darin's contrast ratio thread - AVS Forum
Forum Jump: 
Closed Thread
 
Thread Tools
post #1 of 754 Old 09-27-2005, 12:40 AM - Thread Starter
AVS Special Member
 
dokworm's Avatar
 
Join Date: Sep 2003
Location: Grattai, Australia
Posts: 2,749
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 10
If only it would work on a projector - the images are well worth a look.

http://www.tomshardware.com/hardnews...23_170519.html

DON'T feed the troll.
dokworm is offline  
Sponsored Links
Advertisement
 
post #2 of 754 Old 09-27-2005, 03:12 AM
AVS Special Member
 
mp20748's Avatar
 
Join Date: Nov 2000
Location: Maryland
Posts: 6,339
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 13
Man, it's been over two hours since someone has posted the words "contrast ratio", and Darin has yet to chime in.

He's either on vacation, or there must be something wrong with the special device that alerts him of these posts.

It's all about the performance... Got Marquee!

 

High Performance Marquee Video chain modifications.  Now available!

mp20748 is offline  
post #3 of 754 Old 09-27-2005, 03:53 AM
AVS Special Member
 
kschmit2's Avatar
 
Join Date: Oct 2002
Location: Heidelberg, Germany
Posts: 2,224
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
$49k is all I can say
kschmit2 is offline  
post #4 of 754 Old 09-27-2005, 03:55 AM
AVS Special Member
 
techman707's Avatar
 
Join Date: Nov 2002
Location: Florida & New York
Posts: 9,348
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 20
Quote:
Originally Posted by mp20748
Man, it's been over two hours since someone has posted the words "contrast ratio", and Darin has yet to chime in.

He's either on vacation, or there must be something wrong with the special device that alerts him of these posts.

LOL - :p :D :p

"The greatness of a nation and its moral progress can be judged by the way its animals are treated." -Ghandi
techman707 is offline  
post #5 of 754 Old 09-27-2005, 10:13 AM
AVS Club Gold
 
Art Sonneborn's Avatar
 
Join Date: Jun 2000
Location: Battle Creek,MI USA
Posts: 22,289
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 4 Post(s)
Liked: 42
Yea very impressive. If you measure on off CR using direct measurements from the tube, a good LC CRT like the G90 is probably close to this as it is.

Art

My HT


iRule rules my theater
 

"If she's amazing she won't be easy,if she's easy she won't be amazing"

 

Bob Marley

Art Sonneborn is offline  
post #6 of 754 Old 09-27-2005, 11:47 AM
Senior Member
 
Outsider787's Avatar
 
Join Date: Jan 2005
Location: Toronto, Canada
Posts: 253
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
It's all great that we can make devices that have a huge contrast ratio, but what about the material to be displayed?
Currently, we can only capture 5-6 f-stops with either film or digital, which is likely the amount of dynamic range most display devices are currently capable of.

But I guess if the display industry starts the push towards more dynamic range, the recording industry will likely follow.

If it's worth doing, it's worth doing to excess!
Outsider787 is offline  
post #7 of 754 Old 09-27-2005, 12:12 PM
AVS Special Member
 
KenLand's Avatar
 
Join Date: Sep 2000
Location: Sugar Land, TX, USA
Posts: 3,628
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 13
Darin,

I think you are dismissing Outsider's line of reasoning too quickly.

I design and build scanners not cameras, but I'm sure there is some minimum brightness below which the movie cameras won't register.

I think it would be a useful exercise to determine the minimum light sensitivity of modern HD cameras and use that as a practical goal for projector black level. It's not going to be zero light no matter what the spec says.

Why spend a bunch of money reproducing what the best movie cameras can't capture?

I'll bet Wigggles knows this off the top of his head.

Ken
KenLand is offline  
post #8 of 754 Old 09-27-2005, 12:20 PM
Advanced Member
 
atomiccow's Avatar
 
Join Date: Jul 2005
Posts: 711
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
They must look better in person because my perception of the photos is that the high contrast display looks terrible. The wizard has poor color and the whites are blooming an insane level of blue.
atomiccow is offline  
post #9 of 754 Old 09-27-2005, 12:40 PM
AVS Special Member
 
KenLand's Avatar
 
Join Date: Sep 2000
Location: Sugar Land, TX, USA
Posts: 3,628
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 13
Just as a quick random google check. Here is the spec on a JVC 720P HD camera

http://www.smsprod.com/hdcinema/aj-hdc27f4.html

Minimum Illumination
0.7 lux (F1.4, +36 dB gain 59.94p)

So I'm assuming anything not illuminated to this level is lost. (going to be very noisy with 36 dB gain!)

How would this translate to black level at the viewing position of a FP displaying this camera's output?

Ken
KenLand is offline  
post #10 of 754 Old 09-27-2005, 01:03 PM
AVS Special Member
 
KenLand's Avatar
 
Join Date: Sep 2000
Location: Sugar Land, TX, USA
Posts: 3,628
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 13
Ok 16 is black and can vary. Obviously there is some limit to this. You wouldn't want 100 footlamberts of black most likely.

17 must be brighter than 16 and should be where the sensitivity of the camera is dovetailed into your brightness scale.

It seems unnatural for 16 to represent 0.000 lux and 17 to jump to .700 lux. (the min illuminance that will be descernable)

Maybe not, but I think its a useful discussion. Maybe thats exactly what we want, but although I can't find the words, it seems that there is some relevance that no information between 0.000 lux and .700 lux is ever present in the material.

Assuming for arguments sake that .7 lux is typical.

Ken
KenLand is offline  
post #11 of 754 Old 09-27-2005, 01:51 PM
Senior Member
 
lamestlamer's Avatar
 
Join Date: Jan 2003
Location: Seattle, WA
Posts: 342
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
CCDs used in astronomy can detect single photons. I don't know of any display that has that sort of precision. Of course, any pixel that recieves more than 2^16 photons will begin to blead into the surrounding pixels, resulting in the cross shaped glare that is so common in those beautiful 3-color images of deap space.

My guess is that they have a OLED stacked with a standard LCD. That would explain why the ANSI CR is lower (as seen in the white box's halo). If I wanted to drive my 24" CRT monitor as hard as possible, I bet I could get a similar image.

Michael
lamestlamer is offline  
post #12 of 754 Old 09-27-2005, 02:06 PM
AVS Special Member
 
deronmoped's Avatar
 
Join Date: Apr 2001
Location: San Diego, Ca.
Posts: 3,483
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
The guy doing the transfer owns a LCD PJ and hates the fact that CRT's can do absolute black, so every chance he gets he sets the threshold for absolute black at what his LCD can do.

There are just mean people out there:)

Deron.

CRT Videophile
deronmoped is offline  
post #13 of 754 Old 09-27-2005, 02:09 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 KenLand
It seems unnatural for 16 to represent 0.000 lux and 17 to jump to .700 lux. (the min illuminance that will be descernable)

Maybe not, but I think its a useful discussion. Maybe thats exactly what we want, but although I can't find the words, it seems that there is some relevance that no information between 0.000 lux and .700 lux is ever present in the material.
I understand that, but here is where a person converting from film to digital (doing the telecine) can dither between adjacent pixels with 16s and 17s to get values that our eyes will see as levels between them and probably one reason you want a person with some intelligence and skill doing the telecine. And because ANSI CRs aren't perfect the 16s in mixed scenes won't actually end up getting zero light even if the on/off CR of a projector was infinite. I could see some day in the future where a display device has so much ANSI CR that some people complain that the blacks are too black in many mixed scenes, but I don't think we are there yet and it would also take screens or rooms that could hold super high ANSI CR.

Quote:
Originally Posted by deronmoped
The guy doing the transfer owns a LCD PJ and hates the fact that CRT's can do absolute black, so every chance he gets he sets the threshold for absolute black at what his LCD can do.

There are just mean people out there:)
This is kind of funny, but it is somewhat frustrating that things near black are encoded wrong in so many cases for projectors that can go very dark. While this work is generally done on CRTs, they are CRT monitors (not front projectors) and my understanding is that there is light in the room and I believe the monitors are calibrated to between about 32 and 36 ft-lamberts. Much different than what I believe most people here have for their setups.

--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 #14 of 754 Old 09-27-2005, 02:46 PM
AVS Special Member
 
KenLand's Avatar
 
Join Date: Sep 2000
Location: Sugar Land, TX, USA
Posts: 3,628
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 13
Darin,

But there is no information between 16 and 17 if 16 and below is reserved for the blackest black and 17 is the first value the camera can see.

So the perfect projector displaying real material from our real camera would jump from 16 = 0.000 lux to 17 = .700 lux.

There is no infomation below .7 lux. You can shift it all down to a certain extent, but you're not going to want the camera's .7 lux output to map to 0.0 lux.

So it's debatable how far below first light you need or want to go.

Along the same line. This HD camera will measure up to 2000 lux, so max CR is 2000 / .7 = 2857:1

So if this were a typical movie camera then movie content would contain no more than 3000:1 On/Off contrast.

I think this information is useful for setting projector performance goals and calibrations. If you know your content has 3K:1 CR do you want to expand it to 15k:1?

I'm not sure.

Ken
KenLand is offline  
post #15 of 754 Old 09-27-2005, 04:41 PM
AVS Special Member
 
KenLand's Avatar
 
Join Date: Sep 2000
Location: Sugar Land, TX, USA
Posts: 3,628
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 13
Do whatever you want with 16 (and below) and 235 (and above), but the camera can only capture .7 to 2000 lux and has 17-234 to map the measured levels. Or perhaps 16-235.

Whatever the brightness you pick to map the dimmest detail possible to it was .7 lux in reality. I think its fine for black or below black to dip below this, but to say it has to expand the CR from the signals maximum of 3K:1 to 200k:1 is not a good idea unless you know what you're doing.

If you for instance make the bottom of your gamma curve 0.0000 fL then you will loose shadow detail or at least severely distort the scene captured by the camera, because anything above absolute black is at the equivalent of .7 lux.

Basically its not a good idea to map your whopping 220 possible levels to onto a 200K:1 CR. That would be an incredible gap between any two adjacent values.

A better setup would map the real 3K:1 in the signal onto say 17-234/235 and reserve 16 for the blackest black your projector can muster.

Ken
KenLand is offline  
post #16 of 754 Old 09-27-2005, 05:12 PM
AVS Special Member
 
cmjohnson's Avatar
 
Join Date: Nov 2003
Location: Sharply focused on sharper focus.
Posts: 5,972
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Interesting. If I interpret the description correctly, what they have there is actually a stacked display. The LED backlight is a full color active matrix display in and of itself,
or at the very least it's a monochrome (black and white) active display, if I interpret this right,
and in front of it is the LCD assembly. So the backlight is a variable light source on a per-pixel basis.


Brute force thinking. About as subtle as a hand grenade in a barrel of oatmeal.

I've thought of that one before. A pity I didn't have any way to turn it into a real idea.

CJ
cmjohnson is offline  
post #17 of 754 Old 09-27-2005, 06:15 PM
AVS Special Member
 
KenLand's Avatar
 
Join Date: Sep 2000
Location: Sugar Land, TX, USA
Posts: 3,628
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 13
Darin,

"On/off CR is video 235 to video 16, not video 235 to video 17. That is a different matter and if 235 to 17 is 3k:1 it does not mean that 235 to 16 could not be 200k:1."

Ok, we're in agreement. Every value other than the two extremes should be in the neighborhood of reality.

My jury is still out on whether you want such a large gap between 17 and 16.

If 235 is 12 fL then 17 is .004 fL and 16 is .00006 fL. (3K:1 and 200K:1)

I'm not sure how that would look but its has to be better than crushing your blacks by anchoring your gamma curve at .00006 fL.

Ken
KenLand is offline  
post #18 of 754 Old 09-27-2005, 06:24 PM - Thread Starter
AVS Special Member
 
dokworm's Avatar
 
Join Date: Sep 2003
Location: Grattai, Australia
Posts: 2,749
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 10
Yeah we have touted an 'active backlight' a few times in these threads, and I have seen a product using and active backlight (not this one) and the effect is astoundingly good. I'd put money on this screen outperforming CRT monitors. You need to see this gear in person as the average digital camera can't record it.
Yup it is 50 grand, but people paid that for a plasma about that size only a few years ago, so it isn't out of reach to the high end guys.


Just on the camera issue, there is a bit of confusion here I think. The lux measurement isn't the cameras 'black level'.

The low lux measurement for the camera means that is the lowest level of light it can detect. If the light level drops below that then the camera 'see nothing' and records black.

i.e. it doesn't mean that if a camera is rated to 0.7lux and you record a completely black room (while sitting under a black velvet lead-lined blanket) that you get a recording of 0.7lux - you get pure black i.e. 16 for NTSC in the US, or 0 for the rest of the world (most cameras just record zero anyway and it gets stepped up to 16 later by an output device)

What is does mean is that any light in the room under .7lux will be recorded as absolute black, so if you had a bunch of very dark areas say .3 lux areas and .5lux areas and .6 lux areas you wouldn't see them in your recording - they would just all be black.

DON'T feed the troll.
dokworm is offline  
post #19 of 754 Old 09-27-2005, 06:34 PM - Thread Starter
AVS Special Member
 
dokworm's Avatar
 
Join Date: Sep 2003
Location: Grattai, Australia
Posts: 2,749
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 10
Plus you guys are missing a vital point, this is not an 8 bit display, so you are not stuck with the 16 to 235 steps (which is only valid for US NTSC anyway, as a monitor in 8bit you get 0-255)

You get a massive increase in bit depth with a display that can go higher than 8bit, so instead 256 available levels of R, G or B, in a 16 bit display you get 65,536 levels per RGB color channel.
i.e. instead of a 16,777,216 colourspace in 8bit per channel RGB, you get a 281,474,976,710,656 colourspace with 16bits per channel.

This is why film still tends to look so good, the colour depth far outshines any current CRT devices. It is also why in the film industry we work in at least 12bit colour when manipulating film digitally before recording it back to film again.
If anyone is interested in HDR stuff, there is a nice link page on their site.

http://www.brightsidetech.com/tech/resources.php

DON'T feed the troll.
dokworm is offline  
post #20 of 754 Old 09-27-2005, 07:21 PM
AVS Special Member
 
KenLand's Avatar
 
Join Date: Sep 2000
Location: Sugar Land, TX, USA
Posts: 3,628
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 13
I would like to see this display and talk to them about calibration and setup.

No one said that the camera's minimum light sensitivity was its black level. Only that you don't want to waste too many levels between this and black, because the camera has recorded no, none, nada, zilch information at those levels.

Ken
KenLand is offline  
post #21 of 754 Old 09-27-2005, 09:47 PM
AVS Special Member
 
Ohlson's Avatar
 
Join Date: Oct 2001
Location: Falun, Sweden
Posts: 5,434
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
I am sure you know this but the leds in this monitor is far fewer than the lcd pixel resolution. I do not know or remember how mane led regions there are in this display. It works out well since our eyes tend to smooth things out. The boundary effects that I am sure are minimized by the technology is further reduced by our eyes.

Mattias Ohlson
Ohlson is offline  
post #22 of 754 Old 09-27-2005, 11:47 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:
Plus you guys are missing a vital point, this is not an 8 bit display,
But presumably the source video is 8-bit, so it doesn't really matter that much what the bit depth of the display is as long as it's, what, basically 12bit for a linear display?
ChrisWiggles is offline  
post #23 of 754 Old 09-27-2005, 11:56 PM - Thread Starter
AVS Special Member
 
dokworm's Avatar
 
Join Date: Sep 2003
Location: Grattai, Australia
Posts: 2,749
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 10
No, the source is not 8 bit, that is the whole point of one of these displays, they are for people who need to use HDR imagery.

Nothing to stop you talking to them Kenland, their number is on their website.
With a greater than 8bit display, the number of levels available are so huge, that the problem of the number of steps wasted between absolute black and the cameras lowest light register becomes a non-issue. There are plenty of increments left over in a 16bit per channel system :)

DON'T feed the troll.
dokworm is offline  
post #24 of 754 Old 09-28-2005, 12:01 AM
 
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
Well, then I guess I take that back, I figured it would be used with more standard video ;).

Quote:
There are plenty of increments left over in a 16bit per channel system
mmmm....16 bits per channel..... :)
ChrisWiggles is offline  
post #25 of 754 Old 09-28-2005, 02:24 AM
AVS Special Member
 
KenLand's Avatar
 
Join Date: Sep 2000
Location: Sugar Land, TX, USA
Posts: 3,628
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 13
Yeah, 16 bits per channel :)

Ken
KenLand is offline  
post #26 of 754 Old 09-28-2005, 11:10 AM
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 dokworm
No, the source is not 8 bit, that is the whole point of one of these displays, they are for people who need to use HDR imagery.

Nothing to stop you talking to them Kenland, their number is on their website.
With a greater than 8bit display, the number of levels available are so huge, that the problem of the number of steps wasted between absolute black and the cameras lowest light register becomes a non-issue. There are plenty of increments left over in a 16bit per channel system :)
Do you know how they can get more than 8 bit data into the display? I see DVI, but that is limited to 8 bit.

--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 #27 of 754 Old 09-28-2005, 11:26 AM
AVS Special Member
 
cmjohnson's Avatar
 
Join Date: Nov 2003
Location: Sharply focused on sharper focus.
Posts: 5,972
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
The idea behind crushing blacks is very simple: To have a PERFECT black level, that is, no light output AT ALL, the very next displayable step in black level will certainly be too dim to be seen.

To get a fully visible dark grey scale, you HAVE to settle for a very slight elevation of the total black level. And this is quite acceptable, if you try it.

I'm finicky about black levels, myself, but have found that I can accept a slightly elevated black level in favor of dark grey scale performance more easily than I accept a dead black
black level but at the cost of deepest grey scale resolution. It's more important to me
that I see all that is there in the picture.

CJ
cmjohnson is offline  
post #28 of 754 Old 09-28-2005, 12:20 PM
AVS Special Member
 
cmjohnson's Avatar
 
Join Date: Nov 2003
Location: Sharply focused on sharper focus.
Posts: 5,972
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
But to get the full benefit of that, you'd need to be in a viewing room that is otherwise COMPLETELY dark. No light AT ALL except from the projected image. ANY other light in the
room is going to result in your ultimate black level being slightly washed out.

I'm not sure that that's even a good idea, to try for an utterly dark viewing room.

Also, consider this: The lower limit of how little light you can perceive may be different than
what I can, for physiological reasons. As we get older, the lenses in our eyes lose a bit of their transmissivity, which is a prime cause of night blindness. And then there's the
thing known as the cataract, in some people. So, the ideal setup will STILL have to be
very subjective. Eyes aren't uniform in their performance.


CJ
cmjohnson is offline  
post #29 of 754 Old 09-28-2005, 03:10 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:
REC.709 calls for an absolute black level that is darker than your 200k:1 level (since it calls for no light)
Darin: I'm looking at 709, and I'm not seeing quite where it says this except if you're getting this from where it describes the gamma function+tail. In which case, due to my having sworn off mathematics, could you show the calculations to get to what you're showing? :)
ChrisWiggles is offline  
post #30 of 754 Old 09-28-2005, 04:15 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:
but doesn't the gamma+tail start at zero?
I think so, but it doesn't really spell it out that I see, just that 0 is less than or equal to luminance less than or equal to 1. So I guess you'd intepret that to say that black should be 0 luminance, though in reality I don't think it ever is on anything that's calibrated. I definitely agree that ideally that's what we should be striving for.

Anyway, that wasn't really my Q, I was interested in the CR numbers. Thanks for the link, I think I glanced over that quickly before.

I'm curious though, why hasn't anyone put out a LUT for a PC for instance with a linear tail, or even what you were talking about quite a while ago of pushing all 16 values down to say 0 or something. I'd be interested to see what would happen on a CRT.
ChrisWiggles is offline  
Closed Thread CRT Projectors

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