schematic block diagram of a column driver of an exemplary display system utilizing a dataramp signal.

FIG. 2 is a timing diagram illustrating column brightness nonuniformity.

FIG. 3 is an exemplary gradient image produced on a display.

FIG. 4 is an exemplary gradient image displayed with a display that produces a nonuniform column. FIG. 5 is a first display image produced using a modified dataramp signal.

FIG. 6 is a second display image produced using a modified dataramp signal.

FIG. 7 is a timing chart and block diagram illustrating signals used to produce a display image according to the first representative embodiment.

FIG. 8 is a display image having nonuniform column brightness in displaying a gradient image.

FIG. 9 is a third display image based on a modified dataramp signal.

FIG. 10 is a set of timing diagrams illustrating the measurement of column offset according to the second representative embodiment.

FIG. 11 is the set of timing diagrams from FIG. 10 after an offset correction value has been applied to the video signal.

FIG. 12 is a block diagram illustrating two exemplary, alternative dataramp signals as might be used in the second representative embodiment.

FIG. 13 is a block diagram illustrating the third representative embodiment.

Detailed Description

Disclosed below are representative embodiments that should not be construed as limiting in any way. Instead, the present disclosure is directed toward novel and nonobvious features and aspects of the various embodiments of the column nonuniformity measurement methods described below. The disclosed features and aspects can be used alone or in novel and nonobvious combinations and sub-combinations with one another. Although the operations of the disclosed methods are described in a particular, sequential order for the sake of presentation, it should be understood that this manner of description encompasses minor rearrangements, unless a particular ordering is required. For example, operations described sequentially may in some cases be rearranged or performed concurrently. Moreover, for the sake of simplicity, the disclosed methods typically do not discuss in detail the various ways in which particular methods can be used in conjunction with other methods. Additionally, the detailed description sometimes uses terms like "determine" and "obtain" to describe the disclosed methods. These terms are high-level abstractions of the actual operations that are performed. The actual operations that correspond to these terms will vary depending on the particular implementation and are readily discernible by one of ordinary skill in the art.

General Considerations

The various methods described generally concern the problem of column nonuniformity that can occur in projection display architectures that employ a dataramp signal for charging pixel array capacitors. One exemplary architecture is a liquid-crystal light valve architecture that utilizes a two-stage chopped-ramp system for charging capacitors. As more fully described below with reference to FIG. 1, this type of display architecture buffers video voltages for each display row in a bank of so-called sample-and-hold ("S/H") capacitors instead of charging the pixel capacitors directly. During the subsequent row time, comparator circuitry associated with each column uses the video levels stored in the S/H capacitors combined with a linear ramp signal to "chop" a dataramp signal that is used to charge all of a row's pixel capacitors simultaneously. As used herein, the dataramp signal may have a number of different shapes and sizes, but can generally be described as a periodic waveform used to set one or more pixel voltages. The dataramp signal is typically controlled by a gate or control signal that selectively connects and/or disconnects the dataramp from associated pixel(s) of the column. The ramp signal may also have a variety of different shapes and sizes. Further, for purposes of this disclosure, it is assumed that a greater pixel voltage results in a darker pixel, but this relationship may be altered such that a larger pixel voltage results in a lighter pixel (or in a pixel having a greater magnitude of some optical characteristic).

FIG. 1 shows a representative display system 100 that utilizes a two-stage chopped-ramp system. The display system 100 includes pixels arranged in one or more rows and one or more columns. FIG. 1 shows only a column 165 and rows 150, 151 that include representative pixels 160, 161. Columns of pixels are not shown. A typical display system includes 200-2000 rows and 200-2000 columns of pixels. The pixels 160, 161 include FETs 135, 138, pixel capacitors 136, 139, and pixel electrodes 137, 140, respectively. The pixel electrodes 137, 140 are situated to provide image-dependent pixel voltages to a liquid crystal or other display element with respect to a voltage applied to a backplane electrode 170 that is common to some or all pixels .

A dataramp source 102 supplies a dataramp voltage, such as a time- dependent voltage 103 to a buffer 104. The buffered dataramp voltage is then delivered to a series of column FETs, such as the exemplary column FET 106. The dataramp signal may produce an alternating positive and negative voltage that is used to charge the pixel capacitors. The display system 100 typically includes additional column FETs corresponding to each column of pixels. A ramp source

110 provides a ramp voltage, such as a time-dependent voltage 109, to a comparator

111 that also receives voltages corresponding to image picture elements (pixels) from a sample and hold (S/H) module 112. The S/H module 112 includes sample capacitors 114, 115 that receive image voltages from a video input 118 from a video source such as a gradient image source 105 or other image source (not shown in FIG. 1) via sample input switches 116, 117. The S/H module 112 also includes sample output switches 119, 120 corresponding to sample capacitors 114, 115. The switches 116, 117, 119, 120 are generally configured so that one of the capacitors 114, 115 charges to a sample voltage corresponding to a pixel voltage via the corresponding switch 116, 117, respectively, while a pixel voltage stored on the other of the capacitors 114, 115 is delivered to the comparator 111 via the corresponding switch 119, 120. As a specific example, the switch 116 is closed to permit the capacitor 114 to charge and the switch 120 is closed to permit the voltage on the capacitor 115 to be delivered to the comparator 111. The switches 117, 119 are open. After charging the capacitor 114 and delivery of the voltage on the capacitor 115 to the comparator 111 is complete, the switch states are reversed so that the capacitor 115 charges to a pixel voltage corresponding to another pixel and the sample voltage on the capacitor 114 is delivered to the comparator 111. The S/H module 112 includes the sample capacitors 114, 115 that acquire and store pixel voltages for pixels in a single column and additional modules can be provided for the remaining columns. In a representative example, the display columns are divided into eight groups and eight video inputs (such as the video input 118) are sequentially switched to S/H modules associated with the columns. For example, a first video input is sequentially switched to S/H modules for columns 1, 9, 17, . . ., a second video input is sequentially switched to S/H modules for columns 2, 10, 18, . . . , and other video inputs are similarly switched. For convenience, only one S/H module is shown in FIG. 1.

The delivery of data to the row 150 begins with the column FET 106 and a scanner output 128 configured so that a voltage on the pixel capacitor 136 follows the dataramp voltage. The sample capacitor 114 is charged to a voltage determined by a video signal applied to the video input 118. At a switching time Ts, the comparator is switched off in response to the ramp voltage and the voltage on the sample capacitor 114. As a result, the column FET 106 is also turned off and a dataramp voltage associated with the switching time Ts remains on the pixel capacitor 136, and the voltage on the pixel capacitor 136 does not follow additional changes in the dataramp voltage. Pixels of other rows and columns are addressed in a similar fashion by controlling a switching time at which pixel capacitors stop following the dataramp voltage.

Comparison of a voltage on a capacitor (such as the capacitor 114) with the ramp input 109 converts a pixel voltage from a video input voltage to a switching time Ts at the comparator 111. The switching time T_s controls the column FET 106 to select a voltage applied to a pixel by the dataramp input 103. This procedure can be regarded as conversion of a pixel voltage to a pixel-dependent switching time that is then reconverted into a pixel voltage.

As with other displays, this type of chopped-ramp architecture can suffer from some degree of inter-column variation in the delay introduced by the comparator circuitry 111. An exemplary timing chart illustrating this delay is shown in FIG. 2. In particular, FIG. 2 shows a video signal 210 as it interacts with a ramp signal 220. As described above, the two signals are combined in a comparator to determine a switching time T_s at which an associated column FET is turned off. When the column FET is turned off at switching time T_s, the dataramp signal 230 is set and delivers a particular voltage that charges, or discharges, the respective pixel capacitor. A column enable signal 250, which is produced by the comparator and controls the column FET, is shown as going high when the ramp signal 220 is greater than the video signal 210, and as going low when the ramp signal 220 is less than the video signal 210 at T_s. The dotted line 240 illustrates the theoretical (i.e., the ideal) time at which the column enable signal 250 goes low, thereby switching the column FET off and chopping the dataramp voltage 230.

The comparator, however, does not typically perform in an ideal manner and may either delay or speed up the switching time T_s. As shown in FIG. 2, for instance, the comparator may create a delay in the switching time T_s, such that the actual switching time lies within a range of times (shown as the shaded region of the column enable signal). As a result of this delay, the dataramp signal 230 in FIG. 2 is chopped later than it should be, resulting in a greater voltage delivered to the corresponding pixel capacitor. Consequently, all of the pixels in a given column may have a different gray-level-to-brightness function than their neighbors. Therefore, when displaying a single shade, the screen may appear to have vertical striping. This striping results from the different columns of pixels having slightly different shades and from the pixels in a given column being closer in displayed value to each other than to those in surrounding columns.

In other display systems, the general cause of this problem is that analog signals are delivered along column conductors to the pixel by column-dedicated circuitry. The differences in the column circuitry results in differences in the signal levels delivered to all of the pixels in the associated column. To correct this problem, an average delay for the column comparators can be computed, and the dataramp signal can be time-shifted according to this average delay. A time-shifted dataramp signal 232 is shown in dashed lines in FIG. 2. As can be seen, the dataramp signal 232 allows the correct voltage 260 to be applied to the pixel capacitor.

Shifting the dataramp signal 232, however, still does not account for the individual variations from average that the columns may exhibit. As a result of these variations, the relative column brightness may be nonuniform. These variations can be reduced or compensated on a column-by-column basis by introducing an offset into the digital data being sent to each column in a display. For this correction system to work correctly, however, a method of measuring the amount of column-to-column variation must be employed.

The following representative embodiments disclose methods by which column nonuniformity can be measured and methods for generating gray-level offset data that can used by a correction offset system.

First Representative Embodiment

A first representative embodiment is based on displaying a gradient image and using a modified dataramp signal to drive the individual columns of a display. For example, the gradient image is formed such that it produces a range of gray-level values, which may vary from implementation to implementation. For example, the range may include a full range of gray level values (e.g., from 0 to 255 for an 8-bit architecture) or some smaller range (e.g., 150-190). The difference between each gray-level value in the range may also vary. In one particular embodiment, the gradient image produces a vertical band of gray values in a given pixel column. The number of pixels displaying a particular gray level may vary depending on the display and implementation. For example, a display with 600 rows can display two pixels for each gray level in a given column (i.e., 256 x 2 = 512 < 600). For any remaining rows, black bands may be displayed so that the number of pixels for each gray level remains constant for the entire column. In this example, the gradient image can be displayed as either one continuous gradient, or as two gradient bands with single pixels dedicated to each gray level. FIG.3 shows an example of a gradient image 300 formed on an exemplary display. In FIG. 3, the dataramp signal driving the display is in a normal, operational mode. The exemplary display of FIG.3 has no column nonuniformity. Thus, the displayed gradient appears smooth and uniform along all columns of the display. By contrast, FIG.4 shows an image 400 of the same gradient image data on an exemplary display that exhibits column nonuniformity. As can be seen, the columns in image 400 appear uneven and exhibit variations in brightness between neighboring columns. The actual amount of variation will depend on the particular light valve architecture used and the column-driving circuitry for the particular display. The first representative embodiment includes modifying the dataramp signal that drives the columns. More particularly, the dataramp signal is modified to a signal that produces contrasting values. For example, the modified dataramp signal may be a square wave (e.g., a 50% duty cycle square wave) that produces either a full voltage or no voltage at the corresponding pixel or pixels. The actual values produced by the square wave, however, may vary so long as some measurable differences exists.

When the dataramp signal is configured in this manner, it produces either a high pixel value (e.g., a black pixel) or a low or zero pixel value (e.g., a white pixel) on the display. Further, when the dataramp signal is controlled by the gradient image and the ramp signal, it produces bands of alternative high and low pixel values (e.g., bands of black and white on the resulting display).

For example, FIG. 5 shows an image 500 produced with a display driven using a square dataramp signal. As can be seen in FIG. 5, the dataramp signal produces high contrast images of alternating black and white bands. In particular, the dataramp produces a black pixel for all gradient values below a mid-gray-level value (e.g., < 128) and a white pixel for gradient values above the mid-gray-level value (e.g., > 128). As in FIG.3, the image 500 of FIG.5 exhibits no column nonuniformity. Thus, the bands appear solid and uniformly transition from one to another from column to column. In contrast, FIG. 6 shows an image 600 that exhibits column nonuniformity when driven by a square dataramp signal. As can be seen, the gray values at which the dataramp signal produces a white pixel instead of a black pixel vary on a column-by-column basis. This process of using a gradient image to drive a modified dataramp signal is more fully illustrated in FIG. 7. In FIG. 7, an exemplary declining gradient image signal 710 (or a declining image video signal) is shown. A ramp signal 720, which is combined with the video signal 710 in a comparator, is used to determine a switching time T_s at which the value of a square dataramp signal 730 is taken and transferred to the pixel capacitor. Because of the gradually declining value of the gradient image signal 710, the 5 time T_s gradually shortens. Consequently, the time at which the dataramp signal 730 is "chopped" within the period of the signal similarly decreases. The resulting image is shown as the display 750, which shows two bands of black and two bands of white being produced. The display 750 shows only nine pixels per column for illustrative purposes only. As seen in the display 750, certain columns (columns 751 , 752, and 753) are not

10 uniform with the other columns, representing a problem in the liming of the column- driving circuitry (e.g., a delay in the comparator).

Various methods may be used to measure the column nonuniformity displayed via the above-described techniques. For example, in one implementation, optical power or brightness measurements of screen brightness are obtained for the columns. In one

15 particular implementation, an integrating sphere fitted with optical detectors is used to obtain these measurements. The optical detectors maybe connected to an optical power meter used to measure the overall intensity of the light entering the integrating sphere. With this system, an average of the light emitted from a single light-valve-under-test may be measured. For example, a projector and display may be configured to activate each

20 column individually so that each light valve can be separately tested.

In one particular implementation, optical power or brightness measurements are obtained when the column is driven white, then to the gradient image, and then to black. The number of white pixels (or black pixels) in a column may then be calculated. The following equation may be used to determine the number of white pixels in a column: nr n pixels /column p ^D p p ^gradient ' white black wherein P represents optical power when the column is driven white (P_w/κ/e)₅ black {Pbiack), mάPgradient is a gradient, and n_pixehlcobmn is a total number of pixels in the column.

As more fully described above, a column uniformity error conesponding to one 30 gray level of brightness causes the square dataramp signal to be cut off by the gradient image either one gray level sooner or one gray level later. For example, if there are two pixels per gray level (as in the example described above for a display having 600 rows), each gray level of error corresponds to a ±2 change in the number of white pixels displayed in the column.

To determine an appropriate offset correction value to apply to the display, the number of white pixels in each column can be subtracted from the average (mean) of all the number-of-wbite-pixels measurements. The result is a variation from the average number of white pixels, which can then be divided by the number of pixels per gray level to determine a deviation from the average gray level. This deviation is an offset for each column and can be compensated for by adding it to the digital video data for each column before the data is converted to analog voltages.

As noted above, the gradient image and the size of the bands produced may vary from implementation to implementation. For example, if column nonuniformity is constrained within a certain number of gray levels, the range over which the gradient image varies can be reduced, correspondingly increasing the number of pixels for each gray level. Increasing the number of pixel for each gray level can help improve measurement sensitivity, since one gray-level brightness error will cause a greater change in the number of white pixels. For example, FIG. 8 is a non-uniform display 800 showing gradient bands ranging from a gray level from 90 to 150 when being driven in a normal dataramp mode. FIG. 9 shows an image 900 produced in a square (or test) dataramp mode.

The disclosed embodiment may be modified in a number of ways. For example, optical sensors tuned for different wavelengths of light may be used to measure the emitted light. By employing different sensors for different colors, and filtering each of the sensors with a pass-band filter for its color, it is possible to reduce the background level of the measurement and improve its accuracy. This method can be used to perform conections on a single channel or for performing simultaneous corrections on red, green, and blue channels of a projector. Further, a single cell may be placed in the entrance port to the integrating sphere and illuminated with a suitable light source. A fixture can then be produced that is used for taking calibration data on a single light valve before installation in an end product, such as a projector. Alternatively, the integrating sphere can be replaced with a pair of cylindrical convex lenses at right angles to each other so that the image of all the pixels can be placed on a single small region of a sensor. Such a sensor can then respond to the entire light valve output in the same manner as the integrating sphere and also be used for single light valve measurements. Moreover, instead of the optical system described, a machine-vision system having, for example, a charged-couple device (CCD) can be used to count the number of white pixels or to measure the relative size of the black and white bands in each column.

Prior to performing the methods according to this or any other embodiment disclosed herein, a back-plane voltage dataramp timing for the display may be set to reduce, compensate, or minimize an average flicker exhibited by the display.

Second Representative Embodiment

The dataramp signal described above may have a variety of forms that create unique pixel-charging techniques. A particular dataramp signal may also be inverted to create a substantially identical, but opposite, signal. The second representative embodiment utilizes inverted dataramp signals to produce an offset measurement. Two exemplary inversion modes that can be utilized in a display architecture having a ramped signal, such as the dataramp described above, are the so-called "opposing-out" and "opposing-in" inversion modes. In opposing-out mode, the dataramp signal ramps from the white voltage to the black voltage over time. Consequently, gray-level values nearer to black cause the dataramp signal, and thus the pixel-charge values, to be chopped at a later time. Correspondingly, in opposing-in mode, the dataramp signal ramps down from the black voltage to the white voltage over a row time. Thus, gray-level values nearer to black cause the dataramp signal to be chopped earlier. While operating in this mode, the video signal may similarly be inverted, such that the resulting display is theoretically identical in either mode. When utilized in the projection display architecture described above, the opposing-out mode causes a column's pixels to charge to a higher value when a delay is present in the comparator. Therefore, the resulting column has a darker appearance. By contrast, the opposing-in mode causes a column's pixels to discharge farther, resulting in a column having a brighter appearance. Similarly, when the comparator operates faster than expected, the opposing-out mode creates a brighter column, whereas the opposing- in mode creates a darker column. In certain specific embodiments, the offset correction values are determined by measuring the difference between a column's brightness when operating in opposing-out mode and when operating in opposing-in mode. This process is illustrated in FIG. 10, which shows a first timing chart 1000 having a video signal 1010, a ramp signal 1012, a column enable signal 1014, and an opposing-out dataramp signal 1016. A switching time T_s is illustrated in FIG. 10 as the line 1020. As discussed above, comparator delay may cause the column enable signal to "chop" the dataramp signal at a time after the desired switching time (shown as dashed line 1030). Consequently, a pixel voltage 1040 is created that has a larger value than a desired pixel value, indicated as dashed line 1060. A second timing chart 1002 in FIG. 10 includes the video signal 1010, the ramp signal 1012, the column enable signal 1014, and an opposing-in dataramp signal 1018, which is the inverse of the opposing-out signal 1016. The second timing chart illustrates that the opposing-in dataramp signal 1018 creates a pixel voltage 1050 that is less than the desired amount 1060. Because the dataramp signal 1018 is the inverse of dataramp signal 1016, pixel voltage 1050 differs from the desired pixel voltage 1060 bythe same amount as pixel voltage 1040, but in the opposite direction. In one particular embodiment, the difference between pixel voltage 1040 and pixel voltage 1050 is determined. Based on this difference, an offset correction maybe applied to the column, and the process repeated. This iterative process may continue until an offset correction value is found that results in a minimum difference or reduced difference between the pixel voltages resulting from application of the opposing-out dataramp signal 1016 and the opposing-in dataramp signal 1018. For example, FIG. 11 shows the resulting timing charts 1100, 1102 after an offset correction has been applied to the video signal 1010 (for illustrative purposes, the new signal 1010' is shown as a solid line, and the old signal is shown as a dashed line 1010). As can be seen, the offset correction selected results in two pixel voltages 1040' and 1050' that are substantially identical. Thus, no further offset corrections are necessary and the selected column is properly adjusted.

In one specific implementation, an appropriate offset correction is predicted based on the initial difference measured. More specifically, an appropriate offset correction can be selected and used for the next set of pixel voltages based on offset corrections determined for previously measured columns. For instance, a predictive algorithm can be derived using the following measurements obtained from one or more previously measured columns: (1) the initial error measurement; (2) a measurement made after applying an estimated correction; and (3) a final measurement. In certain embodiments, the process terminates when the measured error is within a certain error tolerance (e.g., ± 1 gray value). This technique of using a predictive algorithm and terminating the process when the difference is within a selected error tolerance can substantially reduce the time required to find offset correction values for each column.

In another embodiment, the technique described above is used to measure column flicker instead of column brightness. For example, in certain embodiments, the display is operable in two additional inversion modes, hi general, the additional inversion modes alternate between opposing-in and opposing-out dataramp waveforms on a frame-by-frame basis. In a first mode, the so-called "parallel-down" mode, the opposing-in mode is used for positive excursion frames (i.e., frames in which the relevant pixel voltage is set is with a positive voltage), and the opposing-out mode is used for negative excursion frames (i.e., frames in which the relevant pixel voltage is set with a negative voltage). In a second mode, the so-called "parallel-up" mode, the opposing-out mode is used for positive excursion frames, whereas the opposing-in mode is used for negative excursion frames. FIG. 12 illustrates these two additional inversion modes. Dataramp 1200 of FIG. 12 shows an exemplary parallel-down waveform, whereas dataramp 1210 shows an exemplary parallel-up waveform. When successive frames of a column are driven using one of these inversion modes, the resulting pixel voltage will change from frame-to-frame, or flicker, if the comparator timing is offset. For example, consider a display operating in parallel-down mode with a delay in the comparator signal. During the positive excursion portion of the parallel-down dataramp, the delay will cause the resulting voltage to be less than the expected value, but during the negative excursion portion of the parallel-down dataramp, the delay will cause the resulting voltage to be greater than the expected value.

The degree of flicker observed (i.e., the peak-to-peak brightness variation measured between successive frames) is related to the amount the comparator timing is offset from the average comparator time, and is thus directly related to the offset correction value needed to correct the timing. As described above, an iterative process may be used to find the correction value that results in the lowest flicker. However, in certain embodiments, the first offset correction value selected can be in an arbitrary direction because the polarity of the error is not known initially. After selecting a first offset value in the arbitrary direction, a second measurement is taken and a determination is made whether the second measurement yields a lower error. If a lower error results, then the assumed initial direction was correct; otherwise, a new offset value is selected in the opposite direction. The algorithm may temiinate when a correction value yielding the minimum amount of error is found. For example, a final correction value is found when a greater flicker is measured for correction values that are both greater than and less than the current value.

Third Representative Embodiment

In a third embodiment, a voltage-to-transmission response curve is determined for a selected column. The transmission value of a test column is measured at selected gray-level values and mapped onto the response curve. Using the measured values mapped onto the determined response curve, an offset correction value (in terms of a gray-level value) can be determined.

FIG. 13 illustrates this method in greater detail using an exemplary response curve 1310 and column-under-test 1320 from an exemplary display 1312. The exemplary response curve 1310 may comprise a transmission-to-gray-level curve for any one of the columns or set of selected columns in the display. In one specific embodiment, for example, the curve 1310 is from the middle of the display. The curve can be determined by measuring the brightness of tight transmitted by the column at selected gray-level values between white and black (e.g., from 0 to 255) and using known interpolation techniques. It is assumed that the shape of this curve will be approximately constant for each column in the display and that variations in comparator timing will cause the curve to be shifted left or right (i.e., along the gray-level axis of the curve 1310). In certain other embodiments, however, the axes of the graph are flipped so that the curve shows gray-level-to-transmission values, and comparator timing shifts the curve up and down on the resulting graph.

A column-under-test 1320 is selected from the display. For the column-under- test 1320, multiple brightness measurements may be taken when the column is set for different gray-level values. For example, in one particular embodiment, three transmission measurements are measured: (1) when the column is black (e.g., the gray- level value is maximum); (2) when the column is white (e.g., the gray-level value is minimum); and (3) when the column is at a 50% transmission gray-level value. The 50% transmission gray-level may be found using the response curve 1310. For example, as illustrated in FIG. 12, the 50% transmission level 1314 for the exemplary response curve 1310 is at a gray-level value 1316 of "110." This value is not necessarily equal to half of the gray level range because the response curve 1310 is typically nonlinear. The multiple transmission measurements obtained for the column-under-test 1320 are illustrated in FIG. 13 at 1322, 1324, and 1326. In other embocliments, only two transmission measurements are obtained. In still other embodiments, additional transmission measurements are obtained (e.g., 3, 4, 5, . . . ). In some embodiments, the transmission measurements are obtained for all columns in the display.

As noted above, variations in comparator timing will cause the transmission values measured for a column-under-test 1320 to vary. The multiple transmission measurements obtained for the column-under-test 1320 can be mapped to the response curve 1310 in order to determine the offset correction value. For instance, in one particular embodiment, the transmission for a column-under-test may be mapped to a gray level using a gray-level-to-transmission curve, which is generated as the inverse of the transmission-to-gray-level curve described above. For example, in FIG. 13, three transmission measurements are obtained for the column-under-test 1320: a measurement 1332 at a gray-level value of 0, a measurement 1334 at a gray-level value of 110 (i.e., the 50% transmission gray-level value), and a measurement 1336 at a gray-level value of 255. This process can be repeated for any or all of the other columns in the display, hi one particular embodiment, the mapped gray-level values for all of the columns tested are averaged, and the sets of gray levels determined for the columns-under-test subtracted from the average in order to generate the appropriate offset correction values.

Fourth Representative Embodiment

A fourth representative embodiment includes measuring a brightness of a selected column or set of selected columns in the display. A column offset value is either increased or decreased until the column-under-test has an absolute brightness that most closely matches the original brightness. Thus, the brightness of the selected column functions as a baseline measurement, hi one particular embodiment, every column in the display (except the originally selected column(s)) are measured and adjusted. This method typically requires substantially constant illuminatioh across the part being measured. Further, the method may result in a set of correction values that do not have an average near zero. Accordingly, the values may be shifted by a constant value in order to minimize the mean adjustment.

Fifth Representative Embodiment

A fifth representative embodiment includes determining a gray-level value that results in a selected transmission level for a column-under-test. For example, in one particular embodiment, the gray-level resulting in the 50% transmission level is determined. This gray-level value may be determined, for instance, by measuring the brightness of the column-under-test when it is set to black and to white, and then applying an appropriate search algorithm. The search algorithm may vary, but in one embodiment is a binary search algorithm, which begins by measuring the brightness at the middle gray value (e.g., 128) and comparing the resulting brightness with the known brightness measurements. The process may be repeated for each column in the display to develop a set of gray-level values for the selected transmission level, hi one representative embodiment, correction values are obtained by subtracting the average gray-level values from the individual gray-level values determined. In view of the many possible embodiments to which these principles may be apphed, it should be recognized that the illustrated embodiments are only representative examples and should not be taken as a limitation on the scope of the disclosure. We claim all that comes within the scope of the appended claims.

Claims

What is claimed is:

1. A method of measuring column nonuniformity in a proj ection display, comprising: delivering a gradient image signal to a column driver of the projection display, the gradient image signal comprising at least two gray-level values; comparing the gradient image signal with a ramp signal; driving a column of the projection display using a dataramp signal based on the comparison, wherein the dataramp signal is configured to produce a first pixel brightness in a first set of contiguous pixels and a second pixel brightness in a second set of contiguous pixels, wherein the first set of pixels and the second set of pixels define a boundary between the first set and the second set; determining a location of the boundary; and establishing an offset correction value based on the location.

2. The method of claim 1, wherein the first pixel brightness is associated with a substantially white pixel and the second pixel brightness is associated with a substantially black pixel.

3. The method claim 1, wherein the dataramp signal comprises a square wave.

4. The method of claim 1, wherein the dataramp signal is configured so that the first set of pixels and the second set of pixels include substantially all pixels in the column.

5. The method of claim 1, wherein the gradient image signal is configured to produce at least two sets of contiguous pixels having the first pixel brightness and at least two sets of contiguous pixels having the second pixel brightness.

6. A method of measuring column nonuniformity in a proj ection display, comprising: delivering a gradient image signal to a column driver of the projection display, the gradient image signal comprising more than one gray-level value; comparing the gradient image signal with a ramp signal; driving a column of the projection display using a dataramp signal based on the comparison, wherein the dataramp signal is configured to produce a first pixel brightness in a first set of pixels and a second pixel brightness in a second set of pixels; and establishing an offset correction value based on a number of pixels in the first set of pixels or the second set of pixels.

7. The method of claim 6, wherein at least one of the first set of pixels and the second set of pixels includes at least two pixels.

8. The method of claim 6, further comprising measuring a brightness associated with the column, and the number of pixels in the first set of pixels or the second set of pixels is based on the measured brightness.

9. The method of claim 6, wherein the first pixel brightness is associated with a substantially white pixel and the second pixel brightness is associated with a substantially black pixel.

10. The method claim 6, wherein the dataramp signal comprises a square wave.

11. The method of claim 9, further comprising: measuring optical powers P_Wh e and Pb_kc_k associated with all column pixels having pixel brightnesses associated with white and black pixels, respectively; and

determining the number of white pixels by calculating — ^pp^aix^eel^ss I ^cc^<o"lu"m""n xP gradient ■ '

' white - P bulack wherein P_smdi_enά& the measured optical power when the column is driven with the gradient image, and n_pixehlcolumn is a total number of pixels in the column.

12. The method of claim 6, wherein the gradient image signal is configured to produce at least two sets of contiguous pixels having the first pixel brightness and at least two sets of contiguous pixels having the second pixel brightness.

13. The method of claim 6, wherein the number of pixels is based on a measurement of a sum of pixel brightnesses.

14. A method of measuring column nonuniformity in a projection display, comprising: driving a column of the projection display using a first dataramp signal, the first dataramp signal producing a first pixel brightness at one or more pixels of the column; driving the column of the projection display using a second dataramp signal, the second dataramp signal producing a second pixel brightness at the one or more pixels of the column, wherein the second dataramp signal is an inverse of the first dataramp signal; determining a difference between the first column brightness and the second column brightness; and selecting an offset correction value based on the difference.

15. The method of claim 14, further comprising driving the column based on a substantially constant video signal.

16. The method of claim 14, wherein selecting the offset correction value includes adjusting a stored offset correction value.

17. The method of claim 14, wherein the first dataramp periodically ramps from a substantially zero value to a substantially maximum value, and the second dataramp periodically ramps from a substantially maximum value to a substantially zero value.

18. The method of claim 17, wherein the substantially zero value is associated with a white pixel, and the substantially maximum value is associated with a black pixel.

19. The method of claim 14, wherein the first dataramp is an opposing-in dataramp, and the second dataramp is an opposing-out dataramp.

20. A method of measuring column nonuniformity in a projection display, comprising: driving a column of the projection display using a first dataramp signal, the first dataramp signal producing a first pixel brightness in one or more pixels of the column; and driving the column of the projection display using a second dataramp signal, the second dataramp signal producing a second pixel brightness in the one or more pixels of the column, wherein the second dataramp signal is an inverse of the first dataramp signal, and the first dataramp periodically ramps from a substantially zero value to a substantial value, and the second dataramp periodically ramps from the substantial value to the substantially zero value.

21. The method of claim 20, wherein the first dataramp and the second dataramp have alternating positive and negative portions that produce corresponding positive pixel voltages and negative pixel voltages.

22. The method of claim 20, wherein the substantially zero value is associated with a white pixel, and the substantial value is associated with a black pixel.

23. A method of measuring column nonuniformity in a projection display, comprising: driving a column of the projection display using a dataramp signal, the dataramp signal comprising, a first period that ramps from a first nonzero value to a substantially zero number value, the first nonzero value having a first polarity, and a second period that ramps from the substantially zero value to a second nonzero value, the second nonzero value having a second polarity opposite the first polarity, wherein the first period and the second period are alternately applied to the column in consecutive display frames; measuring a degree of flicker; and obtaining an offset correction value based on the measurement.

24. The method of claim 23, further comprising driving the column based on a substantially constant video signal.

25. The method of claim 24, wherein measuring the degree of flicker comprises measuring a brightness variation of the column.

26. The method of claim 23, wherein the offset correction value is obtained based on a stored value of the offset correction values.

27. A method of measuring column nonuniformity in a projection display, comprising: determining a voltage-to-transmission response curve for a selected column; measuring transmission values of a column-under-test at selected gray-level values; mapping the measured transmission values to the determined response curve in order to determine corresponding gray-level values for the column-under-test; and determining an offset correction value based on a difference between the selected gray-level values and the corresponding gray-level values of the selected column.

28. The method of claim 27, wherein the selected gray-level values comprise a minimum gray-level value, a maximum gray-level value, and a mid- gray-level value between the minimum gray-level value and the maximum gray- level value.

29. The method of claim 28, wherein the mid-gray-level value corresponds to a 50% transmission value from the voltage-to-transmission response curve.

30. The method of claim 27, wherein the selected column is from substantially the middle of the display.

31. A method of measuring column nonuniformity in a proj ection display, comprising: measuring a brightness of one or more baseline columns; measuring a brightness of a column-under-test; and adjusting an offset correction value associated with the column-under-test until the brightness of the column-under-test is substantially equal to the brightness of the baseline columns.

32. The method of claim 31 , further comprising repeating the steps for multiple columns-under-test of the display.

33. The method of claim 31 , further comprising shifting the offset correction values for the multiple columns-under-test so that the offset correction values have an average that is substantially zero.

34. A method of measuring column nonuniformity in a projection display utilizing a dataramp signal for setting pixel voltages, comprising: determining a gray-level value that results in a selected transmission level for multiple columns-under-test of the projection display; averaging the gray-level values determined for the multiple columns-under-test; and deternrining an offset correction value for a selected column-under-test by subtracting the average gray-level value of the multiple columns-under-test from the gray-level determined for the selected column-under-test.

35. A display system, comprising: a gradient image signal source configured to provide a gradient image signal to a column driver of a projection display; a comparator configured to compare the gradient image signal with a ramp signal; and a dataramp signal generator configured to produce a first pixel brightness in a first set of contiguous pixels and a second pixel brightness in a second set of contiguous pixels based on the comparison.