yeah i guess the halo thing would best be addressed by implementing this:
http://people.csail.mit.edu/sparis/publi/2011/siggraph/
(if i wasn't so lazy..)
-jo
On Mon, Feb 15, 2016 at 9:39 AM, Matthieu Moy
wrote:
> David Vincent-Jones writes:
>
>> Thanks for the ideas: yes ... I was aware of us
David Vincent-Jones writes:
> Thanks for the ideas: yes ... I was aware of using a second instance
> and I do use this approach with several modules.
>
> The low pass filter looks interesting will also give that a try.
Some real-life examples on harry durgin's recent screencasts:
https://w
Thanks for the ideas: yes ... I was aware of using a second instance and
I do use this approach with several modules.
The low pass filter looks interesting will also give that a try.
David
On 14/02/16 09:08 PM, Matthieu Moy wrote:
> David Vincent-Jones writes:
>
>> I think that the proble
David Vincent-Jones writes:
> I think that the problem would be less difficult if S&H were broken into
> 2 modules and wonder if others might consider this a good idea.
2 modules doing what?
If you want one module for the shadows, and one for the highlights, then
you can already do that: just
I am still learning about how to properly use darktable, but one idea that
comes to my mind here, is to create a new instance of the S&H modules and
apply to them only the effects you need, so you can in this way separate
what you want to change in each. But anyway, not sure how useful this might
b
I use S&H extensively and in most cases I am able to readily mitigate
the 'halo' effect where significantly different tones collide. There are
however times where it is impossible to fully correct the problem. The
best solution that I have found so far is to use a parametric mask to
allow the h
Hans Petter,
When I started this process I forgot I had any TIFF files in the database,
so I didn't think your comment applied to my situation. Apparently there
are some files in there I had forgotten about (thank goodness the database
has a better memory than I do :-) I think you are absolutely c
As I said in a previous reply, I get these all the time when I open tiffs
from Darktable in Gimp. My tiffs are generated from Canon CR2 raw files, so
this is probably not Nikon specific.
Hans Petter
http://hpbirkeland.com
2016-02-14 18:06 GMT+01:00 Pedro Côrte-Real :
> On Sun, Feb 14, 2016 at 5
On Sun, Feb 14, 2016 at 5:22 PM, Lefty Ace wrote:
> So, it appears that TIFF files generated by darktable contain the unknown
> field issue. Is that because they came from NEFs with crazy manufacturer
> fields? And darktable (or a library it is using) is suppressing those
> messages when processin
Maybe this should go into a new thread (and maybe it doesn't matter at all).
I've stumbled across some more information. It appears the messages are
triggered by TIFF files that came from NEFs. Here is what I see:
All of the TIFF files in my database are of the pattern %Arboretum.tif (I
only have
Ah, I see. Well, if I ever figure out which files it was I'll investigate
further, but for now it seems like something I'm safe to ignore.
-Owen
On Sun, Feb 14, 2016 at 11:06 AM, Pedro Côrte-Real
wrote:
> On Sun, Feb 14, 2016 at 5:00 PM, Lefty Ace wrote:
> > Regarding the TIFF issue, I did som
On Sun, Feb 14, 2016 at 5:00 PM, Lefty Ace wrote:
> Regarding the TIFF issue, I did some digging and I actually have a few TIFF
> files in my database that I forgot about from a panorama experiment. The
> messages are most likely not coming from RAW files at all.
That is unlikely actually. TIFF f
Thanks again for your reply, Johannes.
> i hope that's the old configuration and not the new one, right?
>
You are correct, I wrote the wrong thing in my original message. My config
file now says:
never_use_embedded_thumb=TRUE
plugins/lighttable/low_quality_thumbnails=FALSE
Regarding the TIFF
On Sun, Feb 14, 2016 at 4:34 PM, johannes hanika wrote:
>> TIFFReadDirectory: Warning, Unknown field with tag 37398 (0x9216)
>> encountered.
>>
>> Any idea what that means?
>
> no but that should be relatively harmless (probably comes from
> non-standard tags in tiff that is actually a raw or simi
heya,
On Sun, Feb 14, 2016 at 12:17 PM, Lefty Ace wrote:
> Thanks for your reply, Johannes.
>
> I checked darktablerc and both of those options were the opposite of what
> you listed. I changed them, so now I have:
>
> never_use_embedded_thumb=FALSE
> plugins/lighttable/low_quality_thumbnails=TRU
On Sun, Feb 14, 2016, 14:15 Patrick Shanahan wrote:
> * Carlos Maiolino [02-14-16 07:49]:
> > Ok, I finally figured out what is happening:
> >
> > For my understanding, I can't compare the pictures from darktable with
> > another image viewer, instead I do a 1:1 ratio comparison, otherwise
> > t
* Carlos Maiolino [02-14-16 07:49]:
> Ok, I finally figured out what is happening:
>
> For my understanding, I can't compare the pictures from darktable with
> another image viewer, instead I do a 1:1 ratio comparison, otherwise
> the picture can be very different. Depending on the software's
> a
Ok, I finally figured out what is happening:
For my understanding, I can't compare the pictures from darktable with
another image viewer, instead I do a 1:1 ratio comparison, otherwise
the picture can be very different. Depending on the software's
algorithm used to display the image, it will inter
Hola tocayo,
Only for Web, I compare jpeg with Firefox (not for printing, of course).
Firefox is capable of managing color profiles like sRGB, AdobeRGB and
the like.
We'll see what Patrick Shanahan says
Carlos
El 14/02/16 a las 08:57, Carlos Maiolino escribió:
BTW Patrick, how are you compa
BTW Patrick, how are you comparing both images?
I've been using Image Viewer to open jpeg files, not sure if this make
much difference. But I got a bit annoyed with it, mostly because I'm
quite new on photography processing and I have no idea how to debug
this issue :) But I'm still seeing very di
On Sun, Feb 14, 2016 at 5:14 AM, Patrick Shanahan wrote:
> * Carlos Maiolino [02-13-16 19:54]:
>> Definitely something is wrong. I think darktable is not applying
>> denoise modules on top of some other module. I don't have time to
>> figure out not what is going on, but there is something wrong
21 matches
Mail list logo
