AVS Contrast Thread - Now with Dynamic Contrast Results! - AVS Forum
Forum Jump: 
Reply
 
Thread Tools
post #1 of 200 Old 05-25-2007, 01:22 PM - Thread Starter
AVS Club Gold
 
Mark Petersen's Avatar
 
Join Date: Dec 2000
Posts: 4,671
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Last updated: June 1st, 2007 (test pattern and spreadsheet now available for download)

Note: This is an updated thread based on: http://www.avsforum.com/avs-vb/showthread.php?t=781060 The older thread was obsoleted due to the inclusion of additional explanations and dynamic contrast results which couldn't be fit into the earlier thread.

The first 6 posts of this thread contain a detailed explanation of the test patterns and methodology that is being used as part of the AVS Contrast Project. The contrast framework that is described here was applied to two projectors, one native and one with a dynamic iris (JVC RS-1 and Sony VW50). These first 6 posts are intended to be used as an example of the type of information that can be extracted from these simple measurements. It is hoped that more projectors will be measured and their results added to the AVS Contrast Project. The results of this ongoing work can be found after the 6 posts of this initial discussion (starting at post 8 or thereabouts).

Also, please realize that while considerable care has gone into the methodologies, measurements, calculations, and graphs that are contained here, the accuracy of the results cannot be guaranteed and are subject to variances in measurements, accuracy of sensors and unit to unit variance. The information presented here is desiged for rough comparisons only.

This is a work in progress, check back often for additional results and information!


Index

Part I - Overview

Purpose
Terminology
Background
- Sequential On/Off Contrast and ANSI Contrast
- Dynamic and Static Intra-Image Contrast
- What causes Contrast?
- Contrast as a function of Luminance
- Theoretical vs Measured Intra-Image Contrast
Test Pattern Suites - Static Intra-Image Test Patterns and Dynamic Intra-Image Test Patterns

Part II - Static Intra-Image Contrast Example (JVC RS-1 and Sony VPL-VW50)

RS-1 Intra-Image Contrast
RS-1 Contrast vs Throw
Maximum Intra-Image Contrast defined
Usefulness of Maximum Intra-Image Contrast as a Metric
VW50 Maximum Intra-Image Contrast (Iris off and Iris 1)
Examination of Intra-Image white and black levels
Constancy of White Examined
RS-1 White Levels
VW50 White Levels (Iris Off and Iris 1)
Constancy of Black Levels
RS-1 black level vs luminance
VW50 black levels vs luminance
VW50 Iris 1 vs RS-1 relative black level comparison

Part III - Dynamic Intra-Image Contrast Example (JVC RS-1 and Sony VW50)

Dynamic intra-image contrast test patterns
Sony VW50 Dynamic Contrast Examined
Intra-Image Contrast as a combination of static and dynamic contrast mechanisms
Sony VW50 Relative and Absolute Luminance - Iris 1 and Iris Off comparison
Sony VW50 Dynamic Gamma Examined

Part III (Continued)

Low APL greyscale test pattern
Sony VW50 Intra-Image Contrast and relative luminance - Iris 1 vs Iris 2 comparison
JVC RS-1 Intra-Image contrast examined
Source Material with a Dynamic Iris Projector
Sony VW50 Shadow Detail Examined
JVC RS-1 Examined
Examination of Theoretical Parameters of a VW50
Improvement factor of VW50 Iris 1 over Iris Off as a function of stimulus

Part III (Continued)

Comparison of RS-1 and VW50
JVC RS1 with a Sony VW50 DI - A Theoretical Treatment

Part IV - Summary / Acknowledgements / References

Dynamic Iris Performance Summarized
Final Summary
Acknowledgements

AVS Contrast Project Database

AVS Contrast Project Database
Static Contrast Test Patterns
Dynamic Contrast Test Patterns.

AVS Contrast Project - Static Intra-Image Contrast Results

AVS Contrast Project - Dynamic Intra-Image Contrast Results

Contrast Discussion Starts Here
Mark Petersen is offline  
Sponsored Links
Advertisement
 
post #2 of 200 Old 05-25-2007, 01:22 PM - Thread Starter
AVS Club Gold
 
Mark Petersen's Avatar
 
Join Date: Dec 2000
Posts: 4,671
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Purpose:

Contrast has long been a key metric when discussing the image quality generated by video displays, but in this era of dynamic irises, fixed irises, native only, high ANSI contrast but average on/off and high on/off but average ANSI contrast a lot of confusion exists over the pros and cons of different display technologies from a contrast perspective. This project was started as a way to provide a consistent and measurable contrast benchmark that can be applied equally to all display technologies to help understand their contrast strengths and weaknesses. To this end, two sets of arbitrarily generated test patterns are provided which can be used as a framework for analyzing the relative differences in intra-image contrast for both native and dynamic iris projectors alike. Two test patterns were used so that the effects on contrast from both static and dynamic mechanisms can be fully captured.

The hope is that AVS members will use these included test patterns and supplied methodology to measure their display devices and add them to the AVS contrast database as an ongoing effort (hence the reason this thread was made "sticky"). The pages below provide a technical discussion of this contrast benchmark and how it can be applied to video projectors. Two off-the-shelf projectors were used for this example discussion, one is a native projector (JVC RS-1) and the other is a projector that utilizes a dynamic iris (Sony VW50 Pearl). At the end of this discussion is a section that contains the actual ongoing AVS Contrast project data and results which will be continually updated as information is provided.

Terminology:

To avoid confusion lets define some key terms which will be encountered in this thread:

Intra-Image Contrast - This is often referred to as simultaneous contrast and it refers to the contrast within regions of one specific image at the same point in time. ANSI contrast is a well known example of intra-image contrast because the black and white measurements are done while both checkerboxes are displayed simultaneously on the screen. As we will see, virtually all forms of contrast measurements in the AVS contrast project are either a direct measurement of intra-image Contrast or extrapolated from intra-image measurements. The term intra-image contrast can apply to any type of an image but in this discussion it is exclusively used to describe the contrast of test patterns and the term intra-scene contrast is used to describe real-world images from video and film sources.

Static Intra-Image Contrast - In this thread we use this term to describe the intra-image contrast created by projectors that don't perform electronic alterations of the image to boost brightness of portions of an image that are not already at full white. Native contrast (i.e. non-dynamic iris) projectors and fixed iris projectors are the most well known examples of projectors of this category. In this discussion however we will also use the term to include the contrast effects from a dynamically changing iris aperture (but not the brightness boost from dynamic gamma). Most (but not all) DI equipped projectors perform some form of gamma boost so the minority that do not can be fully treated using the static intra-image methodology. As we will see, Dynamic iris projectors that utilize gamma boost can and should be measured for static intra-image contrast which is useful for comparison purposes.

Dynamic Intra-Image Contrast - In this thread we use this term to describe the intra-image contrast created by projectors that can selectively boost the brightness of less than 100% full white pixels in some images. This contrast boost improves shadow detail and is performed by boosting the stimulus of pixels (not already at their maximum) which is currently always done (but doesn't have to be) in combination with a reduction in iris aperture. The boosting of the stimulus of pixels that aren't at their maximum is called Dynamic Gamma. Dynamic Iris projectors are the most well known examples of this technology and benefit the most from dynamic intra-image contrast, although as we will see a non-DI equipped projector can also be measured using the same methodology and test patterns as a DI equipped projector which is useful for comparison purposes.

Maximum Dynamic Range - We will use this term to refer to the maximum range of white to black that a projector is capable of achieving but not necessarily at the same time in the same image. This is usually expressed by the sequential on/off contrast of a projector. As we will see, this isn't the same as the maximum intra-image contrast because DI equipped projectors can not achieve full white and dark blacks at the same time in the same image.

Maximum Intra-Image Contrast - This term represents the largest range of black to white contrast that a projector is capable of achieving in an image at one time. As contrast is a ratio of white to black, determining the maximum intra-image contrast entails using a 100% stimulus white reference level to ascertain the maximum intra-image contrast. As we will show the on/off contrast of a native projector provides a surprisingly close approximation to the maximum intra-image contrast. DI projectors on the other hand do not achieve full white and full black simultaneously in the same image so mechanisms other than simple sequential on/off contrast measurements are needed in order to determine both their maximum intra-image contrast as well as their dynamic contrast benefits.

Intra-Scene Contrast - When used in this thread this term refers to intra-image contrast in video or film sources rather than the images used in a specific test pattern to determine intra-image contrast. Intra-Scene contrast is complicated by the luminance distribution of the image, the spatial relationships between light and dark pixels as well as the gamma used by a projector. Because of these complications, it must be stressed that intra-image contrast and intra-scene contrast are not the same. Measuring intra-image contrast is still useful in providing a basis of comparison between projectors and higher contrast results from this intra-image benchmark generally equates to improvements in intra-scene contrast in much the same way that high on/off contrast equates to improvements in low APL images and improvements in ANSI contrast equates to improvements in higher APL images. Despite the complexities of Intra-Scene contrast, it is possible to compare specific regions of well understood scenes (where both the luminance and spatial relationships are known) and determine roughly how various intra-image contrast metrics may affect these scenes (see this cine4home article for example: http://www.cine4home.de/Specials/ANS...NSIvsONOFF.htm).

Background:

Traditionally two forms of contrast measurements have been used to measure contrast. Sequential On/Off contrast uses alternating full fields of 100% white and 0% black to determine the maximum brightness and darkness that a display device is capable of achieving. ANSI contrast on the other hand uses a 4x4 checkerboard pattern of full white and full black to determine the contrast achievable when 50% of the image is black and the other white. ANSI contrast is often said to be the only metric of intra-image contrast because its contrast is measured with both the white and dark patterns displayed simultaneously. It's important to realize that this is a myth and that both ANSI and on/off contrast play important roles in intra-image contrast. Dark, low APL images (both test patterns and video/film scenes) are more heavily influenced by the on/off contrast while bright content is more heavily influenced by ANSI contrast. On average, movies are inherently a dark medium so on/off contrast plays a large role although ANSI still plays an important role in bright scenes (and sporting events, etc.).

Dynamic and Static Intra-Image Contrast:
Segregating contrast into dynamic and static contrast categories gives the mistaken impression that there are two unique types of contrast. The reality is that there is only one type contrast and that the terms dynamic and static contrast really refer to the mechansims employed by the display device to render contrast in an image. Display devices with dynamic contrast mechanisms have become common only fairly recently. Determining static contrast parameters and also dynamic contrast parameters with these display devices is required because they use a mix of static and dynamic mechanisms.

In this thread we only look at intra-image contrast and we employ two different techniques and two different sets of test patterns to examine the full benefits of contrast from both static and dynamic mechanisms. Because the same intra-image contrast is being measured in both cases (but with different techniques) there is considerable overlap between both discussions and we will see that both technologies can and should be compared using the same techniques and methodologies.

What causes Intra-Image Contrast?
We all know that light from a lamp or light source creates the whites in an image, but we will also examine the constancy of the brightness of white regions in both native and dynamic iris projectors. The real contributor of display contrast however is the black level in regions of an image. The black level of these regions is the sum of the minimum black level of the device (all pixels off) coupled with variance in dark regions of an image as it is influenced by the luminance in the bright regions. Variations in the amount of luminance (ie more pixels turned on or off) or variations in the intensity (brightness) of the luminance will have profound changes on the black level in regions of an image.

The underlying reason for the variance of intra-image black levels is due to leakage and scatter of light penetrating the dark regions of the image. The factors that affect light leakage and scatter are complex and include factors such as the device/pixel characteristics of the display device, the effectiveness of light traps, polarizors and internal light baffling, the use and degree of internal irises and also optical characteristics such as dispersion and distortion in the lens. These factors all combine to reduce the overall contrast of the system. In this project we address only the system contrast measured at the projector and not contrast effects from individual sources such as the room or screen (although effects from the room and screen may be addressed at a later date).

Intra-Image Contrast as a function of Luminance.
As mentioned the causes of intra-image contrast are complex and caused by many factors. This complexity also affects contrast measurements because both the geometry and overall amount of luminance can play a role. As an example, the washout effect on dark regions from bright regions in an image varies as one moves further away from the bright region. We could for example construct two different test patterns with different geometries but with the same overall luminance and measure different contrast results. If the geometries were kept relatively similar however we can see contrast change as both the overall amount and/or intensity of luminance is changed.

This relationship of contrast vs luminance (both in amount and intensity) is an important concept because it helps us to tie together the intra-image contrast benefits from both on/off contrast and ANSI contrast and it shows that they influence Intra-Image contrast in different regions of the luminance range. By using a fixed set of test patterns (even with an arbitrary geometry) that varies by the amount of luminance, a synthetic benchmark of contrast vs luminance can be obtained. Applying this contrast benchmark to projectors allows us to infer relative performance differences between projectors.

This methodology is similar to measuring and applying traditional ANSI contrast but with the added benefit that the contrast performance in the low luminance range can also be ascertained. As we will see, for DI-equipped projectors, utilizing low luminance contrast test patterns is the only way to gauge intra-image contrast with dark material because the low APL of these patterns allows the dynamic iris to close.

It should be stressed that the shape of the contrast curve is affected by the geometry and the amount and intensity of the luminance of the test patterns used. As we will see, the shape of the contrast curve using the test patterns generated for the AVS contrast project is completely arbitrary but it still follows an inverse power curve relationship between contrast and luminance that was predicted by AVS forum member Erik Garci.

Geometrical differences between each of the test patterns within this suite cause some variation in this relationship but overall the measurements correlate surprisingly well to this relationship. This points out a key point however that the precise shape of the static intra-image contrast curves that will be presented is synthetic and influenced by the arbitrary geometry of the test patterns chosen. This is only one such possible benchmark, but it nonetheless allows us to make some interesting comparisons and to infer relative performance differences versus luminance that we can't do otherwise. It is only one possible benchmark but as far as I know it's the only one of its type in the industry.

As we will see these test patterns and associated methodology also has the surprising benefit of allowing us to determine the maximum intra-image contrast of a display device. The maximum intra-image contrast occurs when the overall amount (area) of 100% full white luminance approaches 0 in an image. Interestingly, we will show that the maximum intra-image contrast is not dependent on the geometry of the test pattern being used but using low APL test patterns is the only reliable way to determine it.

To illustrate what static intra-image contrast vs luminance looks like, the graph below shows a theoretical plot of contrast vs luminance which does not take into account geometric factors and was created using a mathematical model by AVS forum member Erik Garci (An online version of this model can be found here: http://home1.gte.net/res18h39/contrast.htm). This model and associated plot shows a classic inverse power function that Erik derived for his model.



As one can see from this graph of theoretical intra-image contrast, traditional ANSI contrast corresponds to the 50% luminance end point and traditional on/off contrast corresponds to the 0% luminance end point.

If we compare this model against the static contrast test pattern suite that we've created we can see a relatively close correlation to this inverse power function relationship even though the test patterns used were constructed independently and without any prior knowledge of the contrast model. The graph below illustrates this comparison. Deviations from this model are likely due to variations in the geometry of each individual test pattern used in the suite and in fact we see this in a slight jaggedness of the curve particularly at the 2% luminance point which also happens to correspond to a test pattern where the geometry happens to place a higher percentage of white content around the dark measurement region.



Rather than plot the x-axis luminance scale linearly as is done above, it's useful to plot the results of each test pattern with an equal spacing so that the results from each test pattern in all luminance ranges can be more easily seen. Plotting the data this way makes it easier to see the full range of intra-image contrast values but it does alter the shape of the curve from the expected inverse power curve relationship. The reader should keep this in mind as nearly all of the graphs and data presented will use this graphing convention even though it alters the shape of the curve.



Test Pattern Suites: Two sets of test patterns have been developed by William Phelps and I for characterizing intra-image contrast. The first set of patterns is designed to measure maximum intra-image contrast and this is accomplished by varying 100% full white luminance by area. Luminance is varied in this way from 50% (modified ANSI) to 0.1% (0.1, 0.2, 0.5, 1.0, 2.0, 5.0, 10, 15, 20 and 50%). For each step on the luminance scale, two test patterns are used to determine the white and black levels for that luminance step. Using two test patterns ensures that the probe placement does not change between white and black readings. A separate white reading is also needed at each luminance step because the constancy of the white levels can not be guaranteed (which is particularly true for a dynamic iris).

All of the patterns are designed for center probe placement which does not change which helps to ensure accuracy. The 50% pattern is a modified variation of the ANSI pattern designed for the same center probe position used by the other patterns. Full field white and Full field black patterns are also included so that traditional on/off contrast can be measured at the same time.

One key point about this set of Intra-Image test patterns is that they measure intra-image contrast by utilizing full white which is needed to determine the maximum contrast of a display device. By utilizing 100% full white these results are also not dependent on projector gamma. 100% full white also allows us to determine the maximum intra-image contrast of a display device. Measured in this way we can see the full benefits to absolute intra-image contrast from both dynamic and static display technologies, but we are unable to measure the full benefits of dynamic contrast which occurs with less than 100% stimulus white and which help to improve shadow detail. It was for this reason that the 2nd set of test patterns was developed.

Black Test Pattern for 50% luminance (modified ANSI pattern designed for center probe position)



White Test Pattern for 50% luminance (modified ANSI pattern designed for center probe position




Black and White Test Patterns for 1% luminance range





The second set of test patterns are designed to measure the full benefit of intra-image contrast from display devices that dynamically enhance the contrast (e.g. dynamic contrast). These test patterns achieve this by using the same geometrical pattern but with luminance varied by intensity rather than by area as is done with the static contrast test patterns described above. Using the same white area in each pattern, the intensity of the white regions are varied from 100% down to 1% (1,2, 5, 10, 20, 30, 40, 50, 60, 70, 80, 90, 100).

In this discussion we will use the naming convention of "% pixel stimulus" as a way to describe the way in which discrete changes in pixel brightness are modulated from 0 to 100% rather than the 0 IRE and 100 IRE convention. This was done to avoid confusion between PC/Video levels, 7.5IRE black enhance, etc. It's worth mentioning however that the test patterns were designed with video levels (16-235) in mind and the proper application of video levels (as well as properly set brightness and contrast controls) are critical to ensure accurate results with this set of patterns.

It should also be pointed out that in this discussion we use the term luminance (either relative or absolute) to mean light that that is being generated by the projector in response to the number of pixels of the specified % stimulus.

We should also point out that display gamma plays a role in the brightness of these dynamic contrast test patterns and therefore contrast is also a function of display gamma for these measurements and this will be explored in more detail later.

In order to provide continuity between the static and dynamic suite of test patterns, the same 0.5% (by area) luminance/low APL white pattern from the suite of static test patterns is used for all off the dynamic test patterns. The 0.5% pattern was chosen because it's the largest pattern which reliably fills the probe area while also being small enough to force a dynamic iris projector to use its smallest aperture and presumably the largest gamma boost settings so that the maximum benefit of dynamic contrast can be seen for all luminance ranges from 100% to 1% stimulus.

Both sets of test patterns were designed for ease of use so that any forum member with access to a relatively inexpensive light meter such as an AEMC CA813 will be able to use these patterns to determine the static and dynamic contrast of their projectors and contribute these results to the AVS contrast project. Both sets of test patterns will contain a small readme file which describes how best to use these patterns. This data will be collected and maintained in an ongoing effort.

One word of caution in using these test patterns: It's expected that there will be considerable variation in the data collected because of unit to unit variability in specific projector models and more so between probe brands and specific models of probes. The reader should be cautioned in drawing conclusions from these comparisons in much the same way as they do in comparing ANSI contrast and on/off contrast from various sources. Relative comparisons between the same probe type and better yet the same probe type from the same forum member are likely to yield more accurate and meaningful results. If enough forum members contribute to the project it may even be possible to get a much better feel on the variability of probe brands by examining the statistics of each.
Mark Petersen is offline  
post #3 of 200 Old 05-25-2007, 01:22 PM - Thread Starter
AVS Club Gold
 
Mark Petersen's Avatar
 
Join Date: Dec 2000
Posts: 4,671
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Static Intra-Image Contrast Results for a JVC RS1 projector:

The graph below shows the contrast results for a JVC RS1 projector using the static intra-image contrast test patterns. The 50% luminance point represents the contrast achieved using the modified ANSI contrast test pattern (which yields nearly identical results to the standard ANSI test pattern within a percentage point or so). Further down the scale we see the results for patterns with decreasing luminance. This particular RS-1 did not yield as much sequential on/off contrast as is typical for the model (~9,600:1 on/off CR at a 2.35 throw and 10,300:1 at longest throw but with 306:1 ANSI CR) but it is nonetheless the sample that we have picked for applying the test patterns and methodology used here to describe the AVS contrast project.



Previously we postulated that the maximum intra-image contrast performance of a native contrast projector is closely approximated by the on/off contrast of the projector. To further test and reinforce this point we've added the 0.5, 0.2 and 0.1 test patterns to the suite which allows the projector to achieve close to the sequential full black reading even though a relatively large number (but small percentage) of pixels are contained in the image and at their maximum (100% stimulus) brightness. From the data for the 0.1% test pattern we see that the black reading is within 6% of the sequential full black reading even though no less than 2,074 full white (100% stimulus) pixels surround the probe area being measured.

Measuring intra-image contrast at values that are well over a factor of 10 higher than the ANSI contrast and close to the sequential full on/full off contrast values may surprise some people and this is one of the points of the static intra-image test patterns in that it vividly shows that intra-image contrast is affected by both the on/off contrast and the ANSI contrast of a projector and that it can be much higher than the ANSI contrast metric alone would suggest.

As previously mentioned, the shape of the curve is somewhat arbitrary owing to the geometry of each test pattern in the suite. This by itself would be relatively uninteresting, but it does have several useful attributes. The first is that it's useful when used as an intra-image benchmark for comparison purposes. The graph below illustrates this point by showing the intra-image contrast of an RS-1 using this test suite for both the longest and shortest throw.



The second and perhaps the most surprising attribute of these test patterns is that it allows one to determine the maximum intra-image contrast of native, fixed iris and dynamic iris projectors a like. This is a value which as we will see is independent of the geometry of the test pattern used.

Definition of Maximum Intra-Image Contrast: As we will see in subsequent plots, the white level in the lowest luminance test patterns does not vary significantly for either native or dynamic iris projectors. The white level in the DI projector is reduced drastically further up the luminance range owing to the iris reduction but in the low luminance range the iris is fully closed and the white level is relatively constant. Because the white level in this low luminance region is constant we can measure this value along with the minimum black level of the projector which are then used to determine the maximum intra-image contrast.

Maximum Intra-Image contrast = (measured white level using low luminance test pattern) / (measured sequential full black)

As we can see from a native projector such as the RS-1 this approximation yields a number very close to the sequential on/off contrast of the projector. The plot below includes this maximum intra-image contrast at 0 luminance (ie as the area of the white region of the test pattern begins to get very close to 0).



Next let's apply these same test patterns and methodology to a dynamic iris projector to determine its maximum intra-image contrast performance. In this example we display two curves in the graph below showing the behavior of iris off and iris 1 modes. Both modes yield the same number in the higher luminance range because the iris aperture is the same (iris open). However, in the lower luminance region we can see the relatively modest static intra-image contrast benefits of a reduced iris aperture. As we will see a dynamic iris has profound effects on both the black level and shadow detail of an image but the intra-image contrast is much less than one would assume based on the on/off contrast metric alone and much less than if a projector natively had the same on/off contrast ratio.

The graph below applies the same maximum intra-image contrast definition that we used earlier to both the iris off and iris 1 modes and it clearly shows that intra-image contrast is unrelated to the sequential on/off contrast of a DI equipped projector. As we will see this is due to the fact that both the full white and black levels are reduced in the low luminance range due to the reduced iris aperture.



This is an important result from this project in that it shows that the only way in which to gauge the maximum intra-image contrast for dynamic iris projectors is through the use of low luminance test patterns that force the iris into it's smallest aperture and then to measure both the white and black levels within this aperture.

The usefulness of Maximum intra-image contrast as a metric: The maximum intra-image contrast metric provides a value for the greatest range of contrast that a projector is capable of rendering in a specific image. The maximum Intra-Scene contrast however is likely lower than this value because of the elevated black levels and complicated luminance histograms and geometries that exist with video or film images. A projector with higher maximum intra-image contrast however will generally render dark images with better contrast in much the same way as higher on/off contrast renders better contrast in dark scenes with a native projector.

Maximum intra-image contrast is not the only metric of contrast however. It should be stressed that that Intra-image contrast performance in higher luminance ranges also plays an important role (which is also captured with this static contrast benchmark). Black level performance also plays a role as does shadow detail (i.e. contrast with less than 100% full white). Black level performance can be gauged using the static contrast test patterns and will be examined below while shadow detail requires the use of the dynamic contrast test patterns which will be covered in the next section.

Examination of Intra-Image White and Black Levels:
Because both the white and black measurements are recorded separately rather than simply the end contrast values, it's possible to examine both values separately in more detail. This is a side benefit to using these intra-image test patterns and methodology.

It's also important to point out that attempting to compare absolute differences of black level and brightness between projectors is notoriously difficult due to measurement errors that occur in even slight differences in probe distances. For this reason all of the black and white readings between projectors in this contrast project use relative values (from 50% luminance - i.e. ANSI CR). The reader is urged not to draw conclusions of the absolute differences in white and black levels but only the relative change from the ANSI (50%) end point. Absolute differences between modes within the same projector can be ascertained so long as the setup (probe position, angle, etc.) is left unchanged. (I should mention here that it may be possible to easily calculate the absolute black and white levels between any luminance point in the set when working backwards from the ANSI contrast measurements and the relative data presented here. This is a work in progress however and the reader is urged to make only relative comparisons until this conjecture has been tested and proven).

Constancy of white: The plot below shows the constancy of 100% full white measured as luminance is varied by area (ie smaller white boxes) in a JVC RS-1. On average the white level is relatively constant and does not deviate more than about 6% across the range. This constancy of white level is an important point which allows us to extrapolate brightness levels and in turn contrast when readings can't be directly obtained.



Close examination of RS-1 White Level


As an interesting aside, this detailed examination of the white level also uncovered one surprising aspect about LCOS which is that windowed (less than full field) white levels are actually slightly brighter than full field white. This is a result peculiar to LCOS that wasn't expected. This phenomenon showed up in the dozen or so separate measurements that I took from the RS-1 and also on another dozen measurements taken from the VW50.

In each case full field white is always slightly less than windowed white by about 5% or so. This phenomenon was also confirmed by William Phelps who said that that every LCOS (both DILA and SXRD) projector that he has measured has this feature. One possible theory that may explain this behavior is that it represents the added brightness that occurs when polarized off state light is shunted back into the lamp housing by the polarizing beam splitter which is then re-polarized and reused. This is only a theory however, the point here isn't to explain this phenomenon but to point out that in depth contrast measurements such as this can yield insight into technology behavior that isn't expected.

The end result is that the white level in the low luminance region of a LCOS projector may be slightly higher than is suggested by the full field white values used in traditional sequential on/off contrast measurements. This implies that for a native LCOS projector that the intra-image contrast in this low luminance region may actually be slightly higher than the on/off contrast metric and in fact we see this with the maximum intra-image contrast result on the RS1 which is slightly higher than the on/off contrast.

This is an important result in that it shows that as far as intra-image contrast is concerned, our definition of maximum intra-image contrast may actually be a more accurate metric of low luminance intra-image contrast than is the traditional sequential full on/ full off metric even for native projectors. This point will be driven further home when we examine the white levels of the VW50.



The plot above shows the constancy of 100% full white in the VPL-VW50. Of note is the significant drop-off in brightness between the 2% and 5% luminance regions which is due to the iris reduction and the inability of the VPL VW50 to boost brightness at 100% full white sufficiently to make up for the light loss due to the iris reduction. Of particular interest is the constancy of the white level from the 0.5% to 2% regions which suggests that the iris is fully closed in this region and that the constant white level in this region is therefore an accurate predictor of the maximum intra-image contrast metric that we have defined.

What is very clear from this diagram is that the iris open white level is not an accurate predictor of intra-image contrast in low APL images once the iris aperture has been reduced. Clearly, the white level in this low luminance region (iris closed) must be directly measured to draw conclusions about intra-image contrast with a DI equipped projector. This is an important point because it validates the approach of using low luminance test patterns to measure intra-image contrast. Without these sorts of test patterns and methodology any discussions of intra-image contrast in low luminance regions is at best an ineffective exercise.

Constancy of black levels:
The plot below shows the relative black level change vs. luminance in a JVC RS-1 projector. As we have previously mentioned the relative black level (like the relative white level) is calculated as a normalized percentage from the ANSI CR point. As can be seen from this diagram the constancy of the black level is radically different from the constancy of the white level and this graphically shows the obvious result that while white levels are relatively well controlled and modulated at a desired value, the black level is not.



The plot below shows a graph of the VW50 relative black level for both iris off and iris 1 modes. This clearly shows the biggest advantage of a dynamic iris projector, a significantly reduced black level of an image where it is needed the most - in the low luminance region. This is a region where the negative effects of elevated black levels are the most apparent and in this region the DI delivers over a 4x reduction in relative black level compared to iris off.

As an interesting side note, when performing this particular comparison the same identical probe position could have been used to determine absolute comparisons between iris off and iris 1 modes but this was not done so only the relative differences are shown. The ANSI contrast and overall brightness is the same in both cases though so that in actual practice a relative comparison such as this will likely yield (in this particular instance) the same results as absolute comparisons.



As we have mentioned, it is difficult to attempt to compare absolute white and black levels on different projectors measured at different times with different spacing between lens and probe sensor. We can however make relative comparisons when the data is normalized and the graph below is an example comparison between the VW50 and the RS-1. In this comparison we examine the low luminance region from 0.1-20%.



As can be seen from this graph the RS-1 (at not quite its furthest throw) achieves a higher initial reduction in relative black level compared to the VW-50 from ANSI up until the iris closes on the VW50 at which time the situation is reversed and the VW-50 has a higher relative reduction in black level. Eventually at the very bottom the RS-1 begins to catch up, and in fact the relative reduction of both projectors is very close at the very bottom of the luminance range.

As we can see from this graph, the dynamic iris is capable of producing some profound changes to both the black and white levels. It's interesting to point out that the shape of the black level curve between both projectors is relatively similar except around the region where the iris aperture closes. In this region the non-linear (ie discrete steps) way that the DI closes differs from the smooth transition of a native projector and it may in some cases and with some content deliver lower black levels in this luminance range than is suggested by it's on/off contrast.

To summarize the results from this section, the dynamic iris provided only a modest increase in static intra-image contrast but with significant improvements in black levels. Interestingly, the sequential on/off CR metric for the VW50 with iris 1 accurately predicts the black level while the same is not true when it is used to predict the maximum intra-image contrast. As we will see in the next section, dynamic contrast also needs to be examined to get a complete picture of the benefits of a dynamic iris and this leads us to part III of this discussion.
Mark Petersen is offline  
post #4 of 200 Old 05-25-2007, 01:22 PM - Thread Starter
AVS Club Gold
 
Mark Petersen's Avatar
 
Join Date: Dec 2000
Posts: 4,671
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
So far we have discussed the contrast results using 100% full white test patterns which traditionally have been used to measure contrast because they provide the maximum stimulus needed to determine the full contrast range of a projector. This provides us with useful information and insight but it doesn't give us the complete picture because it does not take into account the dynamic contrast benefits of DI equipped projectors that are capable of boosting the brightness of pixels that aren't already at their maximum brightness (which includes most DI projectors). By boosting the brightness of these pixels (via dynamic gamma) an improvement in shadow (less than 100% full white) detail is obtained which isn't captured by the static contrast results because only the maximum brightness results are being captured.

To capture the benefits of dynamic contrast we employ a 2nd set of test patterns which use the same geometry as the 0.5% luminance static contrast test pattern, but instead of using 100% full white we vary the intensity of the white from 1 to 100% stimulus in increments of: 1, 2, 5, 10, 20, 30, 40, 50, 60, 70, 80, 90, 100. Utilizing a very low overall luminance test pattern such as 0.5% helps to ensure that the dynamic iris is in its smallest aperture and that the maximum degree of gamma boost is being applied. Using the same geometry from the 0.5% static contrast test pattern also allows us to maintain continuity between both sets of test patterns and between the contrast results of both the static and dynamic contrast tests.

Example 100% stimulus black measurement test pattern (0.5% luminance by area)


Example 100% stimulus white measurement test pattern (0.5% luminance by area)


Example 20% stimulus black measurement test pattern (0.5% luminance by area)


Example 20% stimulus white measurement test pattern (0.5% luminance by area)


Proper video level, brightness and contrast settings are a must for this type of testing. The results of this test pattern are also dependent upon the display gamma used. As we will see the display gamma is an integral factor in dynamic contrast and there is no way to separate the two. Quite the contrary, measuring dynamic contrast also gives us some useful and interesting information about the gamma of the video display device.

VW50 dynamic contrast examined
A dynamic iris is a doubly modulated technology, meaning that the brightness is dependent upon both the modulation of the iris aperture and the modulation of the gamma boost. Rather than try to characterize every possible iris setting and gamma boost, we instead focus on characterizing the two extremes of dynamic iris behavior which in turn define the bounds of all other possible iris and dynamic gamma settings. Using the 0.5% test pattern with the more aggressive iris 1 mode allows us to characterize the upper contrast bound while simply placing the VW50 in iris off mode allows us to characterize the lower bound. As we will see, using the 0.5% test pattern does not characterize the maximum contrast of the VW50 (which happens when one approaches 0% luminance by area) but it still provides for some useful comparisons and the maximum contrast of the VW50 can be easily extrapolated which we will do later.

The chart below shows the dynamic contrast results between iris off and iris 1 modes of the VW50.



Of particular interest is the shape of each curve. The iris off mode follows a typical luminance vs. stimulus curve which is predicted by the VW50 display gamma. The iris 1 mode on the other hand shows a significant boost in contrast early in the lower stimulus range which contributes a significant improvement in intra-image contrast for pixels in this region. The curve flattens out towards the high stimulus end of the curve. This clearly shows the contrast benefits of the gamma boost being applied by the Iris 1 DI mode. Also of interest is the difference in contrast at the 100% stimulus point on the curve. At this region of the curve, no additional boost in brightness is possible because the projector has hit its maximum brightness and additional boost is not possible. At this 100% end point of the curve the contrast results should match the static contrast results using the same 0.5% test pattern (within measurement variance). In this particular case however the static and dynamic VW50 data sets were performed at different times on different projectors but even in this case there is only a 17% variance between both cases.

Viewed from this perspective we can see that dynamic and static intra-image contrast data sets are really two perspectives of the same intra-image contrast with one capturing variations in contrast due to changing amounts of luminance and the other the intensity of the luminance. The graph below illustrates this relationship by showing how we can vary the amount of light (via test pattern luminance) and also the brightness of the light to see how each affects intra-image contrast. In this graph only the dynamic contrast results from the 0.5% pattern were captured and used, but given enough time we could have done the same treatment for each of the geometric patterns for the static test pattern suite and used this information to generate a complete 3-dimensional surface perspective of this relationship



Dynamic Gamma examined:
Because the setup for both the iris on and iris off data sets was identical (probe position, etc), it is possible to directly compare the absolute brightness of both iris modes of the VW50. We can also take this absolute brightness data and normalize it to determine the gamma of both modes. The two graphs below illustrate this point.

Relative Luminance Graph showing Gamma Effect


Absolute Luminance - Iris 1 vs Iris Off


It's interesting to note that the shape of the dynamic gamma used in iris 1 is significantly different than a typical display gamma (usually in the 1.8 - 2.5 range) such as is used in the iris off mode. The dynamic gamma boost in the low stimulus range allows the VW50 to counteract much of the brightness loss due to the iris reduction in the low stimulus range and as we have seen this boost significantly improves contrast in this range. As we can see from the associated brightness curve however not all of the brightness loss due to the iris reduction is compensated for (not even in the low stimulus regions).

Even more interesting, if we calculate the gamma from this luminance information (see graph below) we see that the gamma in the upper pixel stimulus regions (> 50% stimulus) is actually less than 1.0. This is due to the fact that the VW50 cannot boost brightness in this region to counteract the light lost due to the reduction in iris aperture. This extraordinarily low gamma is the root cause of the well known brightness compression artifact that occurs in bright regions of low APL scenes with DI projectors.



Much has been discussed about the benefits of dynamic gamma when coupled with a dynamic iris but this clearly shows that the most serious side effect of a DI - brightness compression - is directly attributable to dynamic gamma and the fact that a well shaped display gamma cannot be maintained simultaneously with gamma boost.

One interesting observation that one can see from this data is that if one looks at the luminance graph in the iris 1 mode there is a pronounced upswing in brightness from 90% to 100% stimulus region. As we will see, this luminance peak exists for the Iris 2 mode as well. Since the luminance curve tapers off as the curve approaches 90% stimulus (just before the big upswing at 100%) and every bit of luminance in this region is critical to allow for more differentiation between steps, one must assume that Sony deliberately added this peak to their gamma profile for both the iris 1 and iris 2 modes. I assume that this was done to give more pop at the very top of the luminance range in order to help break up some of the perception of brightness compression that occurs in this range. By adding this peak however, luminance is reduced even further in the middle stimulus range which makes white levels in this region less differentiated and even more prone to brightness compression.

This is a good example of the tradeoffs that are often made in product design. Invariably when such tradeoffs exist, some users will disagree with the choices that were made and want a method to strike a different balance. This is one of the key benefits of these test patterns and methodology as it allows a person to more closely gauge the effects of the many service mode parameters that exist in many DI products such as the VW100 and VW50 and to therefore better tailor these products to their tastes.
Mark Petersen is offline  
post #5 of 200 Old 05-25-2007, 01:22 PM - Thread Starter
AVS Club Gold
 
Mark Petersen's Avatar
 
Join Date: Dec 2000
Posts: 4,671
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
As an additional verification of the low APL luminance curve shown above, a 20 step greyscale test pattern was generated (included in the dynamic contrast test patterns) which contains two strips of 20 small blocks that vary by increments of 5% stimulus each. One strip goes from 100% to 0% and the other is reversed so that high and low luminance ranges can be viewed at once. The low overall luminance of this greyscale test pattern allows the DI aperture to close and the gamma to be boosted so that the effects of gamma boost can be readily viewed. Here is a picture of the greyscale:



What might not be obvious from this image is that the greyscale is similar to the dynamic contrast test patterns in that it has a low APL. This greyscale test pattern was also designed for use with video levels where 0% black is defined as RGB 16 and 100% white as RGB 235. The reader should be aware that when this image is viewed on a PC or other display device that the video levels and gamma of that display will interact with this image and skew the perception of the greyscale and the whites and blacks may be artificially crushed when viewing this image. It's still useful for illustration purposes however.

When displayed on a DI equipped projector such as the VW50, it's possible to see the effects on the greyscale of the various iris modes as well as iris off. The photo below is an example of one of the modes (I believe it was iris 1 but I can't be certain). Unfortunately I had taken photos of each iris modes but my camera wasn't tripod mounted and the other photos came out blurry.



From this photo the brightness peak at 100% stimulus that we saw from the graph above is readily apparent. Also apparent is the good shadow detail in the low luminance range and unfortunately also the brightness compression in the high luminance range. Unfortunately the contrast range of the digital camera also comes into play in this photograph and the BC is in fact not this bad. When viewed firsthand, about 4 discreet luminance blocks become undifferentiated when the DI is engaged. Since each luminance block is separated by an increment of 5% stimulus, even the merging of two blocks represents significant compression.

A quick examination of iris 1 and iris 2 modes was also performed on the VW50. Unfortunately our time with the VW50 was limited and the measurements taken were somewhat rushed so these results should be considered preliminary. The probe position was also moved during this time so only iris 1 and iris 2 comparisons are shown here as iris off wasn't measured at this same probe position. Hopefully owners of a VW50 can perform a more detailed comparison of iris 1 and iris 2 modes and submit their results to the AVS contrast project. The dynamic contrast and dynamic gamma results for the iris 1 and iris 2 modes are compared in the graph below.





These graph verify what others have noticed which is that iris 1 is the more aggressive mode and delivers better contrast improvements but with more brightness compression.

Native (non-DI) projectors measured for dynamic contrast?
Native projectors can also be measured using the dynamic contrast test patterns and methodologies even though they don't perform gamma boost (and their performance can be predicted by their fixed gamma profile). It is still useful however to perform such a measurement for comparison purposes. The graphs below show the Intra-Image contrast and gamma of an RS-1 (note how the gamma profile of the RS-1 is similar to the VW50 Iris off)





Source Content with a Dynamic Iris:
It should be pointed out that the content of the source material itself plays a very large role in the end result of the image that is displayed with a dynamic iris projector. As we have seen from the absolute luminance chart, gamma charts and greyscale photograph, the benefits of the DI occur primarily at the bottom half (0-50%) of the stimulus range and negative effects (BC) happen at the upper half of the stimulus range. If the source material does not contain bright content in a low APL scene then the negative effects of brightness compression and low absolute luminance are greatly reduced.

Since low APL scenes are by their nature dark and don't contain a lot of bright content, it's likely a safe bet that the tradeoff of enhanced shadow detail in low stimulus regions vs increased brightness compression in high stimulus regions within the same image is a positive one that favors many (but not all) low APL scenes. When viewed this way the overall luminance effects of a dynamic iris projector from the combination of both the iris reduction and the dynamic gamma ends up being somewhat analogous to a low pass filter in that bright content if it exists in the source material is reduced, compressed and altered more so than less bright content.

Dynamic Iris Shadow detail examined:
So far we have compared only the iris 1 and iris off modes of the VW50. By recording the luminance of both the bright and dark intra-image measurements (rather than just the end contrast result) we can easily make some useful and accurate extrapolations of the VW50 if it natively had the same on/off contrast as it does with the DI engaged. By definition this happens when the VW50 has the same white levels as it has with iris off but the black level performance that it has with the iris engaged. Furthermore, now that we know the absolute luminance limit of the VW50 with the iris 1 engaged we can accurately extrapolate what the dynamic iris performance would be if the DI attempted to maintain the same luminance throughout the range as it does when the iris is off. The graph below shows this extrapolation compared to iris off and iris 1 modes.



As can be seen from this example, if maximum gamma boost were applied, the VW50 would be able to counteract the reduction in brightness due to the reduced iris aperture up to about the 50% stimulus region at which point the maximum brightness of the VW50 is reached (with the iris aperture reduced) and no additional brightness can be achieved beyond this point. As we can also see, the bottom portion of this curve (up until the maximum brightness is reached) also corresponds to the contrast performance of the projector if it had the same on/off contrast natively as it does with a DI. It's also apparent from this graph that when maximum gamma boost is used that the entire region from about 50% on up is completely compressed.

As we can see the contrast benefits of a DI is a tradeoff between improved shadow detail and increased brightness compression, the degree of both being controlled by the amount of dynamic gamma applied. The graph above shows that the designers of the VW50 likely have the option (if they chose to) to achieve the full potential of the VW50's on/off CR in the low luminance range but at the expense of 100% brightness compression in the upper ranges.

It's clear from the graph above that the designers of the VW50 have struck a specific balance with the Iris 1 mode between the shadow detail that they wanted and the degree of brightness compression that they were willing to tolerate. When the source content requires maximum boost (such as this test pattern), the Iris 1 mode (which is the more aggressive of the two modes) responds with much less than the maximum amount of boost. There is still some noticeable BC, but it's much less than the worst case scenario.

Unfortunately with this same amount of boost, the contrast/shadow detail throughout the range is significantly less than the maximum possible which in turn means that it is also significantly less than what the projector would do if it natively had the same on/off contrast as it does with the DI engaged (which was measured in this case at 12,500:1). This should put to rest the notion that the shadow detail of a DI projector is a function of its on/off contrast as the reality is considerably more complicated.

It's worthwhile to point out that this information was first extrapolated from static contrast data several months before these dynamic contrast test patterns and methodologies were developed. The post that predicted this behavior can be found here.
http://www.avsforum.com/avs-vb/showt...&&#post9751112

The dynamic test patterns and methodologies have proven the conjecture in the earlier analysis while also providing measurements of the exact degree of gamma boost (and dynamic contrast) for both the iris 1 and iris 2 modes which are both less than the hypothetical case of maximum gamma boost.

We can also note that the contrast performance of the VW50 Iris 1 mode is significantly better (in this same low APL scenario) than the Iris off mode throughout the stimulus range and the benefit is larger in some ranges than others. In fact if we calculate the improvement factor that the iris 1 makes over the iris off mode we can graph this as we have done below. It's interesting to note that the improvement from the iris reduction alone makes a relatively small contribution and we can subtract off this fixed contribution and show the contribution from dynamic gamma alone as is done in the 2nd plot below:



Mark Petersen is offline  
post #6 of 200 Old 05-25-2007, 01:23 PM - Thread Starter
AVS Club Gold
 
Mark Petersen's Avatar
 
Join Date: Dec 2000
Posts: 4,671
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
We have previously displayed the dynamic intra-image contrast results of the RS-1 and shown that the usefulness of doing this (i.e. native projector measured with dynamic contrast techniques) is primarily for comparison purposes. So far we have (for the most part) avoided comparing different brand/model of projectors because the primary purpose of this discussion is to explain the framework that is being presented as a way to characterize contrast of native and DI projectors a like. Collecting contrast data and performing comparisons between projectors is the purpose behind the AVS contrast project that follows however so we will see how a sample RS-1 (with just okay on/off contrast but good ANSI - for an RS-1) compares to a sample VW50 (with good on/off contrast and typical ANSI - for a V50).



As we can see the RS-1 does surprisingly well as far as shadow detail and bests the VW50 throughout the stimulus range even though the VW50 has the higher on/off contrast. Also, as we can see, if the VW50 natively had the same on/off contrast it would best the RS1 throughout the range. Even more interesting, we can see that if the VW50 applied maximum dynamic gamma boost it would best the RS-1 in the bottom of the stimulus range (0-50%).

The situation at the very low stimulus range from 0-20% is also worth examining more closely. The graph below shows the affects on intra-image contrast from the RS-1 gamma modes and also the VW50 Iris 1 and Iris Off modes. As can be seen and is also commonly known, the gamma chosen in the low stimulus range has a direct affect on the contrast/shadow detail of the image (in this range). We can see for example that the VW50 Iris 1 performs very close to the RS-1 and it's only until higher up the stimulus range that the full benefits of the high RS-1 native contrast begins to clearly out distance the VW50.



The Sony DI comes to the RS1:
So far we have been comparing both the static and dynamic measured performance of a ~3000:1 native projector with a dynamic iris (the VW50) with a native projector that has a measured 9600:1 on/off contrast performance (at less than maximum throw). The DI equipped projector with much less native contrast has compared relatively well and in most of the performance ranges (both static and dynamic) the performance is surprisingly close. It's only in low APL images with regions where there is pronounced bright (pixels > 50% stimulus) content where the RS1 has clearly superior contrast performance (up to 2x).

This begs the question then - What would the contrast performance of the same RS1 be if it benefited from the dynamic iris of the VW50. From the data set that we have gathered we can extrapolate the performance from this hypothetical hybrid projector. We know the degree of luminance reduction for both the white and black levels of the VW50 DI and we also know the gamma profile when dynamic gamma is used. Extrapolating this information with the same aggressive iris 1 mode yields the interesting but hypothetical graph below:



The contrast improvements shown above are only one benefit of adding the Sony DI. The other chief advantage is that the minimum black level floor drops by 4x.

The last point that I'd like to make is that so far we have exclusively used the dynamic intra-image contrast results from the 0.5% pattern. As we did with the static contrast discussion it's possible to extrapolate what the dynamic performance would be if the luminance approached 0% which yields the maximum dynamic intra-image contrast graph below

Mark Petersen is offline  
post #7 of 200 Old 05-25-2007, 01:23 PM - Thread Starter
AVS Club Gold
 
Mark Petersen's Avatar
 
Join Date: Dec 2000
Posts: 4,671
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Dynamic Iris Performance Summarized:
Much has been written about dynamic iris performance from a subjective standpoint. At various times for example, there have been claims that a DI performs the same in low APL scenes as it would if it natively had the same on/off contrast performance. At other times people have said that the shadow detail is as good or even better than if it natively had the same on/off contrast. People have also claimed that the use of dynamic gamma allows the DI to completely make up for the light reduction caused by the iris reduction. What we have found is that the contrast performance of a DI is far more complicated than these perspectives and we can summarize the key points here:
  • The maximum intra-image contrast is fully characterized using the framework (static contrast methodology and test patterns) presented here.
  • Unlike a native contrast projector, the on/off contrast metric of a dynamic iris projector has little relevance in determining the intra-image contrast of a DI equipped projector.
  • The intra-image contrast of a dynamic iris projector has to be measured utilizing low luminance test patterns that allow the white levels to be measured when the iris aperture has been reduced.
  • The traditional sequential on/off contrast of a dynamic iris projector is a useful metric only for determining the minimum black level floor of the projector.
  • Black levels (above the minimum black level floor) do not follow the same relationship (vs. luminance) of a native projector because of the non-linear way in which the DI engages (ie discrete steps rather than smooth transitions) so in fact the black level may be significantly better in some luminance ranges than is suggested by it's on/off contrast.
  • The white level for 100% stimulus pixels can vary widely depending on the iris position - down to 1/3rd at smallest aperture of what it is when fully open.
  • Iris reductions alone make a relatively small improvement (~20%) to the maximum intra-image contrast.
  • Dynamic Gamma alone provides a significant intra-image contrast boost (up to 2.2x at 30% stimulus).
  • Dynamic Gamma significantly improves the intra-image contrast in regions with less than 50% stimulus but at the expense of brightness compression above 50%.
  • The Dynamic Gamma implementation in both the Iris 1 and Iris 2 modes of the VW50 apply much less than the maximum level of boost needed to overcome the brightness lost due to the iris reduction. This is true for almost all of the %stimulus range except possibly at the very bottom (<10% stimulus). For this reason the improvements in shadow detail is less than the maximum possible (i.e. less than what it would have if it natively had the same on/off contrast as it has with the iris engaged). Since less than the maximum gamma boost is applied brightness compression with both iris modes is less than it would be otherwise.
  • Despite the fact that many contrast parameters of the dynamic iris don't fully approach the levels suggested by the on/off contrast with the iris engaged, the results are nonetheless impressive when one considers that we're attempting to compare a projector with a 3000:1 native on/off contrast to what it would achieve natively (12,500:1 in this case) and in many cases the differences are not as much as one would expect.

We should also mention that only the maximum intra-image contrast, shadow detail and black/white level performance effects of a dynamic iris have been examined. Other effects such as hysteresis, iris pump, etc have not been examined.

Final Summary:
Using both sets of test patterns and applying the methodology described here is a relatively easy thing to do and it is no more difficult than performing other contrast measurements such as ANSI (although it does require more time and more measurements). It does however give us a framework for characterizing the contrast performance of all projector technologies including DI projectors. This framework allows us to perform objective and consistent measurements of contrast parameters and to fairly infer relative performance differences between projector technologies. In addition we have seen how it provides some useful insights into areas that one wouldn't expect such as small deviations with some technologies (LCOS) with white levels when full field patterns and windowed patterns are used.

There are many key points that have been presented but a few are worth summarizing again here:
  • With the invention of the dynamic iris projector, traditional sequential on/off contrast is no longer a relevant metric for determining maximum intra-image contrast.
  • The maximum intra-image contrast result is not dependent upon the geometry of the test patterns chosen and should serve as a replacement for the sequential on/off contrast metric when discussing intra-image contrast performance.
  • Determining maximum intra-image contrast must be done using low luminance test patterns that allow an iris in a DI equipped projector to close.
  • The static intra-image test patterns provide a relatively arbitrary measure of contrast vs luminance but they are useful for comparison purposes nonetheless.
  • Other arbitrary static contrast test patterns with different geometries could be used but the net effect would only be to change the shape of the curve between the maximum intra-image contrast and the ANSI contrast end points.
  • Intra-Image contrast is a single parameter that is influenced both by the amount of luminance and the intensity of the luminance in an image. Using the two techniques provided we can measure each contribution and also show how they combine to affect the overall intra-image contrast value. The terms static and dynamic contrast really apply to the projection techniques used and do not imply that there are two separate types of contrast. Two separate techniques are needed to fully determine intra-image contrast from both technologies however.
  • Using this framework also gives us insight into technology differences that we may not have expected (such as variations in white levels, black levels, etc.).
  • Due to technology dependent deviations in the white level at low luminance levels our definition of maximum intra-image contrast may yield more accurate results than traditional sequential on/off contrast even for native projectors.
  • Using this framework can help people adjust the performance parameters that affect contrast to better suit their tastes. An example is the service mode parameters in both the VW100 and VW50 projectors.

Acknowledgements/References:
  • I'd like to thank William Phelps (wm) for understanding the need for and value of this project and contributing his time to help develop the test patterns used and also to supply much of the early RS-1 and VW50 contrast measurements.
  • I'd also like to acknowledge the contrast work that Darin Perigo (Darinp2) has done in educating people on the science behind contrast and also for this excellent contrast reference: http://www.hometheaterhifi.com/volu...006-part-1.html and also for his illuminating discussions.
  • I'd also like to thank Erik Garci whose online contrast calculators have proven very useful in this project. These calculators can be found here: (Intras-scene contrast calculator) http://home1.gte.net/res18h39/intrascene.htm. (contrast calculator) http://home1.gte.net/res18h39/contrast.htm
  • I'd also like to thank HoustonHoyaFan for some interesting discussions on DI technology and also for uncovering the reference to this white paper (http://www.etconsult.com/papers/Bla...att%20Cowan.pdf
    which provided much of the initial inspiration in creating low APL test patterns:
  • I'd also like to thank all of the individuals who contributed to the early thread that kick started this project (Mr.Wiggles, Chris Wiggles, lovingdvd, gregr, stranger89, and many others that I apologize for not specifically naming)
    http://www.avsforum.com/avs-vb/showthread.php?t=761806
  • I'd also like to thank Alan Gouger for providing AVS, hosting this presentation and discussion and in general for supporting the community of serious home theater enthusiasts.
  • I should also mention that this thread supersedes the earlier static contrast work which was described in this thread: http://www.avsforum.com/avs-vb/showthread.php?t=781060 Unfortunately the addition of the dynamic contrast work and additional descriptions and explanations proved too large for this earlier thread so a newer thread with room to expand was needed (hence the creation of this one).
Mark Petersen is offline  
post #8 of 200 Old 05-25-2007, 01:23 PM - Thread Starter
AVS Club Gold
 
Mark Petersen's Avatar
 
Join Date: Dec 2000
Posts: 4,671
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
The AVS Contrast Project Database contains both dynamic and static contrast measurements as well as the graphs that are used in this thread. It can be downloaded below (Excel spreadsheet format).

Click here for AVS Contrast Project Spreadsheet

The Static Contrast Test Patterns (in .ZIP format) can be found here.
Click here for Static Contrast Test Patterns

The Dynamic Contrast Test Patterns (in .ZIP format) can be found here:
Click here for Dynamic Contrast Test Patterns

Instructions for using both sets of Test Patterns can be found here:
Click here for Instructions on both sets of Test Patterns

Update:
I forgot to include the 20 step low APL greyscale in the dynamic contrast test pattern zip file above so I will include it below. If DI owners can take photographs of the greyscale with different iris settings we can post them here so that people can get a feel for the differences in the greyscale from the different iris settings.
Here
Mark Petersen is offline  
post #9 of 200 Old 05-25-2007, 01:24 PM - Thread Starter
AVS Club Gold
 
Mark Petersen's Avatar
 
Join Date: Dec 2000
Posts: 4,671
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Note: The static contrast data that is presented here measures intra-image contrast as the amount of full (100% stimulus) white is varied (by changing the area of the test pattern stimulus). This data is updated as part of an ongoing effort. Check back often for additional contrast results on other projectors. For an in-depth discussion of these results please read the earlier posts that describe the static contrast methodology, test patterns and meaning of the results.


Static Intra-Image Contrast Results



Full White Level vs Luminance



VW50 Black Levels (Iris 1 vs Iris Off)



Relative Black Level vs Luminance (RS1 vs VW50 Iris 1)

Mark Petersen is offline  
post #10 of 200 Old 05-25-2007, 01:24 PM - Thread Starter
AVS Club Gold
 
Mark Petersen's Avatar
 
Join Date: Dec 2000
Posts: 4,671
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Note: The dynamic contrast data that is presented here measures intra-image contrast as the intensity of white is varied. Each test pattern in the suite uses the same geometry and only the intensity of the white is varied. A low test pattern APL is used with 99.5% of the pixels off and only 0.5% of the pixels turned on and providing the stimulus. A low APL test pattern such as this manipulates a DI projector to use an aggressive iris and gamma boost setting so that the upper bound of iris contrast performance can be determined.

It should also be pointed out that the contrast numbers at 100% full white correspond to the 0.5% luminance point in the static contrast results posted in the previous section.

This data is updated as part of an ongoing effort. Check back often for additional contrast results on other projectors. For an in-depth discussion of these results please read the earlier discussion which describes the dynamic contrast methodology, test patterns and meaning of the results.



Dynamic Intra-Image Contrast (low APL)



Shadow Detail (Dynamic Contrast with 0-20% pixel stimulus)



Relative Luminance showing Gamma Curve



VW50 Gamma vs Luminance



VW50 Absolute Black Levesl Comparison (low APL)



VW50 Absolute White Levels Comparison (low APL)

Mark Petersen is offline  
post #11 of 200 Old 05-25-2007, 04:45 PM - Thread Starter
AVS Club Gold
 
Mark Petersen's Avatar
 
Join Date: Dec 2000
Posts: 4,671
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Mark Petersen is offline  
post #12 of 200 Old 05-25-2007, 05:54 PM
AVS Special Member
 
mhafner's Avatar
 
Join Date: Nov 2000
Location: Switzerland
Posts: 4,606
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 6 Post(s)
Liked: 28
Good work.
Quote:


From the data for the 0.1% test pattern we see that the black reading is within 6% of the sequential full black reading even though no less than 207,360 full white (100% stimulus) pixels surround the probe area being measured.

0.1% of 1080p = 2073.6 pixels.
Quote:


Measuring intra-image contrast at values that are well over a factor of 10 higher than the ANSI contrast and close to the sequential full on/full off contrast values may surprise some people and this is one of the points of the static intra-image test patterns in that it vividly shows that intra-image contrast is affected by both the on/off contrast and the ANSI contrast of a projector and that it can be much higher than the ANSI contrast metric alone would suggest.

A very familiar result for CRT owners.
Quote:


As we have mentioned, it is difficult to attempt to compare absolute white and black levels on different projectors measured at different times with different spacing between lens and probe sensor. We can however make relative comparisons when the data is normalized and the graph below is an example comparison between the VW50 and the RS-1. In this comparison we examine the low luminance region from 0.1-20%.

Put a DI on that JVC...
Now someone please measure an LC 9 inch CRT.
mhafner is offline  
post #13 of 200 Old 05-25-2007, 06:00 PM - Thread Starter
AVS Club Gold
 
Mark Petersen's Avatar
 
Join Date: Dec 2000
Posts: 4,671
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
So who is going to be the first person to comment on this major addition to the AVS Contrast Project? I apologize for all of the verbiage but an in depth explanation of the methods, graphs and conclusions seemed sorely needed.

The previous work provided a good foundation for characterizing static contrast and this update contains a ton of dynamic contrast data and analysis. The Sony VW50 was chosen to provide an example application of the framework (test patterns, methods, etc.) that is described here. As far as I know this is the first detailed analysis of DI performance that has been publicly released.

The data presented here challenges a lot of conventional wisdom so I suspect the discussions will be lively. Please keep all critiques and comments professional and please don't label this effort as anti-DI or pro-DI. The point of this whole effort is to come up with a framework for characterizing contrast that can be applied equally and fairly to all technologies.

EDIT: It looks like mhafner beat me to it
Mark Petersen is offline  
post #14 of 200 Old 05-25-2007, 06:09 PM - Thread Starter
AVS Club Gold
 
Mark Petersen's Avatar
 
Join Date: Dec 2000
Posts: 4,671
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Quote:
Originally Posted by mhafner View Post

Good work.

Thanks!

Quote:


0.1% of 1080p = 2073.6 pixels.

Yup, looks like I missed a 0. I'll fix it. Thanks for pointing it out.

Quote:


Now someone please measure an LC 9 inch CRT.

Yeah I'd also love to see the results of a 9" LC CRT. I've been interested in measuring a CRT and after finally getting this stuff done I'll have some free time to do so. I'd also like to encourage others to show off the contrast of their projectors.
Mark Petersen is offline  
post #15 of 200 Old 05-28-2007, 05:08 AM
AVS Club Gold
 
bgosselin's Avatar
 
Join Date: Jun 2000
Location: Montreal,quebec,Canada
Posts: 1,078
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Very nice work! It would be interesting to compare those data with a 3 tube (king of onff contrast) and a Sharp XV-Z20000 (king of Ansi contrast)

Keep the good work!
bgosselin is offline  
post #16 of 200 Old 05-28-2007, 03:25 PM - Thread Starter
AVS Club Gold
 
Mark Petersen's Avatar
 
Join Date: Dec 2000
Posts: 4,671
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Quote:
Originally Posted by bgosselin View Post

Very nice work!

Thanks!

Quote:


It would be interesting to compare those data with a 3 tube (king of onff contrast) and a Sharp XV-Z20000 (king of Ansi contrast)

Yup those are two that I would definitely like to see measured too. It'll be interesting to see the contrast results from new projectors going forward too
Mark Petersen is offline  
post #17 of 200 Old 05-28-2007, 03:34 PM - Thread Starter
AVS Club Gold
 
Mark Petersen's Avatar
 
Join Date: Dec 2000
Posts: 4,671
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
I just went through the write up in the first 6 posts and corrected some typos while adding some additional clarifications. I had forgotten a graph that shows how iris 1 gamma varies over %stim and added that. I also realize that some of the charts in the static contrast portion use the term "luminance % (by area)" which is technically incorrect. It should really read something like % stimulus (by area)" or "test pattern stimulus % (by area). Just to clarify we're talking about varying the area of the test patten which in turn governs the amount of luminance that is measured. It's a lot of work to change the text in the graph and rehost them so I'll just mention this caveat here.
Mark Petersen is offline  
post #18 of 200 Old 05-29-2007, 09:33 AM
AVS Special Member
 
shodoug's Avatar
 
Join Date: Dec 2000
Posts: 1,262
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Very impressive.

Thanks to you, and those who helped you put all this together.

Best Regards,
Doug
shodoug is offline  
post #19 of 200 Old 06-17-2007, 06:46 PM
Senior Member
 
cjfrbw's Avatar
 
Join Date: Mar 2002
Location: Pleasanton, CA
Posts: 210
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
This is one of the most helpful and interesting posts I have ever read on AVS forum.
cjfrbw is offline  
post #20 of 200 Old 06-17-2007, 07:28 PM - Thread Starter
AVS Club Gold
 
Mark Petersen's Avatar
 
Join Date: Dec 2000
Posts: 4,671
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Quote:
Originally Posted by Shodoug View Post

Very impressive.

Thanks to you, and those who helped you put all this together.

Quote:
Originally Posted by cjfrbw View Post

This is one of the most helpful and interesting posts I have ever read on AVS forum.

Thanks for the kind words! I'm hopeful that there will be some updates soon with other projector types. Check back often
Mark Petersen is offline  
post #21 of 200 Old 08-03-2007, 04:02 AM
Senior Member
 
Thebes's Avatar
 
Join Date: Jan 2007
Posts: 447
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 22
Hi,

I'm trying to understand this chart :



If I have a good comprehension, you obtain the best result with the gamma "c" on the JVC HD1/RS1.

It seems to me that the gamma "C" is the lowest average gamma on JVC, in my memory it's around 1.9 and the others (Normal, A or B) are higher (around 2.1 or 2.2).

In my mind, the picture "seems" more contrasted when the average gamma is higher and I don't understand your result in this case (but it could be a mistake) ?
Thebes is offline  
post #22 of 200 Old 08-03-2007, 11:23 AM - Thread Starter
AVS Club Gold
 
Mark Petersen's Avatar
 
Join Date: Dec 2000
Posts: 4,671
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Quote:
Originally Posted by Thebes View Post

Hi,

It seems to me that the gamma "C" is the lowest average gamma on JVC, in my memory it's around 1.9 and the others (Normal, A or B) are higher (around 2.1 or 2.2).

In my mind, the picture "seems" more contrasted when the average gamma is higher and I don't understand your result in this case (but it could be a mistake) ?

You are correct about the gamma C mode on the RS-1 being the lowest average gamma on the JVC. What this means is that the luminance curve is flatter than if a higher gamma were used. At the bottom of the %stim range this means it initially is higher (more contrast) than a curve with higher gamma and this is born out in the measurements above. You can see this for youself by pausing on an image that has very dark shadows in it and then cycle through the gamma settings of the RS-1. The image may seem more "contrasty" with the higher average gamma modes on the RS-1 such as "normal", "A" and "B", but if you look only at the dark shadows there is actually more contrast in gamma mode "c". This is a good example on how applying gamma can result in tradeoffs between contrast with brighter details versus darker details.

Incidentally, when talking about average gamma modes there is no guarantee that the gamma doesn't change throughout the range, particularly at the bottom of the curve. Measuring gamma throughout the range as was done here rather than assuming a single average gamma for the range can give a person a much better feel of what is going on throughout the range.

The comparison between the RS-1 and VW50 also show how dominant of a role gamma itself can be as far as contrast in the low part of the curve (ie within dark shadows).
Mark Petersen is offline  
post #23 of 200 Old 08-03-2007, 11:41 PM
Senior Member
 
Thebes's Avatar
 
Join Date: Jan 2007
Posts: 447
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 22
Hi,

Than you.

I understand.

Only one more question, when you use the different gamma mode on the JVC, do you change the setting of contrast and brightness for the measures of "dynamic contrast" on each "mode" ?
Thebes is offline  
post #24 of 200 Old 08-05-2007, 11:40 AM - Thread Starter
AVS Club Gold
 
Mark Petersen's Avatar
 
Join Date: Dec 2000
Posts: 4,671
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Quote:
Originally Posted by Thebes View Post

Hi,

Than you.

I understand.

Only one more question, when you use the different gamma mode on the JVC, do you change the setting of contrast and brightness for the measures of "dynamic contrast" on each "mode" ?

The same brightness and contrast settings were used only the gamma mode itself was changed.
Mark Petersen is offline  
post #25 of 200 Old 08-11-2007, 02:42 PM
AVS Special Member
 
TomHuffman's Avatar
 
Join Date: Jan 2003
Location: Springfield, MO
Posts: 6,441
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 61 Post(s)
Liked: 225
Why is there such a profound discrepancy between what Eric Garci's Intrascene Contrast Calculator predicts for a PJ with the RS1's characteristics (9587 on/off and 297 ANSI) at 15 APL and what you measured at 15 APL?

Plugging in these values and using the Saruman scene (which has an APL of 13.97) returns 4119. However, your measurements show a contrast ratio of only 947 at 15 APL.

I am reading something wrong?

Tom Huffman
ChromaPure Software/AccuPel Video Signal Generators
ISF/THX Calibrations
Springfield, MO

TomHuffman is offline  
post #26 of 200 Old 08-11-2007, 06:07 PM - Thread Starter
AVS Club Gold
 
Mark Petersen's Avatar
 
Join Date: Dec 2000
Posts: 4,671
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Quote:
Originally Posted by TomHuffman View Post

Why is there such a profound discrepancy between what Eric Garci's Intrascene Contrast Calculator predicts for a PJ with the RS1's characteristics (9587 on/off and 297 ANSI) at 15 APL and what you measured at 15 APL?

Plugging in these values and using the Saruman scene (which has an APL of 13.97) returns 4119. However, your measurements show a contrast ratio of only 947 at 15 APL.

I am reading something wrong?

Hi Tom,

I had a discussion with Eric about this in another thread (not sure which one). Basically we came to the conclusion that the gamma coded APL is what is best suited for comparisons purposes with these test patterns. In other words the Saruman scene has a 14% average Luma but if you apply a 2.0 gamma to the image (2.0 is ~ the measured gamma of RS-1 in normal mode) the average luminance is 3.2%. Running Erik's calculator with these numbers yields a maximum theoretical contrast of 3023:1. The RS-1 in question delivered 2909:1 with a 2% pattern (and 2278:1 with a 5% pattern).

Incidentally, extrapolating the contrast of real world scenes based on measurements from the contrast project based on APL alone is going to yield at best only a rough guesstimate. Geometry also plays an important role and the geometry of real world images is unique to each image. The spatial and luminance distribution of real images are very different from the test patterns utilized in this project.

This is an important point that is worth discussing in detail. Contrast in front projection displays is a function of both luminance and geometry. It's relatively easy for example to construct test patterns that vary the geometry while keeping the luminace (APL) constant. The contrast from these patterns will differ significantly even though the APL is the same. This is discussed in the write up at the beginning of this thead but let's pick an example - say we measured the contrast from many different variations of a 50% pattern. They may vary from each other by as much as 2x or 3x in measured contrast. Despite this deviation between patterns with the same APL, it's probably impossible to construct a pattern that could deviate by the 15x factor that a 0.1% test pattern can achieve (using the RS-1 as an example). The 0.1% can yield contrasts very close to the on/off cr which is something that no variation of a 50% pattern can come close to. This is because luminance also plays a role. The static contrast curve as a function of luminace that is displayed earlier in this thread could vary widely between end points based on the test patterns chosen.

Since contrast in a front projector is a function of both luminance and geometry, some have suggested that there is no utility in measuring contrast as a function of luminance. My feeling is that that using this particular suite of test patterns can provide an objective benchmark from which to compare projectors so long as people realize that this is a synthetic benchmark and the exact shape of the static contrast curve is unique to only this set of test patterns. Although I think the general inverse power curve shape will probably apply to most test patterns that are not maliciously created. In addition to providing objective comparisons between projectors, the methodology also provides some very useful metrics such as the maximum intra-image contrast of a projector (which isn't dependent upon geometry and can't be determined using traditional contrast measurements). It's only when trying to extrapolate the performance of real world scenes based on the APL of these test patterns that trouble may arise.

It's also been pointed out that the geometry within the static contrast test pattern suite is fairly random. It has been suggested that using a 4x4 checkerbox with decreasing area white boxes would be a better test pattern. Unfortunately decreasing the size of the white boxes also changes the geometry. For example keeping the center of the box at the same position means that the distance between the checkerbox edge and the probe will increase as the checkerbox get's smaller. In fact it's impossible not to change the geometry when constructing test patterns that use 100% full white and 0% black which vary the luminance by decreasing the size of the white pattern. In the end we end up with arbitrary geometries. Another problem is that as the white area decreases they become so small that it becomes difficult to measure. Combining the white boxes to get larger boxes as was done with the really low APL test patterns has to be done in order to get a large enough area to be measurable.

Another thing to keep in mind is that changing the luminance by varying the intensity of the white region of a test pattern can be done without changing the geometry and this is accomplished with the dynamic contrast test patterns. The results from these test patterns are completely independent of geometry.

With all of these caveats in mind it's still interesting to see that Erik's calculator yields ~3000:1 with a 3.2% scene while the RS-1 in question delivered ~2900:1 with a 2% test pattern. If I do a rough linear interpolation between the RS-1 2% and 5% data points I get about ~2500:1 @ 3.5% which differs from Erik's result by only ~20%. Given all the caveats I mentioned in this post this is an amazingly close correlation.
Mark Petersen is offline  
post #27 of 200 Old 08-11-2007, 06:36 PM
AVS Special Member
 
TomHuffman's Avatar
 
Join Date: Jan 2003
Location: Springfield, MO
Posts: 6,441
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 61 Post(s)
Liked: 225
Quote:


Basically we came to the conclusion that the gamma coded APL is what is best suited for comparisons purposes with these test patterns. In other words the Saruman scene has a 14% average Luma but if you apply a 2.0 gamma to the image (2.0 is ~ the measured gamma of RS-1 in normal mode) the average luminance is 3.2%.

Gosh, you lost me here. I can make a gamma adjustment to get this image down to 3% APL, but doing so converts



into



which bears no resemblance to anything you see on the screen. 3% APL is so dark that you can barely make out anything at all. Virtually no film material is that dark.

I understand the methodology that gets you 900+:1 on 15 APL. What is his methodology that gets him 4000+:1 on 15 APL?

Tom Huffman
ChromaPure Software/AccuPel Video Signal Generators
ISF/THX Calibrations
Springfield, MO

TomHuffman is offline  
post #28 of 200 Old 08-13-2007, 12:45 AM - Thread Starter
AVS Club Gold
 
Mark Petersen's Avatar
 
Join Date: Dec 2000
Posts: 4,671
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
The discussion on the Intrascene Contrast Calculator was here:
http://www.avsforum.com/avs-vb/showt...2&page=1&pp=20

There was never a mathematical treatment that equated test pattern APL to average luminance in a specific scene. After examining the scenes, both Eric and I got the impression (maybe incorrectly) that a better comparison between test pattern APL is not the average luma (before gamma coding) but rather the post gamma coded value (what is called average luminance in the calculator).

A black and white checkerboard is a unique image in that for something like a 50% ANSI image the average luminance is the same before and after gamma coding (both are 50%). Now if we look at something like this nurse image (included in the gamma calculator):



This image has an average luma of close to 50% and it also contains full white and full black pixels so you would think it would be similar to the ANSI test pattern. But after applying gamma the average luminance is really only 27% (using a 2.0 projector gamma). If we look closely at the image and also the histogram you can see why - there isn't a lot of bright ~100% stim pixels. Bright pixels have much more luminance than lower stim pixels and are weighted more because of the gamma (power curve) relationship.

So the question comes down to what test pattern is better suited for comparison purposes with this image, something with 50% luminance or 27% luminance? Using the latter gamma coded values and a 30% test pattern seems to provide the best match.

Fundamentally though I think that comparing the contrast in images to these test patterns will only yield a very rough range of contrast values. The spatial and luminance distribution of each image is unique and won't directly fit any of these test patterns so at best only a rough guestimate can be achieved. As an example Eric's calculator attempts to find the contrast between the brightest and darkest pixels in the image but what if we slightly changed two images and kept the luminance the same but moved the darkest pixel next to the brightest pixels in one image and then moved it as far away as possible from the brighter pixels? If we were somehow able to measure the contrast on these two images the contrast difference might be 2x or even greater. I still think the calculator is a useful tool for illustration purposes though.

As far as the contrast project goes, I Personally I think that it's a better use of time and energy to characterize projector contrast behavior than it is to try to extrapolate contrast values between dissimilar images. I know this was an early goal of the contrast project but such extrapolations can yield only rough guesstimates. I may be wrong though and I would like to hear what others have to say.
Mark Petersen is offline  
post #29 of 200 Old 08-13-2007, 01:04 AM - Thread Starter
AVS Club Gold
 
Mark Petersen's Avatar
 
Join Date: Dec 2000
Posts: 4,671
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Quote:
Originally Posted by TomHuffman View Post

Gosh, you lost me here. I can make a gamma adjustment to get this image down to 3% APL, but doing so converts

which bears no resemblance to anything you see on the screen. 3% APL is so dark that you can barely make out anything at all. Virtually no film material is that dark.

I understand the methodology that gets you 900+:1 on 15 APL. What is his methodology that gets him 4000+:1 on 15 APL?

Now that I think about it, I realize that I didn't answer this particular question in detail. If you look at the calculator the Saruman image has an average Luma of 13.97% but the average luminance that is displayed on the screen is dependent upon the projector gamma. Using the calculators default 2.5 gamma and pressing the "calculate values below" button will yield the average luminance after gamma coding for a 2.5 gamma which is only 1.85% (using 2.0 gamma yields 3%).

In a nutshell it's important when using the calculator to keep separate the concepts of average luma (pre gamma) from average luminance (post gamma). To put it another way: If you have a 50% stimulus pixel the displayed luminance is not going to be 50% but it's going to dependent on the gamma and in the case of something like the RS-1 it's going to be only 25% as bright as a 100%stim pixel because of the gamma non-linearity. If we average all of the pixels after gamma coding this is what the average luminance is. Whereas average luma in the calculator is averaging the %stim. Make sense?
Mark Petersen is offline  
post #30 of 200 Old 09-14-2007, 02:01 PM
AVS Addicted Member
 
stanger89's Avatar
 
Join Date: Nov 2002
Location: Marion, IA
Posts: 17,451
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 151 Post(s)
Liked: 151
Quote:
Originally Posted by TomHuffman View Post

Gosh, you lost me here. I can make a gamma adjustment to get this image down to 3% APL, but doing so converts

which bears no resemblance to anything you see on the screen. 3% APL is so dark that you can barely make out anything at all. Virtually no film material is that dark.

I understand the methodology that gets you 900+:1 on 15 APL. What is his methodology that gets him 4000+:1 on 15 APL?

Tom, remember that what's measured as a result of the test patterns (and what the contrast calculator "predicts" is, "on screen" contrast, and it's talking about "on screen" average luminance, if that makes sense.

"On Screen" luminance is a function of both the luma values in the source, and the gamma applied by the display. The reason the image doesn't look right when you convert it, is because you're effectively applying the gamma correction twice, once to change the average luma of the source, and then your monitor is applying gamma again.

Quote:
Originally Posted by reio-ta View Post


Mark
You were talking about how the RS1 will still have noticeably better "intra-scene" contrast than the Black Pearl. What would be a few scenes from various movies that you could think of which would show show this off the most? You said in your first post intra-scene contrast is hard to show with a test pattern. Would it be easier to demonstrate with a video? What would I be looking for?

Not sure I've got any good scenes off the top of my head, but I can describe what to look for. Scenes where high native On/Off contrast will be "noticeably better" than dynamic iris augmented On/Off contrast, are scenes that are low in average luminance but high in "dynamic range". That means scenes that are dark over all, but have areas that are very bright.

Credits would be a good one. Star Fields, perhaps the scene where they approach the Death Star in Star Wars (Falcon/Death Star take up a small portion of the screen but are brightly lit on a black back drop).

See what an anamorphoscopic lens can do, see movies the way they were meant to be seen
stanger89 is online now  
Reply Digital Hi-End Projectors - $3,000+ USD MSRP

User Tag List

Thread Tools
Show Printable Version Show Printable Version
Email this Page Email this Page


Forum Jump: 

Posting Rules  
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off