[Digikam-users] serious performance problem with large image collections

Benjamin Block bebl at mageta.org
Mon Mar 28 18:00:33 BST 2016


Hello digikam-users,

I don't really expect immediate help, but maybe can give a user-view for future
development.

I have quite a bit image-collection of around 320000 pictures stored on a NAS
(for raid redundancy and such). At the moment I use Google's Picasa to browse
and organize this, which works reasonable well (I have some problems with the
detection of new images and some image-formats are not supported well), but gets
outdated because development for that project stopped quite some time ago AFAIK.
As I don't want migrate ANY of this online, I am still searching for an
alternative, but the competition seems scares to me. So I knew about digikam on
linux and saw on the website and wikipedia that it is also available under
windows and tried it out (version 4.12.0).

The initial import took probably around 6 hours (I added the folder as
"Collection on Network Shares" in the options), after which it still had to
create thumbnails for every image. But the real fun begins when I want to
restart digikam after this.. because it takes a very long time - and I am not
talking about minutes here (maybe around 20 minutes). And even after that it
shows in the status-bar that some kind of other rescan is still running - again
for quite a long time. So, this is obviously completely unusable at this point
(I am sorry if that sounds harsh, but you have to be realistic here).

I have not found any option that would allow me to use a similar deferred scan
as in Picasa (so that digicam doesn't scan everything every time I start it
up-front). Maybe in combination with a feature that would allow me to re-scan a
small subset of this collection and thus also eliminate Picasa's shortcoming
with detecting new images timely. I know this can produce asynchronicities in
the collection (moved folders, deleted pictures, ...) but in such large
collections you only work on smaller subsets at a time anyway and with the added
"sub-scan" you might eliminate such smaller diffs quickly.

But maybe I also overlooked something. I think the project is really great,
because even commercially there is nothing really amazing that would serve me at
this point (which really surprises me) - nothing that I know of. Maybe you find
some time in the future to consider this.

Anyway, I hope this doesn't sound too much like a rant, its not intended as
such. Have a good day.


best regards,
- Benjamin



More information about the Digikam-users mailing list