Jump to content


CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.


Linear calibration vs raw conversion for DSLR images - what is the tradeoff?

astrophotography dslr
  • Please log in to reply
3 replies to this topic

#1 nathanm


    Explorer 1

  • -----
  • topic starter
  • Posts: 56
  • Joined: 28 Jun 2020

Posted 15 July 2020 - 12:44 AM

One way to process a DSLR astro image is to go through a calibration process to produce a linear image.  This involves collecting flats and darks and making master images.   Software advances have made it easier to create the necessary files, but it is still a length and tedious process.


For a unique astro camera attached to a (potentially) unique optical train in a telescope that's what has to be done. 


But for a DSLR that is to a large extent and standard product, using lenses that are also standard products, there is an alternative - shoot in RAW format and then use a high quality RAW converter which includes lens and sensor specific corrections for vignetting and distortion .     Although raw conversion software typically applies a nonlinear tone curve, one can get most software to not do that step.   This generally involves having the software output use a profile that specifies a linear tone curve.  


What you are left with is a calibrated linear file that is much like the output of the astro-style full processing.


A raw converter like Raw Therapee (RT) can do all that, as well as apply darks and flats (although it does not help you make the masters).


Has anybody tested the two approaches head to head recently?   


I was using RT to make linear files for a terrestrial photo project when I came across a discussion of using Raw Therapee by Roger Clark https://clarkvision....th-rawtherapee/  and https://clarkvision....age.processing/ written in 2016- 2018.


He is correct in general that RT is an excellent raw converter, with tons of options you don't get in Adobe Camera Raw or other converters.    That said, I am not sure that Clark's workflow as discussed in those web pages produced fully linear output files because he does not mention using an output profile with RT which is necessary to get fully linear output.   


But regardless of whether his workflow was full linear, he has a point that it is a lot more convenient to use RT.   Have others tried this comparison since then?



#2 whwang



  • *****
  • Posts: 3,174
  • Joined: 20 Mar 2013

Posted 15 July 2020 - 01:24 AM

This issues had been discussed/debated for numerous times on this forum.  Personally I would not want to go through it again.  However, if you intend to find out, I can offer my files to you.  I publicly release all my raw images that are more than one year old, to everybody in the world, for free (but not entirely without conditions).  You can get them here:



Some of the files are of the same kind of Milky Way mosaics that you are planning to do.  You can download the raw files and associated calibration files, and compare the best results you can achieve with real astronomical calibration vs. general raw conversion.  No matter how people argue about it, the best way to find out the answer is to try it yourself, right?




#3 sharkmelley



  • *****
  • Posts: 3,309
  • Joined: 19 Feb 2013
  • Loc: UK

Posted 15 July 2020 - 03:53 AM

Roger Clark's comparison of traditional vs "modern" processing is an excellent example of why you should avoid the "modern" approach.  Look at the blue tinge to the outer parts of Flame Nebula in figure 3b here: https://clarkvision....age.processing/

Subtracting light pollution from non-linear data causes exactly these colour shifts to faint structures and also increases their colour saturation.


However, I agree that if the raw convertor keeps the data linear, as RawTherapee is able to do, then this objection disappears.



Edited by sharkmelley, 15 July 2020 - 03:54 AM.

#4 ChristopherBeere


    Mariner 2

  • -----
  • Posts: 294
  • Joined: 24 Jul 2019
  • Loc: London UK

Posted 15 July 2020 - 06:28 AM

I do all my DSLR Raw calibration in RawTherapee and have been happy with the calibrated debayered output I get from it.


Don't forget that Clarkey does not use a modded camera.


All 3 of my modded cameras have gremlins that need flatting out.


Plus a custom flat gives you a much better vignetting profile than a generic one which is super important for mosaics when you start pushing the data hard.

Edited by ChristopherBeere, 15 July 2020 - 06:39 AM.

  • sharkmelley likes this

CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Recent Topics

Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics