ISOless sensors, read noise and photography - many questions!

Started Oct 3, 2013 | Discussions thread
RussellInCincinnati Veteran Member • Posts: 3,201
many answers. Also notes on convenience of limited ISO.

original poster boardsy wrote: - in order to maximise exposure and DR and minimise noise,

You must mean "in order to decide if you should limit raw in-camera ISO settings (i.e. make intentional underexposure) decisions in cases of low light, to minimize noise and maximize dynamic range,..." You must only be asking about making ISO decisions in low light, because when there's plenty of light, there are no ISO decisions to minimize noise...we all know to use a sensor at its base ISO to optimize noise and dynamic range.

You also are probably only asking about raw files, because when you're using JPEGs, you don't have as much luxury to "underexpose" a scene with a "too low" ISO and then brighten the final print in post-processing.

do you really need to know your sensor's read noise and ISO DR curves?

Excellent question, hundreds of trials over the last couple of months tells me over and over again the answer is no. You don't need to know your sensor's read noise and ISO DR curves to decide whether or not to limit your camera's ISO to about ISO 800 or so, no matter how dim the light. Because you can decide for yourself whether or not "pushing" ISO beyond 800 in post-processing works better than setting the raw camera ISO past 800, with about 5 minutes of experiments.

I.e. it takes less time for a raw photographer to figure out for themselves, with say Lightroom, whether or not it makes sense to use camera high ISOs in low light, or software-brightening post-processing in low light. No high-tech sensor data needed. Definitive experiments with your own equipment takes less time than trying to "puzzle out" (by arguing theory, guessing or consulting a bunch of noise charts) whether your camera firmware can increase your effective image ISO better than you can.

For example, does limiting raw ISO to 800, "underexposing" the image in the field, work well with a Nex C3 (or your camera) if you have access to Lightroom 4? Forget the charts, let's just spend 90 seconds taking an "underexposed" ISO 800 raw image

and move a couple of sliders to brighten it in a $50 dollar copy of Lightroom 4 and see what we get...

Boardsy I don't need to know anything about my Nex C3 sensor noise charts, to decide whether or not limiting it here to ISO 800 in low light works OK or not.

Bobn tells us here that the flat read noise characteristic of the Nikon D7000 makes it effectively ISOless:

If you have a Nikon D7000, you could figure out whether or not you ever "need" to change the raw ISO, for any reason besides previewing convenience, in less time than it takes to argue or theorize about it.

- if the sensor is effectively ISOless then is base ISO always optimal (providing that you can see enough of the image on your camera to compose and focus in a sub-optimal exposure scenario), or are you as well off raising it in camera (providing highlights are preserved) rather than in RAW/pp?

Not only difficult to theorize about (after all some camera could have magic sauce in firmware brightening machinery that Lightroom etc can't match), but also needless (as described above) to speculate about.

The overall calculation is this:

  1. You may or may not get exactly the same dynamic range and noise results when you "underexpose" low-light raw images. But you can find this out in 5 minutes.
  2. You will definitely eliminate all need for exposure bracketing by limiting camera ISO to 800 in low light, and letting the raw file be "under-exposed" until post-processing.
  3. You will also via low-light ISO limiting, get the world's best control of how much bright highlight detail you want captured/not burnt out.

Is it possible or likely that the camera's ISO implementation is better tuned to its sensor and RAW data than your RAW software?

No need for this question since you'll answer it the first 2 minutes you try it, if you have something as easy as Lightroom 4 to use.

Aside from highlight preservation on exposure, are there other benefits to "doing ISO" in post?

Well there are certainly some drawbacks to doing ISO in post-processing. It takes some seconds, and some skill/judgement, per image. And what if the dark parts of your images aren't noise-suppressed as well as your camera firmware can render 'em?

  1. Raw users spend some time and skill per-image anyway. And this limited-ISO workflow topic only applies to raw images as explained above.
  2. The albatross around one's neck of exposure bracketing disappears immediately. Great for people like me who use cheap, not-too-many-frames-per-second cameras like a Nex C3, and/or people who don't like having 66% of their file storage space taken up with pointless bracket images, and/or people who take photos where the scene is changing from moment to moment.
  3. Less low-light camera settings operator errors in the field (my #1 reason for adopting this workflow, other people may be less mistake-prone).
  4. Less time fiddling with low-light camera settings in the field.
  5. You don't just get different highlight "preservation", you get somewhat different recording of highlight detail...that differently-rendered highlight detail becomes addictive, once you realize that, getting your head out of the boring/featureless test-chart world, the better-rendered lightest parts of your interesting images generally are much more visually significant than the possibly-noisy darkest parts of those nice pictures.

Boardsy: for a given sensor may it be more effective to raise ISO to some minimum/optimal level vis-a-vis its read noise?

Too vague, what does "effective" mean? You may like the way your camera firmware handles shadow detail and noise more than the way Lightroom (with all its convenient noise-reduction sliders?) does. But it's definitely less convenient in the field to fiddle with over-800 ISOs, if a dark viewfinder preview and dark in-camera playback doesn't bother you.

And it's not useful to ponder the theory of this question, when the practice is so quick and easy.

how do us non-scientists interpret these sensor graphs? :-)...And could this non-base ISO be valid for all shooting, or just low-light where optimal exposure becomes problematic? From

5 or 10 test images will determine whether it's practical to do as I do with a 'C3, which is (a) always leave limit camera raw ISO to no more than 1/2 or 1/4 the "ideal preview" ISO and (b) don't exceed a raw ISO of 800 in any case.

- finally, could all this be solved for a given sensor by simply shooting a set of fixed aperture/shutter shots at each ISO level, normalising the resultant RAW brightnesses (with the, er, "exposure" slider :-|) and comparing noise (in the shadows, I guess?) by eye?


Post (hide subjects) Posted by
(unknown member)
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark MMy threads
Color scheme? Blue / Yellow