Digital histogram as meter

[Bob McCarthy]

I'm getting back into LF after a 15 year absense.

My old Zone-VI soligor is off being repaired. In its absence I've been playing with using my D2x meter in its place. Has anyone attempted such folly? <G>

As most know the lcd output and histogram is based on a camera processed jpeg. By changing color space and contrast settings, i "think" I can get the camera to match up to the different film characteristics. If I can get the histogram to sync up to film, it could provide very scannable slides/negatives.

Any pioneers out there before I spend a ton of time and money proving myself correct or otherwise?

Bob

[Bill Kantor]

This might be helpful. I've not tried it but it makes a lot of sense to me.
http://www.kenrockwell.com/tech/meters-digicam.htm

[Brian Ellis]

I've never tried it but I've seen a lot of discussion about it and plenty of people seem to do it. If you happen to have access to back issues of Photo Techniques magazine there was on article on this and other uses of a digital camera in "real" photography within the last two years or so.

One reason I've never tried it is that I don't understand how you relate the histogram to film. When using a digital camera to make pictures the standard practice is to select an exposure that moves the histogram as far to the right as possible without blowing out the highlights. That usually results in an overexposed digital image which is then fixed somewhere in the editing process. But if you're using the camera as a meter I assume you wouldn't want to do that so you'd have to figure out how you wanted the histogram to look - would you bunch things up towards the left to make sure you exposed for the shadows even though that isn't what you'd normally do with a digital camera, try to keep the important details towards the center or what? Or would you just select any exposure that doesn't clip the shadows or highlights, on the theory that where the histogram falls is of no importance with film as long as nothing is lost?

I don't know, a meter seems like a much better way to me but maybe I'm just old-fashioned. I do know that you can't pay any attention to how the image looks on the LCD. I've had plenty of images that looked great on the LCD and looked awful when opened in Photoshop. I no longer pay any attention to the LCD except to see the histogram.

[Bob McCarthy]

It's true that one should move the data to to the right with a histogram to minimize noise. And that the digital capture more resembles a transparency in technique.

Also the jpeg based histogram doesn't necessarily resemble the raw capture (this item confuses many). The jpeg can be managed with internal camera settings like contrast, space, etc.

A B&W (and color to a degree) negative is based on exposure and modified by development, and this may be more challenging using the histogram. But the D2x also has a spot meter so this can possibly be managed.

Digital cameras have exposure compensation available and this may be needed to "sync" up the camera meter with the film. The nikon "smart" meter may be more of a challenge than the straightforward spot meter, but it is very capable.

I want to see the spread of light captured by the main sensor as it may be the ultimate meter, rather than relying on the internal meters guestimate.

Digital sensors have very long scale (10 stops+), but noise rears its ugly head at the shadow end of the scale but then I don't care necessarily about noise, just a spread of prodicted densities.

I always carry a digital with me. If it works well with the D2x, I may then try the D70 I own as backup.

The weakness I forsee, is that the camera manufactures all calibrate there cameras (and LCD jpegs) with there own vision of what will provide the expected customer with good results. They apparently don't agree with each other as well as within there own product lines.

I think its worth the effort. I'll soon find out. And will make a report to the forum. I can't start for a week. I'm on a road trip beginning the middle of this week.

Any thoughts or feeback is appreciated.

Bob

[Guy Tal]

This is a bit outdated but may stil be helpful:

http://www.naturephotographers.net/a.../gt0203-1.html

Guy

[Bob McCarthy]

Guy, nice article.

I have found that with a higher end dslr I can manipulate the histogram by up to a couple of stops with contrast and color space adjustments. It may (and I emphasize "may") be possible to adjust chemical development desired by adjusting and previewing the histogram to get the DR.

The concern I have with the histogram is that middle grey is not always middle gray. I may start with polariods first to speed up the initial findings.

I'm trying to sort out my methodology. Do I shoot a grey card first to see where the zones fall, do I work off of highlights or shadows. Can I find a camera setting to emulate velvia or Tri-x (Tmax).

Is a transparency/color negative properly exposed best for scanning, or under(over)exposed with the scanner digging out the info.

I starting to think I'm biting off a lot.

bob

[Rob Landry]

Guy, great article. I've been using my Nikon DSLR with limited success. Like Bob, I find that changes made to the contrast and other image parameters will change the shape and placement of the histogram.

I also find that if I take a handheld incident reading with my light meter, and set this on my DSLR, rarely is the histogram "exposed to the right". I usually have to add another 1/2 to 2/3 of a stop to achieve a perfect histogram; with film, my incident meter yields perfect results. So, this leaves me believing that either I'm doing something wrong, my DSLR is screwed up, or perhaps there are so many settings with a DSLR that getting it to match film out of camera is proving difficult. Maybe a point-and-shoot digital is better suited to this task, seeing that they are set up from the factory to provide the best possible image directly from the camera for the majority of the public. Regardless, my results so far have been less than promising, so I've been staying with my meter and/or grey card for metering with LF.

[Greg Miller]

I also find that if I take a handheld incident reading with my light meter, and set this on my DSLR, rarely is the histogram "exposed to the right". I usually have to add another 1/2 to 2/3 of a stop to achieve a perfect histogram;

This is to be expected. Your incident meter is trying to make the overall scene middle gray. "Exposing to the right" is not trying the make the scene middle gray but is trying to capture the scene at the highest exposure without losing highllight detail.

For using a histogram to judge exposure for film, you would first want to determine where the tones should fall on the histogram naturally, and then add or subtract exposure to place them where you want the tones to be recorded on film (would be different depending on positive or negative film).

For eaxample, if my primary subject is middle gray, then I would expect the see that item to appear in the midlle of the histogram (assuming a correct exposure). If I want to record that item as light gray then I would want to add exposure. I could change my exposure and retake the shot, and verify that I moved the histogram appropriately.

[Brian Ellis]

[QUOTE=Rob Landry]
I also find that if I take a handheld incident reading with my light meter, and set this on my DSLR, rarely is the histogram "exposed to the right". I usually have to add another 1/2 to 2/3 of a stop to achieve a perfect histogram; with film, my incident meter yields perfect results. So, this leaves me believing that either I'm doing something wrong, my DSLR is screwed up, or perhaps there are so many settings with a DSLR that getting it to match film out of camera is proving difficult.

You're not doing anything wrong and your camera isn't screwed up. You're using your meter to determine a "correct" exposure. When you move everything as far as possible to the right of a digital camera histogram without clipping the highlights you're usually over-exposing, which isn't what you would normally do with your meter. So you often have to "force" the overexposure by increasing the exposure beyond what your meter says is a "correct" exposure.

[Rob Landry]

Greg, Brian,

Thanks for the responses, makes sense I guess. Shooting years of film (trannies mostly), this digital realm is proving harder to figure out than I thought. I guess I'm just used to getting great results with my various meters and was perplexed that digital should be so different in this respect. For example, the other day while shooting, taking an incident reading gave spot-on results for my trannies, but my DSLR constantly needed a little more exposure or the images appeared muddy upon opening in Capture. I guess it will take more getting used to but it appears as though it will be of limited usefulness in previewing the exposure for the LF camera. Maybe a P&S digital would be better suited for the task.