Hello guys!
There was a Darktable code revision 2 years ago following the bug I reported https://redmine.darktable.org/projects/darktable/repository/revisions/b939ed9d49322aa7d5eb27309677419e4fc31d10
When it was fixed I had compiled v2.3+xxxxxx and continued using it until now. This time, two years later, when I upgraded to 2.6, hoping for good, DT showed me the old error again ‘cannot read Nikon J1 camera balance’. The stranger part is that WB is perfectly read for neighbor images of the same photo series. It fails to read WB like for every third-fifth file or so.
In order to test, I removed my library.db and data.db, as well as ./config/darktable, suspecting my old configs to be corrupted and pulled DT v2.6 from Ubuntu repository, installing new configs, but, no, it wasn’t them configs - the error repeated itself.
I wonder where did do wrong? Was the fix transferred to subsequent DT versions >2.3?
You did not attach the sample, thus it is impossible to answer.
I can guess that it might be either https://redmine.darktable.org/issues/12209, which is fixed in 2.6.0 or https://redmine.darktable.org/issues/12208 (fixed in darktable git master, not yet fixed in 2.6.x)
(Also, people, do report bugs in the bug trackers, in dt’s case it’s either [1] or [2] (yep, who needs one issue tracker when you can have two, am i right?))
You always are, Roman!
In my case this is the same old camera, for which I reported the bug two years ago - Nikon J1. Nevertheless, take this sample please. _DSC8140.NEF (8.3 MB)
I own two cameras and have had more, but this bug affects just this old little buddy Nikon J1.
Really, really strange issue - a flashback with the very same camera, same lens, setup, everything for which you had produced a bug-fix Roman. Honestly, initially did not even believe it is not mine, but DT error.
I will log a bug-report too. Thanks for taking time to look into this.
Filed as #12528
Will try to build it from the master and, once you releasde 2.6.1, upgrade it.
Hold on. dang, that is a sweet link you gave! Did not know about this release platform. Do you think Ubuntu repository is worth being kept on the sources list until you release version >2.7 ?
If you feel like, you could contribute that affected sample to https://raw.pixls.us/,
to my surprize there is now only two samples with such an issue there.
There are no releases with the odd versions (x.y.z; y % 2 != 0)
But sure, if you are ok with using up-to-date bleeding edge - why not.
I had contributed to your raw files DB from the very same camera back in the day. Was it lost? Nevermind, now will do the same to raw.pixls.us.
Got it. Downloaded what it offered for 18.04 - something like 2.7 and, in fact GDebi installer says it’s the same 2.6.0 version. Nevertheless, I have installed it and the issue remains there.
It’s … slightly more complicated than that.
Back then the call was for ‘just some sample’.
And we do have ‘just some sample’ on RPU.
(on RPU, we only take CC0 samples, and back then we did not ask for any license release,
so even if we did not have ‘just some sample’ on RPU, it wouldn’t be possible to just upload it there)
But here i’m specifically interested in such broken sample.
And as i said, we only have a small handful of these, one more would be good.
As I mentioned above, there already was another sample available. I have supplied the new submission with a commentary that DT cannot read white-balance from this one.
I was about to write yes, but then I used Software Center to install the deb package and it appeared that Gdebi did not recognize the version right. Indeed, it’s 2.7 now and I confirm that DT 2.7+xxx is digesting the WB perfectly.
Could you give us a heads up about 2.6.1 release?
