This thread is an adjunct to the AVS contrast thread, and also to the recent discussion of Cine4Home's DI article. The inspiration of this thread is that contrast is one of the "highest profile" image attributes out there, and with the help of the AVS Contrast Project, we are gaining a greater understanding of how differing projectors and projector technologies perform in this area.

However despite having that great new info, we have a rather significant problem, in that we don't really a good way to relate that information into the "real world".

So the idea of this thread is to shed light on the question:

"What are the characteristics of the 'typical' movie?"

To expand on that, what this thread is trying to do is collect data from a number of real movies and collect things like the distribution of APL, Peak Level, Min Level, and perhaps any other interesting/useful information.

So how are we going to go about doing this, well, over the course of a couple different threads, we've come up with a method that seems to work pretty well, it requires only two free tools:

DGIndex
AVISynth

The first step is to open up all the movie's VOB files in DGIndex, and then save the project (no need to save the audio).

The second step is to run the following AVISynth Script on the movie:
mpeg2source("dgindexprojectfile.d2v")
filename = "output filename"
ConvertToRGB()
Levels(16,0.4545,235,0,255,coring=false)
ConvertToYV12()
Crop(0,54,720,360)
Selectevery(24,0)
AssumeFps(240)
colon = " : "
WriteFile(filename, "current_frame", "colon", "YPlaneMin", "colon", "YPlaneMax", "colon", "AverageLuma" )
WriteFileStart(filename, """ "AverageLuma for Raiders8" """)
WriteFileStart(filename, """ "Frame : MinLuma : MaxLuma : AverageLuma" """ )
WriteFileEnd(filename, """ "End" """)
ConvertToYUY2()

A couple comments, mpeg2source should point the the project file you saved from DGIndex.
filename should point to the filename/path of the output file you want.
Crop should be adjusted to remove any black bars so they don't "contaminate" the results (could probably use autocrop).
SelectEvery is used to sample only once per second.
AssumeFPS is there to simply speed processing.

As posted, it's preferable to normalize the results into % by the following forumula:

normvalue = ( actualvalue - 16 ) / 219

I posited in the DI thread that the Levels command could be used to apply display gamma to the above script to convert from %stim to APL.

So, with that, I'll open this up for discussion, and for posting of results of the Script.

Results

Indiana Jones and the Raiders of the Lost Ark
Overall Average Luminance: 9.63%

ADL - proportion of frames below that ADL (and above the previous)
0.10% - 0.64%
0.20% - 0.74%
0.50% - 4.38%
1.00% - 6.31%
2.00% - 13.6%
5.00% - 23.4%
10.0% - 16.1%
20.0% - 7.55%
30.0% - 11.8%
40.0% - 5.01%
50.0% - 0.72%
100% - 0.52%
See attached image for graph.

Thanks for starting this thread. I think there are two issues that need to be addressed. The first is determining the correct methodology for calculating APL and the second is how best to try and perform a rough comparison of an actual film image to a checkerboard type test pattern based on the amount of luminance alone.

To me this second issue is really the rub. I think at best only the roughest of ranges might apply because precision more than that requires taking into account the other factors of contrast such as the geometry of the images and histogram distribution. It might be better then to just try and break it down to three ranges: low, middle and high APL and try to find a rough determination on what range of checkerboard patterns might apply (based on percent area) to these three ranges.

The discussion of relating APL to film images has come up in several other threads. Including this one:
http://www.avsforum.com/avs-vb/showthread.php?t=765552

It was from this thread that Eric Garci introduced the intrascene calculator:
http://home1.gte.net/res18h39/intrascene.htm

As was mentioned in the earlier threads. Checkerboard patterns are unique in that the average luminance is the same even after gamma coding has been applied. For an ANSI pattern for example it's 50% before and after gamma coding. At 50% I also speculate that its much too bright of a pattern to try and compare with what is happening in intra-image contrast with real scene information.

To see how gamma can affect the average luminance calculation we can take an image of a specific scene such as the starfield image in Eric's calculator. The average luma before gamma coding is 4.29% but after gamma coding it's only about 0.12%. So let's see where that particular scene would fall with the measured projector data below:

http://public.csusm.edu/kai/mark/avs/image001a.gif

From this data we can see that 0.1% places it at the extreme end of the chart at close to the on/off limit of the projector. This seems to fit well with the general perception that on/off plays a huge role in dark scene performance and it would seem to suggest a rough match.

In fact if we take any of the images in the intrascene calculator with the exception of the nurse scene they all tend to be very dark and tend to fall in the region of projector performance that suggests the importance of on/off contrast.

The nurse scene is an interesting example. Most people would characterize it as a bright scene. It contains an average luma of 50% similar to the ANSI pattern but after gamma coding it only has an average luma of 21%. If we compare it to the projector data above it suggests that this region would favor ANSI performance. What's interesting is that even though this is a bright scene it still contains only about 1/2 of the luminance of the ANSI pattern and at this place along the curve the benefits of high ANSI is reduced. In other words a projector that has say a 3x advantage in ANSI contrast might only have a 2x or less advantage in intra-image contrast at this point along the curve. This might partially explain why people seem to have a hard time seeing differences in images based on ANSI CR alone. It's possible that the ANSI CR metric may be implying larger intra-image contrast benefits that don't exist with typical scenes, even scenes that most people consider to be bright.

EDIT: I should also point out that the AVS Contrast thread that is referenced in this thread can be found here:
http://www.avsforum.com/avs-vb/showthread.php?t=852467

Thanks for starting this thread. I think there are two issues that need to be addressed. The first is determining the correct methodology for calculating APL and the second is how best to try and perform a rough comparison of an actual film image to a checkerboard type test pattern based on the amount of luminance alone.

That's basically the goal of the thread, though more for the first issue than the second.

To me this second issue is really the rub. I think at best only the roughest of ranges might apply because precision more than that requires taking into account the other factors of contrast such as the geometry of the images and histogram distribution. It might be better then to just try and break it down to three ranges: low, middle and high APL and try to find a rough determination on what range of checkerboard patterns might apply (based on percent area) to these three ranges.

Oh, absolutely, but so far we can't even do that. I think the goal is not to be able to say "this scene would have a contrast of x based on the test pattern" since as you say, that can't be done due to geometry and other issues. I think what would be optimal to come from this thread is to gain a good understanding of what "average" is in terms of APL, where most movies fall in the spectrum between On/Off and ANSI type luminance, and maybe come up with some good examples of images/movies that would showcase each "type" of contrast.

The discussion of relating APL to film images has come up in several other threads. Including this one:
http://www.avsforum.com/avs-vb/showthread.php?t=765552

It was from this thread that Eric Garci introduced the intrascene calculator:
http://home1.gte.net/res18h39/intrascene.htm

Yup, I actually concieved this thread when that one was going on but was too lazy to start it.

As was mentioned in the earlier threads. Checkerboard patterns are unique in that the average luminance is the same even after gamma coding has been applied. For an ANSI pattern for example it's 50% before and after gamma coding. At 50% I also speculate that its much too bright of a pattern to try and compare with what is happening in intra-image contrast with real scene information.

One thing to consider (and I think this is born out by Eric's calculators) is that being so bright, the ANSI pattern effectively bounds performance on the upper end, with almost no content exceeding it. Likewise On/Off bounds it on the other end.

FWIW, I had some success using Levels() in AVISynth to attempt "Gamma coding" the frames before pulling the min/max/average data, so I'll try and post that tonight and we can see how it compares to what we'd expect.

Would you guys please explain just what you mean by "gamma coding?" In video, gamma is the non-linear relationship between input signal going into the display and the output signal you actually see on the screen.

So, for example, a 2.2 gamma has this relationship of input to output:

Input - Output
10% - 0.63%
20% - 2.90%
30% - 7.07%
40% - 13.3%
50% - 21.8%
60% - 32.5%
70% - 45.6%
80% - 61.2%
90% - 79.3%

When it comes to film material, I am having a hard time understanding what is the input and what is the output. Clearly, what you see on the screen or on a monitor is the output, but wehat was the input. Not signal voltage.

Also what does gamma have to do with determining APL? Hadn't we already decided that the correct methodology was (average luma -16)/219? Where does gamma come into this?

Would you guys please explain just what you mean by "gamma coding?" In video, gamma is the non-linear relationship between input signal going into the display and the output signal you actually see on the screen.

So, for example, a 2.2 gamma has this relationship of input to output:

Input - Output
10% - 0.63%
20% - 2.90%
30% - 7.07%
40% - 13.3%
50% - 21.8%
60% - 32.5%
70% - 45.6%
80% - 61.2%
90% - 79.3%

When it comes to film material, I am having a hard time understanding what is the input and what is the output. Clearly, what you see on the screen or on a monitor is the output, but wehat was the input. Not signal voltage.

The basic issue is that the values stored on the disc do not map linearly to the actual luminance on the screen. The display device applies a gamma to the input before displaying it. So the input would be the values on the disc, and the output would (normally) be what's displayed on the screen.

As Mark pointed out, the contrast test patterns avoid Gamma issues because they use full white/black thus are unaffected by Gamma. Now we want to collect data representative of what's displayed on the screen from "real world" content, but it's completely impractical to do that using a sensor and a display so we need to find an alternative means.

One possiblity is to "simulate" a display, applying a display Gamma to the source data on the disc, and collecting metrics on the result.

Also what does gamma have to do with determining APL? Hadn't we already decided that the correct methodology was (average luma -16)/219? Where does gamma come into this?

That's the correct methodology for normalizing the source data, but the problem is the source data is still luma, or % Stim, not APL or what we see (or measure) on the screen.

Poynton has a good explanation of Gamma and how it fits into the system:
http://www.poynton.com/PDFs/GammaFAQ.pdf
The intensity of light generated by a physical device is not usually a linear function of the applied signal. A conventional CRT has a powerlaw response to voltage: intensity produced at the face of the display is approximately the applied voltage, raised to the 2.5 power. The numerical value of the exponent of this power function is colloquially known as gamma. This nonlinearity must be compensated in order to achieve correct reproduction of intensity.

In a video system, linear-light intensity is transformed to a nonlinear video signal by gamma correction, which is universally done at the camera.

So what we have (on the disc) is a non-linear representation of what the camera saw, and what is to be displayed. What the AVISynth script above captures is luma, which is a weighted average of nonlinear R'G'B' components, denoted as Y'.

What we need, in order to relate source data to the measurements by Mark and others is luminance (Y), which is the weighted sum of the linear RGB components.

So, to get average luminance (which I believe is what APL describes) we need to undo the gamma correction done by the camera, which can be done with the following equation:
http://www.poynton.com/notes/colour_and_gamma/R-nonlin-709-2-seg.gif

Unfortunately that's quite hard to do with AVISynth, but we can approximate it with the levels filter:
ConvertToRGB()
Levels(16, 0.45, 235, 16, 235)
ConvertToYV12()

We need to convert to RGB so levels can apply the gamma correction to each RGB component, then we need to convert back to YV12 so WriteFile's Luma parameters can read the values. This has the (unwanted) side effect of clamping the signal to 16-235, thus obliterating BTB and WTW, but for our purposes I think that's acceptable.

Ok I'm going to show my ignorance here. But, one of the problems with the ANSI checkerboard is that it uses full white for the checker board. Now, as I understand it using less than full white would be better except the projectors gamma gets in the way. So, why not instead of using gray pixels for the white square make it a mixture of alternating single pixels of full white and black. Since the sensors available average groups of pixels you would in essence be measuring a gray field on average and the projectors gamma would not play a part.

FWIW, I had some success using Levels() in AVISynth to attempt "Gamma coding" the frames before pulling the min/max/average data, so I'll try and post that tonight and we can see how it compares to what we'd expect.

It'll be really interesting to see these results.


Would you guys please explain just what you mean by "gamma coding?"


It just means the output after gamma is applied to the input. Since it's non-linear the avg. calculation has to be done with every point in the image after gamma is applied rather than take the average input and then apply gamma.


Also what does gamma have to do with determining APL? Hadn't we already decided that the correct methodology was (average luma -16)/219? Where does gamma come into this?


I originally thought that APL was the average image luminance after gamma coding but there is a lot of literature to suggest that it's just average luma before the display gamma is applied. For the purpose of this discussion though I don't think the definition of APL is important so long as we consistently apply the gamma coded forumula in discussions about contrast as I think that's whats really relevant.

For what it's worth Joe Kane Productions defines APL as:
"APL (Average Picture Level)
The video signal level, during the active picture part of each horizontal line, is mathematically averaged over the period of a frame to come up with APL. A 100 IRE Window pattern has a lower APL than a 100 IRE Flat Field pattern. Television program material is said to have a 15% average picture level over a long period of time. The Window patterns used for setting gray scale in all three titles have a 16% APL so that they will closely represent normal program material in their current drain on the high voltage power supply. The width of the Window was enlarged the extra 1% to accommodate the shape of the Philips PM5539 Color Analyzer which was on the market at the time of A Video Standard. That wide aspect ratio shape was passed on to Video Essentials. The computer world uses a much higher number for APL over time, that of 35%. That is reflected in the ANSI Checkerboard Patterns, VE T17 C27. "

I also found some literature where others have done this sort of APL study before. Here is a reference from IEEEXplore:

"Distribution of average picture levels in television programmes
Quinn, S.F.; Newman, P.M.
Electronics Letters
Volume 1, Issue 9, November 1965 Page(s):261 -
Digital Object Identifier 10.1049/el:19650235
Summary:An analysis of measurements made over 14½ hours of television programmes is presented, to show the distribution of average picture levels as a percentage of total time. It has been found that for 90% of the time the average picture level is between 20 and 70% and that the mean value of the distribution is close to 40%."

Here is another definition:
"average picture level (APL): In video systems, the average level of the picture signal during active scanning time integrated over a frame period; defined as a percentage of the range between blanking and reference white level."

And another:
APL - Average Picture Level
By: Bruce Bahlmann


The average signal level with respect to blanking during the active picture time. APL is expressed as a percentage of the difference between the blanking and reference white levels. APL represents the video signal level, during the active picture part of each horizontal line, is mathematically averaged over the period of a frame to come up with APL. A 100 Institute of Radio Engineers (IRE) Window pattern has a lower APL than a 100 IRE Flat Field pattern. Television program material is said to have a 15% average picture level over a long period of time. "

OK, but if the average luminance values that we capture on the disc is the input, then gamma coded APL output will be smaller (50% input gets you 21.7% output @ 2.2 gamma).

Why then does this white paper show the relationship in the opposite direction? Look at pages 10-12. He shows the gamma coded output APL as LARGER than the linear average luminance input.

http://www.etconsult.com/papers/Black%20Paper%20Matt%20Cowan.pdf

Which is it?

Mark:

What Kane reports
"Television program material is said to have a 15% average picture level over a long period of time."

and what the Quinn and Newman paper reports
"It has been found that for 90% of the time the average picture level is between 20 and 70% and that the mean value of the distribution is close to 40%."

cannot both be true.

I don't think that we are working with a clearly defined set of terms.

The data I collected showed a fairly consistent average luma in the low-to-mid 20s. How does that relate to these quotes regarding APL? There can't be a gamma difference, because the numbers are too similar.

Ok I'm going to show my ignorance here. But, one of the problems with the ANSI checkerboard is that it uses full white for the checker board. Now, as I understand it using less than full white would be better except the projectors gamma gets in the way. So, why not instead of using gray pixels for the white square make it a mixture of alternating single pixels of full white and black. Since the sensors available average groups of pixels you would in essence be measuring a gray field on average and the projectors gamma would not play a part.

This is an interesting methodology and something that I've thought about before. There are a lot of issues that need to be resolved in order to gleen something useful out of this though. The MTF of the display device will play a big role in any alternating on/off pixel patterns. There is also the issue of trying to ramp luminance by using a pattern like that. Does one keep the same pattern and use larger/smaller checkerboards or alter the underlying on/off pattern for example.

Fundamentally though I don't think that there is a problem using full white and full black checkerboxes. The challenge is trying to understand how to interpet the information that they convey to images based on APL alone and I think you'll have the same challenge with other patterns like these alternating on/off ones.

Checkerboard patterns are unique in that the average luminance is the same even after gamma coding has been applied.That is true for checkerboard patterns where the brightest rectangles are 100 IRE. For a checkerboard that is 50/0 in %stim or IRE the ft-lamberts or luminance (post gamma) would be about 20% of that 100/0 checkerboard.
Ok I'm going to show my ignorance here. But, one of the problems with the ANSI checkerboard is that it uses full white for the checker board. Now, as I understand it using less than full white would be better except the projectors gamma gets in the way. So, why not instead of using gray pixels for the white square make it a mixture of alternating single pixels of full white and black. Since the sensors available average groups of pixels you would in essence be measuring a gray field on average and the projectors gamma would not play a part.Going along with what Mark said about MTF, if you mean a single pixel checkerboard then I think this would be a problem with LCOS projectors. At least the latest JVC ones. I don't know the specifics, but they have some trouble with single pixel checkerboards. I think it is that the white pixels end up being smaller than they should be, with the average luminance in that section being dimmer than it should be. Using alternating white and black lines in either direction may or may not cause problems as far as trying to hit 50% ft-lamberts or luminance. Calculating based on the current ANSI checkerboard may actually have less problems with some technology.

--Darin

OK, but if the average luminance values that we capture on the disc is the input, then gamma coded APL output will be smaller (50% input gets you 21.7% output @ 2.2 gamma).

Why then does this white paper show the relationship in the opposite direction? Look at pages 10-12. He shows the gamma coded output APL as LARGER than the linear average luminance input.

http://www.etconsult.com/papers/Black%20Paper%20Matt%20Cowan.pdf

Which is it?

That paper has tripped me up before too. Someone suggested that the use of the term gamma coding means camera gamma rather than display gamma. It has the note (see on waveform monitor) which implies that this notion is correct. So I think they are comparing film luminance to post camera gamma coded APL. Since camera gamma has an inverse shape from display gamma it may explain why the relationship seems reversed in the white paper from what we would expect. What we really want to do is compare film luminance to projector luminance after the display gamma has been applied.

I think people use the term APL to mean different things depending on what part of the display pipeline they are talking about. To be clear maybe what we should do is coin our own term and call it display APL.

The average luma before gamma coding is 4.29% but after gamma coding it's only about 0.12%.
CORRECTION: The average luma before gamma coding is 4.29% but after gamma coding the average luminance is only about 0.12%.
It contains an average luma of 50% similar to the ANSI pattern but after gamma coding it only has an average luma of 21%.
CORRECTION: It contains an average luma of 50% similar to the ANSI pattern but after gamma coding it only has an average luminance of 21%.

Note that average luma is calculated before the display's gamma is applied, and average luminance is calculated after the display's gamma is applied.

That is true for checkerboard patterns where the brightest rectangles are 100 IRE. For a checkerboard that is 50/0 in %stim or IRE the ft-lamberts or luminance (post gamma) would be about 20% of that 100/0 checkerboard.

Yup, 100%stim checkerboards like the ANSI pattern are the ones that I meant although I should have clarified what I meant.

CORRECTION: The average luma before gamma coding is 4.29% but after gamma coding the average luminance is only about 0.12%.

CORRECTION: It contains an average luma of 50% similar to the ANSI pattern but after gamma coding it only has an average luminance of 21%.

Note that average luma is calculated before the display's gamma is applied, and average luminance is calculated after the display's gamma is applied.

Hi Erik, yes I should have said average luminance to be more clear.

To be clear maybe what we should do is coin our own term and call it display APL.
I think the term "display APL" might be an oxymoron. I typically use the terms "average luma" and "average luminance." Since luma and luminance sound similar, the terms "average signal level" and "average display luminance" might be better. Maybe ASL and ADL.

Erik,

Thanks for stopping by, would you care to comment on my take on the situation:

1) The ANSI pattern is both 50% luma, and 50% luminance (due to being made from 100%/0% boxes), the same is true of Mark's static test patterns, ie their %luma = %luminance.
2) What we care about, in attempting to relate real-world images to these test patterns is luminance (vs luma), which I/we have been referring to as APL.
3) Applying a per-pixel, per-frame, inverse-camera gamma correction will successfully approximate luminance.

I think the term "display APL" might be an oxymoron. I typically use the terms "average luma" and "average luminance." Since luma and luminance sound similar, the terms "average signal level" and "average display luminance" might be better. Maybe ASL and ADL.

"ADL" and "ASL" are perfect :)

This may be true.

What I am trying to figure out is what the data on nearly 20,000 frames from 3 films means?

The white paper reports this image

http://home.comcast.net/%7Etlhuffman/apl/party.jpg

as having an average luminosity of 95.5 and an APL of 37%. This relationship corresponds almost exactly to the relationship between the average luma and the normalized values that I reported for the 3 films I sampled.

Yet, I gather that you believe that there is yet some further calculation required to account for gamma coding. What would that be? Applying a 2.5 gamma to .37 (if that is to be understood as the % stim) results in 8.3. I can't believe that you mean to say that we should interpret the display APL (or whatever you choose to call it) of this relatively bright image as 8.3 APL.

I really don't see where this is going.

That paper has tripped me up before too. Someone suggested that the use of the term gamma coding means camera gamma rather than display gamma. It has the note (see on waveform monitor) which implies that this notion is correct. So I think they are comparing film luminance to post camera gamma coded APL. Since camera gamma has an inverse shape from display gamma it may explain why the relationship seems reversed in the white paper from what we would expect. What we really want to do is compare film luminance to projector luminance after the display gamma has been applied.

I think people use the term APL to mean different things depending on what part of the display pipeline they are talking about. To be clear maybe what we should do is coin our own term and call it display APL.

1) The ANSI pattern is both 50% luma, and 50% luminance (due to being made from 100%/0% boxes), the same is true of Mark's static test patterns, ie their %luma = %luminance.
Technically, for those patterns, the average luminance may be slightly greater than the average luma, depending on the display's on/off CR, but the difference is usually less than 1% so it can be ignored.
2) What we care about, in attempting to relate real-world images to these test patterns is luminance (vs luma), which I/we have been referring to as APL.
In typical usage, the term APL refers to the average level of the signal before the display's gamma is applied. If you use the term APL to refer to something else, then you run the risk of confusion.
3) Applying a per-pixel, per-frame, inverse-camera gamma correction will successfully approximate luminance.
I think it will probably be good enough for most images.

In typical usage, the term APL refers to the average level of the signal before the display's gamma is applied. If you use the term APL to refer to something else, then you run the risk of confusion.OK, that explains why the image below is reported as it is.

Thus, I assume that this means that the methodology of deriving APL from a normalized average luma would be correct absent any futher calculation would be the correct approach.

I thought it an interesting exercise to try and relate some of the checkerboard data from the AVS contrast project to the scenes and calculations from Erik's calculator.

The comparison uses the measured values below that correspond to the same data set that is shown in the graph at the beginning of this thread. These numbers were then plugged into Erik's calculator.

RS-1 on/off - 9587
ANSI - 397
gamma 2.2 (not measured directly - this is a "guestimate")
20K on/off 5000
ANSI 560
gamma 2.2 (also not measured)

I picked two images from Erik's calculator where the calculated ADL falls close to where we have measured data. The two patterns are the star field (~0.2%) and Saruman(~2.5%). With these two patterns I looked up the intra-image contrast data at the ADL location on the curve (or close to it) and if the image doesn't contain 100% stim pixels I extrapolated what the contrast would be using a 2.2 gamma. Here are the results:

Star Field:
RS1 (extrapolated data) - 4984 RS1 (Erik's calc) - 5402
20K (extrapolated data) - 2877 20K (Erik's calc) - 3073

Saruman:
RS1 (extrapolated data) - 2850 RS1 (Erik's calc) - 3482
20K (extrapolated data) - 2879 20K (Erik's calc) - 3334

Both of these match much closer than what I would have expected.

It would be interesting to collect more images where the ADL falls close to where we have measurement data and see if this rough relationship holds for other images. If the relationship holds this could go along way in giving us a rough idea how the strengths and weaknesses of different projection technologies play a role in contrast with real images. I'm guessing that this will also validate the importance of high on/off as the ADL in most of the images will tend to be very low.

...What would that be? Applying a 2.5 gamma to .37 (if that is to be understood as the % stim) results in 8.3. I can't believe that you mean to say that we should interpret the display APL (or whatever you choose to call it) of this relatively bright image as 8.3 APL...The image is APL 37%, applying a 2.5 degamma would yield a luminance of 8.3% ( of the all white, 100% value). If an all white image was 10 ftL, that image would be 0.83 ftL. That is the nonlinearity, and why high on off is important. In order to get a 100:1 CR between a 10% object and black in said image would require a 30,000:1 pj

I think I am finally starting to see what you are doing. So is the calculated ADL just the APL with gamma applied? If so, then yes the ADL will certainly be very low, rarely above 15.

If so, then the only patterns from the contrast project that will be relevant will be the 15% of full field white and below, right? I say this because it looks like APL above 50 is pretty rare and 50 post-gamma is somewhere between 13-17, depending on the gamma used.

I thought it an interesting exercise to try and relate some of the checkerboard data from the AVS contrast project to the scenes and calculations from Erik's calculator.

The comparison uses the measured values below that correspond to the same data set that is shown in the graph at the beginning of this thread. These numbers were then plugged into Erik's calculator.

RS-1 on/off - 9587
ANSI - 397
gamma 2.2 (not measured directly - this is a "guestimate")
20K on/off 5000
ANSI 560
gamma 2.2 (also not measured)

I picked two images from Erik's calculator where the calculated ADL falls close to where we have measured data. The two patterns are the star field (~0.2%) and Saruman(~2.5%). With these two patterns I looked up the intra-image contrast data at the ADL location on the curve (or close to it) and if the image doesn't contain 100% stim pixels I extrapolated what the contrast would be using a 2.2 gamma. Here are the results:

Star Field:
RS1 (extrapolated data) - 4984 RS1 (Erik's calc) - 5402
20K (extrapolated data) - 2877 20K (Erik's calc) - 3073

Saruman:
RS1 (extrapolated data) - 2850 RS1 (Erik's calc) - 3482
20K (extrapolated data) - 2879 20K (Erik's calc) - 3334

Both of these match much closer than what I would have expected.

It would be interesting to collect more images where the ADL falls close to where we have measurement data and see if this rough relationship holds for other images. If the relationship holds this could go along way in giving us a rough idea how the strengths and weaknesses of different projection technologies play a role in contrast with real images. I'm guessing that this will also validate the importance of high on/off as the ADL in most of the images will tend to be very low.

I think I am finally starting to see what you are doing. So is the calculated ADL just the APL with gamma applied? If so, then yes the ADL will certainly be very low, rarely above 15.


Yes, ADL is calculated from the luma after display gamma is applied. It will be very low and maybe this is the value people are quoting when they say APL is a dark medium and over time it averages 15 or less. I don't think the definition of APL that others use is all that important so long as we are precise in using ADL as that's what's really relevant to this discussion of display contrast.


If so, then the only patterns from the contrast project that will be relevant will be the 15% of full field white and below, right? I say this because it looks like APL above 50 is pretty rare and 50 post-gamma is somewhere between 13-17, depending on the gamma used.

Not necessarily. the nurse scene has an ADL of ~25%. I would have done the same calculations for the nurse scene as the star field and saruman but the closest checkerboard pattern we have is at 20%. If I use the 20% values we'll have higher numbers than the 25% predicted by Erik's calculator and in fact that's what we see.

I think that with enough images we'll see that many of the images will fall in a region that is heavily influenced by on/off while most of the brightest images will fall well below the ANSI pt. The Saruman is a good example of an image that falls close to the crossover point between the RS-1 and 20K where the ANSI CR advantage of the 20K is mitigated by the on/off advantage of the RS-1.

I think this also shows pretty clearly why current on/off and ANSI measurements are poorly suited for trying to gauge differences in intra-image contrast. For example, if we plug in on/off values into Erik's calculator for the VW50 and the Star Field image, the result will yield intra-image contrast values that are far above the native contrast limit of the VW50. On the other side of the spectrum, the ANSI cr measurement suggests intra-image contrast differences that are larger than they actually will be with the vast majority of images.

Not necessarily. the nurse scene has an ADL of ~25%. I would have done the same calculations for the nurse scene as the star field and saruman but the closest checkerboard pattern we have is at 20%. If I use the 20% values we'll have higher numbers than the 25% predicted by Erik's calculator and in fact that's what we see.My point was that scenes as bright as the nurse scene seem fairly rare. 49 APL is outside the 95% range for all of the films I have sampled so far.

For example, if we plug in on/off values into Erik's calculator for the VW50 and the Star Field image, the result will yield intra-image contrast values that are far above the native contrast limit of the VW50.
The calculator does not take into account the effect of a dynamic iris, so the result would be more accurate if you plug in the native on/off CR.

I've updated the first post with the Average Luminance results for Raiders, using the method above to do a per-pixel gamma correction. The resulting average Luminance for the movie is just under 10%.

Interesting perhaps, is that there are really two peaks in this movie, one at about 5% (pretty dark) and another at around 30% (rather bright).

I categorized the results around the same values as the contrast results (though I added a few more steps between 20 and 50%) so that it's easy to compare the two. For example, on balance Raiders is brighter than the "crossover point" between the VW60 and the Z20k in either configuration, but mostly darker than the same point between the RS1 and Z20k in the Med Contrast config, and basically right on top of the crossover point between the RS1 and the Z20k in High Contrast config.

My point was that scenes as bright as the nurse scene seem fairly rare. 49 APL is outside the 95% range for all of the films I have sampled so far.

Tom, I think we're on the same page. If you were to pick 3 patterns based on ADL to represent dark, average and bright ranges would you pick?

The calculator does not take into account the effect of a dynamic iris, so the result would be more accurate if you plug in the native on/off CR.


Yup that's my point. It's not that the calculator is inaccurate it's that on/off as a predictor of intra-image contrast doesn't work with a DI.

Rather than using native on/off (iris open), it would be even more accurate if a person plugged in measured values with the smallest pattern where measurements could still be reliably done because this would include the iris benefits when the iris is in the smallest aperture position. In fact it's pretty easy to extrapolate the native on/off (iris closed) from that data by taking the measured white level with the smallest checkerboard (iris in smallest aperture) and then dividing by the full off black level. In the AVS contrast project I've termed this extrapolated value "maximum intra-image contrast". In some cases this may be the only way to measure static on/off if iris off measurements cannot be performed (or in the case of DLP where the iris alone can make a huge impact to contrast).

Just to dramatize how erroneous the numbers can be let's run the numbers for for the VW50 and the Star Field scene:

Assumptions: ANSI - 263 (measured). Gamma 2.2 (guestimate)

a) Iris 1 on/off - 12550 (measured)
b) Iris off on/off - 3089 (measured)
c) Iris 1 on/off - 4000 (note: extrapolated from measured max intra-image contrast metric)

Contrast numbers from the calculator:
a) contrast - 6747
b) contrast - 1863
c) contrast - 2379

Actual VW50 Iris 1 measured contrast at .2% = 2280 (white level corrected to reflect 80%stim at max luma).

So you can see how much more accurate we get using the value derived from the max intra-image contrast metric.

I've updated the first post with the Average Luminance results for Raiders, using the method above to do a per-pixel gamma correction. The resulting average Luminance for the movie is just under 10%.

Interesting perhaps, is that there are really two peaks in this movie, one at about 5% (pretty dark) and another at around 30% (rather bright).

I categorized the results around the same values as the contrast results (though I added a few more steps between 20 and 50%) so that it's easy to compare the two. For example, on balance Raiders is brighter than the "crossover point" between the VW60 and the Z20k in either configuration, but mostly darker than the same point between the RS1 and Z20k in the Med Contrast config, and basically right on top of the crossover point between the RS1 and the Z20k in High Contrast config.

Interesting data. If I were to pick three checkerboard (full white) test patterns to try and use to gleen intra-image contrast comparisons based on this data I would probably pick 30% as the upper limit, 5% for an average and the smallest pattern that could be reliably measured (0.1% or 0.2% for the lower limit). I would use these as a replacement for the ANSI and on/off metrics. I would also supplement it with a test pattern to show the full benefit of a DI (probably a very small area pattern with a low %stim grey like 20%). What do you guys think?

Rather than using native on/off (iris open), it would be even more accurate if a person plugged in measured values with the smallest pattern where measurements could still be reliably done because this would include the iris benefits when the iris is in the smallest aperture position.
Couldn't you just set the iris to the smallest aperture in Manual mode and then measure on/off CR in the usual way? The VW50 allows this, but I'm not sure about other projectors.

Tom, I think we're on the same page. If you were to pick 3 patterns based on ADL to represent dark, average and bright ranges would you pick?If I understand your ADL calculation, it requires specifying performance criteria for a targeted display, so let me answer this in terms of APL, which is display independent. You can then do the conversion.

Low: 8 APL
Medium: 22 APL
High: 40 APL

If I understand your ADL calculation, it requires specifying performance criteria for a targeted display, so let me answer this in terms of APL, which is display independent. You can then do the conversion.

Low: 8 APL
Medium: 22 APL
High: 40 APL

Actually, looking at the data more closely, I would adjust this a little.

Low: 10 APL
Mid: 22 APL
High: 35 APL

Couldn't you just set the iris to the smallest aperture in Manual mode and then measure on/off CR in the usual way? The VW50 allows this, but I'm not sure about other projectors.

There are no guarantees that the iris aperture will be the same. There are also other DI implementations to think about along with other dynamic technologies like dynamic lamp. As an example, is there a manual mode for dynamic lamps?

By comparison the determination for max intra-image contrast is easy to do (easier than the ANSI measurement). Since it's based on direct ultra low APL measurements it can be done independently without any concern of the dynamic implementation that was used.


If I understand your ADL calculation, it requires specifying performance criteria for a targeted display, so let me answer this in terms of APL, which is display independent. You can then do the conversion.

Low: 8 APL
Medium: 22 APL
High: 40 APL


Unfortunately there isn't an exact, direct conversion from APL to ADL because the gamma needs to be applied to every pixel in the image and then the average taken.

This does bring up a good point though that ADL does vary slightly between different display devices (even if the gamma is the same) so a checkerboard pattern that may best apply in one case may not be the best for the other (for example on one display the ADL may be 2% and with the same image 3% on another display).

Unfortunately there isn't an exact, direct conversion from APL to ADL because the gamma needs to be applied to every pixel in the image and then the average taken.OK, here are three sets of images taken from Manhattan. First Low, second Mid, and last High.

These won't look exactly right. For some reason I cannot fathom, when I take a screen shot, it darkens the image a little.

http://home.comcast.net/%7Etlhuffman/apl/dark.jpg

http://home.comcast.net/%7Etlhuffman/apl/mid.jpg

http://home.comcast.net/%7Etlhuffman/apl/high.jpg

These are roughly
10 APL
22 APL
35 APL

which are each right in the center of the each of the 1/3 APL ranges for the films I have sampled.

OK, here are three sets of images taken from Manhattan. First Low, second Mid, and last High.

These won't look exactly right. For some reason I cannot fathom, when I take a screen shot, it darkens the image a little.

These are roughly
10 APL
22 APL
35 APL

which are each right in the center of the each of the 1/3 APL ranges for the films I have sampled.

Erik do you mind adding these to the intrascene calculator? It'll be interesting to see what we get as far as ADL and if any of the brighter images happen to fall on a place where we have measured contrast data.

Erik do you mind adding these to the intrascene calculator? It'll be interesting to see what we get as far as ADL and if any of the brighter images happen to fall on a place where we have measured contrast data.
I did a quick approximation of the ADL assuming a 2.5 gamma:

1.8%
1.5%
2.4%

5.9%
6.1%
5.2%

10.8%
13.8%
12.4%

I did a quick approximation of the ADL assuming a 2.5 gamma:

1.8%
1.5%
2.4%

5.9%
6.1%
5.2%

10.8%
13.8%
12.4%

Wow, those numbers are all over the place. The APL is very consistent within the images of the 3 sets. After I take the screenshot, as I said, this alters the image somewhat, for reasons unknown, but even then the histogram shows that they are very similar.

29
27
26

57
57
64

87
91
88

In any case, the 1.8%, 5.2%, and 10.8% ADL samples are reasonably close to the points provided by your contrast data.

After seeing how these came out, it occurs to me that you might want some even darker images. The dark images before were around 10 APL, which are darker than approximately 84% of the material I sampled. Here's one 8 APL and two 5 APL images. They are darker than approximately 90% and 95% respectively of the images I sampled.

http://home.comcast.net/%7Etlhuffman/apl/verydark.jpg

http://home.comcast.net/%7Etlhuffman/apl/extremelydark.jpg

http://home.comcast.net/%7Etlhuffman/apl/extremelydark2.jpg

As you can see, one of the problems with the 5 APL images is that now we've gotten so dark it becomes difficult to make out much detail in the image.

Here's one 8 APL and two 5 APL images.
The approximate ADLs are:
1.0%
1.7%
0.6%

More data.

Shindler's List

# of Frames: 11,288
Ave. APL 22.4%

http://home.comcast.net/%7Etlhuffman/apl/chart_shindler.gif

Casablanca

# of Frames: 6,130
Ave. APL 26.1%

http://home.comcast.net/%7Etlhuffman/apl/chart_casablanca.gif

Jaws

# of Frames: 7,377
Ave. APL 30.3%

http://home.comcast.net/%7Etlhuffman/apl/chart_jaws.gif

I appreciate all the data gathering. Not trying to be pushy, but if anybody wants to try it, I think Sunshine could be a very interesting one. There are some pretty bright scenes with the sun and at least some with mixes of dark and bright (although I don't remember how many really dark scenes there are). As far as dark movies, I wonder how dark Alien, Dark City or a movie in that class would be compared to some of these others.

--Darin

Tom, just so we're clear are those % Luma or % Luminance? I'm guessing % Luma by the "high" numbers.

Tom, just so we're clear are those % Luma or % Luminance? I'm guessing % Luma by the "high" numbers.I am normalizing what AVISynth reports as AverageLuma by (AverageLuma-16)/219. I am just trying to determine APL, which I think we have decided is the number prior to gamma coding.

Alien

# of frames: 6628
Average APL: 17.9%

http://home.comcast.net/%7Etlhuffman/apl/chart_alien.gif

Alien

# of frames: 6628
Average APL: 17.9%
Thanks. Close to 55% of images at under 15 APL (before gamma) is amazing IMO and goes a long ways toward explaining why on/off CR is so important for a movie like that. I don't know what the ADL is since each color for each pixel would need to be gamma corrected before the average is done, but with all those under 15 it seems very likely that at least half the scenes in Alien have an average luminosity off the screen of under 5% of a full screen 100 IRE image (and the median ADL is likely lower than 5%). That is compared to an average of 50% for the ANSI CR image.

I may be stating the obvious, but if a projector is putting out 1000 lumens for a full screen 100 IRE image and white is steady, then the ANSI CR image is about 500 lumens, while the median image in Alien is likely between about 20 and 50 lumens, and under around 60 lumens for Indiana Jones and the Raiders of the Lost Ark. I wonder if we'll ever find a movie where the median image is even half that of the ANSI CR image (so 25% of the ANSI lumens or 250 lumens in the above case). And yet there seem to be some who still believe that ANSI CR is the only one that matters because ANSI CR is done with a mixed image and on/off CR isn't.

I should be getting a Panasonic AE1000 by Monday and I'm hoping I can learn some things with its waveform monitor (like the highest value for any pixel color in some of these scenes, including the one from Manhattan with the girl reading below the lamp).

--Darin

I posited in the DI thread that the Levels command could be used to apply display gamma to the above script to convert from %stim to APL.I just want to make sure. Does the levels command work on each color individually (like would a 100% green with 0% red and blue stay at 100% green with 0% red and blue)?
See attached image for graph.Maybe my brain just isn't working, but what is the Max Luminance there?

--Darin

Alien

# of frames: 6628
Average APL: 17.9%


Can an AviSynth script be written to generate an ADL curve for specific display parameters? It would be really interesting to compare a specific movie like Alien with different iterations of known display parameters:

For example:
15000:1 on/off (native) 300:1 ANSI 2.2 gamma
30000:1 on/off (native) 300:1 ANSI 2.2 gamma
30000:1 on/off (native) 300:1 ANSI 2.5 gamma
8000:1 on/off (native) 600:1 ANSI 2.2 gamma.

And yet there seem to be some who still believe that ANSI CR is the only one that matters because ANSI CR is done with a mixed image and on/off CR isn't.


Exactly. This data would tend to suggest that when it comes to real images that it's all about the on/off, although I think black levels and shadow detail also play a huge role which makes a DI important.

But if you think about it this jives with real-world perceptions. Most people seem to easily see the differences that on/off alone provides but ANSI differences seem to be much more difficult to spot and the differences seem to be more arguable.


I may be stating the obvious, but if a projector is putting out 1000 lumens for a full screen 100 IRE image and white is steady, then the ANSI CR image is about 500 lumens, while the median image in Alien is likely between about 20 and 50 lumens, and under around 60 lumens for Indiana Jones and the Raiders of the Lost Ark.


Just curious how you came up with a rough estimate for output lumens. Did you take the final APL calculation and then apply gamma to it?


the one from Manhattan with the girl reading below the lamp).


That's a great scene. What I like about it is that it's a very well composed low APL scene that shows exactly what the director is trying to convey. Scenes like that don't happen by accident and I think it's important for projection display technology to get scenes like that right.

Speaking of scenes. I watched the western "The Professionals" last night. It has a lot of day for night scenes and the bonus content mentioned that the reason for this was to create a sense of depth and distance in the night scenes by allowing the landscapes to be more visible. The interesting thing that I noticed is that it mutes the brightest whites which suggests that a DI would perform exceptionally well in scenes like that. I mentioned before that it would be interesting to determine where on a APL or ADL curve that a DI begins to operate and then write an Avisynth script to take just those frames in that range and build a histogram of max brightness for all of those scenes. I'm guessing that in the majority of those low APL scenes the brightest pixels all tend to fall fairly low in the IRE range. The histogram would be a good way to validate the dynamic gamma tradeoff in a DI.

I just want to make sure. Does the levels command work on each color individually (like would a 100% green with 0% red and blue stay at 100% green with 0% red and blue)?

From my reading of the documentation, if the input to Levels is RGB, then it works individually on each channel. However if the input is component, then it only works on the Luma channel. So for the data I posted, I first converted to RGB for the Levels filter, and then back to YUV for the data collection.

Maybe my brain just isn't working, but what is the Max Luminance there?

It's just the distribution of maximum Luminance values, same technique as Average Luminance.

Can an AviSynth script be written to generate an ADL curve for specific display parameters?

What sort of calculations are required for that? You can do just about anything in AVISynth. Though some things probably take more than just passing parameters to existing functions.

Just curious how you came up with a rough estimate for output lumens. Did you take the final APL calculation and then apply gamma to it?For Indiana Jones and the Raiders of the Lost Ark I just looked at the ADL data in the first post in this thread where about 49% of the frames had an ADL of 5% or under.

For Alien I looked at the data that showed APL (pre-gamma) was under 15 for over half the images. While it is true that in the most extreme case of 30% of the image being white and the rest black the median image would go as high as 300 lumens, that would be a rare extreme, Alien is likely darker than Indiana and if those images were more like half 30 IRE and half 0 IRE, the lumens would be about 35 at 2.2 gamma (from .3^2.2 * .5) and about 25 at 2.5 gamma (from .3^2.5 * .5). If they were like full screen 15 IRE then they would be about 8 to 15 lumens (from .15^2.5 and .15^2.2). Although I probably should use a power function and not just the x^y. Since Erik does this better, in order to get better estimates I probably should have used his contrast calculator here:

http://home1.gte.net/res18h39/contrast.htm

instead of x to the gamma power. He does correct the values for poor on/off CR (not put all the error from low on/off CR at the low IREs, but calibrate the projector to spread it out), but I would put in the CRs and gamma I want for projector A and/or B, then enter 30 for one of the checkerboard values and hit the "Calculate values below" button. I then look at the percentage for white for each case. With a million to one on/off CR and 2.5 gamma he reports around 5% of white and with 2.2 gamma around 7% of white for 30%stim, which is close to my estimates.

--Darin

It's just the distribution of maximum Luminance values, same technique as Average Luminance.What is the value on the vertical access then? I see it go from 0 to 1800. If that was 0 to 180.0 then I could see the Max Luminance being higher than 50% in about 81% of images. Also, is the Max Luminance after gamma?

One thing I think would be interesting is to do something like take all the images that have ADLs between 2% and 5% and map just their maximum color luminances, where that is the maximum for any of the 3 primaries which exists in that image. Then do the same thing for all the images that have ADLs between 20% and 30%. The reason I think this would be interesting is that it could tell us what a DI could do that doesn't do any brightness compression.

--Darin

snippage...

One thing I think would be interesting is to <snippage>. The reason I think this would be interesting <snippage>.

--Darin

One thing I think would be interesting would be to actually watch some movies without obsessing...:cool:

You guys are getting further and further out...:D

Time for some two-wheeled therapy!

Cheers,

Vertical axis is just the count of frames that fall between the steps. Unforunately I goofed the calculation so a bunch of frames weren't counted for MDL (Maximum Display Luminance). I setup the formula to compare just <, rather than <= so all the frames that had a max Luminance of 1 were missed.

But basically all it shows is that 85% of frames have a Maximum Luminance of 100%

I updated the picture in the first post to remove the confusing (and not very useful) Maximum Luminance data.

That's a great scene. What I like about it is that it's a very well composed low APL scene that shows exactly what the director is trying to convey. Scenes like that don't happen by accident and I think it's important for projection display technology to get scenes like that right.Yea, that's Woody Allen imitating German impressionist film makers. That type of Chiaroscuro composition is very stylized and not often used in commercial film, but it does make for a dramatic visual effect.

Yea, that's Woody Allen imitating German impressionist film makers. That type of Chiaroscuro composition is very stylized and not often used in commercial film, but it does make for a dramatic visual effect.

"Chiaroscuro". That's one I hadn't heard before, I had to look it up on wikipedia: http://en.wikipedia.org/wiki/Chiaroscuro#Cinema_and_photography


What sort of calculations are required for that? You can do just about anything in AVISynth. Though some things probably take more than just passing parameters to existing functions.


This is a good question for Erik. I understand how gamma must be applied and how the relative weight of each color in a pixel needs to be taken into account but I'm not sure how the on/off and ANSI cr is used. Erik would you care to enlighten us?


One thing I think would be interesting would be to actually watch some movies without obsessing...

You guys are getting further and further out...

Time for some two-wheeled therapy!

Cheers


Hey, I'm not obsessive! Believe it or not I do have a life. In fact I just got back from the taping of the VH1 Honors "The Who" which will air on Thursday (July 17th). :D I also enjoy two wheel therapy but of the off-road dirt bike kind (Yamaha WR-250). :D :D

But seriously though do you think that discussions of APL are really that far out there? If you check the literature (IEEE, SMPTE, SID, etc.) this is something that comes up a lot although what makes this thread interesting is that I've never seen it applied to projection displays before.

This is a good question for Erik. I understand how gamma must be applied and how the relative weight of each color in a pixel needs to be taken into account but I'm not sure how the on/off and ANSI cr is used. Erik would you care to enlighten us?
To compute the ADL, my calculator uses a gamma function that takes the on/off CR into account. The ADL is not affected by the ANSI CR.

There is an interesting SMPTE article that is related to this thread called, "The Importance of Contrast and its Effect on Image Quality By: Segler, D. J., Pettitt, Gregory, Kessel, Pete van Issue: 2002 11 November". Unfortunately it doesn't seem to be available publicly so I can't post a link but it's available in the SMPTE library.

Most of the article discusses the perception of contrast, but what I thought most interesting was a discussion of contrast as a function of luminance. The authors discuss two methodologies that they used and a few charts of their results. Rather than using a mix of small full white checkerboxes to ramp luminance as was done in the AVS contrast project the authors use a solid grey background that is ramped from 0% to 100% stim while also including two small checkerboxes in the foreground with full white and full black that are used to measure contrast. A chart of the data is provided that shows contrast as a function of "ALL" (average luminance level). The data was obtained from a DMD based RPTV and a Sony WEGA CRT. The crossover point between the two was around 1%.

The other methdology used a set of real images from the movie Titanic that, "were chosen to represent an ALL range from ~1% to ~40%, while maintaining the requirement that each image contain both an area of almost peak white and an area of almost full black. The images allowed the contrast ratios of the two televisions sets to be compared as a function of true video image ALL".

Time for some two-wheeled therapy!
Cheers,
says the ANSI but not On-Off champion. :D

Mark,

I also do a lot of Dual Sport riding...have an '06 DRZ400S...some great trails and forest service roads all over N.Ga, E.Tn and W.Nc.

No, I don't think it's that far out there, just that with front projectors, they are part of a "system" that consists of the projector, screen and the room. Unless you are in a totally light absorbtive room, every room will give you a different response in terms measureable or perceived CR vs APL. The point at which a higher ANSI CR rating becomes imperceptible, and therefore irrelevant, in various rooms is going to be all over the map.

Seems to me that trying to arrive at some sort of function to evaluate how a unit will perform in a real environment is not going to be practical without modeling the entire visual environment. Not at all unlike the complexity of the acoustic modeling I used to do when designing commercial sound reinforcement systems for theaters and such.

Like I said, rather watch the movie or go ride...:D

Seems to me that trying to arrive at some sort of function to evaluate how a unit will perform in a real environment is not going to be practical without modeling the entire visual environment.

FWIW, that wasn't my intent with this (and I doubt it was Mark's with his thread). The purpose was to gain a greater understanding of where real world content falls in terms of ADL. Since we've been gaining a greater understanding of how projectors perform across the Luminance range (thanks to Mark's efforts), it seemed appropriate that we should try to better understand where real world content falls on that "curve".

Oh, and I'll be off for a drive as soon as I tweak my front fender so I can turn again :eek:

Mark,

I also do a lot of Dual Sport riding...have an '06 DRZ400S...some great trails and forest service roads all over N.Ga, E.Tn and W.Nc.


Fun :) Most of the riding out this way is desert and mountains. If you're out this way in the winter time drop me a line and I can show you some really interesting desert rides.


No, I don't think it's that far out there, just that with front projectors, they are part of a "system" that consists of the projector, screen and the room. Unless you are in a totally light absorbtive room, every room will give you a different response in terms measureable or perceived CR vs APL. The point at which a higher ANSI CR rating becomes imperceptible, and therefore irrelevant, in various rooms is going to be all over the map.

Seems to me that trying to arrive at some sort of function to evaluate how a unit will perform in a real environment is not going to be practical without modeling the entire visual environment. Not at all unlike the complexity of the acoustic modeling I used to do when designing commercial sound reinforcement systems for theaters and such.


It's very true that room effects play a huge role. When I first started the AVS contrast project the idea was to have two parts. The first part would measure projector performance and the second part would use screen measurements to see how various rooms also played a role. Ric (Lovingdvd) had started a thread along those lines where many people measured ANSI at the screen and some experimented with different room treatments to see how it affected their ANSI readings. Unfortunately the whole idea was a little too grandiose and it ended up being about a full time job just to capture some insights into projector performance.

Even still though I think both of these threads (the original thread and this adjunct) provide some useful information. As an example, most people agree that the room plays a huge role in final (screen) ANSI performance and less of a role in on/off. But what does this mean to the majority of scenes that people watch in their HT? Is the luminance generated from a ANSI pattern typical of most scenes or is it rare? What's the average luminance for typical movies and in dark scenes what is the norm? That's what this thread discussion is about. Armed with that information it's easier to relate projector performance and also room effects.

It's an interesting state of affairs that we have two measures of contrast (on/off and ANSI) and neither are particularly good at measuring contrast in the range of luminance for the vast majority of images that most people use. As stranger89 has mentioned these two threads help supply some information about the in between range. I realize that this sort of discussion isn't for everyone though :)

Isn't there a big fat problem in your AVIsynth script?

Shouldn't you be using the Decomb filter (http://neuron2.net/decomb/decombnew.html) (telecide and decimate) to turn the interlaced footage back into the progressive frames with a real 24FPS frame rate? Also, why are you sampling only once per second, and cropping of the black bars? Unless you're doing this for a CIH setup the black bars are part of the image that the projectors use for DI calculations.

Ok, this is a few years late, but ...

APL, as the term is used by a video engineer, stands for “average picture level,” and refers to the mean of the Y' component (luma) component across the image area of a frame (or sequence, or movie, or whatever).

Average luminance – more correctly, average relative luminance – is the mean of the “true” CIE linear-light luminance across the image area of a frame or sequence, relative to reference white luminance.

The math that relates the two isn't simple, and it can't be done at all absent all three components R’, G’, and B’ or Y’, CB, and CR. Apart from the special case where R', G', and B' are all equal to either zero or one – notably, 100% colour bars – APL and average luminance differ, and sometimes dramatically.

At a display, video R', G', and B' signals (for our purposes, scaled to 0=reference black and 1=reference white) are each raised to a power of between 2.0 and 2.5 to yield linear-light RGB components that are directly related to luminance. That act maps 0.5 on the video scale to about 0.18 in relative luminance.

So: Don't mix up the two!

The APL of broadcast video is very roughly 50%; its average luminance is very roughly 18%.

The average luminance of a movie as presented in the cinema - and not necessarily a dark movie - is roughly 10%. I have this on extremely good authority, from a studio/DI guy who has access to DCI movie data and ran some perl scripts. (Sorry, I know that's gross, but he's an old awk/sed/grep kind of guy.)

The average luminance of a movie as transferred to DVD or Blu-ray is up for discussion, but I'd make a guess somewhere between 10% and 18%, say 14%.

- Charles

Hi Charles,

Just want to say thanks for the information and nice to see you post. I talked to you for a while a couple of months ago about some different things (like how things are setup for mastering) and appreciate your time.

As far as this:
The average luminance of a movie as transferred to DVD or Blu-ray is up for discussion, but I'd make a guess somewhere between 10% and 18%, say 14%. I've been wondering whether any of the stuff that is ending up on Blu-ray has had all of its gamma related stuff (not counting bit depth changes or color gamut type changes) done for DCI and then not get adjusted for Blu-ray. That is, aimed more for 2.6 display gamma (you can correct me if I have that wrong for DCI), and this could impact the average luminance. It seems to me that if DCI content is converted to Blu-ray in a way that takes the different gammas into account then the average luminances should end up in the same ballpark on playback, although I haven't thought about it too deeply.

--Darin