Histogram response for "Sunny f/16" exposure

[scottbel]

If I shoot a clear blue (cloudless, hazeless, pollution-free, etc.) north sky at mid-day to attempt to determine what my optimum "Sunny f/16" exposure is for ISO 200, what should I be looking for in the in-camera histogram? Do I want to see a tall spike centered in the middle of the scale?

(I performed this test with a Nikon D200 with 70-200mm zoom @ 200mm, focused at infinity at a 45 degree angle to the horizon, with nothing in the frame but the sky. The ISO 200 exposure that seemed to peg the histogram spike in the middle was either 1/100 @ f/16 or 1/125 @ f/16.)

[digitaldog]

If I shoot a clear blue (cloudless, hazeless, pollution-free, etc.) north sky at mid-day to attempt to determine what my optimum "Sunny f/16" exposure is for ISO 200, what should I be looking for in the in-camera histogram? Do I want to see a tall spike centered in the middle of the scale?

(I performed this test with a Nikon D200 with 70-200mm zoom @ 200mm, focused at infinity at a 45 degree angle to the horizon, with nothing in the frame but the sky. The ISO 200 exposure that seemed to peg the histogram spike in the middle was either 1/100 @ f/16 or 1/125 @ f/16.)
[a href=\"index.php?act=findpost&pid=130932\"][{POST_SNAPBACK}][/a]

You shooting Raw or JPEG? If Raw, the Histogram you see is way, way off, its based on the JPEG.

The only stuff useful in a histogram happen on either end, highlight or shadow clipping. The height and all the stuff in the middle isn't very useful information.

[scottbel]

You shooting Raw or JPEG? If Raw, the Histogram you see is way, way off, its based on the JPEG.

The only stuff useful in a histogram happen on either end, highlight or shadow clipping. The height and all the stuff in the middle isn't very useful information.
[a href=\"index.php?act=findpost&pid=130998\"][{POST_SNAPBACK}][/a]

I am shooting RAW.

So you're saying that where the tonal values fall on the in-camera histogram scale aren't useful in making judgments of proper exposure?

[digitaldog]

I am shooting RAW.

So you're saying that where the tonal values fall on the in-camera histogram scale aren't useful in making judgments of proper exposure?
[a href=\"index.php?act=findpost&pid=131001\"][{POST_SNAPBACK}][/a]

For Raw, yup. It has no directly (some would say useful) correlation to the Raw data.

I can 'over expose' 1.5 stops over my external meter, but get full tonal data from the Raw using Lightroom's rendering controls but if the shot is 'over exposed' a half stop, the Histogram shows clipping (based on the JPEG it will not build). Its messy. For Raw, the histogram is nearly useless.

[scottbel]

For Raw, yup. It has no directly (some would say useful) correlation to the Raw data.

I can 'over expose' 1.5 stops over my external meter, but get full tonal data from the Raw using Lightroom's rendering controls but if the shot is 'over exposed' a half stop, the Histogram shows clipping (based on the JPEG it will not build). Its messy. For Raw, the histogram is nearly useless.
[a href=\"index.php?act=findpost&pid=131003\"][{POST_SNAPBACK}][/a]

Thanks for the feedback. I must say, the deeper I get into this stuff, the more confusing it seems to get...

[digitaldog]

Start here:

http://www.luminous-landscape.com/tutorial...ose-right.shtml

Then:

http://wwwimages.adobe.com/www.adobe.com/p...renderprint.pdf

[Ray]

I am shooting RAW.

So you're saying that where the tonal values fall on the in-camera histogram scale aren't useful in making judgments of proper exposure?
[a href=\"index.php?act=findpost&pid=131001\"][{POST_SNAPBACK}][/a]

You can try experimenting with the in-camera jpeg settings which will affect the behaviour of the histogram. For example, a high contrast setting will tend to result in a histogram which shows a lot of clipping of highlights when in fact, if you are shooting RAW, there is no clipping.

I've set my 5D, in Picture Styles, Landscape, to minimum contrast. If I see just a small area of sky clipping (flashing on the LCD) I'm confident I've exposed fully to the right. If a large area of sky flashes, I know I've probably blown the red channel.

[digitaldog]

I've set my 5D, in Picture Styles, Landscape, to minimum contrast. If I see just a small area of sky clipping (flashing on the LCD) I'm confident I've exposed fully to the right. If a large area of sky flashes, I know I've probably blown the red channel.
[a href=\"index.php?act=findpost&pid=131016\"][{POST_SNAPBACK}][/a]

Me too but as I indicated, half stop over with this picture style shows clipping. I don't get actual hightlight data clipping until I go over 1.5 stops recommended by incident meter. So the Histogram, even with a custom setting is a mile off.