Some people are still being told that 200:1 is all we can see.... - Page 16 - AVS Forum
Forum Jump: 
Reply
 
Thread Tools
post #451 of 505 Old 01-10-2007, 01:27 PM
Senior Member
 
sisaacs's Avatar
 
Join Date: Jul 2003
Location: Abilene, TX
Posts: 391
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
It occurred to me while reading this thread that possibly you all are discussing two different things. There's the people discussing the contrast within an image and if you can see it or not, example "driving at night"

This is the question of "can you see"

The answer is obviously yes to this.

But there is another question, if you took the image that you have in your head of seeing at night time, and reduced it's contrast in the scene to 300:1 or 100:1 or whatever you want to agree on, would it actually look different to you?

That's a much more difficult question to answer, and at least to me does not have a clearly obvious answer to it.

So while you can "see" perhaps 1,000,000:1 images just fine, do these 1,000,000:1 images actually look any different to you than the same image that has had it's contrast reduced by some method to 300:1 or some other low number.

Edit: Also I don't believe the putting your hand in the shadow of ANSI test is quite the right test for this, since that changes the picture and your mind may be equalizing the contrast in some way to actually "see" this difference, perhaps the white was decreased visually to you to accomplish this?

Also the visual tricks of how putting light colors by dark colors changing the percieved lightness etc supports that this can be happening.
sisaacs is offline  
Sponsored Links
Advertisement
 
post #452 of 505 Old 01-10-2007, 01:37 PM
AVS Addicted Member
 
darinp2's Avatar
 
Join Date: Sep 2003
Location: Seattle, WA
Posts: 21,165
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 43
Quote:
Originally Posted by tbrunet View Post

Dr. Soneira answered that already, yet you still?
http://www.avsforum.com/avs-vb/show...83&page=3&pp=30
Quote:
Originally Posted by Raymond Soneira View Post

So what happens when the display only has a Contrast Ratio of 2000? For digital signal levels restricted to 16 to 235 as in Rec.709, signals below level 24 will be indistinguishable from the signal level 16 for black.

I know Thomas wants to continue his games and find any way possible to keep from answering:
Quote:
Originally Posted by darinp2 View Post

With standard 8 bit video (like on DVDs, HD DVDs, Blu-ray, etc), and considering just pure grays, if video 235 gives 10 ft-lamberts, then what should each of the following give:

video 180: ??
video 126: ??
video 71: ??
video 38: ??
video 17: ??

but I don't want people to be misled. I sent an email to Dr. Soneira asking why he stated that the levels below 24 would be indistinguishable from level 16 on a display with a contrast ratio of 2000:1 when that isn't true with the majority of 2000:1 displays that have been calibrated correctly. And also that this wouldn't show that 8 bits isn't enough even if they weren't distinguishable. It would support that 2000:1 isn't enough.

He responded that his point was that as CR decreases more and more digital levels near black become obscured and that this depends on how the lookup tables are calibrated at the factory. When I asked if 2% stim bars would have been invisible on the displays he tested with 2000:1 on/off CR, I didn't get a response. I personally find it strange that somebody who has taughted the advantages of CRTs having 30k:1 on/off CR also said that the performance in the low luminance area isn't important, but I didn't push that one.

So, please don't be misled. If you have a 2000:1 on/off CR display and level 24 and level 16 don't look different, then you either have calibrated the display incorrectly, or there is a problem with it (which could be that the factory settings were bad and don't allow the shadow detail they should). Those with 2000:1 on/off CR displays like the Sharp 10k can do some testing with test patterns and see that it isn't hard at all to distinguish level 24 from level 16 after proper calibration with room lights off.

However you cut it, those levels do have values and Thomas is using stalling tactics to avoid actually answering the question.

--Darin

This is the AV Science Forum. Please don't be gullible and please do remember the saying, "Fool me once, shame on you. Fool me twice, shame on me."
darinp2 is offline  
post #453 of 505 Old 01-10-2007, 01:42 PM
AVS Addicted Member
 
darinp2's Avatar
 
Join Date: Sep 2003
Location: Seattle, WA
Posts: 21,165
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 43
Quote:
Originally Posted by sisaacs View Post

But there is another question, if you took the image that you have in your head of seeing at night time, and reduced it's contrast in the scene to 300:1 or 100:1 or whatever you want to agree on, would it actually look different to you?

That's a much more difficult question to answer, and at least to me does not have a clearly obvious answer to it.

One think I have basically proposed here is put up the 1% white and 99% black image that William Phelps posted, raise the Brightness setting to the point that the simultaneous CR is 300:1 and swap that with what a projector like the Sharp 20k or new JVC can do with that image (well over 2000:1 simultaneous CR) in a proper room with a screen with a black velvet border. Then have somebody leave the room and come in each time to tell you what image they think is up. I don't think it would be hard at all for them to tell you which is which as 300:1 in that image would look pretty poor. The testing could also be done including lowering the Contrast setting to get the same simultanteous CR.

--Darin

This is the AV Science Forum. Please don't be gullible and please do remember the saying, "Fool me once, shame on you. Fool me twice, shame on me."
darinp2 is offline  
post #454 of 505 Old 01-10-2007, 01:46 PM
Senior Member
 
sisaacs's Avatar
 
Join Date: Jul 2003
Location: Abilene, TX
Posts: 391
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by darinp2 View Post

One think I have basically proposed here is put up the 1% white and 99% black image that William Phelps posted, raise the Brightness setting to the point that the simultaneous CR is 300:1 and swap that with what a projector like the Sharp 20k or new JVC can do with that image (well over 2000:1 simultaneous CR) in a proper room with a screen with a black velvet border. Then have somebody leave the room and come in each time to tell you what image they think is up. I don't think it would be hard at all for them to tell you which is which as 300:1 in that image would look pretty poor. The testing could also be done including lowering the Contrast setting to get the same simultanteous CR.

--Darin

That would be an interesting experiment, I don't own a projector that has a super high simultaneous CR to try it out personally or I think I would try that out on my wife hehe.
sisaacs is offline  
post #455 of 505 Old 01-10-2007, 02:23 PM
Newbie
 
matttrent's Avatar
 
Join Date: Oct 2006
Location: Vancouver, BC
Posts: 11
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by sisaacs View Post

But there is another question, if you took the image that you have in your head of seeing at night time, and reduced it's contrast in the scene to 300:1 or 100:1 or whatever you want to agree on, would it actually look different to you?

That's a much more difficult question to answer, and at least to me does not have a clearly obvious answer to it.

I think it obviously does. Go find any photograph of an oncoming car at night. Here's one:



When you look at that, can you make out both the detail in the headlights and the surrounding areas? Do you have a strong urge to avert your eyes or blink due to the brightness?

While not whole story, you can look at it from a luminance quantization perspective too. Across a 10,000:1 dynamic range, you see on the order of 1000 separate brightness levels. Across a 100:1 dynamic range, you see on the order of 250. Any way you cut it, if you scale down the dynamic range, you don't perceive 75% of intensities. That doesn't equate to a 75% information loss overall, but I don't see the basis for any argument that there is no difference.
matttrent is offline  
post #456 of 505 Old 01-12-2007, 04:17 AM
 
tbrunet's Avatar
 
Join Date: Jul 2005
Location: Savannah, GA
Posts: 2,750
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 12
Quote:
Originally Posted by darinp2 View Post

I sent an email to Dr. Soneira asking why he stated that the levels below 24 would be indistinguishable from level 16 on a display with a contrast ratio of 2000:1

Then just to be consistent with gremmy, and the prevailing logic, we should take away Dr. Soneira's ability to post, considering his position regarding CR does not agree with Darin, Chris, ect. We don't want to 'mislead' anyone

*Note to self: seems when I'm able to find empirical evidence/information to support my conjecture, the obligatory reaction is "we should take away his ability to defend said bogus information"

Thomas Brunet
Savannah College of Art & Design
Film Television And Sound Engineer
tbrunet is offline  
post #457 of 505 Old 01-12-2007, 06:24 AM
Senior Member
 
sisaacs's Avatar
 
Join Date: Jul 2003
Location: Abilene, TX
Posts: 391
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by matttrent View Post

I think it obviously does. Go find any photograph of an oncoming car at night. Here's one:



When you look at that, can you make out both the detail in the headlights and the surrounding areas? Do you have a strong urge to avert your eyes or blink due to the brightness?

While not whole story, you can look at it from a luminance quantization perspective too. Across a 10,000:1 dynamic range, you see on the order of 1000 separate brightness levels. Across a 100:1 dynamic range, you see on the order of 250. Any way you cut it, if you scale down the dynamic range, you don't perceive 75% of intensities. That doesn't equate to a 75% information loss overall, but I don't see the basis for any argument that there is no difference.

Though if your eyes scaled things back non-linearly it may be percieved any differently, and as many biological functions are non-linear it would make sense to me that this is a possibility.

In fact the whole edge experiement seems to support this also even accepting Chris's HDR number of 100,000:1 instantaneous this is considerably less contrast than if you applied the 150:1 limit between each pixel on a fairly low rez 1024x768 monitor, if each pixel was 150 times brighter than the one to it's left, you would have 150^1024 contrast across the entire panel, which is way more than 100k, so it's apparent that there is something going on.
sisaacs is offline  
post #458 of 505 Old 01-12-2007, 07:11 PM
Member
 
armadillo's Avatar
 
Join Date: Jul 2005
Posts: 108
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
So Matt, we are still waiting on your squares test done on your super-display. Can you simultaneously see a 1% bright square next to a 100% square on that display? Or how about 2%?
armadillo is offline  
post #459 of 505 Old 01-13-2007, 04:43 AM - Thread Starter
AVS Special Member
 
Gary Lightfoot's Avatar
 
Join Date: Jun 2000
Location: Surrey, UK
Posts: 4,452
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 16 Post(s)
Liked: 38
I would think the headlight example would cause a lot of iris adaption and reduce contrast capability due to veiling glare, so it's not indicative of the range available under normal conditions like that in the Gladiator image. It's like looking into the Sun.

When we move the fovea away from a certain part of an image, the rods are brought into play and are more sensitive then the cones, so we can still see a large range of contrast under normal conditions IMHO.

If there is chemical adaption going on during normal viewing (isn't that really for night vision though?), and it being within the time span of refocussing when moving the eye from one place to another, the perception is visibly instantaneous, so contrast perception is visibly quite large and at least over 1000:1 using the Gladiator image as an example.

Gary

Quote:
Originally Posted by elmalloc
Who says Cameron is "right" and why do we care about him so much - lol!

I trust Gary Lightfoot more than James Cameron.
Gary Lightfoot is offline  
post #460 of 505 Old 01-13-2007, 05:38 AM
 
tbrunet's Avatar
 
Join Date: Jul 2005
Location: Savannah, GA
Posts: 2,750
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 12
Hello Mr. Kuang, I understand in advance I may not get a reply from you regarding my question. I also understand that absolute perception thresholds and limits are subjective. Below I quoted you:
http://www.cis.rit.edu/mcsl/about/ch...e/spring06.pdf
Jiangtao (Willy) Kuang, PhD Imaging Science Candidate
Quote:
our visual system can readily perceive details in scenes that span the range of 4-5 orders

My conjecture is empirical data suggest we can't perceive more than 150:1 or ~2.2 orders of luminance at a contrast boundary. While I agree we can see over a large dynamic range, but i.e. the relative small angle in which we see image detail makes us parse a given visual frame, and our perception is possibly constrained by the noise floor of the human eye. My question is, would it be wrong to suggest that the simultaneous dynamic range of our HVS is ~200:1 or 2.2 orders of magnitude? It is also my conjecture, that it's our brains processing power, coupled with iris adjustments, and adaptation which enable us to combine the stimuli over time, and thus we think we perceive 4-5 orders?

Response:
Quote:
I think you are partially right. The output responses of the retina have a contrast ratio of around 100:1, and the signal-to-noise ratio of individual channels in the visual pathway (from retina to brain) is about 32:1, less than 2 orders of magnitude. However, the adaptation in human visual system play a leading role in HDR perception. Photoreceptors respond linearly to a range of about 3 log units, and the response curves would shift according to different background intensity levels. When the eye moves (saccades) it re-adjusts it sensitivity by these adaptation. Therefore, by this mechanism, visual system can readily perceive details in scenes that span the range of 4-5 orders of luminance through local adaptation.

I hope this helps.

-Willy Kuang

tbrunet is offline  
post #461 of 505 Old 01-13-2007, 06:00 AM
Member
 
armadillo's Avatar
 
Join Date: Jul 2005
Posts: 108
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
The response pretty much sums it up. I couldn't have said it any better.
armadillo is offline  
post #462 of 505 Old 01-13-2007, 12:53 PM
 
ChrisWiggles's Avatar
 
Join Date: Nov 2002
Location: Seattle
Posts: 20,730
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 14
I think that response is yet another vindication of my statements in this thread.
ChrisWiggles is offline  
post #463 of 505 Old 01-13-2007, 02:13 PM
Member
 
armadillo's Avatar
 
Join Date: Jul 2005
Posts: 108
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Chris: Cool, so you now agree that "However, the adaptation in human visual system play a leading role in HDR perception." and that "When the eye moves (saccades) it re-adjusts it sensitivity by these adaptation."? I already told you so many times that eye movement does not equate to instantaneous CR but involves adaptation. Can you now for the record state that without eye movement, the instantaneous HVS CR is ~100-300:1? That would be greatly appreciated and close this thread.
armadillo is offline  
post #464 of 505 Old 01-13-2007, 02:57 PM - Thread Starter
AVS Special Member
 
Gary Lightfoot's Avatar
 
Join Date: Jun 2000
Location: Surrey, UK
Posts: 4,452
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 16 Post(s)
Liked: 38
I've just done another experiment - I paused the Gladiator image which is fed via my HTPC and moved the mouse pointer in between two dark parts of the stomach armour. My light meter registered 57.4 lux for the pointer, and 0.04 to 0.05 for the darkest part and 0.10 for the lighter part around it. I could see both parts as well as the pointer.

The CR difference was between 1148:1 and 1435:1 for the brightest/darkest parts (difference between a 0.05 and 0.04lux reading), and 574:1 for the lighter part. Moving the focus/fovea away made the darker parts more visible due to the extra sensitivity of the rods in the peripheral vision.

Adaption must be happening on an individual cone/rod level ('local' adaption?) to enable me to still see all those things at once in the same focal area, as well as beyond that point.

So if the lower numbers are to be believed, it must be a case of the sum being greater than the parts as I can see considerably more than 200:1 simultaneously, no matter which parts of the eye I am using.

I can enclose a pic of the part of the image I was testing if anyone wants to try this themselves.

Gary

Quote:
Originally Posted by elmalloc
Who says Cameron is "right" and why do we care about him so much - lol!

I trust Gary Lightfoot more than James Cameron.
Gary Lightfoot is offline  
post #465 of 505 Old 01-13-2007, 03:08 PM
Member
 
armadillo's Avatar
 
Join Date: Jul 2005
Posts: 108
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Gary: I don't believe it. Of course you can see the bright spot (as can your spotmeter), but you cannot say whether it is 1000x brighter and you won't be able to see anything in its vicinity that's 1000x less bright. Do the squares test, that will give you your answer.
armadillo is offline  
post #466 of 505 Old 01-13-2007, 03:20 PM - Thread Starter
AVS Special Member
 
Gary Lightfoot's Avatar
 
Join Date: Jun 2000
Location: Surrey, UK
Posts: 4,452
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 16 Post(s)
Liked: 38
Why don't you believe it? It's a real image with real numbers and I could see the darker objects as well as the bright spot. The reason I can is because the mouse pointer (being at least 1148 times brighter) is not causing iris adaption/compensation to crush shadow detail.

The squares test will overload the eye just as will looking into headlights or the sun, and bring the iris into play so the blacks will probably become less visible, so that test is testing other criteria such as veiling glare perhaps.

My test was more real world and normal for an everyday image, and the numbers prove that over 1000:1 is possible. Try it yourself, and you will see. If I put a much larger white image into that scene, then I agree I won't be able to see the dark parts any more.

I did do the glove puppet test while I was there and the results were as predicted.

Gary

Quote:
Originally Posted by elmalloc
Who says Cameron is "right" and why do we care about him so much - lol!

I trust Gary Lightfoot more than James Cameron.
Gary Lightfoot is offline  
post #467 of 505 Old 01-13-2007, 03:38 PM
 
ChrisWiggles's Avatar
 
Join Date: Nov 2002
Location: Seattle
Posts: 20,730
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 14
Quote:
Originally Posted by armadillo View Post

Gary: I don't believe it. Of course you can see the bright spot (as can your spotmeter), but you cannot say whether it is 1000x brighter and you won't be able to see anything in its vicinity that's 1000x less bright. Do the squares test, that will give you your answer.

Do the test.

As far as I see it the only people willing to do actual testing that is relevant to the issue are reaching similar conclusions.

If you have un-substantiated statements on the one hand, and then on the other you have statements from fairly reputable sources and testing mirrors those statements, it seems to me the onus is clearly on you to substantiate your claims. Otherwise you're wasting people's time making statements that have no support to them and are contrary to the science.
ChrisWiggles is offline  
post #468 of 505 Old 01-13-2007, 04:08 PM
AVS Club Gold
 
GlenC's Avatar
 
Join Date: Jun 2003
Location: Los Angeles - Whittier, CA
Posts: 2,555
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by Gary Lightfoot View Post

...The squares test will overload the eye just as will looking into headlights or the sun, and bring the iris into play so the blacks will probably become less visible, so that test is testing other criteria such as veiling glare perhaps.

This is the issue I was trying to make, there are situations that can reduce the ability to see "all the detail" with out focusing the eye away from the bright areas.

Glen Carter
Home Theater Calibration
www.ISFHT.com
GlenC is offline  
post #469 of 505 Old 01-13-2007, 04:18 PM
AVS Addicted Member
 
darinp2's Avatar
 
Join Date: Sep 2003
Location: Seattle, WA
Posts: 21,165
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 43
Quote:
Originally Posted by GlenC View Post

This is the issue I was trying to make, there are situations that can reduce the ability to see "all the detail" with out focusing the eye away from the bright areas.

I think we all agree that this can happen. But the issue as far as I see it is what the upper limit is, not the lower limit. If a person picks a case that is more limiting for the eye then they aren't finding the upper limit and shouldn't claim to be talking about the upper limit.

--Darin

This is the AV Science Forum. Please don't be gullible and please do remember the saying, "Fool me once, shame on you. Fool me twice, shame on me."
darinp2 is offline  
post #470 of 505 Old 01-13-2007, 04:27 PM
AVS Club Gold
 
GlenC's Avatar
 
Join Date: Jun 2003
Location: Los Angeles - Whittier, CA
Posts: 2,555
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by darinp2 View Post

I think we all agree that this can happen. But the issue as far as I see it is what the upper limit is, not the lower limit. If a person picks a case that is more limiting for the eye then they aren't finding the upper limit and shouldn't claim to be talking about the upper limit.

--Darin

I totally agree, there is total range and the upper limit and the lower limit. The issue I have had difficulty getting across is the eye has variable range depending on the situation, and in any given situation, we can see somewhere from 100:1 to well over 1000:1 CR.

Glen Carter
Home Theater Calibration
www.ISFHT.com
GlenC is offline  
post #471 of 505 Old 01-13-2007, 04:42 PM
Member
 
armadillo's Avatar
 
Join Date: Jul 2005
Posts: 108
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Guys: I already gave you the spaceship example. So if you are in a dark-adapted state say at night and you look at the stars, what you see is the stars as tiny little dots hitting some of the rods on your retina. If you then add a star or two, it won't make a difference, because you won't move out of the dark-adapted state. So according to your theory, you will be able to see distinguish between the intensities of 1000s of different stars. You won't. And the same is true for the pointer. All you are doing is saturating a few photoreceptors. If the mouse pointer was made up of pixels of 99% and 100% brightness, would you be able to see those? Hell, if you did a blind test with a mouse pointer of 100% brightness and 99% brightness, would you be able to say which it was in the Gladiator image? That is the question. And the answer is NO.
armadillo is offline  
post #472 of 505 Old 01-13-2007, 05:02 PM - Thread Starter
AVS Special Member
 
Gary Lightfoot's Avatar
 
Join Date: Jun 2000
Location: Surrey, UK
Posts: 4,452
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 16 Post(s)
Liked: 38
You're talking about dark adaption and then about what's the minimum you can see (100% against 99%), not the maximum under normal conditions which I was proving in my test.

A blind test of 100% and 90% is hardly a test of contrast perception, but rather a test of memory of light intensity perception. As it is, we need about 18% of the original luminance before we perceive what we think is a drop of 50%, so you're right, the difference between 90% intensity and 100% intensity in a blind test is hard to quantify since we don't see in a linear fashion. It doesn't prove anything with regards to the maximum range of contrast perception IMHO.

Gary

Quote:
Originally Posted by elmalloc
Who says Cameron is "right" and why do we care about him so much - lol!

I trust Gary Lightfoot more than James Cameron.
Gary Lightfoot is offline  
post #473 of 505 Old 01-13-2007, 05:02 PM
 
ChrisWiggles's Avatar
 
Join Date: Nov 2002
Location: Seattle
Posts: 20,730
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 14
Quote:
Originally Posted by armadillo View Post

Guys: I already gave you the spaceship example. So if you are in a dark-adapted state say at night and you look at the stars, what you see is the stars as tiny little dots hitting some of the rods on your retina. If you then add a star or two, it won't make a difference, because you won't move out of the dark-adapted state. So according to your theory, you will be able to see distinguish between the intensities of 1000s of different stars. You won't. And the same is true for the pointer. All you are doing is saturating a few photoreceptors. If the mouse pointer was made up of pixels of 99% and 100% brightness, would you be able to see those? Hell, if you did a blind test with a mouse pointer of 100% brightness and 99% brightness, would you be able to say which it was in the Gladiator image? That is the question. And the answer is NO.

I've given up trying to convince you otherwise.
ChrisWiggles is offline  
post #474 of 505 Old 01-13-2007, 05:31 PM
Member
 
armadillo's Avatar
 
Join Date: Jul 2005
Posts: 108
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Gary: It wasn't me who took the mouse pointer extreme. I'm asking you whether you would see intensity differences in the small bright spot and I'm saying, you won't, because the photoreceptors that are hit by the bright spot are saturated.

Chris: Wow. I never thought this moment would come
armadillo is offline  
post #475 of 505 Old 01-13-2007, 05:41 PM
 
ChrisWiggles's Avatar
 
Join Date: Nov 2002
Location: Seattle
Posts: 20,730
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 14
I'm going to go outside and look around with my eyes. Apparently, I have super-human abilities to navigate around in the dark on a snowy night with all these extremely bright streetlights and headlights in my view. It's unbelievable. I am so remarkably rare that I can do this. I am truly blessed with an incredible talent and ability. Clearly nobody else has these abilities because nobody else ever goes out at nighttime because we just can't see in those kinds of high-contrast situations, it's just too dangerous because regular people can't see at all! That's why no humans ever go out at night.
ChrisWiggles is offline  
post #476 of 505 Old 01-13-2007, 05:59 PM
Member
 
armadillo's Avatar
 
Join Date: Jul 2005
Posts: 108
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Chris: You said it yourself "I'm going to go outside and look around with my eyes." No further comment needed, because you simply cannot wrap your mind around the concept of adaptation.
armadillo is offline  
post #477 of 505 Old 01-13-2007, 07:09 PM
AVS Addicted Member
 
darinp2's Avatar
 
Join Date: Sep 2003
Location: Seattle, WA
Posts: 21,165
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 43
Quote:
Originally Posted by armadillo View Post

So according to your theory, you will be able to see distinguish between the intensities of 1000s of different stars.

You seem to be confused about the difference between the range and how many levels can be differentiated.
Quote:
Originally Posted by armadillo View Post

You won't. And the same is true for the pointer. All you are doing is saturating a few photoreceptors. If the mouse pointer was made up of pixels of 99% and 100% brightness, would you be able to see those? Hell, if you did a blind test with a mouse pointer of 100% brightness and 99% brightness, would you be able to say which it was in the Gladiator image? That is the question. And the answer is NO.

I covered some of this quite a while ago. For my testing I have at least two levels that are fairly bright, but not necessarily real close in intensity and try to count the transition point. And you seem to be stuck on using 2 levels that are very close (like 99% and 100%) as if it is your goal to come up with a smaller range so that you can claim a smaller range. You shouldn't be working so hard to come up with a smaller range by using examples that are more limiting if the goal is to find the upper limit. Use something more like 80% and 100%. Seriously, are you trying to find the upper limit or are you trying to come up with situations that are more limiting so that you can claim a lower limit?

--Darin

This is the AV Science Forum. Please don't be gullible and please do remember the saying, "Fool me once, shame on you. Fool me twice, shame on me."
darinp2 is offline  
post #478 of 505 Old 01-13-2007, 07:43 PM
Member
 
armadillo's Avatar
 
Join Date: Jul 2005
Posts: 108
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Darin: I am using the different levels to demonstrate that you cannot actually tell what levels they are. So if you say you have a CR of 10^5 then should be able to differentiate these levels, otherwise there is no point in increasing the image resolution. If you compress a jpg enough, you'll get posterization. So if your point is that high CR display can let you see a CR that is larger than say 300:1 you should be able to see all of the intensity levels and distinguish amongst them. I guess the point is that if we accept the elevator model (which I do) you can distinguish and see everything within the cabin. You can of course accept more or less brightness outside the cabin, but it will crush into black and white. If a photoreceptor is saturated or the light intensity is too low to be registered, that's it. You can only appreciate those other levels after the cabin moves into range.
armadillo is offline  
post #479 of 505 Old 01-13-2007, 08:01 PM
AVS Addicted Member
 
darinp2's Avatar
 
Join Date: Sep 2003
Location: Seattle, WA
Posts: 21,165
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 43
Quote:
Originally Posted by armadillo View Post

Darin: I am using the different levels to demonstrate that you cannot actually tell what levels they are. So if you say you have a CR of 10^5 then should be able to differentiate these levels, otherwise there is no point in increasing the image resolution.

No, if you say that people can see improvements beyond x:1 in one image, that does not mean that they can differentiate that number of levels, like the stars in your example. They aren't the same thing. Small differences can get crushed and become invisible. That doesn't mean that all high CR differences do. And I disagree that you have to be able to tell what level they are. Only that there was an improvement in an area where you wouldn't have been able to see an improvement if it had really gotten crushed.

And we aren't talking about improving the image resolution. We are talking about improving the image CR. One can have some affect on the other to some degree, but they aren't the same thing.
Quote:
Originally Posted by armadillo View Post

If you compress a jpg enough, you'll get posterization. So if your point is that high CR display can let you see a CR that is larger than say 300:1 you should be able to see all of the intensity levels and distinguish amongst them.

I thought we had been over this, but it doesn't mean they can distinguish any intensity levels that you want to come up with. Just like your 1% difference above. And it is difficult for us to differentiate between the levels in stars in a starfield. The things we can't differentiate there (like 2 stars not next to each other) don't mean we can't see some higher range in some other situation. Being able to distinguish steps that are close is about the smallest steps we can see, not the largest.
QUOTE=armadillo]I guess the point is that if we accept the elevator model (which I do) you can distinguish and see everything within the cabin.[/quote]I agree with the elevator model but disagree that you have to be able to see everything within cabin. It should be possible to have things within the cabin that are very close together that we can't differentiate, but that is different than what we can do at the top and bottom of the cabin and whether we could tell if the cabin got bigger, for instance.

Would you please address the test I posted way back in this thread in post #165 here.

http://mirror3.avsforum.com/avs-vb/s...&&#post9313612

I only used 4 steps because we don't have to be able to see x steps to see an improvement that is outside of a range of x:1. What if some visual system could only see 4 steps max (and sometimes things would look like 3). Would you say that that visual system couldn't see beyond 4:1 CR?

--Darin

This is the AV Science Forum. Please don't be gullible and please do remember the saying, "Fool me once, shame on you. Fool me twice, shame on me."
darinp2 is offline  
post #480 of 505 Old 01-13-2007, 08:23 PM
Member
 
armadillo's Avatar
 
Join Date: Jul 2005
Posts: 108
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Darin: I'll grant you that I am not 100% sure that we are absolutely incapable to see something outside the cabin (see p15 post 425 and following). I wish there was a reference and just how "sensitive" the area outside of the cabin might be. But I would say that our primary sensitivity is really quite limited and not exceeding 300:1. Also remember that our empirical tests are of limited value, because we have no control over pupils and saccades (subtle eye movements). So there might be truth in all of the posts here, but we simply cannot locate a reference that would clear it up for us. And once again, high CR projectors are very important no matter what. So please, don't put me in the camp that says, because of the limited instantaneous CR of the HVS we don't need higher CR displays.
armadillo is offline  
Reply Digital Hi-End Projectors - $3,000+ USD MSRP

User Tag List

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


Forum Jump: 

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

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