[darinp2]

I'm starting this thread with a video I made to show an aspect of contrast ratio with front projectors, but figured others might be interested in adding their own videos or pictures to demonstrate other things related to contrast ratio.


In the past there have been some good threads here (IMO) about contrast ratio with graphs and analysis, such as this one by stanger89:

http://www.avsforum.com/forum/24-di...s-apl-study-adjunct-avs-contrast-project.html



I decided to make a video to try to demonstrate something I have been showing in person for a long time. It probably works better in person, but it seems like it works okay in video form too.


I get the impression that there are still people in the industry who believe that ANSI CR tells us the black floor for a system and this video is related to that. In person I try to show people that the black level they are getting when the ANSI checkerboard is displayed with a projector would not generally make a good black floor for the system, even though it can look dark while that checkerboard is displayed.


I posted the video to youtube, since that seemed like a good way to make it easy to access:

If you can't hear the audio, basically what I did was show an ANSI checkerboard, then covered all of image from the projector except for the center of one of the "black" rectangles, then switched the display to an all black pattern (still blocking most of the screen), then took the black velvet board away to show the whole black image (which the camera on my iPhone6 couldn't pick up).

--Darin

 

[darinp2]

This is a set of 3 pictures of a low APL 4x4 checkerboard with my room setup in 3 different ways, where I have added measured values to each rectangle. The "black" rectangles look more gray or more dark in the 3 images mostly because they really were brighter or darker.

The checkerboard used is video level 52 and video level 16, which results in around 1% Average Display Luminance (ADL) after a reasonable display gamma is applied. If people look at the analysis for some different movies I think they will find that images close to this darker or darker are not that uncommon. For instance, in Raiders of the Lost Ark it looks like close to 20% of the images are around 2% ADL or lower, based on some analysis by stanger89.

This is annotated on the images, but the room conditions as setup resulted in the following measurements off the screen for system ANSI CR and system sequential (on/off) CR:

A: 50:1 system ANSI CR and 74:1 system sequential CR
B: 84:1 system ANSI CR and 160:1 system sequential CR
C: 44:1 system ANSI CR and 2100:1 system sequential CR

For this 1% ADL checkerboard the intra-image CR I measured for each case was about:

A: 3:1
B: 5:1
C: 22:1

--Darin

 

[darinp2]

I took a look on YouTube to see if there were any good demonstration videos for contrast ratio that I should link to here.

I see videos that clearly had a fair amount of effort put into them, but I was pretty disappointed in the messages overall.

The videos I found range from claiming that people should just pick the display with the highest number for CR from the manufacturer to multiple videos essentially telling people to just use ANSI CR because it includes both black and white at the same time.

There are some people from Australia who seem very proud that others repeat their claim that if you have even 1 candle's worth of light in your room then 5000:1 isn't any different than 500:1. I wonder if any of these people ever test for themselves before repeating claims.

I think one of the people is the same one I recall would claim each year that if the manufacturers gave us better on/off CR the Human eye wouldn't even be able to see it, and this was back when projectors had less on/off CR than now. I'm thinking 5000:1 or less native from the Epson's when this guy was making those claims.

I realize that CR can be a complicated subject, but I think some people from here could make some better videos than what I found.

--Darin

 

[Wizziwig]

Which projector is being used in that video and is a dynamic iris involved?

 

[Kelvin1965S]

Nice to see some 'science' on this forum again, rather than just the other 's' word. It is a more complex subject than some seem to think, especially whoever from 'down under' who think that we don't need more contrast as we can't see it! If that were true I'd still have my first projector...

Just choosing any display based on a claimed (or even measured) contrast figure seems a bit silly to me, so I'll still finalise any purchase with my eyes. However, having an understanding of what does and doesn't affect the image quality wrt contrast is certainly useful. It's also more interesting reading than someone just offering to sell a product.

 

[darinp2]

Which projector is being used in that video and is a dynamic iris involved?

The video was done with a JVC RS20 (which doesn't have a dynamic iris). I believe I had the manual iris open and based on measurements from a while ago I expect that the projector was probably doing around 20k:1 and 30k:1 sequential CR. I could see the projector lighting up the screen in person with the all black image, but the camera didn't seem to pick it up. Maybe a different camera would have.

The pictures with 3 different room setups were done with a Panasonic AE8000 with the dynamic iris disabled and a Panasonic AE100 acting as a secondary light source for the top two images (to represent lights on or light coming into the room situations).

--Darin

 

[Dominic Chan]

I decided to make a video to try to demonstrate something I have been showing in person for a long time. It probably works better in person, but it seems like it works okay in video form too.

To use a video to demonstrate the checkboard black (or lack thereof), the video camera must have a manual exposure control that is fixed when taking the comparison pictures. With automatic exposure control, the camera would try to brighten the black square when the rest of the image is covered.

 

[darinp2]

To use a video to demonstrate the checkboard black (or lack thereof), the video camera must have a manual exposure control that is fixed when taking the comparison pictures. With automatic exposure control, the camera would try to brighten the black square when the rest of the image is covered.

How is the automatic exposure control different than what the human eye does, which is what I really want to demonstrate, since that is what I can demonstrate in person?

If I made the camera keep the "black" part of the image at say encoded level 0 or 1 both when the full checkerboard is shown and when only the center of one of the darker rectangles is shown, then that would not demonstrate how the human eye works when I do this demonstration in person. While not perfect, it seems to me that what I showed better demonstrates what happens in person.

The effect of the dark part of the image looking brighter was even more pronounced in person than what I see in the video. In person the center of the dark rectangle looked pretty bright. Definitely wouldn't be a good black floor.

If I am missing something here please let me know what that is.

Thanks,
Darin

 

[George Kouzev]

Best approach is in the middle. Many cameras have auto exposure algorithms that try to capture an object as if it was lit well, even when it is very dim. They will likely do a combination of lowering the video shutter speed and bumping ISO. In any case, the object may look like it was in daylight while your eyes will see differently, even after you get used to the low light. So having manual exposure helps.

To Darin's point though, I would tweak the manual exposure to match what my eyes see for each pattern, assuming the LCD/vewfinder is true to what is being recorded.

 

[Ericglo]

DP,
Did you ever get an OLED? I would be curious to see how the theoretical ultimate ANSI and on/off cr of the OLED would compare to a projector.

 

[darinp2]

Did you ever get an OLED?

I haven't. Still tempted though.

--Darin

 

[Ericglo]

Tiger Direct is closing some of their stores, so I decided to stop by there last week. I was with my family and told them if it was under $2K then I was buying it. Unfortunately, it was $2600.

I think I might have actually asked my brother and sister to go in with me and give it to my dad for his bday. Next year it will probably be $1500, but I may grab one before then.