There is an important (but unavoidable) bias in the methodology too...
in your test you processed the image under lightroom then try to reproduce
the same result with DT
this has sevral unfavorable consequences for DT (the same problem would be
reverted if you went the other way round, which is why I say it's
unavoidable)
* the method under DT will always be more complicated because DT doesn't
provide the same tools so you have to use multiple tools and usually the
complex/low level ones to get a similar result. Moreover you need to really
master them well since you need to have an idea what LR does "under the
hood" to do something similar in DT. Thats probably one of the reason why
Johanes (which is one of the guy with the best knowledge of image
processing around this ML) did a better job than you...
* the result will most likely be less flattering for DT, again because we
are trying to achieve someone else's result with different tools, so we can
only be an aproximation.
If you do the opposite (develop a low light image in DT, reproduce in LR)
the results would probably very different and LR would (legitimately) be
the one struggling...
I would be particularly be interested to see how well LR can reproduce what
our "lowlight vision" plugin does. I personally think it gives very natural
looking results for night images because it reproduces "what the eye really
sees" (by oposition to "what the sensor sees") and rings a bell in our
brain...
On Thu, Jun 13, 2013 at 7:19 AM, ary brami <[email protected]> wrote:
> **
> Hello Matthias,
>
> I'm agree with you, about the noise reduction, unfortunatelly. I was sure
> before that test, that DT was closer than that...
> I think that it's also due to the ISO level. Particularly on this image
> where ISO 6400, DT don't achieve the job as good as LR. But on ISO level
> lwer, DT is very good.
>
> We decide with two other bloger to continue this kind of test LR vs DT
> once or twice a month, in order to know better if in some picture the
> difference is realy smaller or if DT is even better than LR.
>
> Ary
>
> Le mercredi 12 juin 2013 à 07:58 +0200, Matthias Bodenbinder a écrit :
>
> Am 11.06.2013 10:31, schrieb ary brami:
> > Hi everybody,
> >
> > I just publish on my blog a post about developement of the same image,
> with lightroom and darktable.
> > The picture was VERY noisy (ISO 6400, at night), and on that point, I
> must admit that I didn't achived to reduce the noise in Darktable as well
> as in Lightroom.
> > If somebody want to try to do better, the link to download the raw file
> is in the post here :
> http://aventurereflex.wordpress.com/2013/06/10/darktable-vs-lightroom-le-logiciel-dadobe-est-il-un-must-pour-les-photographes/
> >
> > We will renew this type of developement both in LR and in DT, sure that
> some time, DR will be closer or even better thant LR.
> >
> > Ary
> >
>
> Hello Ary,
>
> that is interesting. And I am actually shocked about the big difference
> between Lightroom and Darktable. I find two aspects very odd.
>
> (1) DT has a noise profile for the 600D at iso3200. Even with that it is
> almost impossible to reduce the noise properly. And Denoise (non-local
> means and bilateral) do not do any better. I thought that noise profiles
> is the best you can get in terms of noise reduction.
>
> (2) The color cast in the background is weird. The blue background of the
> DT image looks artificial while the grey background of the Lightroom
> picture looks much more natural.
>
> And by the way, rawtherapee is not doing any better than DT.
>
> Kind Regards
> Matthias
>
>
> ------------------------------------------------------------------------------
> This SF.net email is sponsored by Windows:
>
> Build for Windows Store.
>
> http://p.sf.net/sfu/windows-dev2dev
> _______________________________________________
> Darktable-users mailing list
> [email protected]
> https://lists.sourceforge.net/lists/listinfo/darktable-users
>
>
------------------------------------------------------------------------------
This SF.net email is sponsored by Windows:
Build for Windows Store.
http://p.sf.net/sfu/windows-dev2dev
_______________________________________________
Darktable-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/darktable-users
