thx for the pointer! will look into it!
On Mon, 30 Mar 2015 at 17:05 johannes hanika wrote:
> i guess you want to take a look at dt_colorspaces_get_darktable_matrix
> (src/common/colourspaces.c) where we use these values.
>
> especially near the top where the rows are normalised and the first ma
i guess you want to take a look at dt_colorspaces_get_darktable_matrix
(src/common/colourspaces.c) where we use these values.
especially near the top where the rows are normalised and the first matrix
is constructed. the rest is a bradford adaptation to D50, not sure you want
that.
-jo
On Mon, M
hahayou guessed right...i actually shot using a 5D Mark II but the
values from colormatrices.c are all INT and i need float...can I convert
those INTs somehow?
thx for your support
Oli
On Mon, 30 Mar 2015 at 14:04 johannes hanika wrote:
> On Mon, Mar 30, 2015 at 1:17 PM, Oliver Markowski >
On Mon, Mar 30, 2015 at 1:17 PM, Oliver Markowski
wrote:
> seems you guessed right! hdrmerge says i can input the inverse
> of Exif.Image.ColorMatrix2 on the command line but it seems only the Adobe
> DNG converter stores that matrix.
>
> Is there a way to extract Exif.Image.ColorMatrix2 using da
seems you guessed right! hdrmerge says i can input the inverse
of Exif.Image.ColorMatrix2 on the command line but it seems only the Adobe
DNG converter stores that matrix.
Is there a way to extract Exif.Image.ColorMatrix2 using darktable too?
thx
Oliver
>
---
my educated guess is that you're looking at a merged raw file, without any
white balance applied to it. i would try to copy/paste a history stack of
an individual raw in dt on top of one of the merged exr, to get the white
balance right. possible that you're also lacking the colour matrix, unsure
i
Thx Johannes! I just tried hdrmerge and it seems to be the right tool for
my task.
Unfortunately it gives me a massive greenish tint...did anyone experience
this issue with hdrmerge before? (i know this is darktable-devel and I will
send a mail to Wenczel as well)
cheers from munich
Oliver
On Mon
heya,
maybe wenzel's command line merger would be better suited for your
particular job? the internal workings of the algorithm are quite similar to
ours (merge raw bayer, not a reverse-engineered linear buffer derived from
a processed jpg).
https://github.com/wjakob/hdrmerge
-jo
On Mon, Mar 3
