Is there a human-readable, textual format that describes changes (style/presets) that you can apply on the image? Then this format can be read in multiple photo editing softwares and easily shared by ppl.


This is a trivial example, but you will get the picture ;)

Name: MyCoolContrastOnBlue
Contrast: +15            <- increase of the contrast
Hist.blue: (0.7, 0.7)    <- added a point in histogram
Brightness: -2
Saturation: -32.1
Noise: -2

and so on. Then we can share sets of presets/styles, and make an open community of it.

Of course, some algorithms depend on software (like noise reduction, blur), but we can add modifiers like:

Noise: -2
Noise<CaptureOne.9>: -2.1

So if the software is CaptureOne, we actually use -2.1 noise reduction etc.

Is there something like this in this world?

  • What is Hist.blue: (0.7, 0.7) supposed to do?
    – mattdm
    Jan 19 '16 at 14:41
  • @mattdm if we imagine histogram as axes, from 0-1, and by default we have two points in (0,0) and (1,1) - then every new dot will be added with its coordinates, and the line gets interpolated through these dots. So in this case, curve goes from 0,0 -> 0.7,0.7 -> 1,1
    – igor
    Jan 19 '16 at 14:45
  • Ahhh. Generally in photography software, we'd call that a curves adjustment. You can think of a histogram as underlying the curves tool, but it's maybe better thought of as a representation of its result.
    – mattdm
    Jan 19 '16 at 14:47
  • True, this was just a quick-and-dirty example, my bad. Thanx @mattdm.
    – igor
    Jan 19 '16 at 14:49
  • Yeah, I did get the basic point. :)
    – mattdm
    Jan 19 '16 at 14:50

No. There is currently no such standard. The closest thing is Lightroom sidecar XMP files as a defacto standard. These aren't exactly human-readable (because they're XML rather than the simple key-value format you propose), but at least they're text. And they're not standardized, but some other programs (like Darktable) can at least attempt to interpret them.

For example, Darktable says that crop, rotation, flip, and tags can be understood completely; while exposure, some aspects of the tone curve, local contrast, and a few other things can be at least attempted, but aren't accurate.

  • Maybe we can work on making this standardized? wdyt @mattdm?
    – igor
    Jan 19 '16 at 14:46
  • Standards are hard. In order to be really meaningful, you'd have to convince 800lb gorilla Adobe. But, you could start with Rawtherapee, Darktable, Lightzone, Ufraw, and other open source converters, and then maybe convince some of the more niche proprietary players to join in.
    – mattdm
    Jan 19 '16 at 14:49
  • I didn't ask this because it is easy :))) And even gorillas like small bananas! Anyway, i find this cool., but maybe Im wrong, so would like to see what community thinks.
    – igor
    Jan 19 '16 at 14:52
  • and this can start in baby (gorilla) steps, that is the nice part.
    – igor
    Jan 19 '16 at 14:53
  • In general with an idea like this, thinking of it is cheap — code does the real convincing.
    – mattdm
    Jan 19 '16 at 15:13

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.