3

I have been thinking about the Zone System and its goal of maximising image quality given the limitations of film sensitivity and the necessity to choose specific developing parameters.

Clearly you can't extract information that was never recorded or was burnt out but given that raw images can be tweaked in post processing, how critical is it get things exactly right at the outset? Does for example +/- 1 stop exposure compensation have any value beyond the quality of a jpeg (if you shoot raw+jpeg)?

Perhaps what I am asking is how much latitude do I have with post processing raw images? (Nikon D90 and P7100). I have zero experience with raw so far (in case you hadn't guessed).

| improve this question | |
4

It's very important, but depending on the dynamic range of the scene you're shooting. If you have bright skies, but also deep shadow areas, then the dynamic range may exceed what your sensor can capture. You will lose detail in the shadow or in the highlights, or both, even shooting in RAW. Shooting RAW means you retain more of the original information, but it doesn't extend the dynamic range of the camera.

If the scene is lower contrast you may have leeway to under- or overexpose by a stop and still not blow highlights or lose shadow details.

You can get an approximation of this by looking at the histogram (although it is derived from a JPG version of your image, so isn't 100% reliable).

There is more information and a higher signal to noise ratio in the higher exposure values than there are in the shadows. So if you do have leeway to adjust exposure, you're generally better off to "expose to the right" (moving the histogram to the right hand side) and then you can adjust the exposure down in post processing if you want to. If you have most of the histogram to the left, and try to raise the exposure in post processing, you're more likely to end up with noise or banding.

More about expose to the right here and here

| improve this answer | |
  • The theory behind expose-to-the-right is very simple: sensors are basically pixel counters, and each doubling of the number of pixels is significant. When you count from 1 to 2, that's just 1 value in whole doubling; when you count from 64 to 128, that's still one doubling but with 64 discrete values. A higher number means more photons to count, which is better in so many ways. (Like this.) – Please Read My Profile Jan 17 '12 at 20:54
  • 2
    So that's all good. But even with RAW, if you've blown out your highlights, that's difficult to recover, and difficult to make look right if it's a part of the image you care about. Blocked-up shadows are often easier to work with, and modern sensors have incredibly low shadow noise, so keep this in mind — the rule for ETTR fully-stated is: expose as far as you can to the right, without blowing highlights. And it's often easier to play that safe rather than worrying about histogram shape. – Please Read My Profile Jan 17 '12 at 20:56
  • That's what I meant by saying "if you have leeway", that you can expose "left" or "right" a bit without losing detail at either end. – MikeW♦ Jan 17 '12 at 21:15
5

Exposure is always critical. How critical is a matter of personal choice but I would not tell any of my students to be sloppy because of their choice of file format or anything else for that matter.

Great photos come out by applying the best decisions and process possible at every step. If you are using RAW as a way to avoid decisions and apply yourself to the best of your abilities, then your images will miss the mark. Sure, it is reasonable to have a safety net, but if you get the exposure perfect at the time of capture, your images will have greater potential.

| improve this answer | |
  • 1
    You are utterly correct, I suppose I was wondering if RAW was a get out of jail free card, so much to learn. Do you have any pointers to resources on learning how to make the best decisions? – epo Jan 18 '12 at 14:24
  • 2
    The classic is Understanding Exposure by Bryan Peterson. Its available online at Amazon & Chapters at least and probably in libraries too. – Itai Jan 18 '12 at 14:29
1

It really depends on many factors.

On a 5d Mark II when shooting at ISO 200, I can happily push an image close to two stops without noticeable effects. At ISO 800, 1 stop is already a lot more noticeable on the same camera.

Pushing dark shadows more easily leads to noise than mid-tones.

The luminous landscape site has a useful article: http://www.luminous-landscape.com/tutorials/expose-right.shtml

The age of the camera is another matter to consider - I would not want to push a 12bit RAW file from a 400D as much as the RAW file from a 5D Mark II.

As a general rule, I would say 0.5 stops should be fine on any camera made in the last 5-6 years. But how much you can push an image on your model, only experience will tell you. The RAW editor you use is also going to impact your ability to develop the image.

| improve this answer | |
  • Thanks forsvarir for fixing my typing (Mainly capitalization) - tablets are useless for writing. – DetlevCM Jan 18 '12 at 8:52
-1

It's no less important than getting the exposure correct when shooting on any other medium, be it E6, C41, PAN, or whatever.
IOW it is important :) In fact it may be more important than with some films as digicam sensors can't capture nearly the same dynamic range as many consumer films can, so there's less room for fiddling with the results afterwards to correct for blown highlights or blacked out shadows.

| improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.