Features of OMERO.movie

Good morning,

we started recently to use OMERO.movie in our omero instance and pretty like the provided functionality. A few issues arose after the first tries:

  1. For us it would be useful to be able to change or decide about the position of the placed labels. Maybe this is something for future versions?
  2. Are there any recommendations about needed resources for the encoding process? We’ve observed a pretty high CPU load when creating movies from bigger files (e.g. 3D-movies). Are the encoding processes properly parallelized on multiple cores or do we not to be afraid, that too many parallel usages could basically freeze our system?
  3. Is it possible to kill specific encoding process via omero-web or on OS level? In case a process doesn’t seem to come to an end for example or is taken too many resources?

Thanks in advance,
Anna

Hi Anna,

Thanks for the feedback. The Make_Movie functionality is handled by a server-side script.

  1. I’ve created an issue at https://github.com/ome/omero-scripts/issues/177 but it is unlikely we’ll get around to this soon, so you might be able to have a go at this yourself? The code within the script to add labels is under add_plane_info() and add_time_points() https://github.com/ome/omero-scripts/blob/4a6eec0221bc9a5d2f33afd21c2f7e5a95179f87/omero/export_scripts/Make_Movie.py#L224 It’s quite easy to change the placing of those labels (and the scalebar) and to add the position parameter to the script, but the parameter and UI changes would also have to be added to the webclient or Insight UI that launches the script. (are you using webclient or Insight)?
    You could remove that script from the settings at https://docs.openmicroscopy.org/omero/5.6.0/sysadmins/config.html#omero-client-scripts-to-ignore to allow the webclient or Insight to show the script in the scripts menu and to generate a UI from the script parameters. This would allow you to use any new ‘label position’ parameter without needing any code changes to the clients.

  2. The script uses mencoder to build the movie from jpg or png frames but as far as I know there’s no logic to parallelise the process.

  3. The script will run on the server-side (not omero-web) and should be killable, but I would defer to @joshmoore for specific details on this.

Regards,
Will.

:+1: @ahamacher , if you want to go this route, see if you can find the process on the omero-server machine using ps. It should be under a parent process with Processor-0 in the command line:

[jamoore@ome-training-3 ~]$ ps auxw -H | grep Processor
jamoore  27070  0.0  0.0 112708   980 pts/0    S+   08:33   0:00           grep --color=auto Processor
omero-s+  3266  0.0  0.2 2003296 96516 ?       Sl   Oct14   1:52     python -m runProcessor --Ice.Config=/opt/omero/server/OMERO.server-5.6.3-ice36-b228/var/master/servers/Processor-0/config/config
            ^___ This is the parent process of any scripts.

~Josh

Good morning,

one more thought on omero.movie - it would be lovely if the generated movies would contain the defined ROIs. Or at least if the user could select if the ROIs should be included or not. That would be pretty useful for presentations. Or is there already a way and I just didn’t figure it out?

Thanks,
Anna

Hi Anna,

The Make_Movie scripts doesn’t support ROI overlays I’m afraid.
To add that we could maybe reuse some code in the OMERO.figure export script, which does much the same thing. Again, I created an issue at Make_Movie overlay ROIs · Issue #186 · ome/omero-scripts · GitHub to keep track of this feature request.
I could imagine this feature being quite large - e.g. users might want to choose particular ROIs instead of including them all. This reminds me of a long-standing idea to support movie export from OMERO.figure Export as movie · Issue #51 · ome/omero-figure · GitHub, but that would be an even bigger feature!
Regards,
Will.

1 Like