Found this here -- wanted to hear your reactions....

http://www.projectorreviews.com/advice/dlpvslcd/hometheater.php

"I had the opportunity to speak with a Hollywood colorist some months back.(He is responsible for getting the balance right for transferring film to DVDs. DVDs of course have different properties than the original movie film. I can't confirm his numbers, but he indicated that almost all non-animated movies are still shot with film ( a couple of Star Wars, and others excepted). Further, he indicated that the contrast ratio of the color films today is somewhere in the 6000:1 range. That raises the question - if the film only has 6000:1, what benefit would you get if your projector really had 15,000:1 contrast. The answer - I would conjecture, none or very little, unless you are dealing with a movie or other material, that bypasses the film process. Even then, the digital movie cameras would have their own maximum contrast. Basically you want to reproduce faithfully with your projector, not monkey around with settings that change the way the material was meant to be viewed."

If this is true, it would explain a lot of things....

To be clear, you're asking what the contrast ratio of film is?

I'm sure it varies with format size, ISO and whatever other factors there are, much like still photo film. Medium format and black and white film have highly increased CR ratings over 35mm color film. Velvia, however has managed to beat, say, Kodak in CR for 35mm.

What are the max/avg/low CR values are for film?

To be clear, you're asking what the contrast ratio of film is?

And if it tops out somewhere around 6,000:1

And if that limits digital reproduction (projection) to top out around 6,000:1

Found this here -- wanted to hear your reactions....

http://www.projectorreviews.com/advice/dlpvslcd/hometheater.php

"I had the opportunity to speak with a Hollywood colorist some months back.(He is responsible for getting the balance right for transferring film to DVDs. DVDs of course have different properties than the original movie film. I can't confirm his numbers, but he indicated that almost all non-animated movies are still shot with film ( a couple of Star Wars, and others excepted). Further, he indicated that the contrast ratio of the color films today is somewhere in the 6000:1 range. That raises the question - if the film only has 6000:1, what benefit would you get if your projector really had 15,000:1 contrast. The answer - I would conjecture, none or very little, unless you are dealing with a movie or other material, that bypasses the film process. Even then, the digital movie cameras would have their own maximum contrast. Basically you want to reproduce faithfully with your projector, not monkey around with settings that change the way the material was meant to be viewed."

Well he or she is dead on. And considering that digital cameras have even less CR than film kinda pushes the point even farther! But the truth is most projectors have less then 5000:1 if measured honestly.

Here's something else I found while surfing.....


"In our experience, people using monitors or digital projectors to simulate film will need something like a 1500:1 contrast ratio. There seems to be a point somewhere a bit beneath 2000:1 where the blacks come convincing, and the viewer will accept the simulation. There is some point about 1200:1 where the blacks stop looking convincing, and start looking grey.

If you are trying to match a display to a projector, it is nice to have another factor of two, so you can match the absolute brightness without having to go to the display white. You may want to get this because you sometimes have to drive the RGB channels beyond the white point to get bright and clean looking pastel colours.

You will want to have a continuous tone curve. Field-emission devices will have a cube-type power law down to a point, and then they will cut off exponentially. This may give good-looking greys down to a point, and then plunge into black, crushing all the shadow detail. That does not look as nasty as 'milky' shadows, but it is not that much better.

So - about 3500:1 is good for simulating colour film.

http://science.slashdot.org/article.pl?sid=05/10/03/1049254

Devils' advocate here. This guy certainly can't claim that the goal of a projector is to reproduce the color of film, because his entire job is to rebalance the color for DVD. Well, the same thing happens for grey scale, too. And they're doing that rebalancing not on a projector, but on a calibrated CRT with true black.

So it seems to me that the goal isn't to reproduce the film experience, but to reproduce the DVD master.

Finally, there is the experience of people here who are quite anal about contrast (both on/off and ANSI). I don't know of anyone here who wishes they had less.

But the truth is most projectors have less then 5000:1 if measured honestly.

I have been reading reviews of various projectors and I find it really interesting that people have reported anecdotally that a projector with a [measured] on/off CR of around 4,000:1 had CR's equal to that of a projector that supposedly measured far greater in on/off CR. The former was a Planar PD8150, the later a JVC RS1.

I personally watched a projector with a [measured] CR of 7,000:1 and thought it had more convincing blacks than one [measured] at 30,000:1.

The former was a C3X1080 and the later a JVC RS2.

However, these were on different screens, too.

The C3X1080 was on a Firehawk.

This leads me to believe that most people are not getting anywhere near the claimed measured CR's in their actual rooms. Also leads me to believe that after a certain point, other things besides on/off CR come HEAVILY into play.

For just about any projector, the only time you're really going to exercise on/off CR fully is when you fade to black. Once you have some projected image, the instantaneous contrast limits of your projector and room tend to take over. Some of these 3-chip DLP projectors in particular have stunning ANSI CR numbers, and they throw a more convincing black when there are other things going on in the scene. I don't know how JVC is doing these days but their ANSI CR numbers used to be pretty disappointing.

But the truth is most projectors have less then 5000:1 if measured honestly.


That's demonstrably false, at least if you're talking on/off CR which I assume is what you mean.

"Basically you want to reproduce faithfully with your projector, not monkey around with settings that change the way the material was meant to be viewed."I've asked this before and never seen a good answer that supports the above. In Finding Nemo there is a scene where the characters go down into the abyss and one of them is clearly saying that he cannot see anything, while the images are encoded as video black. Was this meant to be viewed with lighting coming off the screen just because some projectors can't actually provide no light? In Cast Away there is a sequence showing Tom Hanks going in and out of consciousness, using blackouts. Should those be grayouts just because some film projector can't block the light for the parts where Tom Hanks is supposed to be blacked out?

If some vehicle only gets 30 mpg, does that mean the owner wanted it that way? When there is a limitation to the technology there is just about no way to know for sure whether the person responsible for it wanted it that way or whether they hit the limit of the capabilities, but a person can make an educated judgement and encode higher than video black or not.

Also, as Michael pointed out, what we get is mastered for us, and not with film projectors. The mastering is mostly done on CRT monitors. From what I've heard those monitors generally have more than 6k:1 on/off CR. If the people encoding want more light for absolute black then they can encode the material higher. I've never heard of any of these places purposely raising the black level of the CRTs in order to match film, which in the past has been quoted as 2k:1. We had these discussions in the past where people kept saying that anything higher than 2k:1 was wrong. Now I see higher numbers for film and have been told that there has been film stock in the past that could actually black out (had some pretty thick stuff on it).

One thing I find interesting is that of the people in the industry who say that on/off CR shouldn't be above what film can do (not sure what conditions they want that to be under or whether they count any silver emulsions), I can't think of a single one who purposely degrades ANSI CR with digital displays to match film. I think some of them revel in having better ANSI CR than film.

If digitally encoded material is supposed to be played back with absolute video black at 1/2000th, 1/5000th, 1/6000th, of white, or whatever the number of the day is attributed to film, why not set the standard(s) so that video black calls for that level of light compared to reference white?

As far as what cameras can capture, capture and display are 2 different things. Cameras can capture no light. They just can't differentiate it from a higher level of light. So, during encoding a person doing that job can choose to set things to video black or to set them higher if something higher than video black is desired.

I realize that there are limitations with the bit depths and unfortunately the displays used to master may not be up to what some consumer equipment can do now or later, but when it is obvious during a section that a complete blackout is what would be called for, I don't see how setting it so that the screen is lit up matches the intent just because with film there is basically no choice due to light tending to leak through the film material. And of course with video material (not film) the limitations of film projectors aren't really relevant to what video can do.

--Darin

Another Interesting tid-bit....

"Projectors claiming 6,000:1 to 10,000:1 in reality are under 1000:1 if you use "Real World" tests."

"I have yet to see a Home Theater room that didn't decrease the actual contrast ratio due to reflected light; Black walls, carpets and all."


http://www.ausmedia.com.au/projector%20contrast%20ratio.htm

Another Interesting tid-bit....

"Projectors claiming 6,000:1 to 10,000:1 in reality are under 1000:1 if you use "Real World" tests."

"I have yet to see a Home Theater room that didn't decrease the actual contrast ratio due to reflected light; Black walls, carpets and all."

http://www.ausmedia.com.au/projector%20contrast%20ratio.htmI wrote my article about CR here:

http://www.hometheaterhifi.com/volume_13_2/feature-article-contrast-ratio-5-2006-part-1.html

partially because of how ignorant that guy is. He got some stuff right, but a fair amount wrong. I'm sure he means well, but his article has been misleading people about this subject for years. He uses CSF (the Contrast Sensitivity Function) to try to show how large CR doesn't matter. When I pointed out to him that the CSF measures the lowest CR a person can see he added the part that starts, "I'm not an ophthalmologist so feel free to do your own research, some will disagree with figures here ..." and still left most of the useless CSF stuff there. Would you put much stock in what somebody said about the upper limits of CR before we couldn't see improvements if they tested whether a person can see off-white against another off-white, instead of black and white, to come to their conclusion about the most CR? A score of 300 in the CSF test is a CR of something like 1.01:1, not 300:1.

As far as his:
Also called "Checkerboard" it's a more realistic way to test a projector than On/Off but not favored because of marketing concerns.ANSI isn't a more realistic way, it is measuring something different. On/off CR tries to measure the whole range while ANSI tries to get the washout effect most.

The guy says that when watching with "reflected and/or ambient light", "We say the eye can see an average of 300:1 CR". On/off CR isn't affected by reflections in the room if there isn't other lighting and other than very minor color issues. On/off CR needs to be high to come close to real life because this is the kind of stuff we can see according to this NASA data:
http://msis.jsc.nasa.gov/sections/section04.htm#_4.2_VISION


Example Scale of Luminance (millilamberts) Effect
Sun's Surface at noon 10^9 Damaging
10^8
10^7 Phototopic
Tungsten Filament 10^6
10^5
White paper in sunlight 10^4
10^3
10^2
Comfortable reading 10
1 Mixed
10^-1 Scotopic
White paper in moonlight 10^-2
10^-3
White paper in starlight 10^-4
10^-5
Absolute threshold 10^-6

Reference: 337, p. 26; NASA-STD-3000 194
And for those who want you to think that the eye takes so long to adapt that 2000:1 or 6000:1 on/off CR is enough, I have shown with measurements that I can see across a range of over a million to one in 3 minutes (the time long it took me to measure bright stuff outside, get inside and do shadow puppets). It is true that our eyes take some time to adapt to the biggest ranges, but we can see over a range of a heck of a lot more than 6000:1 in a very short time.

And as far as this argument about real world, there are real world images where the brightest thing in the image is 10 IRE (10%stim) or less. Because of the way gamma works that ends up being about .5-.6% of white. 6000:1 means the most CR you will get in that image if 10%stim is even up at .6% of the amount of light of white is about 36:1. So, if you can see 300:1, then you are way below that if you are limited to 6000:1 on/off CR and you end up with an image that ranges from 0 to 10 IRE. For those who might think we are never in situations in real life where the brightest thing is as dark as 10 IRE, if the white level is about what that chart shows for comfortable reading level (10 millilamberts, which I believe is around 9 ft-lamberts) then 10 IRE is brighter than what the chart shows for white paper in moonlight. So, if you are under a full moon in real life with that being the only light and have various things around with the brightest being white paper, good chance that everything you are seeing is darker than 10 IRE should be in a normal theater.

If the writer really wanted to cover "Real World" tests he would test some dark scenes, not just some guy with a white shirt standing in front of a black background. I'm not sure how "Real World" became no dark scenes. "Real World" should range over different kinds of images that happen with real content. As far as what you quoted, if the writer really knew what he was talking about I don't think he would have made that bogus claim. If he knew what he was talking about he probably would have said that in order to get good CR in real world tests of dark images it takes a lot of on/off CR (since the darkest images tend to only use a portion of the on/off CR available and so on/off CR must be big for the CRs in those darkest images to be good).

--Darin

It makes sense to master with a highly specified display. Padding is always good. However, is there a benefit to a display at home having a higher contrast ratio than the display the source was mastered with? Is the CR fish fried? Are chromatic aberration or sensor noise or other image quality projects better places to put corporate development dollars?

ANSI isn't a more realistic way, it is measuring something different.

This seems controversial. In surfing around the net, I find quite a few with opinions similar to this writers'.

If he knew what he was talking about he probably would have said that in order to get good CR in real world tests of dark images it takes a lot of on/off CR

Again, there is not universal agreement on this. This writer seems to be saying that "real world" includes mixed scenes -- that's hardly controversial. He's also saying most rooms compromise CR -- even ones that are all black. I've read this in several places as well.

(since the darkest images tend to only use a portion of the on/off CR available and so on/off CR must be big for the CRs in those darkest images to be good).

This is what he wrote...

"So unless your room is TOTALLY black and you're watching very
dark videos don't stress too much about contrast ratio!"

So, he covered dark images, but it seems he doesn't place as much importance on the "darkest" images. He does measure a pretty dark image and still finds the CR shrinks to a fraction of the on/off. But, he's not ony talking about the projector -- he's also talking about the room.

I don't want to be offensive, but I find that JVC owners tend to be extremely sensitive on this topic because they own a projector with a high measured on/off CR. Don't you own own a JVC RS2? Do you think this could be causing you to want to discount opposing opinions to the extent that you cannot even seem to concede his point about mixed scenes or room reflections?

rsbeck, it's kind of risky to criticize someone's arguments as being unduly influenced by what they own. But it is particularly so in Darin's case, because that dude has personally auditioned and owned an incredible number of projectors. Heck, he's owned and used more than one at the same time. I don't know which one (ones?) he has settled on now, but I think it's safe to say he's familiar with far more than what he uses---and that the ones he likes are informed by his preferences, not the other way around.

Thank you. I know Darin's well earned rep. And I'm sorry, but I still think it is a fair question. I am not the reverent sort. I am the type more likely to believe that we are all human -- well, most of us -- and no one is beyond question. Having read Darin quite extensively, I am sure he will have no problem responding to my question in an objective manner.

Unless he gets cranky. :-)

Fair enough. Certainly, Darin can take care of himself. But I've never got the impression that Darin dismisses the importance of instantaneous contrast. The fact that he has a black hole for a theater suggests he's interested in maximizing both types of contrast.

As a former older-generation JVC owner I appreciated its weakness in the area of ANSI CR. I don't know how good the latest models are.

That's a fair point, but having a "black hole" for a theater could also cause sensitivity to the writer's point that even "black hole" theaters compromise CR.

The author of the article writes....

"This doesn't mean that a projector with a CR of 2000:1 or more isn't better than one rated at 400:1, however the perceived advantage is a lot smaller than optimistic claims the manufacturers would have you believe."

And....

"I haven't yet seen a Home Theater room that didn't decrease the actual contrast ratio due to reflected light. Black walls, carpets and all."

http://www.ausmedia.com.au/projector%20contrast%20ratio.htm

There is certainly a lot of good advice in there for people who for other reasons (multi-purpose rooms, WAF factors, etc.) really aren't interested in creating black hole theaters...

But I think statements like "ANSI is the only fair way to test contrast ratio" are seriously flawed. ANSI doesn't provide the full view of a projector's contrast performance any more than on/off does by itself. You really do need to know both numbers, ideally measured in your room, to construct a good picture of the overall contrast performance of your system. His one-sample "real world contrast ratio" idea is a good one, but it's just a sample as opposed to a spectrum.

But I've never got the impression that Darin dismisses the importance of instantaneous contrast.


I don't know. Have you read this thread?

http://www.avsforum.com/avs-vb/showthread.php?t=1035229

In this and other threads, I have found that the debate falls along pretty predictable lines with JVC owners usually minimizing the importance of ANSI (an area where their projector does not enjoy superiority in measured numbers) and maximizing the importance of on/off (the area where their projectors do enjoy superiority in measured numbers).

As a former older-generation JVC owner I appreciated its weakness in the area of ANSI CR.

That would make you the rare bird.

But I think statements like "ANSI is the only fair way to test contrast ratio" are seriously flawed.

Maybe I missed it, but I don't think the author of the article wrote that.

You really do need to know both numbers, ideally measured in your room, to construct a good picture of the overall contrast performance of your system.

Right on.

His one-sample "real world contrast ratio" idea is a good one, but it's just a sample as opposed to a spectrum.

Agree.

I don't know. Have you read this thread?
No I haven't! I gave up my projector dreams moving into this new house so I've not been following it lately. Thanks for the pointer.

On first glance though I'd sure like to know what the measured ANSI CR was for those projectors. I mean, 250:1 is better than many rooms would achieve, so in such a room 790:1 would be a diminishing return. My JVC was easily worse than either of these! (EDIT: OK, they did do room measurements. I'll go back to reading.)

That's a fair point, but having a "black hole" for a theater could also cause sensitivity to the writer's point that even "black hole" theaters compromise CR.

The author of the article writes....

"This doesn't mean that a projector with a CR of 2000:1 or more isn't better than one rated at 400:1, however the perceived advantage is a lot smaller than optimistic claims the manufacturers would have you believe."

And....

"I haven't yet seen a Home Theater room that didn't decrease the actual contrast ratio due to reflected light. Black walls, carpets and all."

http://www.ausmedia.com.au/projector%20contrast%20ratio.htm

The writer is mixing sequential contrast numbers with ANSI contrast effects. Anyone who has seen a projector like the old HD2K or Qualia which hit around 2000:1 sequential contrast and compared them to an RS2 will see an incredible difference in image depth due to this difference.

Although I will agree that maunfacturers are taking full advantage of the improvement numbers the effect is real and dramatic not all hype.

Art

Wow, Mark Petersen's graphs are interesting. I really have been missing out on some developments lately. Of course, they're kind of moot for me now :)
http://www.avsforum.com/avs-vb/showthread.php?p=13995765#post13995765
Yes the RS1 washes out blacks with bright content and a DLP like the Sharp 20K washes out blacks with dark content. If you pick pretty much any test intra-image test pattern with full black and white the RS1 will dominate at the dark end and the high ANSI projector will dominate at the bright end. Here is an example that I've posted before:http://public.csusm.edu/kai/mark/avs/image001a.gif

I find all of this fascinating. It would be interesting to have a CR meter next to the screen while watching to see where ANSI -- and/or other factors -- comes into play.

Actually, you would need two screens, one for the high ANSI projector and one for the low ANSI projector, so you could then compare.

It would also be interesting to determine a "real world" % of dark scenes in "real world" home theaters where higher on/off gives a better *perceivable* result than higher ANSI.

I think this is the writer's point and this is why this will remain controversial until on/off ratios become high enough across the board to eliminate it as a topic of discussion. The value placed on these aspects of projector performance seems to depend on the particular viewer's preferences and taste and the trade-offs with which he/she is willing to live.

Ultimately, though -- I still think the writer has an interesting point -- and that is, "how much of what is *measured* is *perceivable* in an actual home theater during regular viewing?

In audio, we know that we can measure differences that cannot be heard, that are outside the range of audibility, etc. We also know that, even though testing shows that measured differences are below the threshold of audibility, some will still claim to have "golden ears" and will claim to be able to hear them.

I suspect the same is true in video.

I think this is the writer's point and this is why this will remain controversial until on/off ratios become high enough across the board to eliminate it as a topic of discussion. The value placed on these aspects of projector performance seems to depend on the particular viewer's preferences and taste and the trade-offs with which he/she is willing to live.



I think the writers point is well taken but he is defending his position of on/off not being as important as manufacturers claim with ANSI issues.

It doesn't take a golden eye to see this just a couple of minutes with a few demo sections and it is obvious.

Art

I think the writers point is well taken but he is defending his position of on/off not being as important as manufacturers claim with ANSI issues.

Excellent point. I guess it's more the point that I am taking from his (and other) articles rather than his entire article.

My take is that having a higher ANSI CR goes a good way towards minimizing the negative effect of having lower on/off, but having a high on/off is preferable to having a high ANSI CR and a low on/off.

My reasons are:

1.) Unless you have a black velvet covered room with no light pollution, your ANSI is already severely compromised.
2.) In a room without light pollution but with light walls & ceilings, ANSI CR is negatively affected much more than on/off.
3.) I have never been bothered (personally) by the deleterious effects of lower ANSI CR but I have noticed and been quite annoyed by grey blacks.
4.) CRT's have been praised for their incredible depth. However, their ANSI CR is ~ 125-150:1 on the high end IIRC.

I own an RS1, but it is my fifth projector in the last 8 years and I also owned a CRT projector for a while. Dismiss my opinion as a fanboy if you like, but I would gladly trade projectors for one with better performance in areas I find critical.

I also think it is in poor taste to question Darin's objectivity as he is one of the forum's members that would be the least likely to engage in fanboyism.

Art - If you could get the on/off of your G90 stack with the sharpness and light ouput of your HT5000, would you give up the ANSI CR performance of the HT5000? Assume the ease of setup of the two were the same (ie, just consider image quality parameters).

Art - If you could get the on/off of your G90 stack with the sharpness and light ouput of your HT5000, would you give up the ANSI CR performance of the HT5000? Assume the ease of setup of the two were the same (ie, just consider image quality parameters).

Yes.

Art

This seems controversial. In surfing around the net, I find quite a few with opinions similar to this writers'.Mass ignorance doesn't make something right. ;) Unfortunately, there is a lot of misinformation out there about this subject matter, some even worse than some of this guys. For instance, ISF trainers telling people that on/off CR doesn't matter and ANSI CR is all that matters. I've tried to figure out how some people got so confused and I think part of it comes down to them thinking that ANSI CR is a mixed image and on/off CR isn't, so therefore ANSI CR must be what affects mixed images. The truth is they each affect most mixed images in different ways. If one is weak the point at which its weakness overcomes the weakness of the other to dominate the CR is certain images will move higher or lower.
Again, there is not universal agreement on this.If they think that on/off CR and ANSI CR measure the same thing then they are confused. If they think that one affects mixed images and the other doesn't, then they are confused. If they think that there should only be one kind of CR measured then they don't understand the advantages of measuring toward one extreme and toward another. If somebody wants to use real world images, then they should use more than one, because image content varies and so what is needed varies. There is an absolute black floor and there is a washout effect. With only a single "Real World" image you cannot measure when the absolute black floor is the most important factor and your can't measure when the washout effect is the most important factor. You can just try to compromise and grab some of each with that one image, but that won't tell you what happens with images more toward the extremes since you don't know how much of the CR in that "Real World" image is degraded by the raised black floor and how much it was degraded by the washout effect.
This writer seems to be saying that "real world" includes mixed scenes -- that's hardly controversial.The key is "scenes", not "scene". They measured exactly one scene for that from what I can tell. There are a lot of mixed scenes that are way darker than their example and so much more dependent on on/off CR than there one example.
He's also saying most rooms compromise CR -- even ones that are all black. I've read this in several places as well.Most rooms compromise ANSI CR, but unless they have other lighting they don't compromise on/off CR. By compromising ANSI CR they move the point down where on/off CR matters more than the resulting ANSI CR, but it is backward to say that rooms affect ANSI CR and so ANSI CR from the projectors is more important than on/off CR from projectors. If the room affects ANSI CR, then it makes more sense to claim that ANSI CR from the projector is the one that matters less, but I see this argument about rooms as if it supports that ANSI CR is the more important CR.

Turning lights on hurts both kinds of CRs. It is important that people know that turning lights on or having light coming into the room makes the CRs from the projectors much less important, because now the CRs off the screen are more dominated by the light from the projector for white divided by the black level from the lights on in the room or light coming into the room (assuming that is higher than the black from the projector).
This is what he wrote...

"So unless your room is TOTALLY black and you're watching very
dark videos don't stress too much about contrast ratio!"And here is an example of where I think he and others mislead people. I think most would come away thinking that the walls need to be black when he says, "TOTALLY black", when the truth is that on/off CR can matter greatly in a white room. It is ANSI CR that doesn't matter nearly as much in that room.

I've been in plenty of rooms that would be darker than even the black level from 100k:1 on/off CR if the lights were turned off and it was at night, even though the walls are light colored. I'm not sure when real world became having lights on as some of these people seem to imply. Is a white room without lights on or light coming in that exceeds that level from 100k:1 somehow non-real world? I can understand the argument that black walls aren't real world, but I think the argument that rooms have to have enough light to kill even 10k:1 to be real world, even if they are white rooms, is a ridiculous argument. Getting a room that dark at least under some conditions at night shouldn't be that difficult for most people, unless they have a street light shining right in or something like that. I have light coming around some blackout blinds in my off-white living room from a light outside at night and it still doesn't light up my screen as much as a 100k:1 on/off CR projector would on a blackout screen.
So, he covered dark images, but it seems he doesn't place as much importance on the "darkest" images. He does measure a pretty dark image and still finds the CR shrinks to a fraction of the on/off.I wouldn't call that image with the guy in the white shirt a pretty dark image. It is a long ways from many of the dark images we get in real content. Given the way that gamma works, a window that is 80 IRE and takes up 10 percent of the image with the rest being at black has about the same amount of total light in it as a full screen 30 IRE image.
But, he's not ony talking about the projector -- he's also talking about the room.Yes, and if the room has other lighting then it can kill CR (even his "Real World" CR), but if it just has light colored walls then it kills the ANSI CR, then one which many argue is the more real world one, even though the real world white room kills ANSI CR and doesn't kill on/off CR unless there is other lighting.

A person can play around with this calculator:

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

that estimates things for checkerboards with different projector values and different room reflections values.

When I wanted to find out how these things worked, besides thinking about it, I went and tested. For instance, at a friend's house who had a G70 CRT in an off-white room that we estimated at about 700k:1 on/off CR and I think he was getting around 70:1 ANSI CR off the screen. The higher on/off CR definitely helped in that room, although it probably would have helped more if the ANSI CR hadn't have been so low.
I don't want to be offensive, but I find that JVC owners tend to be extremely sensitive on this topic because they own a projector with a high measured on/off CR. Don't you own own a JVC RS2?It is a fair question and I do own an RS1, but it is the opposite way. I own the RS1 because of what it brings to the table and will likely get an RS2 if I don't find something I will like better. They could definitely be improved, but the on/off CR does bring something and would even in my off-white living room where I have a 10' 8" wide High Power and show things to larger groups than in my dark theater room.
Do you think this could be causing you to want to discount opposing opinions to the extent that you cannot even seem to concede his point about mixed scenes or room reflections?I'm not going to concede things that aren't true and those who have read my posts for a while know that I will trash things I own. I figured the times I get accused of being a DLP shill, an LCOS shill, a CRT shill, etc. are good because they tell me maybe I'm more objective about the whole thing and willing to talk about the stronger points of one technology, whether I own it or not.

I'm sure not going to ignore dark mixed scenes and just go with some scene with a guy in a white shirt. That is about as ignorant as it gets with on/off CR since the darker the scenes get the more on/off tends to matter. It turns out to be convenient to ignore the darkest scenes if trying to put down on/off CR, but it doesn't make a lot of sense. The scenes like the guy in the white shirt matter, but so do a lot of other scenes.

And room reflections without other lights hurt ANSI CR as I said. I explained much of this in my article about CR. The screen matters too though, and with a Firehawk screen in my off-white living room I think I got over 100:1 ANSI CR from a prime viewing position. With a matte white screen it might only be 40:1 and with my High Power I think it is more like 80:1. I don't have one to measure with, but based on the samples I have I think that with a Black Diamond screen I could probably get 200:1 ANSI CR or more off the screen even in that off-white room.

As I believe I said earlier, the author did get some stuff right. He just got a fair amount wrong.
That's a fair point, but having a "black hole" for a theater could also cause sensitivity to the writer's point that even "black hole" theaters compromise CR.Yes they do. ANSI CR. Even if I had a projector with 100k:1 on/off CR my dark theater wouldn't even come close to adding as much light to absolute black as the projector is putting out. I've tested this with ND filters. As I have posted before, I think I am probably able to get around 400:1 ANSI CR in my dark theater with a projector at 500:1 or more, but it gets difficult to measure these high ANSI CR values. I use a black velvet light pipe on a Minolta LS110.

As far as having a "black hole", I have that as well as my living room setup that is off-white, as I explained.
"I haven't yet seen a Home Theater room that didn't decrease the actual contrast ratio due to reflected light. Black walls, carpets and all."Yep. The ANSI CR. A person can use that calculator to try to see how this affects other images along the scale.
I don't know. Have you read this thread?

http://www.avsforum.com/avs-vb/showthread.php?t=1035229

In this and other threads, I have found that the debate falls along pretty predictable lines with JVC owners usually minimizing the importance of ANSI (an area where their projector does not enjoy superiority in measured numbers) and maximizing the importance of on/off (the area where their projectors do enjoy superiority in measured numbers).While it is a fair question, by your logic that thread should support that I would be arguing the DLP side, since the DLP in the comparison was mine (the 20k) and the JVC was Kris Deering's (he has the RS2, while I have an RS1 which wasn't part of this comparison). I sold the 20k not because I own a JVC also, but because after that comparison I decided I would rather have an RS2 (or something else maybe after CEDIA if a new projector comes out from whoever that I will want more). Cause and effect. I've lost track of how many DLPs I've owned over the years. In that thread I reported on what I saw. Some of it surprised me as I thought we would see more advantage with the 20k in certain images than we saw, although I did mention some of the limitations of those testing conditions. I sure didn't go in hoping to build up the projector Kris owned and put down the one that was mine, but to see what we saw.

People who do lots of comparing (as Kris and I do) tend to end up owning what we prefer, so while some may think that we must prefer what we do because of what we own, it is way more the other way around.
Ultimately, though -- I still think the writer has an interesting point -- and that is, "how much of what is *measured* is *perceivable* in an actual home theater during regular viewing?

In audio, we know that we can measure differences that cannot be heard, that are outside the range of audibility, etc. We also know that, even though testing shows that measured differences are below the threshold of audibility, some will still claim to have "golden ears" and will claim to be able to hear them.

I suspect the same is true in video.With on/off CR it is pretty easy to show that people can see beyond what the best digital projectors we are discussing can do. It is more like audio differences where just about anybody could pick them out if the test was actually done. One of the problems here is that we are on the internet. If we were in a room trying different things it would be much more clear. Also, some seem to want to test small differences in on/off CR and say that people can't see a difference as if that shows that big differences wouldn't be visible, when that logic isn't correct. I know over the last few years I've had people tell me that humans couldn't see anything beyond about 4k:1 on/off CR and there have been plenty who have argued that on/off CR was irrelevant even when digitals were 2k:1 or less. But yet the projectors keep getting better in this area and the differences keep being visible. How much a person values it is up to them, but whether big on/off CR improvements are visible or not pretty much isn't up for reasonable debate IMO and fortunately I think there are less people arguing that they weren't even visible than in the past. Some may want to use a single image with some bright stuff to try to show that some on/off CR improvements wouldn't be visible, but even if they weren't visible in that one image with bright stuff, it wouldn't show that they aren't in many other images.

Sorry that this is so long already, but going back to that checkerboard calculator here (http://home1.gte.net/res18h39/contrast.htm), a person could try something like 2k:1 on/off CR with 500:1 ANSI CR and see what they get under different conditions. For instance, try a room gain of .001 and the ANSI CR off the screen goes to about 400:1, while the CR for a 10/0 IRE checkerboard goes to about 15:1. Now change the room gain to .016 to get the ANSI CR down to about 100:1 and the 10/0 IRE checkerboard goes to about 13:1. Now leave those huge room reflections and change the on/off CR to 30k:1. The ANSI CR stays at about 100:1 (I could have increase the ANSI CR for the projector a little for the lower black floor, but didn't) and the 10/0 IRE checkboard goes to about 57:1. Quite a ways from 13:1. And as I explained earlier, according to that NASA chart, the 10 IRE is brighter than white paper in moonlight.

--Darin

it comes down to them thinking that ANSI CR is a mixed image and on/off CR isn't

That misunderstanding probably stems from the nature of the respective tests.

If they think that on/off CR and ANSI CR measure the same thing then they are confused.

Maybe I missed it, but I didn't see that in any of the articles posted.

If they think that one affects mixed images and the other doesn't, then they are confused.

Again, I didn't read it that way -- I read him as saying that on/off CR shrinks when a scene becomes mixed, whether it is mixed a little or a lot. I don't think you would argue with that.

If somebody wants to use real world images, then they should use more than one, because image content varies and so what is needed varies.

I would love to see a more complete study, but -- just my take -- I think the guy only intended to print an example of what he was talking about, not a study. So, I didn't see him as trying to load the deck.

There is an absolute black floor and there is a washout effect.


Maybe this is why I perceive you to be leaning towards the JVC Owners position; you seem a lot more concerned about black floor and pretty unconcerned about wash-out.

There are a lot of mixed scenes that are way darker than their example and so much more dependent on on/off CR than there one example.

Not that I don't trust you, but this is pretty anecdotal. I'd love to see some actual figures, some way to quantify.

Most rooms compromise ANSI CR, but unless they have other lighting they don't compromise on/off CR.

That's interesting. I'd like to read more about that -- can you provide a link or source?

Turning lights on hurts both kinds of CRs.

I think Art noted that, too.

I figured the times I get accused of being a DLP shill, an LCOS shill, a CRT shill, etc. are good because they tell me maybe I'm more objective about the whole thing and willing to talk about the stronger points of one technology, whether I own it or not.

If you're getting attacked by all sides, you're probably on the right track. No good deed goes unpunished including, perhaps, objectivity. :-)

I'm sure not going to ignore dark mixed scenes and just go with some scene with a guy in a white shirt. That is about as ignorant as it gets with on/off CR since the darker the scenes get the more on/off tends to matter.

Oh, I'm sure that, with a little effort, it could get even more ignorant. I think you've conceded the points to which I referred, so -- as the kids say -- it's all good.

It turns out to be convenient to ignore the darkest scenes if trying to put down on/off CR, but it doesn't make a lot of sense. The scenes like the guy in the white shirt matter, but so do a lot of other scenes.

Just as convenient to focus on dark scenes, though.

I sold the 20k not because I own a JVC also, but because after that comparison I decided I would rather have an RS2 (or something else maybe after CEDIA if a new projector comes out from whoever that I will want more). Cause and effect. I've lost track of how many DLPs I've owned over the years. In that thread I reported on what I saw.

Fair enough. I suspected you would give a reasoned answer and you didn't disappoint -- Thanks!

Again, I didn't read it that way -- I read him as saying that on/off CR shrinks when a scene becomes mixed, whether it is mixed a little or a lot. I don't think you would argue with that.



Darin and I have looked a lots of mixed stuff here when I had my G90 stack. On/Off has a substantial even dramatic effect in mixed material but generally low APL mixed material to a much greater extent. To say that on/off shrinks in mixed material is a gross over simplification IMO.

Art

To say that on/off shrinks in mixed material is a gross over simplification IMO.

I prefer to think of it as stating the obvious. It just means that whatever the on/off number, when you mix light and dark, that number is going to go down. It is unlikely to stay the same or increase. To me, it should be an easily conceded point -- then -- context can be added, arguments fleshed out, etc.

Maybe I missed it, but I didn't see that in any of the articles posted.My comment was mostly along the lines of their, "Also called "Checkerboard" it's a more realistic way to test a projector than On/Off but not favored because of marketing concerns." They measure different things, so how is the checkerboard contrast ratio more realistic? I think this may go back to that confusion about ANSI CR is a mixed image and we mostly watch mixed images, so therefore ANSI CR must be the more realistic way to test a projector. While that may look like it makes sense to some people, it is just wrong.
Again, I didn't read it that way -- I read him as saying that on/off CR shrinks when a scene becomes mixed, whether it is mixed a little or a lot. I don't think you would argue with that.Mixed images tend to end up with only a portion of the on/off CR, which is one of the reasons that on/off CR needs to be so big. So that when you use only a portion, you still have good CR in the mixed image. I'm sure it isn't the best analogy, but if somebody told you that you could have 1/8th of a pizza and you wanted a big piece, you should want the pizza to be big. That you only get a portion of the pizza isn't a good argument that the pizza shouldn't be big.

Since images are affected by both the black floor and the washout effect, yes the washout effect will tend to reduce the CR in the mixed images. I gave an example of how huge washout effect could still leave on/off CR as the dominant factor in many images and that was only considering full screen checkerboards in that example, not where a smaller portion of the screen is non-black (so even less washout, but still the same absolute black floor). One reason you want a good absolute black floor is so that when washout comes into play the absolute black floor isn't causing things to be worse. And vis versa, you want the washout effect to be small both so that when it is the dominant factor the resulting CR is good and so it doesn't cause too much of a problem when the on/off CR should be the dominant factor.
I would love to see a more complete study, but -- just my take -- I think the guy only intended to print an example of what he was talking about, not a study. So, I didn't see him as trying to load the deck.Their stuff here:

http://www.homecinemacentral.com.au/contrast_review.html

looks to me like they completely replaced the ANSI CR and on/off CR measurements with that one image they call "Real World". BTW: I hope that image doesn't actually have the markers for where to measure on it, but it wouldn't surprise me if it did. That would be a pretty dumb thing to do given that the markers themselves for where to measure the black would raise that black level there on a lot of projectors.

And serious question for you. Do you see any advantages to their one image over the ANSI CR test? Looks to me like the advantages are that it has a person in it, so they can call it "Real World" and it is easier to measure. The combination of the on/off CR test and the ANSI CR test gives a lot more useful information than their one test they have listed there. If their goal is to measure white and black at the same time, the ANSI CR test already does that. And their excuse about what manufacturers can do doesn't make much sense to me. They are the ones doing the review. They could do just like Greg Rogers and just make sure the on/off CR test is done correctly.
Maybe this is why I perceive you to be leaning towards the JVC Owners position; you seem a lot more concerned about black floor and pretty unconcerned about wash-out.I have said many times that both matter. However, after lots of testing I would say that at current levels improvements to on/off CR matter a lot more to me. It has been over 2 years since I wrote the "Future Improvements" section in my article here:

http://www.hometheaterhifi.com/volume_13_2/feature-article-contrast-ratio-5-2006-part-5.html

and I think time is supporting what I said.

I can't think of a current model of HT projector that is under 200:1 ANSI CR. There might be some, but I can't think of any. At the 2000:1 the author there said was fine for light controlled rooms, I would say that ANSI CR would probably have to be well under 100:1 for me to consider it as weak as far as its perceived effects as 2000:1 on/off CR is. I said maybe a month or two ago that I didn't know what I would choose between 60k:1 on/off CR and 300:1 ANSI CR or 30k:1 on/off CR and 600:1 ANSI CR, but after recent testing I'm leaning toward thinking I would take the 60k:1/300:1 if all else was equal. That isn't for love of any manufacturer or projection technology and that is even with a black room that can retain a lot more of that 600:1 ANSI CR than a lot of rooms can.
Not that I don't trust you, but this is pretty anecdotal. I'd love to see some actual figures, some way to quantify.I don't see where they provided the actual values for their image, but you could look at it and just grab some dark movies and look at those, but the first post in this thread:

http://www.avsforum.com/avs-vb/showthread.php?t=1046712

shows an attempt to map that average luminance for Raiders of the Lost Ark. I'm sure more movies will get mapped over time. I'm not saying the average frame is darker than their one test image, but there are a lot that are.
That's interesting. I'd like to read more about that -- can you provide a link or source?No source that isn't my own article. But it is just simple math. Pick any reasonable reflections and washout back to the screen amount you want. Since on/off CR is just a white screen divided by a black screen the white doesn't wash the black out like it does in the ANSI CR test. The simple math is that with on/off CR being x/y, if there were even 10% reflection effect from the room back to the screen, the new math would be 1.1x/1.1y, but 1.1x/1.1y is the same as x/y and so the ratio stays the same. I've mentioned that there might be very minor color effects only because of situations like the white and black images having different color balances and the room being a non-neutral color, like if video black is very bluish and the room is blue, but that is a minor factor with correctly calibrated and designed projectors.
Just as convenient to focus on dark scenes, though.I focus on where I think the most improvement can be reasonably had (and of course that I care about), but that isn't the same as using one image that is far from an endpoint, or far from both endpoints. I think measuring both on/off CR and ANSI CR makes sense because they both matter and give extreme cases that help determine things that are toward them as well as things closer to average. But the human eye has a lot more capability to benefit from high CR over time than at one instance and after much testing, I believe that with the projectors that have things like 300:1 ANSI CR or more, it is still on/off CR where I expect that most people could see the most improvement. Much like the music thing you mentioned, except wanting them to concentrate more on the thing that more people can perceive. From where they are now dark scenes are where I think digitals can show the most improvement as far as CR issues go. For bright scenes I think there are much bigger fish to fry (like getting some blank time to hopefully help motion perception).

As I think somebody basically already mentioned, the mighty G90 CRT was only something like 130:1 ANSI CR. So, unless there are some models I'm missing, we are talking about projectors that kill the G90 for ANSI CR, if we are talking current models. That doesn't mean that ANSI CR doesn't matter, just that along the scale of good to bad, the digitals are mostly already in the upper range there. Set up that G90 properly with a good circuit some of the CRT guys use for high on/off CR and put it in either a white colored room or a black room against a digital with the 2000:1 that the author said was fine for a light controlled room and regardless of the ANSI CR of that projector (even 500:1 or more), I have little doubt that the vast majority would pick the G90 as doing the better job overall for black performance after watching lots of different content on each. Of course the above is without turning lights on or having light streaming into the room. Do that and everything changes, as discussed earlier.

When on/off CR gets high enough I may start putting more importance on ANSI CR even from current levels for things like wanting video black backgrounds to look just as dark as the black velvet border on my screen. I would take more ANSI CR, but when forced to make a choice I will try test and decide which I think brings the most to the table.

--Darin

Mixed images tend to end up with only a portion of the on/off CR, which is one of the reasons that on/off CR needs to be so big. So that when you use only a portion, you still have good CR in the mixed image.[

But, the projector with the highest on/off CR is not the projector with the highest ANSI (mixed image) CR.

And serious question for you. Do you see any advantages to their one image over the ANSI CR test?

No, but I didn't read it as a substitute, I saw it as just one example to illustrate the point.

The combination of the on/off CR test and the ANSI CR test gives a lot more useful information than their one test they have listed there.

Agree.

From where they are now dark scenes are where I think digitals can show the most improvement as far as CR issues go.

I getcha -- This explains your position very well -- in this context, I understand your posts much better -- thanks!

But, the projector with the highest on/off CR is not the projector with the highest ANSI (mixed image) CR.Maybe this is getting nitpicky, but if you mean the projector with the highest ANSI CR then I agree currently and that they don't have to follow each other. But if you mean the projector with the highest mixed image CR, then I would say that from the current digitals they probably are the same model right now as long as we score them based on their best intra-frame CR for any mixed images. The model I think would take the crown is the RS2. As long as you include real images like the 2nd one down here:

http://www.avsforum.com/avs-vb/showthread.php?p=14256608#post14256608

and that includes some video black. To be clear, I'm talking about the image at night with very small white lights and a mostly dark image. Those lights are small enough and the other stuff dark enough that if the lower right corner has any stuff encoded as video black, I would put my money on the RS2 to have the highest CR between the brightest white light(s) and that dark area over things like the Sharp 20k and Marantz 11S2. And I'm sure there are images that would favor the higher on/off CR projector more than that one (I know there are if we are allowed to make test patterns as I would make one with just one or a few white pixels and the rest black, kind of like one frame that shows up during Fight Club, but might have been an accident with a hole in a film frame that was meant to be black).

In general, when one projector has the higher on/off CR and another the higher ANSI CR there are scenes/images where each wins the intra-image CR crown. That is why I've said in the past that I didn't agree with saying that the projector with the higher ANSI CR had the better intra-scene CR. Since it depends on the specific scene and there are basically billions of intra-scene CRs (if we don't limit things to integers).

--Darin

Maybe this is getting nitpicky, but if you mean the projector with the highest ANSI CR then I agree currently and that they don't have to follow each other.

You wrote,

"Mixed images tend to end up with only a portion of the on/off CR, which is one of the reasons that on/off CR needs to be so big. So that when you use only a portion, you still have good CR in the mixed image."

Yes, I mean that the RS2 starts out with the highest on/off CR, but does not retain more CR when subjected to the ANSI test. Other projectors that start out with lower on/off CR retain more CR when subjected to the ANSI test.

The model I think would take the crown is the RS2. As long as you include real images like the 2nd one down here...

Aren't you guilty of providing only one example -- just like the guy you criticized?

And I'm sure there are images that would favor the higher on/off CR projector more than that one (I know there are if we are allowed to make test patterns as I would make one with just one or a few white pixels and the rest black

And that's a real world test of the average scene that we might see on a regular basis while watching the average film?

kind of like one frame that shows up during Fight Club, but might have been an accident with a hole in a film frame that was meant to be black).

Wow -- that's what I would call real stretching. Like the one frame that might have been an accident?

It seems to me like you are going to quite an extreme in trying to make the JVC not only the on/off champ but the intra-scene champ as well. If the guy whose article you criticized was trying to load the deck in favor of his argument, I think it is only fair to charge you with your own criticism.

We have constructed a black tunnel for the purpose of taking contrast measurements. However, even the black velveteen fabric we used to line the tunnel is reflective to some degree. Our tunnel therefore replicates a very dark, but not perfectly black and nonreflective viewing space. Accordingly, our measurements will always be lower than the theoretical maximums quoted in vendor specifications. They are more likely to approximate what you'd actually achieve in fully darkened dedicated theater room.

We measured the RS2's Full On/Off at 13,300:1, which is quite a bit lower than the official spec. Nevertheless, this is the highest native contrast reading we've measured from the projectors we've tested thus far. On the other hand, the RS2 measured 209:1 ANSI contrast, which is the lowest ANSI reading we have found among the high performance 1080p models we've tested.

http://209.85.141.104/search?q=cache:Bt8yWG2fdGsJ:www.projectorcentral.com/contrast_ratios.htm+reflected+light,+ANSI&hl=en&ct=clnk&cd=6&gl=us

You wrote,

"Mixed images tend to end up with only a portion of the on/off CR, which is one of the reasons that on/off CR needs to be so big. So that when you use only a portion, you still have good CR in the mixed image."

Yes, I mean that the RS2 starts out with the highest on/off CR, but does not retain more CR when subjected to the ANSI test. Other projectors that start out with lower on/off CR retain more CR when subjected to the ANSI test.That does apply to the ANSI CR test, which is a very bright test compared to average images in movies. My point in what you quoted was basically that when an image only contains something like 0 to 20 IRE, which looks like 1/5th of the whole range, in ft-lamberts or lumens out of the projector it is only using basically 2-3% of the on/off CR range before the washout effect is included. In other words, the bottom 20% of the 220 different values from video black (16) to reference white (235) only encompasses 2-3% of the amount of light. So, on/off CR needs to be very large for 2-3% of it to be large enough. And it gets worse with the bottom 10% only encompassing about .6% of the amount of light in 100 IRE things. And since the bottom 20% or 10% of encodings mean so little light, the washout effect has to be pretty large before it limits those kinds of images more than on/off CR does, unless on/off CR is very large. That effect can be seen with this calculator:

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

You are right that the current champion for on/off CR is different than the champion for ANSI CR. So, there is a crossover point between those 2 tests where each projector takes over as far as simultaneous CR, compared to the other projector. For projectors without dynamic irises the one with the higher on/off CR is probably going to be the one that is capable of the highest mixed image CR overall, but that wouldn't come in a very bright image (compared to most film) like the ANSI CR image, it would come in something like a small amount of white on black (a low APL, but with some 100 IRE pixels). The RS2 should be able to do multiple thousands to one with a small amount of white on black. For instance, with 1080p, if the 100/0 IRE checkerboard was 192 pixels by 108 pixels and the rest was video black, that calculator estimates that the RS2 could do about 8500:1 for that image if it has 30k:1 on/off CR and 300:1 ANSI CR, with a room gain of 0.01 (a light enough room to drop the off the screen ANSI CR all the way from 300:1 to 120:1).
Aren't you guilty of providing only one example -- just like the guy you criticized?I don't think so. I didn't propose replacing those tests or say that was more real world. Their stuff about how they were going to do their reviews sure looks like that was the only CR they were going to report and the proof seems to be in their AX100 review here:

http://www.homecinemacentral.com.au/panasonicax100.htm

where they listed the CR for that exact test they called "Real World" with no report of on/off CR or ANSI CR.
And that's a real world test of the average scene that we might see on a regular basis while watching the average film?Of course it isn't. Nobody ever said it was an average scene and the ANSI CR test is a long ways from an average scene, so given that you used the ANSI CR test for your argument it would seem that you are guilty of your own criticism if that is the way you want to take it. You said:
But, the projector with the highest on/off CR is not the projector with the highest ANSI (mixed image) CR.You put the mixed image in there and the ANSI CR test is a fairly extreme example of a mixed image (way brighter than average mixed images). If you meant that it is "not the projector with the highest mixed image CR" then the one that can achieve the highest mixed image CR seems like the one that can achieve the highest mixed image CR to me and if a person means the CRs over multiple mixed images, then it would be CRs. If you mean a specific mixed image, then that is a different case and if a person specifically means an average mixed image, then they can call for an average mixed image. That is what I was pointing out. If you look at the cross section of all the mixed images out there and go with the one that can achieve the highest value overall, the RS2 is going to beat the Sharp 20k or Marantz 11S2 for that. Do you disagree with that? Obviously, the 20k and 11S2 win for the ANSI CR test, which is a specific example of a mixed image CR with an image that is very bright compared to average film frames.

Why did you put the "mixed image" in your statement? The ANSI CR test isn't an example of anything close to an average mixed scene that we might see on a regular basis while watching the average film and yet that was your argument against the example I used of where the highest mixed image CR would tend to happen (and where it should have been obvious that it wasn't meant as an average mixed scene since it is the one for the highest).
It seems to me like you are going to quite an extreme in trying to make the JVC not only the on/off champ but the intra-scene champ as well. If the guy whose article you criticized was trying to load the deck in favor of his argument, I think it is only fair to charge you with your own criticism.You can charge whatever you want, but if you look at all the intra-scene CRs out there and look at the best any of the projectors can do across those, then from those the RS2 is going to be the champ. If you pick a specific scene, then it depends on which scene you pick. As I pretty clearly stated, there will be a crossover point (or more than one crossover point depending on how you look at the data). As I'm pretty sure I made clear long ago (including in my article about CR), with real world images the intra-image CRs will vary over a large range. This is one reason just using one scene in the middle makes so little sense. And please don't accuse me of using just one scene like that guy (or those guys) did in their reviews. I have said repeatedly that it makes sense to give both the on/off CR and ANSI CR as what they indicate is far apart and gives a good idea about both the extremes and the midpoints, at least as well as I think can be expected if just using 2 tests.

--Darin

We have constructed a black tunnel for the purpose of taking contrast measurements. However, even the black velveteen fabric we used to line the tunnel is reflective to some degree. Our tunnel therefore replicates a very dark, but not perfectly black and nonreflective viewing space. Accordingly, our measurements will always be lower than the theoretical maximums quoted in vendor specifications. They are more likely to approximate what you'd actually achieve in fully darkened dedicated theater room.

We measured the RS2's Full On/Off at 13,300:1, which is quite a bit lower than the official spec. Nevertheless, this is the highest native contrast reading we've measured from the projectors we've tested thus far. On the other hand, the RS2 measured 209:1 ANSI contrast, which is the lowest ANSI reading we have found among the high performance 1080p models we've tested.

http://209.85.141.104/search?q=cache:Bt8yWG2fdGsJ:www.projectorcentral.com/contrast_ratios.htm+reflected+light,+ANSI&hl=en&ct=clnk&cd=6&gl=usI'm pretty sure I'm the one who suggested to them that they make a light pipe because I saw that they were going to start doing CR stuff and was concerned that they would do things incorrectly. That isn't meant as anything personal against them, but just prior experience. I would tend to trust Greg Rogers more for measuring ANSI CR, but maybe they got a unit with lower CRs across the board. The on/off doesn't surprise me as it depends on the zoom used (Greg pointed that out) and can probably vary from unit to unit. The ANSI CR does seem to be lower than what most people have measured though. Greg measured 310:1 for his modified ANSI CR and that is usually lower than the real ANSI CR (since it measures the rectangles near the center that usually have more washout).

Using their numbers in the contrast calculator with the example checkerboard of 192x108 and against 8000:1 on/off CR with 800:1 ANSI CR with a room gain of 0.01, the calculator estimates that the 13,300:1/209:1 projector would have slightly higher intra-image CR there at 5760:1 to 5330:1. That is probably quite a bit lower than the margin of error for a case like that as it is just estimating and placement of things comes into play quite a bit with something like that.

And as far as their claim about the black tunnel, that doesn't apply to on/off CR unless there is other lighting for the very simple reason I laid out. The math is extremely easy, but yet it seems that most people think reflections hurt on/off CR for some reason.

--Darin

If their black velvet lined tunnel isn't real world, I don't know what is. They admit their numbers might be lower than in a room that is ideal -- the fact that they get lower numbers in a black velvet lined room supports the author previously cited who claimed that even the average all black theater compromises CR.

JVC DLA-HD100

Contrast Ratio:
Full-On/Full-Off —12,790:1
ANSI—227:1

http://www.hometheatermag.com/frontprojectors/408jvcdila/index2.html

And as far as their claim about the black tunnel, that doesn't apply to on/off CR

They use a black velvet tunnel and they claim the RS2 has "the lowest ANSI reading [they] have found among the high performance 1080p models [they've] tested.

If their black velvet lined tunnel isn't real world, I don't know what is. They admit their numbers might be lower than in a room that is ideal -- the fact that they get lower numbers in a black velvet lined room supports the author previously cited who claimed that even the average all black theater compromises CR.They "admit" to something that is BS as far as on/off CR, unless they have other lighting. The math is very easy. Did you understand it? Not matter what they believe, their light pipe having some reflections does not mean they will get lower measurements for on/off CR unless they have other lighting (which I sure hope they wouldn't if they are trying to do responsible reviews).
Greg has addressed this topic in other threads. He takes a different tact for measuring ANSI -- he does not follow the usual protocol. He calls his M-ANSI, or Modified-ANSI.His method will tend to get lower numbers because he measures the center rectangles and they tend to have the lowest CRs (because the dark rectangles there are surrounded by white rectangles, where the dark rectangles at the edges of the screen are not).

BTW: Back to this:
No, but I didn't read it as a substitute, I saw it as just one example to illustrate the point.I can see how you thought it wasn't meant as a substitute, but I think the stuff on their site just about the test and then the AX100 review only using that test shows that the guy you are defending did use that one image as a substitute. Do you disagree that this is what he did?

As far as the RS2, what is your point? If it is that average RS2s are getting less than what Greg's got, then I could believe that. But as I showed, even using projectorcentral's lower numbers, that calculator estimated that the RS2 would end up with higher simultaneous CR in that image with a 192x108 checkerboard of 100 IRE and 0 IRE.
JVC DLA-HD100

Contrast Ratio:
Full-On/Full-Off —12,790:1
ANSI—227:1

http://www.hometheatermag.com/frontprojectors/408jvcdila/index2.htmlThey say:
After calibration, and using a full-field 100-IRE white (12.79 foot-lamberts) and a full-field 0-IRE black (0.001 ft-L), the contrast ratio was 12,790:1.If they measured 0.001 ft-L off the screen with Minolta LS100 then the margin of error in that 12,790:1 is huge. According to Minolta, the margin of error is basically bigger than the CR. .001 is as low as that meter goes and according the specs for the meter, it is accurate to plus or minus 2%, plus or minus another 2 points in the lowest digit.

Just looking at the:
The black level in the normal lamp mode was also 0.001, although the overall light output was lower.in the same paragraph should point out the problem. Different lamp modes shouldn't really have different CRs unless color balance is different (and even then only significant if the projector is recalibrated). But they were at the limits of the meter where the meter couldn't differentiate between 0.0006 ft-L and 0.0014 ft-L (the difference between 9000:1 and 21000:1) because it could only display 3 digits to the right of the zero. I think that is the magazine I am trying to convince to buy a meter that I think would likely be more accurate for both on/off CR and ANSI CR measurements for them (a Minolta T-10).

--Darin

They use a black velvet tunnel and they claim the RS2 has "the lowest ANSI reading [they] have found among the high performance 1080p models [they've] tested.They also said:
We have constructed a black tunnel for the purpose of taking contrast measurements. However, even the black velveteen fabric we used to line the tunnel is reflective to some degree. Our tunnel therefore replicates a very dark, but not perfectly black and nonreflective viewing space. Accordingly, our measurements will always be lower than the theoretical maximums quoted in vendor specifications. They are more likely to approximate what you'd actually achieve in fully darkened dedicated theater room.That doesn't say just their ANSI CR will be lower, it says "our measurements". They may just not understand something as simple as that without other lighting, their on/off CR measurements should not be lower for the reason they gave (reflections).

They got lower numbers for both on/off CR and ANSI CR than Greg. The light pipe doesn't explain both of those unless they have other lighting.

And if you were referring to only the ANSI CR here:
They admit their numbers might be lower than in a room that is ideal -- the fact that they get lower numbers in a black velvet lined room supports the author previously cited who claimed that even the average all black theater compromises CR.why did you say "numbers"? Was it your claim that their lower on/off CR they reported supported the author who claimed that even the average all black theater compromised CR? We've been over this, it compromises ANSI CR. The guy doesn't need any support for that as it should be obvious. Anybody who thinks it compromises on/off CR without other lighting doesn't understand the subject matter.

--Darin

BTW rsbeck: I'm curious. I'm pretty sure I've seen you make multiple accusations against JVC owners, unless my memory is just off. Maybe I've just missed it, but have you made the same kind of accusations here against owners of other projectors that they are probably just saying they value some thing because they own that projector. Like have you posted about DLP owners saying that ANSI CR matters because they own DLPs? Is that what you do? Are you looking for ways to support this author/site that did use only what they call their "Real World" image and didn't report on on/off CR or ANSI CR for the AX100 review, because you own a DLP? :)

It seems that you don't agree that they should have just substituted that one test that was more in the middle for the extreme tests that are on/off CR and ANSI CR, but this seems like what they did and it also seems like you are looking for ways to support that author. I'm wondering if you are going to continue to look for ways to support that author even after this site did that and if so, why you would do that.

--Darin

You are right that the current champion for on/off CR is different than the champion for ANSI CR. So, there is a crossover point between those 2 tests where each projector takes over as far as simultaneous CR, compared to the other projector.

Okay.....

I didn't propose replacing those tests or say that was more real world.

Fair enough.

You put the mixed image in there and the ANSI CR test is a fairly extreme example of a mixed image (way brighter than average mixed images).

It is an example of a mixed image and the ANSI tests show that the RS2 does not fare well in this mixed image test, so this need to be accounted for in your argument -- and you have accounted for it (see above).

If you look at the cross section of all the mixed images out there and go with the one that can achieve the highest value overall, the RS2 is going to beat the Sharp 20k or Marantz 11S2 for that. Do you disagree with that?

I would need to see the data. If the data supports it, I go with the data. However, this would still only mean that there are measured differences -- you'd need to go on to prove these are perceivable differences in real world home theaters while watching the average film, etc. etc.

Obviously, the 20k and 11S2 win for the ANSI CR test, which is a specific example of a mixed image CR with an image that is very bright compared to average film frames.

You can add other projectors, too. This also means that many people who've seen the RS2 still purchase other projectors, so it seems clear that on/off may be the area that you find most important, but others pass on the projector with the highest on/off and seem to prefer others, so this is not universal.

but if you look at all the intra-scene CRs out there and look at the best any of the projectors can do across those, then from those the RS2 is going to be the champ.

This is strange wording, IMO. You've already conceded there is a crossover point where other projectors are superior, so if the RS2 is the champ, it is the champ in some scenes, but not in others. Why call it "the champ?"

If you pick a specific scene, then it depends on which scene you pick.

I don't know. Maybe it's me, but you seem to be going back and forth here.

And please don't accuse me of using just one scene like that guy (or those guys) did in their reviews. I have said repeatedly that it makes sense to give both the on/off CR and ANSI CR as what they indicate is far apart and gives a good idea about both the extremes and the midpoints, at least as well as I think can be expected if just using 2 tests.

Yes you have. Fair enough. I'll take that one back.

BTW rsbeck: I'm curious. I'm pretty sure I've seen you make multiple accusations against JVC owners

I refer to JVC owners earlier in this thread and in the ANSI thread.

Like have you posted about DLP owners saying that ANSI CR matters because they own DLPs? Is that what you do?

Feel free to scan my posting record. If you see such a pattern, feel free to post an example.

Are you looking for ways to support this author/site that did use only what they call their "Real World" image and didn't report on on/off CR or ANSI CR for the AX100 review, because you own a DLP? :)

Hmmm.......I guess anything is possible, but if you notice, this isn't how the thread started, it seems that this was the issue that "stuck." But, I am sure that I see things through my personal prism just like anyone else.

It seems that you don't agree that they should have just substituted that one test that was more in the middle for the extreme tests that are on/off CR and ANSI CR

The point I took from the article was that when subjected to the ANSI test, on/off CR's shrink and that real world home theaters compromise contrast ratio. Since then, we've become more specific. You've stressed that it is ANSI CR that is compromised by room reflections, while On/off is compromised by ambient light. If I have paraphrased you badly, feel free to correct me.

it also seems like you are looking for ways to support that author.

Or, at least a few of the things he wrote.

I'm wondering if you are going to continue to look for ways to support that author even after this site did that and if so, why you would do that.

There is a logical fallacy in this notion, but I can't recall the latin term for it.

If I say 2+2=4 and the moon is made of cheese, 2+2 still equals four, the cheesy moon statement doesn't invalidate everything else. Since you already conceded that the author got some things right, I think we should just leave it at that and not go in this direction.

That doesn't say just their ANSI CR will be lower, it says "our measurements". They may just not understand something as simple as that without other lighting, their on/off CR measurements should not be lower for the reason they gave (reflections).

Or, it might just mean that they intend to measure other projectors in the future, so they are simply adding an "s" to indicate they will be getting more than one measurement, thus the plural.

I would need to see the data. If the data supports it, I go with the data. However, this would still only mean that there are measured differences -- you'd need to go on to prove these are perceivable differences in real world home theaters while watching the average film, etc. etc.

I suggest you watch films yourself in different conditions with different projectors and you will see what the differences are. If you have I don't understand this extended discussion. The visibility and relevance of (very) high On-Off contrast is easily demonstrated with suitable rooms and sources.

From reading this back and forth I think part of the confusion arises from a conflation of contrast measures with the various physical phenomena that affect them. It really isn't accurate to say that "on/off CR affects an image", for instance; it's the phenomena whose effects that on/off CR measures that impact an image. Understanding what those phenomena are can make it more clear and intuitive why those two measurements are useful, and what their limitations are in judging content. And it can also make it easy to understand what happens in real video images.

For instance: on/off CR performance is impacted by:
--- any ambient light sources in the room; e.g., LEDs on your A/V equipment
--- the projector's native "black floor", as determined by fixed irises, polarizers, chip-level contrast, etc.
--- dynamic iris behavior, which can lower the projector's native black floor depending on the content of the image.
--- projector brightness and screen gain, but only if there are ambient light sources in the room

ANSI CR performance, on the other hand, is determined by:
--- light scattering within the projector's optical path
--- secondary reflections from the room, which is determined by the material and colors present on walls, furniture, etc.
--- screen gain and directivity

This list isn't necessarily exhaustive, and it's probably imprecise. But one thing to note is that there really isn't much overlap in these phenomena, at least not to first order. Wall color, for instance, doesn't impact on/off CR. And small amounts of ambient light don't impact ANSI CR, though sufficiently large amounts certainly could. So by focusing on only one of these measures you're going to miss out on understanding your projector's importance.

So now it shouldn't be hard to see how all of these factors combine to determine the achievable contrast for a fixed video image. Let's assume no dynamic iris for the moment. Then ambient light and the projector's native black floor fix an absolute minimum black level, which are independent of the content. Added to this are the projector and room scattering effects, whose impacts are approximately proportional to the average brightness of the image. So at some point, when the video images is dark enough, these proportional effects can get buried beneath the absolute black floor, and that is a situation where on/off CR is said to "dominate" the contrast performance.

Where is that cutoff, where the absolute black level and the proportional scattering effects are approximately equal? The on/off and ANSI measures combine to give us an idea of this. The higher the on/off CR, the darker this cutoff level is, of course. But what may not be immediately obvious is that the higher the ANSI CR, the brighter this cutoff level is. That's right: the better a projector's ANSI CR performance, the sooner its contrast will be degraded as the image gets darker.

Finally, note that these are all numbers. There probably is a point of diminishing returns for ANSI CR. Maybe that number is, say, 200:1, I really do not have any idea. But if that's the case, there's no sense in worrying if a projector with 800:1 ANSI CR suffers this degradation effect until the image contrast drops down below 200:1.

The visibility and relevance of (very) high On-Off contrast is easily demonstrated with suitable rooms and sources.

I don't think this is in dispute. Although this is not very specific since there are no numbers attached, in general, I see no problem saying that if a room and projector were capable of very high on/off, I have no doubt it would be visible versus a room and projector capable of only very low on/off.

IMO, questions still remain about what is actually achieved in the average home theater, what is seen in the average film, what level of on/off, ANSI, and lumens it takes to produce a Just Noticeable Difference.

As I mentioned early in the thread, I watched an RS2 on a ST130 in a room with medium earth tone colored walls, ceiling and carpet and I also watched a C3X1080 in an all-white room at OB's house, with a fair amount of ambient light leaking in through the windows on a Firehawk screen.

I thought the C3X1080 had more convincing blacks and the better picture.

Note -- I didn't own a C3X1080 at the time, I made the decision to purchase the C3X1080 based on auditioning the two projectors.

The RS2's measured on/off is around 30,000:1 according to Greg Rogers and the C3X1080's measured on/off is around 7,000:1 according to Coldmachine.

I don't know. Maybe it's me, but you seem to be going back and forth here.I don't think I am at all. Maybe what I think is a simple example will help. If somebody asked me which person has the longest javelin throw, I would look for the person who has thrown the furthest. If they asked for the average I would look for something different. And the person who has the longest javelin throw overall is not necessarily the one who has the longest average or the longest throw at one specific meet or even the average meet. If you were trying to figure out who had the longest javelin throw and somebody just asked that in general (without reference to average or a specific meet), would you look for the average or an average meet, or would you just look for the person who's best is the longest against other's bests? I think the Guinness Book of World's Records would go with my approach. That is, if one ranged from 3:1 to 9000:1 for its mixed image CRs and another ranged from 5:1 to 8000:1 for its mixed image CRs, the first one would have the highest mixed image CR, even if its average was lower or there were specific mixed images where it was lower.

But it is semantics and as I also said, it depended on what you meant with the addition of "(mixed image)".

--Darin

As I mentioned early in the thread, I watched an RS2 on a ST130 in a room with medium earth tone colored walls, ceiling and carpet and I also watched a C3X1080 in an all-white room at OB's house, with a fair amount of ambient light leaking in through the windows on a Firehawk screen.

I thought the C3X1080 had more convincing blacks and the better picture.

Note -- I didn't own a C3X1080 at the time, I made the decision to purchase the C3X1080 based on auditioning the two projectors.

The RS2's measured on/off is around 30,000:1 according to Greg Rogers and the C3X1080's measured on/off is around 7,000:1 according to Coldmachine.One nice thing about the high lumens of the C3X1080 is that you can use it with a dark screen to help ANSI CR retention from reflections and the extra light from the projector helps retain both on/off CR and ANSI CR in the presence of other lighting. With a Firehawk on a white wall you could end up in a situation where the light that is lighting up the white wall more actually helps the perception of blacks in the images overall in much the same way that backlighting does. Although the Firehawk has higher than 1.0 gain at its brightest spot, overall it is a negative gain screen and if light is coming from off angles it does a good job of killing it.

There are many situations where the RS2 can't really compete with the lumens advantage of the C3X1080.

--Darin

Finally, note that these are all numbers. There probably is a point of diminishing returns for ANSI CR. Maybe that number is, say, 200:1, I really do not have any idea.

It may be that at a certain point, it takes a quantum leap in ANSI contrast to produce a noticeable difference, or that at a certain point, additional ANSI contrast is beyond human perceptibility.

On the other hand, if even a black velvet lined room can compromise ANSI -- it may also be that few people ever really get to see high ANSI performance in real world conditions.

I know that audio guys are often easily offended when it is pointed out that what they claim to be hearing is beyond the range of human audibility. I've had guys in audio forums tell me they can hear .1 db of attenuation, hear the superiority of cables that remain linear way outside the range of human audibility and I've been told by these guys many times that if I were to just listen to these cables I would hear it, too.

So, I am always skeptical, maybe more than necessary, but I really wonder how much of what is measured in video is actually perceptible. Just like in audio, very few people seem to make their observations from DBT's, including me -- I saw the RS2 and C3X1080 in different rooms and I knew the identity of the projector I was watching, etc.

Having said all of that, I am still very interested to hear all of the positions and arguments out there regarding this video stuff and in trying to sort through it all. I find it very interesting and I appreciate the time people like yourself, Darin and others put into these discussions.

I think the Guinness Book of World's Records would go with my approach.

Yeah, but what do *they* know? :-)

But it is semantics and as I also said, it depended on what you meant with the addition of "(mixed image)".

Excellent explanation. With (mixed image) I just meant to indicate that ANSI is an example of mixed image -- let's not go back there -- I think we've cleared that up.

There are many situations where the RS2 can't really compete with the lumens advantage of the C3X1080.

That's very helpful -- thanks!

Well, where the rubber meets the road: you got yerself a kick ass projector. Screw the numbers.

I don't think this is in dispute. Although this is not very specific since there are no numbers attached, in general, I see no problem saying that if a room and projector were capable of very high on/off, I have no doubt it would be visible versus a room and projector capable of only very low on/off.

IMO, questions still remain about what is actually achieved in the average home theater, what is seen in the average film, what level of on/off, ANSI, and lumens it takes to produce a Just Noticeable Difference.

As I mentioned early in the thread, I watched an RS2 on a ST130 in a room with medium earth tone colored walls, ceiling and carpet and I also watched a C3X1080 in an all-white room at OB's house, with a fair amount of ambient light leaking in through the windows on a Firehawk screen.

I thought the C3X1080 had more convincing blacks and the better picture.

Note -- I didn't own a C3X1080 at the time, I made the decision to purchase the C3X1080 based on auditioning the two projectors.

The RS2's measured on/off is around 30,000:1 according to Greg Rogers and the C3X1080's measured on/off is around 7,000:1 according to Coldmachine.

There is simply no way that a C3X1080 can out perform the RS2 is low APL. I know that my HT 5000 has a similar on /off to the C3X1080 and in bright material IMO it does better than an RS1 I saw at similar fL . My room is dark colors but in low APL it just is unable to hold black compared to any RS1 or RS2 I've seen. Of course one can say that the amount of time in the majority of films when this is obvious is relatively small it is none the less a fact.

I personally prefer what my projector does to an RS2 for several reasons but low APL performance isn't one of them.

Art

The RS2's measured on/off is around 30,000:1 according to Greg Rogers and the C3X1080's measured on/off is around 7,000:1 according to Coldmachine.
And the difference is easy to demonstrate in a room that degrades neither projector's performance. If you start adding ambient light and do not calibrate for the specific lighting and room situation anything goes, though.

I don't want to be offensive, but I find that JVC owners tend to be extremely sensitive on this topic because they own a projector with a high measured on/off CR. Don't you own own a JVC RS2? Do you think this could be causing you to want to discount opposing opinions to the extent that you cannot even seem to concede his point about mixed scenes or room reflections?

As a former older-generation JVC owner I appreciated its weakness in the area of ANSI CR.

That would make you the rare bird.

Sorry to take this OT, but I think this little exchange sort of summed up AVS and every other forum on the Internet. The key to AVS being a cornucopia of knowledge is learning who the 10% of the people are that are the rare birds that are knowledgeable and have very little interest in championing a particular technology or product because that's not what it is about for them. Then you learn whose opinions you give value to and can increase the signal to noise ratio of threads exponentially.

Michael, Darin, Art, Glimmie, mhafner, mrwiggles....all rare birds :).

Well that's very nice, Q, thanks. Only I had to come back on this thread to post a correction. You see, technically I didn't own a JVC, I owned a TAW, which as you know was a JVC + Panamorph repackage... I thought that for the sake of simplicity it was no big deal to say JVC since we were talking about them...

...except that the TAW was a Hitachi+Panamorph rebadge :(

Oops. Anyway, it was, like the JVC, an LCOS projector, and shared similar ANSI CR issues of other LCOS projectors at the time, as well as a few issues of its own :)

Anyway, sorry for the error.

There is simply no way that a C3X1080 can out perform the RS2 is low APL. I know that my HT 5000 has a similar on /off to the C3X1080 and in bright material IMO it does better than an RS1 I saw at similar fL . My room is dark colors but in low APL it just is unable to hold black compared to any RS1 or RS2 I've seen. Of course one can say that the amount of time in the majority of films when this is obvious is relatively small it is none the less a fact.

I personally prefer what my projector does to an RS2 for several reasons but low APL performance isn't one of them.

Art

Sorry for my ignorance, but what's APL ?

For the OP...

Ive found that the better low APL performance of an RS2 comes into play is less than 5%. Even then its not that straightforward, for a couple of reasons

1.Fades to black can be trashed by JVCs potentially horrific uniformity and especially bright corners. This also means that the CR figures are actually dependent on where they are measured. Screenshot from RS2 below.

2.Those great blacks suddenly washout in the presence of any bright objects. There is also frequently a halo or veil around objects.

stirring the pot coldmachine...:D

Seriously though, I thought this was a thread about contrast ratios and not about individual projectors. I realize the OP brought up JVC as a way to challenge the objectivity of those who think on/off is important, but I rather like threads that discuss purely technical issues.

stirring the pot coldmachine...:D

Not at all LJ. Just passing info to the OP.:)

The 2 points I made above are technical issues that effect CR on certain machines. They also usually get conveniently overlooked, in the same manner that some people try to portray a soft image as "film like" whilst never having seen what film is capable of. Anyway, I digress.......

Well done rsbeck, Micheal nailed it when he said "you got yerself a kick ass projector. Screw the numbers."

Of course one can say that the amount of time in the majority of films when this is obvious is relatively small it is none the less a fact.

What % of the time that you spend watching films would you say this becomes obvious?

I personally prefer what my projector does to an RS2 for several reasons but low APL performance isn't one of them.

What are some of those reasons?

Seriously though, I thought this was a thread about contrast ratios and not about individual projectors.

It began with some quotes I found regarding the CR limits of film, film cameras and reproduction via projectors.

It continued and developed with some points about the capabilities of real world home theaters and the effects of mixed scenes on contrast ratios.

I wanted to hear what forum members thought with regard to these issues and how they relate to home theater reproduction.

Early on, I also brought up specific projectors as a reference point to question the perceptibility of some of these measured differences.

I realize the OP brought up JVC as a way to challenge the objectivity of those who think on/off is important

I assume everyone thinks on/off is important. Where I question the objectivity of JVC owners is when they seem to want to prevent the discussion from developing past measured on/off CR as a yardstick of projector performance and of measured projector on/off as a yardstick of a typical home theater's performance.

One point that remains curious to me is this;

On my journey to find a projector, I was briefly curious about the Planar PD8150.

Anecdotal reports from professional and customer reviews were that the CR and black levels of the Planar were perceived as equal to the JVC RS1.

Then, the first review with measurements appeared and the Planar measured around 1,700:1 native on/off CR, around 4,000:1 with a Dynamic Iris engaged and around 6,000:1 with the DI engaged and in "film mode." The 6,000:1 measurement was not listed in the measurements with the other figures, you had to find it in a paragraph with a qualifier that the reviewer didn't prefer "Film Mode." Still, the reviewer stressed the quality of the Planar's black levels

One of the non-professional reviews equating the RS1 and Planar's black levels and CR was from AVS poster Tryg. I don't believe you will find anyone on AVS who prizes their JVC projector and its measured on/off more than this poster, so for him to say that he perceived the Planar's black levels to be equal to the RS1 is notable.

To me, this points up the question of perception and how much on/off it takes to produce a noticeable difference.

Since the JVC is a LCoS projector and the Planar a DLP, it might even bring up questions regarding difference between the two technologies.

I assume everyone thinks on/off is important. Where I question the objectivity of JVC owners is when they seem to want to prevent the discussion from developing past measured on/off CR as a yardstick of projector performance and of measured projector on/off as a yardstick of a typical home theater's performance.


I personally haven't noticed JVC owners preventing discussion from developing past measured on/off CR. You may be confused by arguments stating that huge on/off combined with a higher gamma result in pictures with depth similar to that of a lower on/off, higher ANSI CR projector or the fact that higher amounts of ANSI CR are wasted in most rooms. I don't think either of those points are in question though.

I believe the relationship between ANSI CR and on/off CR is a topic that has been very well covered on AVS and I'm surprised it required another thread for this to be understood.

A very significant point regarding those Planars, like the RS1 hysteria, was gullibility. The company themselves were on the forums asking for opinions and taking subtle side swipes at other companies. That, along with some rather stupid claims, was a red flag to me. Its not the style of a quality operation.

It was also obvious that they would be under performers, especially considering the reluctance to confirm certain details.

Some people were obviously suckered by a rather lame charm offensive designed to impress a J6P audience who were all too willing to guzzle the Kool Aid.


Its not for no reason that Planar, Marantz, Optoma, Infocus and some others cant produce max contrast and max lumens at the same time.

1.Fades to black can be trashed by JVCs potentially horrific uniformity and especially bright corners. This also means that the CR figures are actually dependent on where they are measured. Screenshot from RS2 below.This is definitely a potential issue with the JVCs (and some Sonys) and one that I've mentioned with my RS1 before. The RS2s I've seen have tended to be better than the RS1s in this regard though (at least to my eyes).
2.Those great blacks suddenly washout in the presence of any bright objects. There is also frequently a halo or veil around objects.You may already know this, but people's eyes can cause this effect and the lower black floor could make this issue more noticeable. My suggestion is that if people want to find out if it is the projector or their eyes (or how much of each), they should close one eye, put something in front of their face (like their hand) to block the bright part of the image, and see if the halo or veil is still there. If it is just from the projector then the halo should actually get more pronounced as the person's vision adjusts down to the darker level. If it is from the eyes and the veiling luminance vision tends to have around bright objects, then the halo should disappear when the bright object is no longer in the field of vision. I have seen this myself with an RS2 where we were trying to figure out how much the projector or ceiling were lighting up the black bars of a 2.35:1 movie with the raised black part being closer to a bright object in the movie. When I covered the path from my eyes to the movie part so that I only saw the top black bar and things above it, the raised part mostly went away, which told me the effect was really from my eyes. If it is the projector or the room that is causing this kind of effect, the bright part of the image from the projector can be blocked before it gets to the screen. If the halo is from the projector then the halo should still be there even though the bright part is blocked (although you don't want to block it with something bright if the room is light colored so as to avoid cause reflections from that to the screen.)

--Darin

Sorry for my ignorance, but what's APL ?I believe the letters stand for Average Picture Level, but it is basically that even if one or more of the words isn't exact (like the L might be Luminance). I believe that most use the values in the content as opposed to after gamma is applied (which changes a 50% level to only be about 20% of the amount of light as a 100% level). Since we care about the amount of light some have proposed using ADL, for Average Display Luminance. There is a thread about this with some interesting pictures from real content here:

http://www.avsforum.com/avs-vb/showthread.php?p=14261062#post14261062

--Darin

You may already know this, but people's eyes can cause this effect and the lower black floor could make this issue more noticeable.

That is indeed an issue.

Using back to back PJs in a great room also makes it very noticeable and easy to determine the source. Thats how I first came across it.

You may be confused by arguments stating that huge on/off combined with a higher gamma result in pictures with depth similar to that of a lower on/off, higher ANSI CR projector

This is a different argument and I have no problem separating this argument from the others, but your comment begs several questions.

If ANSI (along with other factors) is unimportant, why would it take "huge on/off" (your words) combined with higher gamma to result in pictures with depth equal those of a lower on/off, higher ANSI CR projector?

Further, in my reading, the higher gamma claims are not universally accepted. For example, I have read where some JVC owners have written that they do not prefer the higher gamma setting because it degrades other aspects of the projector's performance.

Huge on/offs and higher gammas dont stop a PJ washing out black areas when whites are adjacent. The bad ANSI performance remains,no matter what.

Despite this, some will try to claim otherwise.

This is a different argument and I have no problem separating this argument from the others, but your comment begs several questions.

If ANSI (along with other factors) is unimportant, why would it take "huge on/off" (your words) combined with higher gamma to result in pictures with depth equal those of a lower on/off, higher ANSI CR projector?

Further, in my reading, the higher gamma claims are not universally accepted. For example, I have read where some JVC owners have written that they do not prefer the higher gamma setting because it degrades other aspects of the projector's performance.

http://www.avsforum.com/avs-vb/showthread.php?t=1035861&highlight=gamma

Some are worried about crushing shadow detail and some prefer to emphasize shadow detail more than other areas of the picture. By lowering gamma, you give the low IRE images more contrast between each other. Many CRT users employ custom gamma curves to strike a good balance.

Coldmachine is correct in his observations. Whites do wash out blacks. A better lens and an excellent room help this. I think Mark Peterson's Contrast Project thread as well as Darinp's explanations are good examples of when each matters and why.

That is indeed an issue.

Using back to back PJs in a great room also makes it very noticeable and easy to determine the source. Thats how I first came across it.I'm not clear, but did you try blocking parts of the image to isolate the source of the issue? Just comparing the projectors won't prove it because a better CR can make the veiling luminance of the eye come through as the dominant factor where a lower CR wouldn't. Differences in gamma down low can also have an effect on certain things here, so it makes some sense to try to isolate those (especially when a projector has very adjustable gamma).

--Darin

a better CR can make the veiling luminance of the eye come through as the dominant factor where a lower CR wouldn't.

Interesting. I've never seen this discussed before. Could this be a significant factor in the continuum of factors that cause a person to prefer one projector over another? Is this true of projectors only or is this also true in other display technologies like flat panel?

What do you guys consider to be huge on/off?

I'm not clear, but did you try blocking parts of the image to isolate the source of the issue?


Yes. Pardon my brevity.:)

huge on/off combined with a higher gamma result in pictures with depth similar to that of a lower on/off, higher ANSI CR projector or the fact that higher amounts of ANSI CR are wasted in most rooms. I don't think either of those points are in question though.

The second point is not in question. Most rooms we see would struggle to yield an ANSI of 100

The first one is, IMHO, despite the evangelical proclamations we were subjected to recently.

If the PJ in question cannot hold a black in the presence of bright areas then the image depth is ruined. The overall depth and immersion, also requires sharpness, focus, accurate color and a decent fL number.

If anyone honestly believes you can take an RS2 with a gamma of 2.4 and will archive the same image depth as a C3X1080, they are very wrong indeed. Ive back to backed them a number of times and there is just simply a class difference. You don't need to go to the extent of a C3X1080 either. I no longer own it, but my old HT380 was a far superior machine too, in terms of punch and depth, amongst other things.

I think we are getting OT anyway.

cheers cm. here's to something better coming down the pipe...

cheers cm. here's to something better coming down the pipe...

Im sure there will will be LJ.:)

What % of the time that you spend watching films would you say this becomes obvious?


Some films it never comes up. Some 80% (ie Dark City). Honestly I'd say less than 1% of the total film time on average(comparing my set up now to my CRTs which had mutiples better sequential CR than the RS2).

Art

What are some of those reasons?

Comparing the HT5K to the RS2 the HT5K has close to five times the light output, better full field uniformity,unbelievable panel alignment ( no place with more than 1/4 pixel), appears sharper (better MTF), has essentially perfect color at SMPTE and 709,significantly better ANSI contrast ,which at least to my eye, gives it more three dimensional looking high APL performance.

Art

Hi Darin,

Could you do a little test for me with your RS1 if you get the opportunity please? Remember this scene?:

http://www.avsforum.com/avs-vb/showpost.php?p=9124891&postcount=7

The on/off of the pj I was using was around 2700:1 IIRC and the ANSI was over 400:1 (my measured figures). With that scene I could see and measure the dark part of his stomach armour and a bright piece of sky coming through the trees from the top right hand corner of the image. I think I measured around 1100:1 from those two points, and using a mouse pointer as the white level, then measured around 1500:1 (numbers are from memory but I think are ballpark). That's around 55% of the pjs on/off capability and higher than it's ANSI capability.

I'm curious how the RS1 would measure with this scene and what pecent of it's on/off that would be. I've a feeling that the RS1s higher on/off will allow for a much higher contrast within that scene despite it's lower ANSI (compared to the DLP I had at the time).

Cheers

Gary

Hi Darin,

Could you do a little test for me with your RS1 if you get the opportunity please?I'll try to get to it in the next few days. My RS1 has one or more corners that are probably 25% higher than the center on a blackout image, so less on/off CR in those corners. I'll also be getting a Panasonic AE1000 by Monday, which I plan to use the waveform monitor on to try to determine the makeup of some of these scenes.

--Darin

Thanks Darin.

Gary

What do you guys consider to be huge on/off?
1000000:1.
Let me elaborate further:
1000:1 very poor
2000:1 poor
5000:1 poor to mediocre
10000:1 mediocre to good
30000:1 good
100000:1 very good
500000:1 excellent
5000000:1 reference, no need to go further. Even black outs hold up long enough

1000000:1.
Let me elaborate further:
1000:2 very poor
2000:1 poor
5000:1 poor to mediocre
10000:1 mediocre to good
30000:1 good
100000:1 very good
500000:1 excellent
5000000:1 reference, no need to go further. Even black outs hold up long enough
I think you have somewhat of an obsession ;-). How can you give numbers out that are entirely theoretical (i.e. no device comes close)? Do you honestly think there is going to be much of a difference between 100,000:1 and 1,000,000:1? Where did you pull the number 5,000,000:1 out of, as the reference beyond which we need not go further ;-)? I think you are wrong, I think 5,648,012:1 is the magic number.

Vaffanculo! Hafner you now take the cake.

The maximum contrast 35mm emulsion is made by Kodak and it is very rare and expensive and borders on 9k-1, the rest are around 5k. A .98 digital cinema can be modified to 5k. The limitation is the native white of the lamp being lower than D65. There is a new experimental solution that will give another 15-20%. That is right around or higher than 35mm. When you count everything else going on that is the only way to go for me, the high MTF is a function of high ansi contrast. The fact that ansi greatly affects MTF handicaps it in a factor of importance to on/off contrast ratio.

ansi contrast importance=1
ansi contrast importance handicapped for being high MTF vehicle=3

on off contrast importance=1

You have a 3 to 1 advantage, psychovisually.

Everyone knows three (http://www.youtube.com/watch?v=yPzAjiLr5Zw) (link) is a magic number.

Peter has touched on something important that most readers will miss. The real importance of ANSI CR is that it’s a proxy for MTF at higher frequencies.

There are unexplained contradictions looking at ANSI CR only. Most movie images seem to be of relatively low APL where on/off CR dominates image contrast performance. Also, ANSI CR is severely limited by most rooms. So why do many experienced viewers prefer projectors that have a relatively low on/off CR and a high ANSI CR? There are other important variables like color accuracy, but ANSI CR seems to be an important indicator for image quality. The answer is that ANSI CR is these designs correlates with MTF performance at higher image frequencies. (Image frequency is nothing more than spatial resolution, with the highest horizontal “frequency” in a 1920x1080 image being 1920 pixels.”

MTF performance in simple terms is the ability to retain contrast as a function of image frequency (resolution). It’s a curve with image spatial frequency on one axis and a 0-1 scale on the other. An MTF of 1.0 is the highest score. At the limiting horizontal resolution of 1920 for consumer HD, an MTF of 1.0 would mean that if your source has a contrast of 100:1 between pixels in an alternating single pixel wide pattern, the projected 1920 single pixel wide alternating pattern would also have a contrast of 100:1 between the pixels. ANSI CR is essentially MTF at a very low image frequency. However, a high ANSI CR in these projector designs also means a high MTF at the higher image frequencies that are very important for clarity of image detail. High MTF means that the image detail has more contrast, is much clearer, and looks sharper. Low MTF looks soft, as if the image is out of focus. While detail in a 0.1 (10%) MTF image is visible, it looks nothing like the detail in a 0.9 (90%) MTF image. The increased clarity and sharpness in finer detail that comes with high MTF explains the preference for projectors with high ANSI CR.

The above definitions of MTF, image spatial frequency, etc. are technically inaccurate. I don’t want to use complex formulas and terms like relative to low frequency modulation and normalized to low frequency contrast. A technically accurate and more detailed treatment of MTF is at: http://www.normankoren.com/Tutorials/MTF.html (http://www.normankoren.com/Tutorials/MTF.html).

I think you will find that there is a significant number of folks on the 3000 plus forum who believe ANSI and MTF are quite far down the list in importance in image quality.

Art

I think you will find that there is a significant number of folks on the 3000 plus forum who believe ANSI and MTF are quite far down the list in importance in image quality.

Art

I have noticed this and I think that the explanation is that they don't know what MTF is or how ANSI CR relates to MTF. Anyone who designs optical systems will tell you that MTF performance is very important. It combines resolution and contrast and is the best objective definition of the resolution of an optical system.

I think you will find that there is a significant number of folks on the 3000 plus forum who believe ANSI and MTF are quite far down the list in importance in image quality.

Art

True, as are O's comments above. Things like MTF, VP etc don't make it onto a typical J6P spec sheet.

There's also a very heavy dose of self delusion, for obvious reasons, at the the cost of objectivity.

Some people also seem only too willing to follow an agenda driven buffoon, so long as he spews whatever deluded garbage they want to hear.

If you feed people mediocrity for long enough, what do you end up with?........Connoisseurs of shite.

Thats why I value this part of AVS above all others.

True, as are O's comments above. Things like MTF, VP etc don't make it onto a typical J6P spec sheet.

There's also a very heavy dose of self delusion, for obvious reasons, at the the cost of objectivity.

Some people also seem only too willing to follow an agenda driven buffoon, so long as he spews whatever deluded garbage they want to hear.

If you feed people mediocrity for long enough, what do you end up with?........Connoisseurs of shite.

Thats why I value this part of AVS above all others.

I wish I could be so eloquent.:D

Art

I think you have somewhat of an obsession ;-). How can you give numbers out that are entirely theoretical (i.e. no device comes close)? Do you honestly think there is going to be much of a difference between 100,000:1 and 1,000,000:1? Where did you pull the number 5,000,000:1 out of, as the reference beyond which we need not go further ;-)? I think you are wrong, I think 5,648,012:1 is the magic number.
Well, Darin measured 700000:1 for a CRT where the blackout did not hold long enough in actual film, so it's not yet optimal. With a CRT you can simulate numbers from 30000:1 to > 500000:1 so it's not theoretical. The numbers from 2000:1 to 30000: 1 are available from digital projectors you can buy now. Numbers > 30000:1 will be available in the coming years. Concerning reference I'm not sure if it's 5000000:1 or more or less if the criteria is to hold blackouts as long as they exist in actual films. 700000:1 is apparently not enough yet.
And all these numbers refer only to rooms and sources/shots where the problems are visible. With other sources/shots there maybe no problems with far far less than 1000000:1.
I'm overall rather happy with the current ~15000:1 I have at the moment, but I can see clearly when it's not enough. And it happens all the time when watching feature films.

I have noticed this and I think that the explanation is that they don't know what MTF is or how ANSI CR relates to MTF. Anyone who designs optical systems will tell you that MTF performance is very important. It combines resolution and contrast and is the best objective definition of the resolution of an optical system.
ANSI contrast is important too. So let's state the same numbers for ANSI. I have no firm opinions here but I would suggest something like this for starters:
<= 50:1 poor
50:1 - 150:1 mediocre
150-300:1 mediocre to good
300:1-500:1 good
500:1 - 750:1 very good
750:1 - 1000:1 very good to excellent
> 2000:1 reference

The answer is that ANSI CR is these designs correlates with MTF performance at higher image frequencies.I agree that it normally does, but there is no rule that it has to and so it can be a little misleading if people say that high ANSI CR was the difference if it was really MTF (although I understand simplifying things). For instance, a DLP projector could have a lens that provided 300:1 ANSI CR, but have much different MTF than an LCOS projector with 300:1 ANSI CR. I suspect that the reasons CRTs have lower ANSI CR to the level they do are somewhat different than the reason(s) they have low MTF and that a DLP design that only ended up with 130:1 or so (degraded from something other than the chips) would probably still have higher MTF than a G90. Although it would probably depend on what had degraded the DLP to so much lower than DLPs can do.

--Darin

1000000:1.
Let me elaborate further:
1000:1 very poor
2000:1 poor
5000:1 poor to mediocre
10000:1 mediocre to good
30000:1 good
100000:1 very good
500000:1 excellent
5000000:1 reference, no need to go further. Even black outs hold up long enoughI realize this NASA data:
http://msis.jsc.nasa.gov/sections/section04.htm#_4.2_VISION


Example Scale of Luminance (millilamberts) Effect
Sun's Surface at noon 10^9 Damaging
10^8
10^7 Phototopic
Tungsten Filament 10^6
10^5
White paper in sunlight 10^4
10^3
10^2
Comfortable reading 10
1 Mixed
10^-1 Scotopic
White paper in moonlight 10^-2
10^-3
White paper in starlight 10^-4
10^-5
Absolute threshold 10^-6

Reference: 337, p. 26; NASA-STD-3000 194is probably rounded and I'm not sure if there is a more accurate chart/graph, but according to that, if the white level is at the comfortable reading level (about 9 ft-lamberts from 10 millilamberts) than getting the absolute black level to the following takes:

White paper in moonlight: 1,000:1
White paper in starlight: 100,000:1
Absolute threshold: 10,000,000:1

Make the white level 18 ft-lamberts and those numbers double.

--Darin

That is the threshold of pain paper again. I found the fact that it covers that subject fitting considering it did the same to me while reading it.:D

Darin I think your research obsession is severely underfunded. As long as you keep on using toy projectors for your research you are just sampling part of the potential as to what this serious field of endeavor is. If you don't even analyze a vehicle with > 825 ansi cr based MTF, how the hell are you qualified to understand it and much less write definitive articles on it?

You have to plant your butt in front of such a screen with mixed content playing and ask yourself what the hell am I seeing?

I suggest you stop looking at outer space and make your pilgrimage inside the US instead to understand the Expert Viewer's approach Odyssey has been espousing for years.

I agree that it normally does, but there is no rule that it has to and so it can be a little misleading if people say that high ANSI CR was the difference if it was really MTF (although I understand simplifying things). For instance, a DLP projector could have a lens that provided 300:1 ANSI CR, but have much different MTF than an LCOS projector with 300:1 ANSI CR. I suspect that the reasons CRTs have lower ANSI CR to the level they do are somewhat different than the reason(s) they have low MTF and that a DLP design that only ended up with 130:1 or so (degraded from something other than the chips) would probably still have higher MTF than a G90. Although it would probably depend on what had degraded the DLP to so much lower than DLPs can do.

--Darin

I agree that ANSI CR and MTF at higher spatial frequencies can diverge, but I think that the correlation is fairly good if ANSI CR is significantly different for DLP, LCOS, and CRT projectors. That's why I said "in these designs."

I think you will find that there is a significant number of folks on the 3000 plus forum who believe ANSI and MTF are quite far down the list in importance in image quality.

Art

Right Art and now the fire of misinformation is about to be fanned into wildfire proportions that dwarf anything ever seen on this planet.

With Japan inc. (Sony and Matsushita) aided by the prestige that Meridian brings to this marketing triumvirate, 4k (having failed miserably in the battle for DCI) is now going to be dumped on us, the unsuspecting Home Theater market.

The MTF on LCOS is very low, when you compound the additional softness by trying to display 2k content on 4k on a LCOS, the MTF will go even more south. Then some of these guys (Meridian) will drive the MTF south another 25% with the Isco lens. And they are going to charge huge bucks for this. They may have all the funds in the world, but I am bullish as hell on 2k SuperKontrast underdog.

So why do many experienced viewers prefer projectors that have a relatively low on/off CR and a high ANSI CR? There are other important variables like color accuracy, but ANSI CR seems to be an important indicator for image quality. The answer is that ANSI CR is these designs correlates with MTF performance at higher image frequencies. (Image frequency is nothing more than spatial resolution, with the highest horizontal “frequency” in a 1920x1080 image being 1920 pixels.”

MTF performance in simple terms is the ability to retain contrast as a function of image frequency (resolution). It’s a curve with image spatial frequency on one axis and a 0-1 scale on the other. An MTF of 1.0 is the highest score. At the limiting horizontal resolution of 1920 for consumer HD, an MTF of 1.0 would mean that if your source has a contrast of 100:1 between pixels in an alternating single pixel wide pattern, the projected 1920 single pixel wide alternating pattern would also have a contrast of 100:1 between the pixels. ANSI CR is essentially MTF at a very low image frequency. However, a high ANSI CR in these projector designs also means a high MTF at the higher image frequencies that are very important for clarity of image detail. High MTF means that the image detail has more contrast, is much clearer, and looks sharper. Low MTF looks soft, as if the image is out of focus. While detail in a 0.1 (10%) MTF image is visible, it looks nothing like the detail in a 0.9 (90%) MTF image. The increased clarity and sharpness in finer detail that comes with high MTF explains the preference for projectors with high ANSI CR.


Excellent info and explanation -- Thank you.

I'd like to thank the contributors to this thread, it was most enlightening. :)

Obviously, a perfect projector would have good ansi/mtf and on/off. This entire thread (and many others) only exists because there is no such projector...a sacrifice must be made. And that, is where opinions collide.

As illustrated below:


Art - If you could get the on/off of your G90 stack with the sharpness and light ouput of your HT5000, would you give up the ANSI CR performance of the HT5000? Assume the ease of setup of the two were the same (ie, just consider image quality parameters).

Yes.

Art


:D

Obviously, a perfect projector would have good ansi/mtf and on/off. This entire thread (and many others) only exists because there is no such projector...a sacrifice must be made. And that, is where opinions collide.

As illustrated below:



:D

I have my HT 5000 because of light output, high on /off for a three chip DLP, tremendous panel alignment, sharpness and nearly perfect color. The ANSI seems to give it that "punch" in high APL but yes, if I had all those other things but ANSI below 200:1 and 700,000:1 on/off I'd take it.:D It might be true that ANSI and MTF are linked to sharpness so not sure it could happen anyway.

Art

I have my HT 5000 because of light output, high on /off for a three chip DLP, tremendous panel alignment, sharpness and nearly perfect color. The ANSI seems to give it that "punch" in high APL but yes, if I had all those other things but ANSI below 200:1 and 700,000:1 on/off I'd take it.:D It might be true that ANSI and MTF are linked to sharpness so not sure it could happen anyway.

Art

I feel the same way. I have owned crt projectors and loved the on/off, and the low ansi just wasn't an issue. But, if I had to choose it would be DLP any day. Awesome ansi and light output, perfect sharpness, convergence, and uniformity, all make DLP very attractive.

DLP, with tremendous on/off contrast (2.5 million:1)

http://www.avsforum.com/avs-vb/showthread.php?p=14289796#post14289796
http://www.zeiss.de/c12567b00038cd75/Contents-Frame/916a1bd1307bf2acc1257480005213b4

Could you do a little test for me with your RS1 if you get the opportunity please? Remember this scene?:

http://www.avsforum.com/avs-vb/showpost.php?p=9124891&postcount=7

The on/off of the pj I was using was around 2700:1 IIRC and the ANSI was over 400:1 (my measured figures). With that scene I could see and measure the dark part of his stomach armour and a bright piece of sky coming through the trees from the top right hand corner of the image. I think I measured around 1100:1 from those two points, and using a mouse pointer as the white level, then measured around 1500:1 (numbers are from memory but I think are ballpark). That's around 55% of the pjs on/off capability and higher than it's ANSI capability.

I'm curious how the RS1 would measure with this scene and what pecent of it's on/off that would be. I've a feeling that the RS1s higher on/off will allow for a much higher contrast within that scene despite it's lower ANSI (compared to the DLP I had at the time).
Gary,

I meant to get back to you earlier about this. I did some measurements, but the areas of black and bright are so small that I had a problem where I think they were either smaller than the sensor or not bright enough to give very accurate results. There is a spot down by or on his right leg that is dark, but I don't think much of it is encoded all the way down to video black. And I wasn't running a PC and so couldn't use a mouse pointer for white. The bright stuff up between the trees was smaller than the sensor for one meter and I think smaller than the cone for a sensor that reads off the screen. The values I got were in the 500:1 to 600:1 range, but I wouldn't put too much faith in them because of the issues with not getting a uniform area for readings.

--Darin

Hi Darin,

Thanks for trying, I appreciate the effort. :)

Gary