Blacks look nice and black for test patterns, animation, and well lit film sources. Blacks appear gray/green for low light film sources. Is this a grayscale problem or is this just what film does with low light? Anyone else see this?

Thanks!!

Without seeing it firsthand, it definitely sounds like a greyscale problem.

Sounds to me like its a content problem. What movies are you seeing this on? Is it just one or two, or many?

If test patterns don't show this problem, I'm tempted to say that it's not a system problem and suggest it's just a content problem. Because you say it also looks fine on other movies. What specific movies does it look wrong on? It's not like the matrix or anything is it...?

Specifically noticed in National Treasure (opening attic scene) and One Night with the King (various parts throughout).

Grayscale has been calibrated with a SpyderII and Color HCFR. Red and Green track nicely from 20 - 100 IRE. Blue has a hump through the lower IRE's and comes into line around 70 which is a known and uncorrectable issue with my set (Hitachi 51swx20b). All my measurements below 20 are erratic and unlreliable (which I attribute to the Spyder) but I would presume the trend shown on the chart would continue (roughly) from 20 down to 0.

Color decoder has been set with Avia color bars. Primaries are quite good (all points on the CIE chart touch or nearly touch the targets).

I would attribute the gray/green I see to the blue hump but the what I see is more green than blue...like underdeveloped film.

That combined with the fact that blacks in animation and test patterns look really good makes me think its a content problem as well but I wanted to see if anyone else noticed similar things with film based material.

I don't know jack, but is the SPYDER used for service menu calibration?

I have a TV that was professionally calibrated; the factory settings were so lousy that on one of the test patterns for GREYSCALE there was very noticable red colors. Yes, on GREYSCALE. It could be a similar issue, though I'm sure one of the many other posters much more experience than me at this kind of stuff, I thought I'd just throw it out there.

AFAIK the spyder can be used for service menu calibrations (specifically greyscale) but is known to have issues correctly reading low IRE's which is why I can't measure to verify my suspicions.

N.C. timestamp 9:19:55 AM

AFAIK the spyder can be used for service menu calibrations (specifically greyscale) but is known to have issues correctly reading low IRE's which is why I can't measure to verify my suspicions.

not at all true

So do you mean to say the Spyder is accurate at all IRE levels? When I do my measurements, everything below 20IRE on the RGB tracking grid is completely erratic - red spikes WAY up off the grid and green spikes down off the grid and it won't even draw the blue line - what could be causing this?

er...no meter is accurate at 0-20 IRE (depending on the display type).

er...no meter is accurate at 0-20 IRE (depending on the display type).

Which part of my statement above was not at all true then? If no meter is accurate at 0-20 IRE, how can I accurately measure 0-20 IRE?

Thank you for your help BTW :-)

0-20 IRE is not the correct way to specify the lower limit. It should be in luminance or illuminance numerical values. Some devices will fail to accurately measure color below 1 fL others can go down to 0.03 fL. It depends on the quality of the device in question.

0-20 IRE is not the correct way to specify the lower limit. It should be in luminance or illuminance numerical values. Some devices will fail to accurately measure color below 1 fL others can go down to 0.03 fL. It depends on the quality of the device in question.

You can further break this down into component vs. total luminance for the typical tristim meters. For example the spyder2 components (X,Z) are pretty noisy below about 1 cd/m^2 but the Y (total luminance) values are decent down to about 0.1 cd/m^2 (if you are patient). I've gotten decent measures with the displayLT down to ~0.04 cd/m^2. For my display 0.1 cd/m^2 is 0.1% stimulus. Btw, HCFR incorrectly lables it's gray scale measurements as "IRE" when in reality the units vary depending on which display option is chosen, RGB, xyY, etc.

How do I correctly measure the bits on the left side of the chart that are blacker than the bits on the right side of the chart?

Im2_mixed^

How do I correctly measure the bits on the left side of the chart that are blacker than the bits on the right side of the chart?

Im2_mixed^

With a CRT it is unlikely you would ever get a good measurement at 0 IRE. The first thing to realize is to not be a slave to the instrument. Trust your eyes and look at some test images including gray ramps, color bars and reference video images.

With the instrument the first thing would be to look at the numerical values and see if the measured light values are within the instruments minimum limits. Based on these values and visual images is your gamma curve too steep or your brightness too low. If either of these are wrong fix them first. Then remeasure and see where the luminance values fall for 0 to 20 IRE. It is likely you should be able to measure 20 IRE in most cases. 10 IRE is a stretch and 0 IRE is highly unlikely for your technology.

I would look at a different instrument or use your eyes as an optical comparator for the dark readings. Once you have the brighter segments calibrated well it is not terribly difficult to judge the darker areas by eye. You will need to use your eyes for 0 to 10 IRE and possibly 0 to 20 IRE in all likelihood.

If you want to pursue other tools my Philips PM-5639/00 goes down to 0.03 fL well. My Eye-One Pro with AccuCal software goes down to about 0.2 fL with reasonable accuracy. Other instruments are likely to fail near the 0.2 to 1.0 fL mark depending on the technology in question. If you want to spend some serious change a PR-670 will go down to 0.07 fL.

I would look at a different instrument or use your eyes as an optical comparator for the dark readings. Once you have the brighter segments calibrated well it is not terribly difficult to judge the darker areas by eye. You will need to use your eyes for 0 to 10 IRE and possibly 0 to 20 IRE in all likelihood.

This sounds like the right apporach for me...red and green are almost touching the 100% line from 100 down to 20. Blue is on the 100% line at 100 IRE and humps up in the middle but touches the 100% line at 20 IRE again. This blue hump is a known issue for my Hitachi 51SWX20B.

I'll just put up a 10 IRE window and eyeball the RGB cuts as I'm not prepared at this time to invest in higher end gear.

Thank you all for your advice!

You can further break this down into component vs. total luminance for the typical tristim meters. ....

My comments are only related to color measurements. Luminance or illuminance in the case of the Spyder is a little easier task.

My comments are only related to color measurements. Luminance or illuminance in the case of the Spyder is a little easier task.

yes, that is exactly what I have found. Measuring 10% stimulus for gamma correction is not a problem with the spyder but balancing RGB there is a stretch and I end up doing it by eye via the ramp.