how to benchmark brightness?

[js09]

I got a new 27" cinema display and it is spectacular. I'd like to know if there is any software I can get to test the brightness levels, so I can monitor the display over time and see if it fades at all.

I'm in the market for a color calibrator (spyder 3) so perhaps this device can do that as well, I haven't had a chance to research it yet.

thanks

 

[yowen]

I doubt any software knows how bright your monitor is, unless something is communicating that information to your PC somehow, like the calibrator maybe.

 

[Ruahrc]

One method probably could be to take a picture of a white screen using standard/known camera settings. Like 0.5sec and f/8 at ISO 100, which would give you a consistent light exposure with time and you could then analyze the brightness by reading off the rgb values of the screen in the photo.

Theoretically, it should be possible to correlate this to real brightness, I.e. If your monitor is at 100cd/m^2, it would come back as rgb=160 or something. However, there are slight deviations in the actual ISO of cameras compared to the rated settings that it would not be strictly accurate.

But, the best way is simply via colorimeter.

 

[js09]

thanks Ruahrc. I'm a professional photographer so I'll go ahead and do that benchmark

 

[Mr.Pixel]

Keep in mind that you'll want to compare photos in a linear space, which means standard output formats are a no-go because they have gamma correction applied. You will need to either compare raw data values directly, or try to apply an inverse gamma curve before comparing.

 

[js09]

Keep in mind that you'll want to compare photos in a linear space, which means standard output formats are a no-go because they have gamma correction applied. You will need to either compare raw data values directly, or try to apply an inverse gamma curve before comparing.

shooting raw with manual settings will not yield any gamma correction. the main issue is shooting in a controlled environment that can be replicated at a later date

 

[Ruahrc]

Well the point is, if you use the same camera settings every time, the "exposure" will be the same, and so you should be able to track monitor brightness with time.

To a certain extent this method can actually be used to compare monitor brightness between different people using different cameras/computers, since f-stops are f-stops and ISO is ISO. And you could theoretically make a chart that correlates RGB values on a white screen to luminance levels so that someone without a colorimeter could set the correct/desired monitor brightness by adjusting the backlight until a standard picture taken with a standard exposure produces the correct RGB values in the resulting photograph.

But, as I mentioned there are slight deviations in "actual" vs. "set" ISO, as well as each lens has a slightly different t-stop even though they may be used at the same f-stop. Also, you'd have to be very rigorous to factor out camera processing, etc. in order to keep things comparable. Shooting in RAW and processing with a linear response would be the way to go with this.

 

[Mr.Pixel]

shooting raw with manual settings will not yield any gamma correction.

The raw file itself is linear (or very close), but after using a developer like Camera Raw the resulting file usually has a gamma of 2.2 (depending on your output color space and settings). Some developers can output the actual raw values, like DCRaw.

Like Ruahrc said it's hard to compare camera-to-camera, but using the same camera under controlled conditions should give you a good idea if anything is changing. The gamma used also doesn't matter if all you interested in is detecting a change, and not how much.