F-stop comparison using stills from footage?

[John Harvey]

Is there an accurate way of measuring the total luminosity of an imported still in Photoshop vs a differant imported still for comparison, then converting that measured differance into an F-stop number?

I guess I could shoot a white balanced board, then stop down at each increment taking note. Then import a still from each stop down and measure the total luminosity of each in photoshop and have a base line for future comparisons? Then this could be used to evaluate DV stills from between various scenes, and lens attachements?

I think I just answered my own question, but correct me if the logic is flawed.

thanks

jh

[Glen Hurd]

If you shoot a card with a single shade/color (preferably an 18% gray) in a flatly lit environment, the "total luminosity" will be easier to determine, since there should be very little difference in the luminance values across the frame.
Instead of shooting "various scenes" with the DV camera, just reshoot the same gray card in the same lighting, and take frames from each iris setting into Photoshop for comparison to the the reference camera. As you swap out lenses and attachments, this could be a way of evaluating your DV camera's exposure behavior.
The minute you start shooting scenes as a means of evaluating exposure, things get very complicated, because different cameras shooting identical scenes at identical fstops will record different values of a scene due to different contrast/gamma settings (or film characteristics), as well as due to differences in exposure latitude.
I suspect you are really looking for a way of monitoring changes in light transmission, so you'll probably want to keep the scene to being a simple card, during the evaluation process. As you know, 18% Gray is the visual half-way point between white and black, and easier to evaluate when it's clipping at either end of a camera's exposure latitude.
Hope this helps in some way,
G