[darktable-user] Almost perfect colors with Nikon D7200 and Darktable

[Saint Germain]

I have no idea what I am doing... but it works !

After struggling to get the correct color from my Nikon D7200 with
Darktable, I finally managed to get a near-perfect result.

It was frustrating because the Nikon JPEG were doing a much better job.
I even bought an IT8 target to try the colormatch approach:
https://github.com/pmjdebruijn/colormatch
https://www.darktable.org/2016/05/colour-manipulation-with-the-colour-checker-lut-module/

The colormatch trick gave me almost the same JPEG as the default
Darktable (i.e. with a greenish tint).

I almost gave up when I noticed that (while following the colormatch
procedure) changing the input color profile to "RGB rec2020 linear"
suddenly got rid of the greenish tint.

So now with a base curve "Neutral" and an input color profile to "RGB
rec2020 linear", I have almost perfect colors in every conditions (I
checked on many pictures).

Could someone explain to me what I have done and why it works ?

[I. Ivanov]

I stumbled on the following

Never tried it but it does talk about rec 2020

Regards,

B

Post by Saint Germain
I have no idea what I am doing... but it works !
After struggling to get the correct color from my Nikon D7200 with
Darktable, I finally managed to get a near-perfect result.
It was frustrating because the Nikon JPEG were doing a much better job.
https://github.com/pmjdebruijn/colormatch
https://www.darktable.org/2016/05/colour-manipulation-with-the-colour-checker-lut-module/
The colormatch trick gave me almost the same JPEG as the default
Darktable (i.e. with a greenish tint).
I almost gave up when I noticed that (while following the colormatch
procedure) changing the input color profile to "RGB rec2020 linear"
suddenly got rid of the greenish tint.
So now with a base curve "Neutral" and an input color profile to "RGB
rec2020 linear", I have almost perfect colors in every conditions (I
checked on many pictures).
Could someone explain to me what I have done and why it works ?
____________________________________________________________________________
darktable user mailing list

[Tomas Sobek]

Hi,
Just in case anyone is still interested in this thread, I have written up some thoughts and experiences with creating camera input profiles:

* My version of the procedure as per video below, but with newer Darktable version. Also ensuring the colour profiles created are well behaved (see http://ninedegreesbelow.com/photography/well-behaved-profile.html ).
* Experiences from following Elle Stone's great tutorial (see http://ninedegreesbelow.com/photography/well-behaved-camera-profile.html ).
* Couple of my Nikon D7100 profiles for download.

My articles are available here - https://tomassobekphotography.co.nz/articles.php
Please let me know if you notice any inaccuracies.
Cheers, Tomas

Post by I. Ivanov
I stumbled on the following
http://youtu.be/GkBgFaSv1kE

[Saint Germain]

On Thu, 27 Oct 2016 08:04:14 +0200 (CEST), "Tomas Sobek"

Post by Tomas Sobek
Hi,
Just in case anyone is still interested in this thread, I have
written up some thoughts and experiences with creating camera input
* My version of the procedure as per video below, but with newer
Darktable version. Also ensuring the colour profiles created are well
behaved (see
http://ninedegreesbelow.com/photography/well-behaved-profile.html ).
* Experiences from following Elle Stone's great tutorial (see
http://ninedegreesbelow.com/photography/well-behaved-camera-profile.html
).
* Couple of my Nikon D7100 profiles for download.
My articles are available here -
https://tomassobekphotography.co.nz/articles.php Please let me know
if you notice any inaccuracies. Cheers, Tomas

Hello,

Thanks for the info.
I see that we had a similar approach.
I was also unsatisfied with the rendered color with my D7200 and I also
bought the IT8 target and the Colormunki Display to try to correct them.

I have more or less given up on the issue and adopted this approach:
https://www.mail-archive.com/darktable-***@lists.darktable.org/msg01213.html

As D7100 and D7200 are more or less the same, I've tried your color
input profiles but I haven't managed to get any improvement with my
difficult shots (inside shot mainly).

For example:
- JPG: http://dl.free.fr/oveYkaeh3
- NEF: http://dl.free.fr/g2xLlF4lQ

In the above example, pay attention to the white wall on the left and
the color of the head of the bed (light brown).
I haven't managed to correct the colors in order to have something like
the JPEG (which gives the correct color as I see it with my own eyes).

Are you using your color profile on top of a base curve ?

Thanks for your interesting article and for this message.
I feel a little less alone !

[Tomas Sobek]

Hi Saint,


"I was also unsatisfied with the rendered color with my D7200 and I also
bought the IT8 target and the Colormunki Display to try to correct them.
"


Having calibrated your monitor is a good start. This way you should see on
screen a reasonable representation of your image, similar to how other
people with calibrated screens will see it.



As for the camera input profile, I was often tweaking green tint in white
balance settings in the past. This was my primary motivation, to see if I
can get a better starting point with less work. I used the profiles I shared
on a few occasions (all outdoors) and they seem to be working reasonably
well for me. If I were to shoot in difficult light situation, I carry with
me ColourChecker Passport and I would take a shot of it in that specific
light. Then create a profile specific to this situation as I described in
this article
(https://tomassobekphotography.co.nz/articles/create-camera-input-profile-with-ColorChecker-Passport.php)
- Once you have things setup, it only takes a few minutes.


"Are you using your color profile on top of a base curve ?"
I am using the camera input profile only to alter relationships between
colours (hence using Linear Rec709 RGB input/output profiles when processing
target shot during the profile creation). I still apply base curve and all
the other adjustments to produce an image I like. With landscape photography
there is a lot of scope to balance dark and bright parts of the image
(graduated density filter for bright sky for example). Another of my
favourites is zone system, or subtle tweaks with levels on parts of the
image (I learned that technique from this ebook
(https://resources.digital-photography-school.com/download/landscapes-book/)
).

"For example:
- JPG: http://dl.free.fr/oveYkaeh3
- NEF: http://dl.free.fr/g2xLlF4lQ

""In the above example, pay attention to the white wall on the left and
the color of the head of the bed (light brown).
I haven't managed to correct the colors in order to have something like
the JPEG (which gives the correct color as I see it with my own eyes).
"
I tried a quick edit of your image to get reasonably close (without being
pedantic) to the Jpeg from camera. I have included both resulting Jpeg and
xmp file here - https://www.dropbox.com/s/1nzicnmuaazhule/DSC_4860.zip?dl=0
(https://www.dropbox.com/s/1nzicnmuaazhule/DSC_4860.zip?dl=0) - Primarily I
used spot white balance on lighter grey patches of IT8 target, then turned
colour temperature a bit down. Increase exposure. A bit less saturation and
some noise reduction; at ISO 2000 the image seems to have a lot of noise to
my liking. Your lens wasn't recognised on my system so I couldn't apply lens
correction.

I hope you find my comments useful. But I don't claim to have much
experience - I am just learning as I go ;)
Cheers, Tomas

[Saint Germain]

On Fri, 28 Oct 2016 09:52:40 +0200 (CEST), "Tomas Sobek"

Post by Tomas Sobek
"I was also unsatisfied with the rendered color with my D7200 and I
also bought the IT8 target and the Colormunki Display to try to
correct them. "
Having calibrated your monitor is a good start. This way you should
see on screen a reasonable representation of your image, similar to
how other people with calibrated screens will see it.
As for the camera input profile, I was often tweaking green tint in
white balance settings in the past. This was my primary motivation,
to see if I can get a better starting point with less work. I used
the profiles I shared on a few occasions (all outdoors) and they seem
to be working reasonably well for me. If I were to shoot in difficult
light situation, I carry with me ColourChecker Passport and I would
take a shot of it in that specific light. Then create a profile
specific to this situation as I described in this article
(https://tomassobekphotography.co.nz/articles/create-camera-input-profile-with-ColorChecker-Passport.php)
- Once you have things setup, it only takes a few minutes.
"Are you using your color profile on top of a base curve ?"
I am using the camera input profile only to alter relationships
between colours (hence using Linear Rec709 RGB input/output profiles
when processing target shot during the profile creation). I still
apply base curve and all the other adjustments to produce an image I
like. With landscape photography there is a lot of scope to balance
dark and bright parts of the image (graduated density filter for
bright sky for example). Another of my favourites is zone system, or
subtle tweaks with levels on parts of the image (I learned that
technique from this ebook
(https://resources.digital-photography-school.com/download/landscapes-book/)
).
- JPG: http://dl.free.fr/oveYkaeh3
- NEF: http://dl.free.fr/g2xLlF4lQ
""In the above example, pay attention to the white wall on the left
and the color of the head of the bed (light brown).
I haven't managed to correct the colors in order to have something
like the JPEG (which gives the correct color as I see it with my own
eyes). "
I tried a quick edit of your image to get reasonably close (without
being pedantic) to the Jpeg from camera. I have included both
resulting Jpeg and xmp file here
- https://www.dropbox.com/s/1nzicnmuaazhule/DSC_4860.zip?dl=0
(https://www.dropbox.com/s/1nzicnmuaazhule/DSC_4860.zip?dl=0) -
Primarily I used spot white balance on lighter grey patches of IT8
target, then turned colour temperature a bit down. Increase exposure.
A bit less saturation and some noise reduction; at ISO 2000 the image
seems to have a lot of noise to my liking. Your lens wasn't
recognised on my system so I couldn't apply lens correction.
I hope you find my comments useful. But I don't claim to have much
experience - I am just learning as I go ;)
Cheers, Tomas

Thanks for the attempt !
However my point was that I couldn't make such effort for each picture
as it is consuming too much time (especially when the JPEG gives
accurate color directly).
I was (naively) hoping that I could find a 'generic' processing with
Darktable which can be applicable on all NEF and gives 'good enough'
result, but that it is simply not working.
For many pictures, I find myself having to individually work on them
and I simply don't have the time and motivation.

So I ended up just working with the Nikon JPEG and only in some extreme
cases, I am using Darktable.

I'm even considering changing my camera to have one which gives better
default results in Darktable !

Best regards,

[KOVÁCS István]

Hi,

I experience the same thing - Nikon D7000. No matter which input
profile I use (standard, enhanced, and an ICC I have converted from a
RawStudio DCP), no matter which base or tone curve. I don't get the
issue when opening the images in RawTherapee. Ref2020 as an input
profile gives much improved colours. If anyone is interested, I'll
upload several versions (in-camera JPG thumbnail, Darktable,
RawTherapee). Darktable is v2.0.5 from Pascal's PPA, OS is Mint;
display is calibrated (but makes just a small difference, it's quite
close to sRGB, so even without loading the profile colours are
decent). OpenCL enabled/disabled, LittleCMS for output profile
enabled/disabled make no difference in this regard.

Kofa

Post by Saint Germain
I have no idea what I am doing... but it works !
After struggling to get the correct color from my Nikon D7200 with
Darktable, I finally managed to get a near-perfect result.
It was frustrating because the Nikon JPEG were doing a much better job.
https://github.com/pmjdebruijn/colormatch
https://www.darktable.org/2016/05/colour-manipulation-with-the-colour-checker-lut-module/
The colormatch trick gave me almost the same JPEG as the default
Darktable (i.e. with a greenish tint).
I almost gave up when I noticed that (while following the colormatch
procedure) changing the input color profile to "RGB rec2020 linear"
suddenly got rid of the greenish tint.
So now with a base curve "Neutral" and an input color profile to "RGB
rec2020 linear", I have almost perfect colors in every conditions (I
checked on many pictures).
Could someone explain to me what I have done and why it works ?
____________________________________________________________________________
darktable user mailing list

[Saint Germain]

Hello,

D7200 and D7000 should be rather similar, so I think we can help each
other !
I will test with RawTherapee but it seems that D7200 is only supported
with quite new version which is not available on Debian (so I'll have
to compile it from sources).
I am interested in your several examples, can you upload them
somewhere ?

I have made some progress with Darktable and my D7200 but I am still
not satisfied.
Using Ref2020 as input profile is better, but I found later that it
gives a reddish tint on skin color which is not pleasant for everybody.

Regards,

On Wed, 20 Jul 2016 07:16:55 +0200, KOVÁCS István

Post by KOVÁCS István
Hi,
I experience the same thing - Nikon D7000. No matter which input
profile I use (standard, enhanced, and an ICC I have converted from a
RawStudio DCP), no matter which base or tone curve. I don't get the
issue when opening the images in RawTherapee. Ref2020 as an input
profile gives much improved colours. If anyone is interested, I'll
upload several versions (in-camera JPG thumbnail, Darktable,
RawTherapee). Darktable is v2.0.5 from Pascal's PPA, OS is Mint;
display is calibrated (but makes just a small difference, it's quite
close to sRGB, so even without loading the profile colours are
decent). OpenCL enabled/disabled, LittleCMS for output profile
enabled/disabled make no difference in this regard.
Kofa

Post by Saint Germain
I have no idea what I am doing... but it works !
After struggling to get the correct color from my Nikon D7200 with
Darktable, I finally managed to get a near-perfect result.
It was frustrating because the Nikon JPEG were doing a much better
https://github.com/pmjdebruijn/colormatch
https://www.darktable.org/2016/05/colour-manipulation-with-the-colour-checker-lut-module/
The colormatch trick gave me almost the same JPEG as the default
Darktable (i.e. with a greenish tint).
I almost gave up when I noticed that (while following the colormatch
procedure) changing the input color profile to "RGB rec2020 linear"
suddenly got rid of the greenish tint.
So now with a base curve "Neutral" and an input color profile to
"RGB rec2020 linear", I have almost perfect colors in every
conditions (I checked on many pictures).
Could someone explain to me what I have done and why it works ?
____________________________________________________________________________
darktable user mailing list
to unsubscribe send a mail to

[Matthieu Moy]

Hi,

I have the same issue. On some pictures, Rec2020 + neutral base curve is
OK, but in many cases I also find it far too redish. Rec709 RGB gives
better results IMO.

I tried this tutorial to get a color profile directly from Nikon:

Unfortunately, darktalbe is not able to open the generated .icm files
(unsupported input profile has been replaced by linear Rec709 RGB).
Anybody got this working for a D7200?

I ended up making a "color balance" preset, manually, by trial-and-error
+ comparison between dt's output and the in-camera's JPEG on a few
images:

Lift:
Red, Green, Blue: -0.01

Gamma:
Green: -0.01
Blue: 0.02

Gain:
Green: -0.01
Blue: 0.02

I still need to play a bit more with it, but it seems like a sane
default for D7200.

Cheers,

Post by Saint Germain
Hello,
D7200 and D7000 should be rather similar, so I think we can help each
other !
I will test with RawTherapee but it seems that D7200 is only supported
with quite new version which is not available on Debian (so I'll have
to compile it from sources).
I am interested in your several examples, can you upload them
somewhere ?
I have made some progress with Darktable and my D7200 but I am still
not satisfied.
Using Ref2020 as input profile is better, but I found later that it
gives a reddish tint on skin color which is not pleasant for everybody.
Regards,
On Wed, 20 Jul 2016 07:16:55 +0200, KOVÁCS István

Post by KOVÁCS István
Hi,
I experience the same thing - Nikon D7000. No matter which input
profile I use (standard, enhanced, and an ICC I have converted from a
RawStudio DCP), no matter which base or tone curve. I don't get the
issue when opening the images in RawTherapee. Ref2020 as an input
profile gives much improved colours. If anyone is interested, I'll
upload several versions (in-camera JPG thumbnail, Darktable,
RawTherapee). Darktable is v2.0.5 from Pascal's PPA, OS is Mint;
display is calibrated (but makes just a small difference, it's quite
close to sRGB, so even without loading the profile colours are
decent). OpenCL enabled/disabled, LittleCMS for output profile
enabled/disabled make no difference in this regard.
Kofa

Post by Saint Germain
I have no idea what I am doing... but it works !
After struggling to get the correct color from my Nikon D7200 with
Darktable, I finally managed to get a near-perfect result.
It was frustrating because the Nikon JPEG were doing a much better
https://github.com/pmjdebruijn/colormatch
https://www.darktable.org/2016/05/colour-manipulation-with-the-colour-checker-lut-module/
The colormatch trick gave me almost the same JPEG as the default
Darktable (i.e. with a greenish tint).
I almost gave up when I noticed that (while following the colormatch
procedure) changing the input color profile to "RGB rec2020 linear"
suddenly got rid of the greenish tint.
So now with a base curve "Neutral" and an input color profile to
"RGB rec2020 linear", I have almost perfect colors in every
conditions (I checked on many pictures).
Could someone explain to me what I have done and why it works ?
____________________________________________________________________________
darktable user mailing list
to unsubscribe send a mail to

____________________________________________________________________________
darktable user mailing list

[Saint Germain]

On Sun, 14 Aug 2016 16:53:09 +0200, Matthieu Moy

Post by Matthieu Moy
Hi,
I have the same issue. On some pictures, Rec2020 + neutral base curve
is OK, but in many cases I also find it far too redish. Rec709 RGB
gives better results IMO.
http://youtu.be/O1nS3URVSYA
Unfortunately, darktalbe is not able to open the generated .icm files
(unsupported input profile has been replaced by linear Rec709 RGB).
Anybody got this working for a D7200?
I ended up making a "color balance" preset, manually, by
trial-and-error
+ comparison between dt's output and the in-camera's JPEG on a few
Red, Green, Blue: -0.01
Green: -0.01
Blue: 0.02
Green: -0.01
Blue: 0.02
I still need to play a bit more with it, but it seems like a sane
default for D7200.

Hello,

Please see the thread about Nikon and Darktable there:
https://www.mail-archive.com/darktable-***@lists.darktable.org/msg00805.html

I also gave up the Rec2020 trick when I saw the red shift in some
pictures (I miss it the first time).

I managed to get quite good result with my D7200 with the Nikon input
color profile for D7100 given in this thread.
Unfortunately I also had the same problem as you when I tried to
generated my input color profile (see the thread).

Can you generate and upload a .dtstyle that give you good result ?
I tried applying the color balance you gave, but the result is quite
incorrect (perhaps are you using other base curve or something).

Regards,

[Matthieu Moy]

Post by Saint Germain
Hello,

Thanks, I missed this thread indeed.

Post by Saint Germain
I managed to get quite good result with my D7200 with the Nikon input
color profile for D7100 given in this thread.

That's surprising: on some images the D7100 color profile gives good
results at least color-wise, but in many cases it also flattens
completely the contrast. I could get a close-to-camera JPEG result by
drawing a slightly S-shape base curve manually for one picture, but then
it's completely wrong on another one. So, I still don't have a "one fits
all" receipe :-(.

Post by Saint Germain
Can you generate and upload a .dtstyle that give you good result ?

Attached is an XMP file.

[Saint Germain]

On Mon, 15 Aug 2016 11:41:36 +0200, Matthieu Moy

Post by Matthieu Moy

Post by Saint Germain
Hello,

Thanks, I missed this thread indeed.

Post by Saint Germain
I managed to get quite good result with my D7200 with the Nikon
input color profile for D7100 given in this thread.

That's surprising: on some images the D7100 color profile gives good
results at least color-wise, but in many cases it also flattens
completely the contrast. I could get a close-to-camera JPEG result by
drawing a slightly S-shape base curve manually for one picture, but
then it's completely wrong on another one. So, I still don't have a
"one fits all" receipe :-(.

Indeed I am also looking for the single recipe which will allow me to
have a defaut satisfying behavior.
So far I can say that the JPEG given by Nikon gives a satisfying result
(far from perfect but acceptable). With Darktable and the different
presets I have tried, I always have to modify something manually,
which is taking me too much time.

I have to emphasize again, that I am NOT especially trying to duplicate
Nikon JPEG. It's just that it gives an acceptable result and it serves
as a point of reference (especially for color tone) in order to achieve
at least a similar result with Darktable (because I always have a
blueish/greenish tint so far).

With the trick to use the .icm profile generated by ViewNX 2/Capture
NX-D, I agree that the result is a bit "flat" (but with almost correct
colors). I can of course modify it, but this modification cannot be
used on another pictures (same problem as you).

Can I offer to try to work on the same set of pictures ?
I already uploaded an example here (see last comment) :
https://redmine.darktable.org/issues/11053#note-11
Perhaps can you also upload somewhere the pictures which are difficult
to process by Darktable ?

Post by Matthieu Moy

Post by Saint Germain
Can you generate and upload a .dtstyle that give you good result ?

Attached is an XMP file.

This style gives me a rather incorrect result in the example I have
provided.

[Colin Adams]

Perhaps making a custom camera input profile, as detailed at
http://ninedegreesbelow.com/photography/camera-profile-make-target-shot.html is
worth pursuing?

Post by Saint Germain
On Mon, 15 Aug 2016 11:41:36 +0200, Matthieu Moy

Post by Matthieu Moy

Post by Saint Germain
Hello,

Thanks, I missed this thread indeed.

Post by Saint Germain
I managed to get quite good result with my D7200 with the Nikon
input color profile for D7100 given in this thread.

That's surprising: on some images the D7100 color profile gives good
results at least color-wise, but in many cases it also flattens
completely the contrast. I could get a close-to-camera JPEG result by
drawing a slightly S-shape base curve manually for one picture, but
then it's completely wrong on another one. So, I still don't have a
"one fits all" receipe :-(.

Indeed I am also looking for the single recipe which will allow me to
have a defaut satisfying behavior.
So far I can say that the JPEG given by Nikon gives a satisfying result
(far from perfect but acceptable). With Darktable and the different
presets I have tried, I always have to modify something manually,
which is taking me too much time.
I have to emphasize again, that I am NOT especially trying to duplicate
Nikon JPEG. It's just that it gives an acceptable result and it serves
as a point of reference (especially for color tone) in order to achieve
at least a similar result with Darktable (because I always have a
blueish/greenish tint so far).
With the trick to use the .icm profile generated by ViewNX 2/Capture
NX-D, I agree that the result is a bit "flat" (but with almost correct
colors). I can of course modify it, but this modification cannot be
used on another pictures (same problem as you).
Can I offer to try to work on the same set of pictures ?
https://redmine.darktable.org/issues/11053#note-11
Perhaps can you also upload somewhere the pictures which are difficult
to process by Darktable ?

Post by Matthieu Moy

Post by Saint Germain
Can you generate and upload a .dtstyle that give you good result ?

Attached is an XMP file.

This style gives me a rather incorrect result in the example I have
provided.
____________________________________________________________________________
darktable user mailing list
to unsubscribe send a mail to

[Saint Germain]

On Thu, 18 Aug 2016 15:26:53 +0000, Colin Adams

Post by Colin Adams
Perhaps making a custom camera input profile, as detailed at
http://ninedegreesbelow.com/photography/camera-profile-make-target-shot.html
is worth pursuing?

I already tried this approach with tht IT8 target:
https://github.com/pmjdebruijn/colormatch
It is more or less the same as the one described on Darktable website:
https://www.darktable.org/2016/05/colour-manipulation-with-the-colour-checker-lut-module/

The result always has the green tint on it.

[Colin Adams]

Hm. I get a nice result with neutral base curve and rgb rec 2020 linear too.

Post by Saint Germain
On Thu, 18 Aug 2016 15:26:53 +0000, Colin Adams

Post by Colin Adams
Perhaps making a custom camera input profile, as detailed at

http://ninedegreesbelow.com/photography/camera-profile-make-target-shot.html

Post by Colin Adams
is worth pursuing?

https://github.com/pmjdebruijn/colormatch
https://www.darktable.org/2016/05/colour-manipulation-with-the-colour-checker-lut-module/
The result always has the green tint on it.
____________________________________________________________________________
darktable user mailing list
to unsubscribe send a mail to

[Colin Adams]

But if I use Nikon D810 base curve instead of neutral, I get a pleasing
effect also. Decidedly vivid though.

Post by Colin Adams
Hm. I get a nice result with neutral base curve and rgb rec 2020 linear too.

Post by Saint Germain
On Thu, 18 Aug 2016 15:26:53 +0000, Colin Adams

Post by Colin Adams
Perhaps making a custom camera input profile, as detailed at

http://ninedegreesbelow.com/photography/camera-profile-make-target-shot.html

Post by Colin Adams
is worth pursuing?

https://github.com/pmjdebruijn/colormatch
https://www.darktable.org/2016/05/colour-manipulation-with-the-colour-checker-lut-module/
The result always has the green tint on it.
____________________________________________________________________________
darktable user mailing list
to unsubscribe send a mail to

[Saint Germain]

Tried looking at the skin tone of people.
In my case it is a bit too reddish/pinkish in a lot of situations (but
not every situation).

On Thu, 18 Aug 2016 18:51:15 +0000, Colin Adams

Post by Colin Adams
But if I use Nikon D810 base curve instead of neutral, I get a
pleasing effect also. Decidedly vivid though.

Post by Colin Adams
Hm. I get a nice result with neutral base curve and rgb rec 2020 linear too.

Post by Saint Germain
On Thu, 18 Aug 2016 15:26:53 +0000, Colin Adams

Post by Colin Adams
Perhaps making a custom camera input profile, as detailed at

http://ninedegreesbelow.com/photography/camera-profile-make-target-shot.html

Post by Colin Adams
is worth pursuing?

https://github.com/pmjdebruijn/colormatch
https://www.darktable.org/2016/05/colour-manipulation-with-the-colour-checker-lut-module/
The result always has the green tint on it.
____________________________________________________________________________
darktable user mailing list
to unsubscribe send a mail to

[Colin Adams]

Well, I don't have that problem, as I only photograph dragonflies :-)

Post by Saint Germain
Tried looking at the skin tone of people.
In my case it is a bit too reddish/pinkish in a lot of situations (but
not every situation).
On Thu, 18 Aug 2016 18:51:15 +0000, Colin Adams

Post by Colin Adams
But if I use Nikon D810 base curve instead of neutral, I get a
pleasing effect also. Decidedly vivid though.

Post by Colin Adams
Hm. I get a nice result with neutral base curve and rgb rec 2020 linear too.

Post by Saint Germain
On Thu, 18 Aug 2016 15:26:53 +0000, Colin Adams

Post by Colin Adams
Perhaps making a custom camera input profile, as detailed at

http://ninedegreesbelow.com/photography/camera-profile-make-target-shot.html

Post by Colin Adams

Post by Colin Adams

Post by Saint Germain

Post by Colin Adams
is worth pursuing?

https://github.com/pmjdebruijn/colormatch

https://www.darktable.org/2016/05/colour-manipulation-with-the-colour-checker-lut-module/

Post by Colin Adams

Post by Colin Adams

Post by Saint Germain
The result always has the green tint on it.

____________________________________________________________________________

Post by Colin Adams

Post by Colin Adams

Post by Saint Germain
darktable user mailing list
to unsubscribe send a mail to

____________________________________________________________________________
darktable user mailing list
to unsubscribe send a mail to

[Ralf Kerkhoff]

Hi,
I am using the Gimp Lua script from:
https://github.com/wpferguson/lua-scripts/tree/master/contrib
which works fine as far as the image exported to Gimp is somewhere in
/home.
For example /home/ralf/pictures
After editing in Gimp the image is imported into darktable and grouped
with the original image.
Fine, so far.
If the original image is not in the home directory - all my pictures
are on a separate harddisk which is mounted as /work, the export script
works fine too:
image is exported, Gimp launched etc.
But the edited image will not be copied to the original location. 
Seems that there is a problem with the copy process when the original
picture (the one exported to Gimp) is not in a folder within my home
directory.
Not a great problem - but I am wondering if I am missing something.
Cheers
Ralf

[William Ferguson]

Hi Ralf,

Congratulations, you have found an "unintended feature". I use the lua
os.rename function to move the file. It appears that this function will
not work across disk boundaries. I tried an nfs mount and a local mount
and wasn't successful with either. I could write outside the home
directory as long as it was on the same physical disk.

I'm working on it.

Thanks,

Bill

On Thu, Aug 18, 2016 at 1:28 PM, Ralf Kerkhoff <

Post by Ralf Kerkhoff
Hi,
https://github.com/wpferguson/lua-scripts/tree/master/contrib
which works fine as far as the image exported to Gimp is somewhere in
/home.
For example /home/ralf/pictures
After editing in Gimp the image is imported into darktable and grouped
with the original image.
Fine, so far.
If the original image is not in the home directory - all my pictures are
on a separate harddisk which is mounted as /work, the export script works
image is exported, Gimp launched etc.
But the edited image will not be copied to the original location.
Seems that there is a problem with the copy process when the original
picture (the one exported to Gimp) is not in a folder within my home
directory.
Not a great problem - but I am wondering if I am missing something.
Cheers
Ralf
____________________________________________________________________________
darktable user mailing list to unsubscribe send a mail to

[Johann Spies]

Thanks for reference to these scripts. I did not know about them. So I
installed it from github.

Unfortunately the gimp script does not work as described. I get the
message "Launching gimp" , but gimp is not launched and the exported images
are dumped in ~/.local/tmp/

I can edit them with gimp called up manually, but then the rest of the
process promised by the script does not happen.

This is with Darktable 2.0.5 on Debian Testing/Sid.

Regards
Johann

[William Ferguson]

Make sure that the line

require "official/yield"

is also in your luarc file. If that doesn't fix it, open a terminal and
run 'darktable -d lua'. That will print out information in the terminal
about what is happening.

Regards,

Bill

Post by Johann Spies
Thanks for reference to these scripts. I did not know about them. So I
installed it from github.
Unfortunately the gimp script does not work as described. I get the
message "Launching gimp" , but gimp is not launched and the exported images
are dumped in ~/.local/tmp/
I can edit them with gimp called up manually, but then the rest of the
process promised by the script does not happen.
This is with Darktable 2.0.5 on Debian Testing/Sid.
Regards
Johann
--
Because experiencing your loyal love is better than life itself,
my lips will praise you. (Psalm 63:3)
____________________________________________________________________________
darktable user mailing list to unsubscribe send a mail to

[William Ferguson]

I ran into the same problem the other night. I was working on the gimp.lua
file and commented everything else out of my luarc file. I had the debug
going so I saw the problem, but it took a little while to track down the
problem. Not long after that I ran across a post by supertobi, in another
forum, explaining the problem and the solution.


Glad it worked,

Bill

Post by William Ferguson
Make sure that the line
require "official/yield"
is also in your luarc file. If that doesn't fix it, open a terminal and
run 'darktable -d lua'. That will print out information in the terminal
about what is happening.
Regards,
Bill

Post by Johann Spies
Thanks for reference to these scripts. I did not know about them. So I
installed it from github.
Unfortunately the gimp script does not work as described. I get the
message "Launching gimp" , but gimp is not launched and the exported images
are dumped in ~/.local/tmp/
I can edit them with gimp called up manually, but then the rest of the
process promised by the script does not happen.
This is with Darktable 2.0.5 on Debian Testing/Sid.
Regards
Johann
--
Because experiencing your loyal love is better than life itself,
my lips will praise you. (Psalm 63:3)
____________________________________________________________________________
darktable user mailing list to unsubscribe send a mail to