Hi,
I think that this is the bug:
https://github.com/darktable-org/rawspeed/issues/100
BR, Micha.
Gesendet: Mittwoch, 25. Juli 2018 um 11:11 Uhr
Von: "Jean-Luc CECCOLI" <[email protected]>
An: [email protected]
Betreff: Re: [darktable-user] Skulls - yes, once more !
Von: "Jean-Luc CECCOLI" <[email protected]>
An: [email protected]
Betreff: Re: [darktable-user] Skulls - yes, once more !
> Message du 24/07/18 20:58
> De : "Michael Fritze"
> A : [email protected]
> Copie à :
> Objet : Re: [darktable-user] Skulls - yes, once more !
>
> ..
> > Here is a link to download such pictures : https://we.tl/tXAnOAe30G
> > OK.rar contains the photo from the computer where it is correctly displayed,
> > while KO.ra contains the same photo (though renammed) on the comuter where
> > it displays as a skull. Both together with the sidecars created upon import
> > process.
> > I doubt it will show anything, but, who knows ?
> > About having an install problem : how can I check what depends on what ?
> >
> > Regards,
> >
> > J.-Luc
>
> Hi,
Hello,
>
> I downloaded those files. The NEFs are binary equal, the XMPs differ only in
> one line:
> xmpMM:DerivedFrom="21114700.NEF"
> xmpMM:DerivedFrom="DSC_0868.NEF"
> Which is very ok. So there should be no difference between those two packages.
>
> When I open these files in dt 2.4.4 I see the embedded preview. I can't open it
> because it can't read white balance. Sometimes I have the same problem with my
> Nikon files, too. It would be nice to have a solution!
This is exactly what I described in my first post from 2018, 07 13th.
However... I realized I only posted a "non displaying correctly" picture.
I ought have posted a "well displaying" one as well.
Pictures nammed DSC_xxx are the ones that were manually copied from the card to the 14.04 system running DT 2.0.7, while pictures nammed with numbers only are the ones transferred and renammed using RPD on the 16.04 computer running DT 2.4.
So, here is one that displays correctly on both the computers :
https://we.tl/AmsWHUHO6D
while the previous one displays skull on DT 2.4 but displays correctly on DT 2.0.
>
> Here is dt command line output.
> Start:
> $ darktable
> [rawspeed] (DSC_0868.NEF) rawspeed::RawImage
> rawspeed::RawDecoder::decodeRaw(), line 270: void rawspeed::BitStream Cache>::fillSafe() [with Tag = rawspeed::MSBBitPumpTag; Cache =
> rawspeed::BitStreamCacheRightInLeftOut], line 128: Buffer overflow read in
> BitStream
>
> (darktable:18948): GLib-GObject-CRITICAL **: g_object_set_data: assertion
> 'G_IS_OBJECT (object)' failed
>
> (darktable:18948): Gtk-CRITICAL **: gtk_widget_get_has_tooltip: assertion
> 'GTK_IS_WIDGET (widget)' failed
>
> Double click on 1st file:
> (darktable:18948): Gtk-CRITICAL **: gtk_widget_get_has_tooltip: assertion
> 'GTK_IS_WIDGET (widget)' failed
> [rawspeed] (21114700.NEF) rawspeed::RawImage
> rawspeed::RawDecoder::decodeRaw(), line 270: void rawspeed::BitStream Cache>::fillSafe() [with Tag = rawspeed::MSBBitPumpTag; Cache =
> rawspeed::BitStreamCacheRightInLeftOut], line 128: Buffer overflow read in
> BitStream
> [temperature] failed to read camera white balance information from
> `21114700.NEF'!
> [temperature] `NIKON CORPORATION NIKON 1 V1' color matrix not found for image
> [temperature] failed to read camera white balance information from
> `21114700.NEF'!
>
> Double Click on dt:
>
> (darktable:18948): GLib-GObject-CRITICAL **: g_object_set_data: assertion
> 'G_IS_OBJECT (object)' failed
>
> --> I guess it goes back from invisible darkroom to lighttable here ;-)
>
> Double click on 2nd file:
> (darktable:18948): Gtk-CRITICAL **: gtk_widget_get_has_tooltip: assertion
> 'GTK_IS_WIDGET (widget)' failed
> [temperature] failed to read camera white balance information from
> `DSC_0868.NEF'!
> [temperature] `NIKON CORPORATION NIKON 1 V1' color matrix not found for image
> [temperature] failed to read camera white balance information from
> `DSC_0868.NEF'!
>
> For me this is the most annoying feature in latest dt iterations: Not to open
> the file and when simply developing all files in a folder: just stopping at first
> error without noticeable message. Can't dt just develop the next picture and
> put a message into some log file?
>
I'm pretty sure that doing the same with the picture in the new archive will not cause the same behaviour.
But it will not tell why DT 2.4 has problems, and why some pictures display different information than others.
Is it a DT problem ?
Is it a Lensfun problem ?
Is it an installation problem ?
How could it be possible to narrow the list so as to finally find out what causes that ?
As I wrote in my first post :
> For those that are correctly processed, the above fields are displayed :
> - modèle (french for brand) : 1 V1
> - fabricant (french for maker) : Nikkon
> - state : 0...r..ap...r
> And those that display the skull say :
> - modèle : NIKON 1 V1
> - fabricant : NIKON CORPORATION
> - state : 0...r..ap....
> Both the body and the lens are correctly displayded within the lens correction module, which applies both distorsion and chroma corrections.
What do the developers/specialists think about this all ?
Rgrds,
J.-Luc
____________________________________________________________________________
darktable user mailing list
to unsubscribe send a mail to [email protected]
> De : "Michael Fritze"
> A : [email protected]
> Copie à :
> Objet : Re: [darktable-user] Skulls - yes, once more !
>
> ..
> > Here is a link to download such pictures : https://we.tl/tXAnOAe30G
> > OK.rar contains the photo from the computer where it is correctly displayed,
> > while KO.ra contains the same photo (though renammed) on the comuter where
> > it displays as a skull. Both together with the sidecars created upon import
> > process.
> > I doubt it will show anything, but, who knows ?
> > About having an install problem : how can I check what depends on what ?
> >
> > Regards,
> >
> > J.-Luc
>
> Hi,
Hello,
>
> I downloaded those files. The NEFs are binary equal, the XMPs differ only in
> one line:
> xmpMM:DerivedFrom="21114700.NEF"
> xmpMM:DerivedFrom="DSC_0868.NEF"
> Which is very ok. So there should be no difference between those two packages.
>
> When I open these files in dt 2.4.4 I see the embedded preview. I can't open it
> because it can't read white balance. Sometimes I have the same problem with my
> Nikon files, too. It would be nice to have a solution!
This is exactly what I described in my first post from 2018, 07 13th.
However... I realized I only posted a "non displaying correctly" picture.
I ought have posted a "well displaying" one as well.
Pictures nammed DSC_xxx are the ones that were manually copied from the card to the 14.04 system running DT 2.0.7, while pictures nammed with numbers only are the ones transferred and renammed using RPD on the 16.04 computer running DT 2.4.
So, here is one that displays correctly on both the computers :
https://we.tl/AmsWHUHO6D
while the previous one displays skull on DT 2.4 but displays correctly on DT 2.0.
>
> Here is dt command line output.
> Start:
> $ darktable
> [rawspeed] (DSC_0868.NEF) rawspeed::RawImage
> rawspeed::RawDecoder::decodeRaw(), line 270: void rawspeed::BitStream Cache>::fillSafe() [with Tag = rawspeed::MSBBitPumpTag; Cache =
> rawspeed::BitStreamCacheRightInLeftOut], line 128: Buffer overflow read in
> BitStream
>
> (darktable:18948): GLib-GObject-CRITICAL **: g_object_set_data: assertion
> 'G_IS_OBJECT (object)' failed
>
> (darktable:18948): Gtk-CRITICAL **: gtk_widget_get_has_tooltip: assertion
> 'GTK_IS_WIDGET (widget)' failed
>
> Double click on 1st file:
> (darktable:18948): Gtk-CRITICAL **: gtk_widget_get_has_tooltip: assertion
> 'GTK_IS_WIDGET (widget)' failed
> [rawspeed] (21114700.NEF) rawspeed::RawImage
> rawspeed::RawDecoder::decodeRaw(), line 270: void rawspeed::BitStream Cache>::fillSafe() [with Tag = rawspeed::MSBBitPumpTag; Cache =
> rawspeed::BitStreamCacheRightInLeftOut], line 128: Buffer overflow read in
> BitStream
> [temperature] failed to read camera white balance information from
> `21114700.NEF'!
> [temperature] `NIKON CORPORATION NIKON 1 V1' color matrix not found for image
> [temperature] failed to read camera white balance information from
> `21114700.NEF'!
>
> Double Click on dt:
>
> (darktable:18948): GLib-GObject-CRITICAL **: g_object_set_data: assertion
> 'G_IS_OBJECT (object)' failed
>
> --> I guess it goes back from invisible darkroom to lighttable here ;-)
>
> Double click on 2nd file:
> (darktable:18948): Gtk-CRITICAL **: gtk_widget_get_has_tooltip: assertion
> 'GTK_IS_WIDGET (widget)' failed
> [temperature] failed to read camera white balance information from
> `DSC_0868.NEF'!
> [temperature] `NIKON CORPORATION NIKON 1 V1' color matrix not found for image
> [temperature] failed to read camera white balance information from
> `DSC_0868.NEF'!
>
> For me this is the most annoying feature in latest dt iterations: Not to open
> the file and when simply developing all files in a folder: just stopping at first
> error without noticeable message. Can't dt just develop the next picture and
> put a message into some log file?
>
I'm pretty sure that doing the same with the picture in the new archive will not cause the same behaviour.
But it will not tell why DT 2.4 has problems, and why some pictures display different information than others.
Is it a DT problem ?
Is it a Lensfun problem ?
Is it an installation problem ?
How could it be possible to narrow the list so as to finally find out what causes that ?
As I wrote in my first post :
> For those that are correctly processed, the above fields are displayed :
> - modèle (french for brand) : 1 V1
> - fabricant (french for maker) : Nikkon
> - state : 0...r..ap...r
> And those that display the skull say :
> - modèle : NIKON 1 V1
> - fabricant : NIKON CORPORATION
> - state : 0...r..ap....
> Both the body and the lens are correctly displayded within the lens correction module, which applies both distorsion and chroma corrections.
What do the developers/specialists think about this all ?
Rgrds,
J.-Luc
____________________________________________________________________________
darktable user mailing list
to unsubscribe send a mail to [email protected]
____________________________________________________________________________ darktable user mailing list to unsubscribe send a mail to [email protected]
