AVS › AVS Forum › Display Devices › Digital Hi-End Projectors - $3,000+ USD MSRP › Some people are still being told that 200:1 is all we can see....
New Posts  All Forums:Forum Nav:

Some people are still being told that 200:1 is all we can see.... - Page 11

post #301 of 505
1/5/07 at 12:54pm
Quote:
Originally Posted by GlenC View Post

I really don't understand your problem with others giving their opinions. You apparently don't like or agree with what I have said, but it is not wrong.......

Oh man..now you did it, here comes another dissertation
Reply
Reply

AVS Top Picks

post #302 of 505
1/5/07 at 1:03pm
Quote:
Originally Posted by tbrunet View Post

Oh man..now you did it, here comes another dissertation

OOPS
Reply
Reply
post #303 of 505
1/5/07 at 1:09pm
Quote:


I have basically stated that we, as a human factor, may not be able to see what the projector is fully capable of producing.

I know. It's a ridiculous claim. And as an ISFer you should know this intimately.

Quote:


As I stated before, display a full 100 IRE field and look at it. Now replace it with a 0-20 IRE stair pattern (5 IRE steps) and see which step you can see before the eye runs out of range.

It's the DISPLAY that is obscuring the detail via light spill! Goodness.

Quote:


That first step would be your instantaneous CR from the screen fL and percent field of view.

That's the system's CR. Measure it with a light meter.

Are all these people insane?
http://www.avsforum.com/avs-vb/showthread.php?t=781060

Or these:
http://www.avsforum.com/avs-vb/showthread.php?t=761806

Did you try the shadow puppet test? Use the same pattern and try to make shadow puppets in the dark parts where you can no longer distinguish the bars. Or see if you can see the screen boundary as distinct. According to your unbelievable logic seeing that would be IMPOSSIBLE!

I'm sorry if I'm frustrated here, but this is really basic stuff. I can understand novices and newcomers not understanding contrast ratios, but not trained ISF calibrators and professionals. This is bordering on unfathomable to me that otherwise intelligent people can still be making these claims that we can only see 200:1 across a scene. So by this logic nothing beyond 200:1 ANSI or instantaneous CR is useful. Incredible.

DO A TEST. It's not hard. Not a single person has done so except Darin and myself. Gee, I wonder why... because it immediately and unambiguously disproves this nonsense about 200:1 CR being all we can see (or pick your random figure: 130:1, 100:1, 250:1, 300:1, we've seen them all pulled right out of thin air, they're all equally ridiculous).
Reply
Reply
post #304 of 505
1/5/07 at 1:12pm
Glen: can you see the difference between a FP display with an ANSI of say 200:1 and one much higher than that (such as a 1-chip DLP)? I mean come on here, you can't honestly be claiming that there is no difference there.
Reply
Reply
post #305 of 505
1/5/07 at 1:26pm
[quote=HoustonHoyaFan]
Quote:
Originally Posted by mdtiberi View Post

He states clearly one to two orders of magnitude for photoreceptors. QUOTE]

Above you posted

Which do you believe is correct over all > 10^3 or that the photoreceptors (rods and cones) at the light receiving/facing layer of the retina only have a dynamic range of 10 to 100 (10^1 to 10^2)?

Photoreceptors have the ability to capture a range from a few photons to damage level. That's Bio 101.

What I believe is basically irrelevant.

I for one cannot take issue with Dr. Boynton of the Salk Institute or Professor Wandell at Stanford. I do not think that any contributor to this thread so far can do so credibly as well.

If what you are saying is that the limit of the HVS in contrast perception is the weakest link (200:1) in the chain from image acquisition to visual cortex processing then that seems contrary to the expert studies and papers that I have read and cited. For example, the following paper Contrast adaptation and representation in human early visual cortex as well as "The Statistics of Natural Images" by Daniel L. Ruderman, Cambridge University both suggest that although visual pathways are limited in their range of stimulus response (10^1 to 10^2 for photoreceptors), especially compared to the range in natural scenes (10^10), a much greater range in contrast is actually perceived via adaptation and contrast gain. So you cannot just say that a photoreceptor does this and a ganglion does that and therefore the "system" does this. That would be an oversimplification.

All of this so far is leading me to an arguably shaky conclusion that contrast perception may not be as high as some contend but certainly not as low as others are saying.
Reply
Reply
post #306 of 505
1/5/07 at 1:33pm
Quote:
Originally Posted by GlenC View Post

As far as I can see in this thread. is a failure to fully define the meaning of the statement "200:1 contrast is all that we can see" Without that defined, nobody has any authoritative claim on opinions being right or wrong, unless there is clearly an error. There are way too many potential variables involved. As I stated before, display a full 100 IRE field and look at it. Now replace it with a 0-20 IRE stair pattern (5 IRE steps) and see which step you can see before the eye runs out of range. That first step would be your instantaneous CR from the screen fL and percent field of view. This will be different, depending on age, health, individual and many other factors. This would be a valid test if I had said "200:1 instantaneous CR is all we can see in a theater environment."

That is only a test of one kind of limitation.

Here is a thread where simultaneous CRs are being measured and calculated with different projectors and some different patterns:

http://www.avsforum.com/avs-vb/showt...0&page=1&pp=30

As you can see there and the thread it references with William's 1% pattern, there are cases where the simultaneous CR is over 4000:1. Do you think that people would be able to see the difference in a dark room with a screen with a dark border between having that pattern shown with 4000:1 simultaneous CR vs a system that only had 500:1 simultaneous CR for the same pattern? I'm trying to figure out if your position is that 200:1 is all they need to be able to do at one time for us to not see improvements, or not. If somebody out there is telling people that projectors don't need to be able to do more than 200:1 at once because we can't see more than that, then they are confused and confusing others.

--Darin
Reply
Reply
post #307 of 505
1/5/07 at 2:26pm
Quote:
Originally Posted by ChrisWiggles View Post

I know. It's a ridiculous claim. And as an ISFer you should know this intimately.

It's the DISPLAY that is obscuring the detail via light spill! Goodness.

If you can see the detail at the screen (this means that the projector is displaying the image), but you cannot see the detail when you move away from the screen, how can you call that "light spill". All projectors have light spill, some systems perform better than others.

You are missing the point of my test/example. If you suddenly replace a full white field (full screen is 100 IRE white) with a black to 20 IRE stair pattern (very low APL screen) (BTW, with no white there is NO light to spill). If all the steps are displayed, but it takes you a few seconds or minutes to see all the steps, how is that an issue of the projector and not the eye?
Quote:
That's the system's CR. Measure it with a light meter.

You totally missed the point here, no it is not the systems CR it's yours based upon the systems numbers.
Quote:
Did you try the shadow puppet test? Use the same pattern and try to make shadow puppets in the dark parts where you can no longer distinguish the bars. Or see if you can see the screen boundary as distinct. According to your unbelievable logic seeing that would be IMPOSSIBLE!

Yes, you can make shadows, but when I properly calibrate a projector, you can see all the bars. If the bars are not being displayed, black level is set too low.

Quote:
I'm sorry if I'm frustrated here, but this is really basic stuff. I can understand novices and newcomers not understanding contrast ratios, but not trained ISF calibrators and professionals.

It clearly seems that your frustration is from the lack of understanding what others are saying and or your inability to express your specific point to where others understand.
Quote:
This is bordering on unfathomable to me that otherwise intelligent people can still be making these claims that we can only see 200:1 across a scene. So by this logic nothing beyond 200:1 ANSI or instantaneous CR is useful. Incredible.

You still don't get it. It doesn't frequently happen, however the display can, at times, out perform the ability of the eye to adjust to the changing light levels.{quote]DO A TEST. It's not hard. Not a single person has done so except Darin and myself. Gee, I wonder why... because it immediately and unambiguously disproves this nonsense about 200:1 CR being all we can see (or pick your random figure: 130:1, 100:1, 250:1, 300:1, we've seen them all pulled right out of thin air, they're all equally ridiculous).[/quote]Have you done any of the tests here? Try them. I am not subscribing to any global "all we can see" specific CR. I have given clear examples of situations where we don't see beyond the 200:1-300:1 CR, it happens. Is that the limit of our range, absolutely not. It usually takes the eye longer to adjust to dark vision than light. More examples, be in a dark room for 2 hours then walk outside into the bright noon time sunlight, any white clipping? drive in the daylight and drive into a tunnel, do you have a temporary "night blindness" can you instantly see all the detail in the tunnel?, then coming out of the tunnel, a little white clipping again?

If your point is we can see a range more than 200:1 across a screen, I would say yes. But, if you are really talking CR, it should be in close proximity, white next to black. A good test would be an inverse Pluge pattern, full white field with small black stair step window.
Quote:
Glen: can you see the difference between a FP display with an ANSI of say 200:1 and one much higher than that (such as a 1-chip DLP)? I mean come on here, you can't honestly be claiming that there is no difference there.

I haven't been able to do an A/B test. It also depends on what part of the picture you are talking about. My CRT (9500LC) might have a 120-130:1 ANSI and a very high ON/OFF CR, it has better low APL performance than brighter projectors with higher ANSI CR, but lower ON/OFF CR.
Reply
Reply
post #308 of 505
1/5/07 at 2:36pm
Darin,

William and others are testing something else there. They are measuring projector performance, measuring light out of the projector with different APL patterns. The way I see this thread, is more in line with what we, as humans, actually see and what some of the limits are. I think some of those tests are important because they are measuring "bulb" projectors low APL performance where CRT has really been King.

I still see issues like I watch a movie, then turn off the projector and suddenly I cant see anything. In a few seconds, I can see everything in the room. There is an "instantaneous" CR issue.
Reply
Reply
post #309 of 505
1/5/07 at 2:53pm
Glen: Good to have a voice of reason with credentials that cannot simply be laughed away
Reply
Reply
post #310 of 505
1/5/07 at 2:56pm
Quote:
Originally Posted by GlenC View Post

It clearly seems that your frustration is from the lack of understanding what others are saying and or your inability to express your specific point to where others understand.

Well said Glen!

http://www.s3.kth.se/signal/reports/...SB-EX-0225.pdf
The visual system can adapt over a huge range of intensities, of order 10^10, but not simultaneously. The following factors influence the eye's response

1)Brightness Adaptation: the eye changes it's overall sensitivity when focusing on a specific level of intensity. The range of brightness that an eye can perceive when adapted to a specific level is small. Not large Chris, and not 1,000,000:1 Darin!
2)Weber Ratio: The ability of the eye to discriminate between differences of intensity levels varies with the light intensity. The ability of discriminating increases non-linearly with the background illumination
3)Mach Bands: Near boundaries of regions with different intensity levels e.g. black and white stripes, the eye over shoots and under shoots. This means that the eye is more sensitive to contrast than absolute luminance
4)Simultaneous Contrast: Perceived brightness does not only depend on the image intensity. It also is determined by the surrounding intensity. This is the explanation of many optical illusions and is demonstrated in Figure 2 and 3
*Figure 3: This is an amazing illustration of the simultaneous contrast.

Master Thesis Report
Department of Signal Sensors and Systems,
Royal Institute of Technology (KTH)

Unfortunately the human perception system does not discriminate between light levels by a simple function of intensity. Brightness is what the eye perceives when light energy reaches the eye. The brightness is the human response to light and is subjective. The subjective brightness is a logarithmic function of the light intensity incident on the eye according to the Weber Fechner law. Which BTW still to this day is a powerful/accurate tool that can scientifically predict the subjectivity of the HVS perception of contrast.
Reply
Reply
post #311 of 505
1/5/07 at 4:33pm
TGIF!!!

Well, I just want to thank all those who have a lot more time on their hands than myself for stepping up to the table and giving some credence to my original post at the beginning of this long and sordid thread.

I will admit to oversimplifying my original comment and probably should have better defined specific conditions where the <300:1 number applies...needless to say, many of you have very specifically made my point.

Cheers,
Reply
Reply
post #312 of 505
1/5/07 at 4:49pm
Quote:
Originally Posted by tstites View Post

TGIF!!!

Well, I just want to thank all those who have a lot more time on their hands than myself for stepping up to the table and giving some credence to my original post at the beginning of this long and sordid thread.

I will admit to oversimplifying my original comment and probably should have better defined specific conditions where the <300:1 number applies...needless to say, many of you have very specifically made my point.

Cheers,

Actually I have found this thread quite interesting; it may be long but certainly not sordid. After actually reading the many references folks have posted I do not believe your contention is in fact well supported. If anything, it less than clear in terms of a hard number given the papers by experts in the field.

Quote:
Originally Posted by tstites View Post

The dynamic range of the human visual system is up in the range you mention WITH the use of the iris...in a static scene, the range is <300:1, typically far less depending on the factors maddog and I mentioned. I'll post some references for your review in a later post...have to dig them up.

Did I miss those references or were they just not posted.
Reply
Reply
post #313 of 505
1/5/07 at 5:12pm
Quote:
Originally Posted by tstites View Post

TGIF!!!

Well, I just want to thank all those who have a lot more time on their hands than myself for stepping up to the table and giving some credence to my original post at the beginning of this long and sordid thread.

I will admit to oversimplifying my original comment and probably should have better defined specific conditions where the <300:1 number applies...needless to say, many of you have very specifically made my point.

Cheers,

Could you clarify what your point was? Because the way I interpreted it you were incorrect. I would be interested in understanding what you meant. If you're claiming very limited CR capabilities such as 300:3 *at* a boundary, that's very different than what you originally claimed.

your original claim was basically:

Quote:


...in a static scene, the range is <300:1, typically far less

Which is incorrect. I don't know where these figures are coming from, nobody as of yet has cited any papers that claim this kind of a figure for across a scene.

Based on this claim, then projector ANSI CR or other simultaneous CR performance measures beyond 300:1 are unecessary. Are you willing to stand by that claim? That ANSI CR beyond 300:1 is not ever beneficial?
Reply
Reply
post #314 of 505
1/5/07 at 5:20pm
Quote:
Originally Posted by GlenC View Post

Darin,

William and others are testing something else there. They are measuring projector performance, measuring light out of the projector with different APL patterns. The way I see this thread, is more in line with what we, as humans, actually see and what some of the limits are. I think some of those tests are important because they are measuring "bulb" projectors low APL performance where CRT has really been King.

I still see issues like I watch a movie, then turn off the projector and suddenly I cant see anything. In a few seconds, I can see everything in the room. There is an "instantaneous" CR issue.

That's a temporal issue, and as such is completely irrelevant to this thread. When a CRT goes from white to black, that is SEQUENTIAL contrast. Do you understand the difference between sequential contrast and simultaneous contrast?

Do you realize that in the other thread, they are measuring simultaneous contrasts, not sequential contrasts?

I, along with Darin, am still confused with what your point is, because you still seem to be confusing sequential and simultaneous issues together and continue to make references to adaptation over time, especially in your clear references to a CRT's ability to black-out, which is a completely off-topic issue. After 11 pages, I'd hope we'd at least have a basic understanding that we are talking about simultaneous contrast issues and capabilities, not sequential CR which you keep wanting to refer to, which is not relevant to this discussion at all and just confuses what you're trying to say.
Reply
Reply
post #315 of 505
1/5/07 at 5:28pm
Quote:


If your point is we can see a range more than 200:1 across a screen, I would say yes. But, if you are really talking CR, it should be in close proximity, white next to black.

That is my point. We can see orders of magnitude more than that across a scene. CR *at* a boundary, i.e. at close proximity as I have explained many times over in this thread, is a distinct issue, and indeed immediately adjacent to high contrast boundaries we have dramatically reduced CR capability in our visual system. One paper I cited on disability glare uses a figure as low as 150:1 for a high contrast boundary. This is not of interest to me in this thread, nor is it relevant to what we can see across a scene. There is only one display on earth that I have ever seen that can actually exceed the human's capability in these regards and that is the Brightside HDR. There, the display's flare at a high-contrast boundary is below the flare in our eyes. Everything else you're looking at in terms of images (and not real scenes) is well below this threshold. This has been stated also in papers I have cited in this thread.
Reply
Reply
post #316 of 505
1/5/07 at 5:34pm
Quote:


It clearly seems that your frustration is from the lack of understanding what others are saying and or your inability to express your specific point to where others understand.

If you want to know why I'm frustrated, it's because I've been making the same points over and over for 11 pages. I've attempted to be very clear to distinguish temporal adaptation over time (sequential contrast performance and capabilities), our abilities across a scene (simultaneous contrast performance and capabilities), and our simultaneous capabilities in extremely localized and specific situations such as *at* a high contrast boundary. I feel I've been pretty clear at keeping those three things distinct. Unfortunately you and others have had an unbelieveable knack of mixing all those things together in to a giant confused mess. And now here we are again back at square one and you're bringing up sequential contrast issues as if that had any relevance to this thread...

And at this point I'm not sure you understand the difference between simultaneous and sequential CR, so we're in like negative squares here, this thread sort of assumes you understand the difference between these two things. If you don't, then I don't know what to tell you. :-/
Reply
Reply
post #317 of 505
1/5/07 at 6:31pm
Quote:
Originally Posted by ChrisWiggles View Post

That is my point. We can see orders of magnitude more than that across a scene. CR *at* a boundary, i.e. at close proximity as I have explained many times over in this thread, is a distinct issue, and indeed immediately adjacent to high contrast boundaries we have dramatically reduced CR capability in our visual system. One paper I cited on disability glare uses a figure as low as 150:1 for a high contrast boundary. This is not of interest to me in this thread, nor is it relevant to what we can see across a scene. There is only one display on earth that I have ever seen that can actually exceed the human's capability in these regards and that is the Brightside HDR. There, the display's flare at a high-contrast boundary is below the flare in our eyes. Everything else you're looking at in terms of images (and not real scenes) is well below this threshold. This has been stated also in papers I have cited in this thread.

The context of the above is one of the main points of this thread, reasons that limit or cause "reduced CR capability in our visual system." It should be of importance.

If I am not mistaken, the ANSI 4x4 pattern meets this high contrast boundary scenario (white next to black). If so, there is a potential for a projector, with a ANSI CR of 300:1 to out perform the eye (300:1 is greater than 150:1, right?). BTW, the other thread is measuring projector performance, not screen performance and especially not what we are actually able see.

You cannot accurately answer which is best, a projector with 300:1 ANSI CR or a projector with 200:1 ANSI CR and 20000: ON/OFF CR.
Reply
Reply
post #318 of 505
1/5/07 at 6:55pm
This seems like a very simple issue. If you have an image on the screen and when you look at it, you cannot see all the dark detail (the is clearly there if you get close enough), you are dealing with visual CR limits.

If in the above, the level where you begin to see the detail is measured at 200:1, then, in that situation, "there is no need to have a system capable of more than 200:1" would be a correct statement. It would also be correct for me to say my projector can achieve 10000:1 CR and I can see it all. I have a 1" white square in the upper right corner at 10 fL and Just not at the same time. I can see CRs on my screen and I can see the 1" square in the lower left corner that is .002 fL. If the squares were 2' in size the results would be different.
Reply
Reply
post #319 of 505
1/5/07 at 7:09pm
Glen: You are saving me a lot of typing. Keep it up. The irony is that Chris is the one who is confusing simultaneous with sequential contrast.
Reply
Reply
post #320 of 505
1/5/07 at 7:46pm
Quote:
Originally Posted by GlenC View Post

William and others are testing something else there. They are measuring projector performance, measuring light out of the projector with different APL patterns. The way I see this thread, is more in line with what we, as humans, actually see and what some of the limits are. I think some of those tests are important because they are measuring "bulb" projectors low APL performance where CRT has really been King.

They are measuring the simultaneous CR from the projector. Yes, it can be lower from the screen, but in a real proper room the simultaneous CRs off the screen can be close to those for the lower percentage of white image because there isn't that much bright stuff to bounce around the room.
Quote:
Originally Posted by GlenC View Post

I still see issues like I watch a movie, then turn off the projector and suddenly I cant see anything. In a few seconds, I can see everything in the room. There is an "instantaneous" CR issue.

Yes you do, but that does not tell you how much CR you can see at once. In that case you are overwhelming the eye with bright stuff. That is different that smaller amounts of bright and having adjusted to a level. And unless you've measured the CRs you are dealing with there, they could be huge.

I don't understand why you won't or can't answer this question:
Quote:
Originally Posted by darinp2 View Post

Do you think that people would be able to see the difference in a dark room with a screen with a dark border between having that pattern shown with 4000:1 simultaneous CR vs a system that only had 500:1 simultaneous CR for the same pattern?

Do you not know the answer or just don't want to say? That definitely has relevance to the limitations of our visual systems and saying that they are testing something different than what you want to test doesn't change that it is relevant. As I already stated, you are testing a certain kind of limitation. A person who puts a lot of bright up and a little bit of black detail really isn't looking for the most CR before people couldn't see any improvement. As I've discussed, a certain amount of that bright light spills over. If you are looking for the maximum then try a small amount of bright. If you don't want to do a test that looks for the maximum in one image then you can't claim to have found the maximum. Just some other limitation.

And there is absolutely no reason that CR has to be white right next to black. CR is Contrast Ratio of course and it can apply to things the are separated by some distance just as well as things that are very close to each other. The official ANSI CR test measures the center of the rectangles from what I understand, so even there they are measuring with a little bit of distance between the points.

--Darin
Reply
Reply
post #321 of 505
1/5/07 at 7:54pm
Quote:
Originally Posted by tstites View Post

Well, I just want to thank all those who have a lot more time on their hands than myself for stepping up to the table and giving some credence to my original post at the beginning of this long and sordid thread.

I will admit to oversimplifying my original comment and probably should have better defined specific conditions where the <300:1 number applies...needless to say, many of you have very specifically made my point.

So, do you think people would be able to tell the difference between an HD2K doing less than 1500:1 simultaneous CR with William Phelps' pattern with about 1% of the image containing 100 IRE squares and the rest at video black, compared to an HD1 doing over 5000:1 simultaneous CR with the same image? Your position seemed to be that images didn't need more than 300:1 because we couldn't see more than that. If that is your position, then please try to HD2K vs the HD1. I bet you could see the improvement on that image between the two in a proper room. And even if you don't, you could raise the brightness until the simultaneous CR was 300:1 there and you should be able to see how bad that would look. If your position was something else then please tell us, because a projector that was limited to 300:1 simultaneous CR would be a pretty poor performing projector as far as CR goes and your post made it sound like you didn't think there was any reason to make projectors better than that.

--Darin
Reply
Reply
post #322 of 505
1/5/07 at 9:40pm
Darin, I agree and you are right about the HD2K and HD1 CR. If both had ANSI CR of 300:1, you would naturally expect the HD1 to do better. With the ANSI pattern, you measure white and black. From a viewing standpoint, the ANSI pattern on a 16fL system would appear to the eye the same as a 8 fL white field (same lumens to the eye, give or take). we should be able to see a spot down to .03 fL, assuming 300:1. Then 16fL to .03fL = 533:1. If you use the Phelps pattern, where the eye sees .4fL (16fL x 2.5%), then at 300:1 we should be able to see some detail at .001 and 16fL to .001fL = 16000:1. The issue would be could you see a .03fL or .01fL box in the black field if it were placed adjacent to a white 16fL square?
Reply
Reply
post #323 of 505
1/6/07 at 5:49am
Thread Starter 
Hi all,

This certainly has turned into a very interesting read I have to say.

Although there is an awful lot of info here, I do find some of it confusing. I'll probably have to try and read a lot of it a few times to begin to take it all in.

I did try Chris's black squares test, and in a room with some ambient (15lux at the screen measured from the centre of the white square), I could see all the squares shown, including the 2% one. I measured the display and got these results (in lux):

black = 2
grey 1 = 11.5
grey 2 = 25.5
grey 3 = 54
white = 141

So the LCD display is only giving around 70.5:1 so by no means a good test in this instance (and ignoring the effects of ambient light). I know when watching some DVDs there is a lot of grey areas where there should be shadow detail, so the display is definitely limiting the source material.

Going back to what people are saying about the physical limits of certain parts of the eye, I don't think it's fair to use individual (component) limits if the eye functions as a whole, as it appears to be better than the sum of it's parts so to speak.

I still have problems believing the lower numbers when I can see the differences in low on/off projectors against high on/off projectors showing the same material, and the image from Gladiator which Eric measured at around 2500:1 in that particular scene seems easily discernible to me on my projector which can do just 2650:1 on/off. I'll have to pause that scene on the pj and measure it in various places to see what readings I'm actually getting in my room. I can see all levels of the greyscale when displayed too, and that's another thing I should measure along with the hand puppet test (I've a feeling my rooms ANSI will be the limiting factor here though).

I can see the boundary effect in action on some other scenes when looking directly at a bright light, but when looking at the scene as a whole I can see all of it without compromising the dark parts. This sounds like the 'elevator' example so in those scenes I'd have to measure the relative luminance to try and see what contrast I'm really seeing.

If indeed we can see just 150:1 to 200:1 at a boundary, and I can believe that from my experience, I don't think it's a fair limit to place on the eye as a whole when we seem to be able to see so much more when not at a boundary. Most images we see don't have a boundary with such a large range in close proximity, so overall we surely must be seeing a lot more than that (hardly a scientific conclusion I know but intuitively that's the conclusion I come to).

If we truly can see just 200:1 at a specific instant, then the eye/brain combination must be building up an image in our brain so that the the overall range of much higher contrast is rendered as a complete image with all contrasts intact (the brain doesn't have the limits the eye has). That's all I can think of that would allow us to see the complete range which we appear to be able to see in real life IMHO, and why we we need displays capable of above that (200:1).

In that case although the eye may be limiting us to some degree, overall we are perceiving much more. I'm still not convinced of the 200:1 though, and Poynton seems to think we are capable of 1000:1 so would like to know how he derived that figure.

Just my 2c

Gary
Reply
Reply
post #324 of 505
1/6/07 at 7:59am
Quote:
Originally Posted by Gary Lightfoot View Post

the image from Gladiator which Eric measured at around 2500:1 in that particular scene seems easily discernible to me on my projector which can do just 2650:1 on/off.

Gary I apologize in advance for repeating myself, but I am professionally responsible for our production (film/video) workflow. Please read this entire post (thanks in advance). Can you describe how 2500:1 contrast/DR was encoded into in the 'Gladiator Scene' with 8-bit gamma corrected video? Measuring the light output of a given display system has no correlation (reverse engineer itself) to the perceptual mapping of content onto medium of delivery. FWIW 2500:1 is far superior to the best daylight film stock used today.

The problem with really high dynamic range is that the image will look flat and dull if you just map it "linearly" (squeeze it all into normal 709 color space, without the film s-shaped transfer function / gamma curve.

Quote:
Originally Posted by Quantel View Post

A modern camera negative stock will have almost a 10-stop exposure range representing a 500:1 brightness range. However, a typical scene will be lit to give a contrast nearer to 100:1 (approximately 6.7 stops) or, for higher contrast, 300:1. But print film is different and will only show a portion of the scene contrast stored on the negative.

Generally, black to white on the print corresponds to a scene contrast of 200:1 so a 300:1 scene contrast means the Director of Photography (DoP) needs to make a decision on how to print the material. Does he roll off the highlights to show the deep shadow detail or crush the shadows to reveal the detailin the highlights?

http://www.quantel.com/resource.nsf/Files/Digital_Film/$FILE/Digital_Film.pdf

**Please copy and paste Quantel url**

Printed films rarely show a full 10 stops of dynamic range, even if the negative film can handle it. They're bleach bypassed, processed, etc. to about 7-8 stops at most (plus a smooth knee). The value of all this extra dynamic range is NOT that you want it in the final image (it would look washed out, bland, and have so much info in it that the frame would take too long to parse). The value is that you can adjust exposure however you want, and have enough data for exposure compensation, as well as color correction.
Reply
Reply
post #325 of 505
1/6/07 at 8:12am
Thread Starter 
Hi,

The 2650:1 figure is the measured on/off contrast ratio of my projector (I measured it myself with a light meter).

The other contrast figure was derived from a spreadsheet by Erik Garci, so he's in a better position to answer you on that one:

http://www.avsforum.com/avs-vb/showthread.php?t=765552

Gary
Reply
Reply
post #326 of 505
1/6/07 at 8:13am
Quote:
Originally Posted by armadillo View Post

epstewart: You should do the 5-square test and tell us what happens to the faint square at the various CR.

'dillo,

OK, I'll try it, but what is the 5-square test? Is it on Avia, Video Essentials, or Digital Video Essentials, the three test DVDs I have?
Reply
Reply
post #327 of 505
1/6/07 at 8:18am
Thread Starter 
I think it's this one:

http://www.avsforum.com/avs-vb/showt...&&#post9317959

Gary
Reply
Reply
post #328 of 505
1/6/07 at 8:24am
Gary, I agree with what you said. My take on this has always been that the criteria for the 200:1 needed to be further defined. The seeing is believing is a big part of it. With a CRT, you have Low white levels, say 8fL off the screen. At that, my tale on this topic is that on a full white field you should be able to see the difference in a small black square that is part .04fL and .03fL, but not .03fL and .02fL. ff the screen was at 16fL, then the three levels of black would be .08, .07 and .06fL. This assumes that the average level of light the eye is seeing is the full white field. If it were a 1/2 black and 1/2 white screen then boundary issues come into play and the further away from the boundary you get, the more CR you are able to see. This is a key factor in why CRT Projectors perform so good to us. They just shift the whole range into darker blacks where they performance has been the strong point.

This is definitely a moving target, and with out defined criteria, the "200:1 is all we can see" is meaningless.

The test is more along the line of an inverse of the low APL test patterns from William Phelps. Full white field with few small black squares, possibly with a just above black square next to it. At what point, adding additional black squares, lowering APL, could you see the difference in two levels of black.
Reply
Reply
post #329 of 505
1/6/07 at 8:36am
Quote:
Originally Posted by Gary Lightfoot View Post

The 2650:1 figure is the measured on/off contrast ratio of my projector (I measured it myself with a light meter).

I edited my post, I was referring to the 2500:1 number provided by Erik. If you can comprehend my post, I'm trying to illustrate i.e. the "industry" is in total agreement that film out-performs video, do to it's logarithmic nature, and tracks color like our eyes.

Reading my post above, you will see that film captures nominal peaks 300-500:1. Printing further reduces this to ~200:1. Due to the logarithmic density of the medium, film to this day is the benchmark in color and DR.

Despite what is understood on this forum (Chris, Darin), the MAX dynamic range per film transfer to video is 200:1, coupled with the fact that this information is coded linearly into a non-linear color space, explains why video looks comparatively flat to the 200:1 film release.

Thomas
Reply
Reply
post #330 of 505
1/6/07 at 8:53am
Thread Starter 
Hi Glen,

So you're saying we'd see a square that would give us 200:1 and 266:1 against an 8fl white, but not 266:1 and 400:1. So assuming you're saying that the 1st black square is half and half, or a smaller square within a larger square (a bit like the SMPTE test card), we could see the one that's around 200:1 and not the one that has overall higher CR.

I'll have to see how I get on with the tests I'd like to try to see what results I get.

Gary
Reply
Reply
New Posts  All Forums:Forum Nav:
  Return Home

AVS Top Picks

AVS › AVS Forum › Display Devices › Digital Hi-End Projectors - $3,000+ USD MSRP › Some people are still being told that 200:1 is all we can see....