Some people are still being told that 200:1 is all we can see.... - Page 3

Did you expect it to just vanish.

Waving, as in producing motion? Looking as in moving one'e eyes?
What data would one gather from that "test"?

OK gremmy,
How about this... lets toss out all the 'research' done on our HVS over the last hundred years, and replace it with Chris and Darin's observations regarding i.e. "car headlights while driving at nighttime"! Sound reasonable?

Yes.

Here's the deal. I think one has to be very careful when applying this 200:1 figure as an ultimate determiner of PQ. The human visual system is more complicated than this one number would imply.

The number itself is largely irrelevant to me, since I have no particular interest in the various parameters of human visual acuity. What I am interested in is whether or not the human eye is capable of deciphering an improvement given a particular increase in contrast. And I assure you that it can, despite that 200:1 number, regardless of the numbers accuracy pertaining to a particular attribute of visual acuity.

Let's not rely on so-called experts in regard to this issue. That is very disempowering to those of us who have been with digital projectors from the beginning and have seen drastic differences in contrast from the early models with 200 to 1 CR to today's models. I am surprised that people including myself are wasting their time with this thread since the idea or assertion that the human eye can discern no more than 200 to 1 CR has absolutely zero relevance to easily observed differences in home theater displays.

Exactly.

The most important point surrounding this entire discussion. Do not take the HVS CR number and convert into ANSI CR number!

Another key point! Sadly using ANSI CR units we don't know the upper limit, or even what ANSI CR really tells us about overall PQ.

Best LCOS ~ 300:1 ANSI CR
Best LCD PJ~ 300:1 ANSI CR
Best DLP ~ 900:1 ANSI CR
Best Plasma ~1,000:1 ANSI CR
Best LCD Panel ~1,300:1 ANSI CR

I completely agree.

Unfortunately, the ISF guys referenced in the original post seem to have been doing just that. Although it sounds like they had taken it one step further and confused the HVS CR number with on/off CR.

Anyone using the HVS CR number to assert that increases in ANSI or on/off are not visible is barking up the wrong tree.

Interesting that all the (so-called experts) PhD research on the SM, with all their fancy Greek letters and such, did not think of this?

I'm not sure all those experts were making specific reference to the contrast measurements of various display devices.

P.S.

Quick question for you. Do you log onto AVS forum everyday and do a thread search for "contrast?"

Since on/off contrast of this magnitude is easily measureable with the right equipment, I certainly hope that no ISF technician was laughing this off as an unattainable goal. If it was just an attendee, then we can chalk it up to ignorance. But a tech? That person should know better.

Gremmy,

Your quote and the phrase I highlighted in RED are interesting. This is one of the reasons I find most of the CR measurements essentially useless in determining the actual quality of a projectors image. The action of changing your viewing distance is changing the visual angle of the main picture element and your screen border. You seem to be investigating some aspects of that High Contrast Border measurement that was mentioned earlier.

Please keep this debate going! You never know when something you say may trigger an idea that leads to a better understanding or measurement methodology for us HT nuts!

Because your comment betrays your misunderstanding of the point being made, and to dismiss the knowledge of experts on the field without verifying one's correct understanding must be the pinnacle of hubris (not a comment on you Leckian, but rather our culture and its far too common practice of glibly dismissing scientific knowledge). There is a difference between the dynamic range and the threshold of cr perception. You are referring to dynamic range of vision, and nobody of any repute is disputing that. The discussion has to do with instantaneous (i.e. fractions of a second) perception of cr.

Darin, the problem with your test is that you are not able to isolate a critical confounding variable. You provide no means to determine if the iris in the eye has remained static through your test; i.e. any change in the iris would invalidate your findings. In other words, as simple and intuitive as your test seems, in borders on naivete testing in its assumptions of isolating meaningful variables. If you can demonstrate that your proposed test can account for any changes in the eye's iris, then you may have a point. However, as described, your test is just slightly better than the one's suggested by the golden ears in audio wrt differentiating the effects of cables. Although your test can actually provide an objective measure (making someone else create the puppet figures) of difference, thus diminishing the potential of placebo effects, it cannot do so by isolating the actions taking place within the eye (thus precluding any conclusions regarding the threshold of perception in the absence of iris adaptation).

Well, you ignored the context of my experiment. I can see the pointer as visibly brighter than peak brightness of a 250:1 ANSI diaply.

I can see blacks significantly darker than the the blacks on that same 250:1 ANSI display.

This is all within the same field of view at the same time.

The claim here is that those extremes would have been crushed and that I should not have been able to see the differences, no?

Within what time line?

I think it's related to both the angle and the proximity (and therefore intensity) of light bouncing off the border.

Yes. And the only scientific statements that directly bear on the questions at hand come from the folks at Brightside. And they cite 100,000:1 as the instantaneous CR capability of the human eye across a single scene with no iris adjustment.

Talk about hubris all you want. On the other hand, we have multiple numbers being claimed for this capability by numerous lay people with no clarity of where these numbers come from or what they mean. The closest thing to a scientific statement in a paper is one sentence in what Houston linked which doesn't really illustrate much of anything and is pretty ambiguous about what they are talking about.

And we also have some common sense what we can apply to this question to see whether these so-called "scientific" claims are even reasonable. They are not reasonable, and they don't come from scientific sources. Houston refuses to do a simple test which quickly illustrates that the notion that we can only see "a couple hundred to one" at any one instant is ludicrous. Can you see white detail and shadow detail that exceeds the ANSI CR range of the display which on its own, it is being claimed, exceeds our capabilities? Yes you can. Now, apply some very simple logical reasoning to this. You have a variety of confused people asserting some things about the limits of what we can see with little information or evidence to back up this claim. Common sense and a quick test apparently seems to disprove their claims. And there are more informed scientific claims which cite other figures which much better match our natural experiences in the real world and also what we can see in our theaters. I mean, it's either that, or we all believe that we have super-human eyesight.

Oh my goodness! Do the test. I swear, if you really think seeing the shadow is attributable to the "placebo effect" then you're on acid. Really. This is ridiculous. It takes five minutes. Be a groundhog and tell me if you see your shadow. It illustrates in a very simple way that we can see across one scene greater than the claims being made here. You are right that it doesn't account for other adaptation adjustments, but that's immaterial. The question is how much can we see in a single static scene. The claims are that we can only see 100:1, 200:1 or maybe 300:1 depending on whose random numbers you're wanting to believe. This very crude experiment illustrates immediately and unambiguously that these numbers are absurd.

Or just attribute the shadow to the placebo effect. I mean come on. I feel like Andy Rooney here...

Agreed completely. And this is the point I'm trying to make too.

Most of the people arguing about this point don't really care about the adaptation of the iris or other specific complications of human visual acuity. Rather, the important point for most A/V geeks is whether or not an increase in contrast is VISIBLE. And, as you have explained quite succinctly, it IS visible well beyond the 200:1 number being thrown around.

You *should* feel like Andy Rooney.

Do they provide a journal reference for their claims where the specifics are clarified?

You are missing the point. The test you provide CANNOT isolate the confounding variable introduced by the iris, which all recognize as introducing a dynamic range exceeding a cr of 1000,000:1. The test you propose can only have validity if you are able to demonstrate you were able to isolate the effects of the iris. On the surface, the description you provide does not give any evidence of this.

However, all of this argument can be said to be "much I do about nothing", since the argument that contends a 200:1 static, and dynamic range exceeding the 100s of thousands easily suggests that today's projectors do not come close to approaching the limits of visual acuity of your avg human. That is, the dynamic range model recognizes that both the ANSI and On/Off CR currently available through both CRTs and digital projectors are still not within the boundaries of the limits of human perception, so there is no need to argue for a conspiracy that is trying to limit the progress/development of better technology.

I used ANSI CR because it is a pattern that you can find on a test disc and I have explained multiple times that there are patterns were we can see more than that pattern. Not sure why you went off on your "no direct relationship" thing since I didn't say there was any direct relationship there. You also seem to want to ignore questions like whether you could see a difference there or like: "If MrWigggles image was put up with 500:1 simultaneous CR and that was improved to 4000:1, do you think you would be able to see that the 4000:1 was better?"

So, is your claim that people can't see more than 300:1 in a single pixel checkerboard, or is it that people can't see more than 300:1 in any single image, or something else? Please clarify why you mentioned the single pixel thing above and if your claim of 300:1 is for single pixel patterns, why you would apply that to images that are much different?

If you understand, then why not answer the question about MrWigggles image? And you should be able to tell us whether your position is that people would not be able to perceive improvements beyond 300:1 simultaneous CR in any image or not. If that is your position, then it should apply to any image I use as an example, not just some special ones that are more limiting than those.

And do your claims apply to things with movement or only static images? If your point isn't that we are limited to not seeing improvements beyond 300:1 at once, then what is it?

Nope. I have gone over multiple different tests that can be used and even if the iris moves in 1 millionth of a second and I can't tell, it would be those who are saying we couldn't perceive improvements beyond x from a display who would be wrong for applying things incorrectly. When I am looking at the scene from "Sin City" that I talked about in my article about CR with the projector zoomed down to less than the size of the screen (to create a border between the image and the screen edge) and I can make out multiple levels of bright and multiple levels of dark, sure it is possible that there are some extremely quick transitions in the eye that allow me to perceive all of those and not just crush things that measure outside the 100:1 or 300:1 range, but that just validates the point that the images need more than that for simultaneous CR before we could not perceive any improvements beyond that CR. As I pointed out before, anybody looking for the largest range needs to have transitions in both directions and that perceiving even 50:1 below and 50:1 above some reference would mean a total range of 2500:1 there.

I think I'm repeating myself, but we should remember that projection systems can do much higher simultaneous CRs than their ANSI CRs. The ANSI CR test is a more difficult one than some others. A projector with 100:1 ANSI CR and 10k:1 on/off CR should be able to do over 500:1 simultaneous CR in lots of images.

--Darin

If I understand Houston's argument correctly, he does not disagree with your general conclusion; i.e. we can see a much broader range of cr than is presently available through FPs. The question is whether the threshold of vision is limited by the eyes static threshold. Again, my understanding of Houston's claim is that the dynamic range comes into play and that is why we are able to see better than 300:1, even thought the static range limit may be around that number. I have seen people argue that 300:1 is all that we need, since that is ~ the static limit of they eye, but those arguments totally miss the effect of dynamic range of the eye. That is why I stated this is much I do about nothing, unless Houston is willing to argue the dynamic range is not at play, which I don't think he is. Your test clearly demonstrates that we have a dynamic vision that easily exceeds 300:1, but it cannot prove anything about the limits of static vision since you have no way of isolating the effects of the iris (but in reality, who cares since the important factor is can we perceive it in a dynamic manner; i.e. the way we experience the world).

Some interesting optical illusions that seem to demonstrate the phenomenon.

http://www.hutchinsoncenter.umaine.e...ve/DynLum.html

-Dave

After reading HH later post, I think I agree with you. It seems like he is saying the 300:1 figure is an accurate number, but it is an irrelevent number, at least for the purposes of this forum.

It may be somewhat related to ANSI CR, but it is not ANSI CR, and he agrees that the people are wrong when they say that 300:1 ANSI CR is enough for projectors. That is a misapplication of that number.

Is that right?

Yes, I think that's right. I can't speak for Houston, but that's very close to my own take, with the exception that I'm not prepared to make a thesis one way or the other regarding the merit of the HVS number, nor do I care about it.

So it seems it's quite possible that we all agree on a critical point regarding the post that started this thread off. That is, we all agree that FP contrast ratios are still not at the limit of human visual acuity. And furthermore, the ISF techs referenced in the original post were misapplying the numbers. Shame on them. It makes me NOT want to hire one of these guys to calibrate my picture if that's the kind of crap they're spouting.

Thank you.

For a display to equal the eye's dynamic range (on/off) it would have to support a range of > 10^8:1, way beyond the limits of any HDR system on the drawing board. If people do not want to believe that at any given instance the HVS can only discriminate ~10^2 within that range, well thats life!

On the instantaneous side I would speculate (I have no supporting data) that if a display could deliver ~300:1 per picture element, that may be enough. As we all know there is not a device in existence that can come close to delivering that level on ICR.

Is it possible to revoke a person's ability to post links?

Just to be clear, is your position that a display would only need to be able to display 300:1 for any one frame for us to not be able to see improvements to the CR at one time from that system, or not?

By "per picture element" are you talking about per pixel, or per full picture? If your position is that each full screen image only needs 300:1 for us to not be able to see improvements beyond that, then why not answer the question about MrWigggles example image?

I think most of us agree that the system needs a huge dynamic range in order for us to not see improvements in on/off CR, but as far as CR in individual images, is your claim 300:1 as the limit?

--Darin

I removed my own post. Is it possible you have issues with empirical data?

Never mind you answered my question already!