Exposure lattitude of film and digital cameras?

[EricM]

we should all read this...

 

http://clarkvision.com/imagedetail/film.vs.digital.summary1.html

[Uhooru]

Byron, remember at a certain point you're actually looking at what tonal

range paper holds which is different than film or what your scanner can hold which is also

different as compared to what you digicam can do. Maybe 5stops was to conservative, but

essentially a digicam holds similar to what tranny's hold and negative film is significantly

broader. Digicams do not hold 9 stops of exposure lattitude nor does film.

[Uhooru]

Roger, I think you did misunderstand me if you thought I was saying you must use a spot

meter. I just like spot meters, if you like to use and incident meter that's great, they get to

the same place. They both tell you what middle grey will be at the point metered. I just

personally find a spot meter very accurate for when you want to place tones because you

can ..well pick out spot tones. Its easy to measure the difference from your darkest tone

and your lightest and everything in between very precisely. So if your darkest tone that

you would like to have some detail is let say EV 4 and your brightest tone you want detail

in is EV 17 than you are going to not get that range in with slide film or digital. You might

be able to get it by altering developement time with B & W negative film..witch for you

zone system fans is one of the purposes of the zone system, and then its even narrower

when you print in the darkroom.

[byronlawrence]

I never said 9. where do you read nine?.. i pointed out how one might could argue 9 but I said usefully (in my view) 7 is a good range, If I plan to use post processing techniques, then I sometimes will try to capture 8 stops of light range, but that will be mostly to bring in a little shadow detail on a brightly backlit scene so as not to have a pure silhouette, I do try to avoid that though. and yes, 5 is conservative (very conservative IMO). 6 is safe. for transparancies, 6 works much better too (I think).

 

and the question is 'what can the medium capture', not what can you print. you can capture more and in the end it gets compressed and printed, but that doesn't change what you can capture. as you pointed out they are different things.

[EricM]

9 stops for a good dslr is about right. guess no one explored the above link...

 

http://clarkvision.com/imagedetail/dynamicrange2/

[Uhooru]

Byron, first you said "digital (20d) I can get 9 stops of information out of a RAW file. from

normal up I get 3-4 is stops before it blows out and on the shadow side it kind of starts to

compress more and more. there is a point were you can tell 'detail' but it is so noisy and

the detail is so faint that and real picture quality is pretty meaningless to use. I feel

comfortable with calling it 7.5 to 8 stops really. I know a lot of people will disagree."

 

Later you said "my counts are wrong. 4-10 is 7 stops and 3-11 is 9 stops. I call useful

4-10."

 

So I stand corrected. Despite Eric's artical, I'm skeptical that you even get 7 good stops

out of digital, but if people like Eric who shoot it professionaly say they are getting 7

useful stops in usage, not shooting some test chart, but in everyday use, then I will believe

it.

[EricM]

Barry, i'm completely unsure. the above article is the first i've come across that i would on a normal day agree with you and everyone on the 5ish stops of latitude. there's apparently a white paper or two floating around stating that most decent dslr cameras get 8 stops on average. as well as a technical paper from canon stating the 1DmkII gets 11. I've yet to find it. all i know, as many here, that highlights will blow out well before c-41 while at the same time reach farther into the blacks. the proof is in the pussing. so, i don't know if it's a semantic thing or what. all i know, is after three years and 60,000 dslr exposures, the whites freakin blow out before c41...

[Uhooru]

Eric, thanx, that's basically what I've heard from the digital shooters I know, that the hi's blow out a little earlier, and more akin to shooting slide film. I'm sure you become very aware of where that margin is when you're shooting. Do you find yourself sliding your scale down through under exposure a little to get more in the highs at the expense of some shadow detail?

[EricM]

I think perhaps this 8 stops and above reference is when a raw images is duplicated three or four times and then having these copies set at different levels of exposure in raw and then layered or blended to max dynamic range. this doesn't count for me as an out of the camera image, so i'll stay in the 5 stop group.

 

i more or less treat digital like slide film Barry. if i have the time, i hand hold a sekonic were i can for optimum exposure consistency. if i don't have as much time, i use the clipping white display and choose if i want those blown or not and adjust from there. if i have even less time, i just rely on the cameras meter with +1/3 exposure comp. don't know if it's right or not, but works for me in post.

[Spearhead]

<i>a technical paper from canon stating the 1DmkII gets 11.</i><p>

 

I have to say that I have been very surprised by what I can get out of the 1DMkII. I rent one for sports shooting and once accidentally changed the exposure setting for a few shots without noticing. I assumed they were lost, based on my experience with the 10D, completely overexposed. However, I was able to get very usable photos despite something like three stops overexposure. It probably helped to be shooting RAW, though, I'm not sure how it would work with jpgs.

[ken_hughes4]

"I'm laughing my a$$ of at some of the above claims giving conventional B/W film more dynamic range than color neg film. More myth and religion about that medium as per usual."

 

when will you ever learn........

 

normal exposure in the zone system is 11 stops of range from full black to full white, yes it is 11 zones---0 to 10. easily more than color neg film. dilute development can pull in a 15 stop range. it's really not that hard to do.