On May 20, 2005, at 11:45 AM, Kenneth Waller wrote:
Don't know what that means. Either the exposure is right or it isn't. His image didn't appear to have the correct exposure as posted. His histogram seemed to agree with that. There were white shirts in the image and I assume there were black features in there somewhere. I also assumed that he wanted the whites white and the blacks black. If the exposure was properly captured, the whites should look white and the blacks should look black. I don't understand how the capture mode (Raw, JPEG, Tiff etc) has anything to do with proper exposure at time of capture.Don's exposure is pretty darn close for RAW
Can you educate me?
The sensor sees light in a linear gamma space, not as your eye or film sees it. The brightest f/stop's range is 1/2 the total quantization space, etc, so the RAW format file, pre-correction, tends to have most of the interesting values down low on the scale. Processing the RAW data properly is the trick ... RAW data is NOT a rendered image in RGB.
If this scene was processed to this appearance by the default values set in the RAW metadata, it could have stood another stop or so of exposure. Then you use the RAW converter to correct and redistribute the tonal curve into the proper space. It's not too bad though, as it has plenty of data to work with in the dark areas. The several bright areas and possibly the meters response to the difficult lighting mix contributed to fool the meter a bit.
I'd suggest, in the future, making a couple of exposures and checking them on the in-camera display histogram then tweaking the EV compensation setting to compensate if you want to use matrix metering automation. But as it is, you have a good, usable exposure to create a photograph with. It needs color correction and tonal correction, that's all.
Godfrey
