Thanks for the explanation. It makes sense. It is a bit tricky because what I am hearing is that I should try to expose so that I don't blow the highlights, but just barely. So in the case of this photo, which visibly is about where I wanted it, I should have visibly over-exposed it (but not based on histogram) and then during Raw conversion, underexpose it. This would then give me the smoother mid tones that are not present now. Does that sound right?
Basically, during exposure the entire histogram should have been shifted to the right as far as possible. Then during raw conversion, shift it back to the middle like it is now. This would result in a much cleaner image than if I had just exposed it where I wanted it to visibly be. Another way of putting it, perhaps. Feel free to jump in, anyone. -- Best regards, Bruce Friday, August 5, 2005, 10:21:19 AM, you wrote: GD> The histograms show that your exposure isn't anywhere near the GD> highlight saturation level, indicative of underexposure. If you had GD> given the photo another stop or so of exposure, you could have GD> adjusted the RAW converter's gamma curve to allow a much better GD> expression of mid- and low-range values (zone II to IV) without the GD> color noise when doing the conversion from linear gamma to RGB. GD> Jack's rendering shows the noise buildup that I see very clearly, a GD> little bit more exaggerated than simply using Levels or Curves GD> adjustments without masking and feathering like I was doing. GD> Remember that in linear gamma terms, HALF the quantized tonal space GD> is within 1 stop of the overexposure/saturation limit, the next GD> QUARTER of the total tonal space is within the next stop. In an image GD> which should have many tonal variations in the Zone II to IV range, GD> you need to give enough exposure to shove the remaining 1/4 of the GD> tonal values around in RGB conversion while minimizing noise to get a GD> smooth effect. GD> Imposing clipping limits on the 8bit RGB image to extend the tonal GD> space to Zone I and Zone IX the way it ought to be demonstrates the GD> fact that the image is underexposed by revealing the Zone II-Zone V GD> mid-range noise. You might be able to reprocess the RAW file with a GD> better set of adjustments to reduce that underexposure, since your GD> RAW file has quite a bit more data in it than this JPEG rendering. GD> Godfrey GD> On Aug 4, 2005, at 11:39 PM, Bruce Dayton wrote: >> I'm a bit confused. I just looked at the raw histogram with two >> different programs and neither of them show anywhere near >> underexposure. I agree that the histogram doesn't come near the black >> or white clip points. So how could it be underexposed? I'm hoping to >> learn something here for future use. >> >> Here is a link to the histograms. The first is from Capture One, the >> second is from BreezeBrowser Pro. The way I presented it was without >> any exposure compensation. >> >> http://www.daytonphoto.com/junk/index.htm >> >>> I'd have been more comfortable with this if it didn't seem >>> underexposed. The histogram in this one shows little in way of blacks >>> and whites, and clipping it with Levels reveals all the underexposure >>> noise. >>> >>>> This is the last shot of Valley of the Gods that might be worth >>>> showing. Since you all have been kind enough to give me your >>>> feelings >>>> on "The Pyramid", hopefully you will do so here. >>>> This shot uses the haze as a way to convey shapes and distances. >>>> Let >>>> me know what you think. >>>> http://www.daytonphoto.com/PAW/monumentvalley_0479a.htm
