[digiKam-users] Slow raw processing

[Gilles Caulier]

Hi,

Raw processing is delegate in digiKam to libraw (https://www.libraw.org)

Libraw does not use video cards acceleration for compatibility and
stability purposes, but uses openmp to parallelize demosaicing
computation.

openmp support is optional in libraw. If your digiKam is not compiled
with openmp dependency, demosaicing will be processed on one core.

Check in Help/Components Info dialog in digiKam if openmp support is enabled.

Note: if you want to see libraw to support OpenCL and Co for video
card acceleration support, please report this wish on libraw github
project.

Best

Gilles Caulier

Le jeu. 5 nov. 2020 à 20:21, <digikam at 911networks.com> a écrit :
>
> On Thu, 5 Nov 2020 18:40:49 +0100
> Hans-Peter Huth <hans-ph at web.de> wrote:
>
> > Hi,
> >
> > i'm using digikam since years and are more and more using it for
> > everything. However opening a raw file is very slow compared to
> > other raw processors maybe because it uses only one core. I have a
> > Intel Core i5 desktop with on-board graphics and Manjaro Linux.
> > Would a dedicated GPU make a difference (for raw processing, i'm
> > not using face detection)? If so, any recommendations?
>
> I asked the same question a couple years ago and was told that there
> was nothing specific optimized for GPUs like darktable/opencl.
>
> But since I was using KDE (and still do) also on Manjaro, qt should
> realize that I have an radeon amdgpu, it should speedup things....
>
> --
> sknahT
>
> vyS