save mpg from Volume Rendering Plugin

AFNI Message Board
1

Hello,

I am having trouble saving accumulated 3d rendered images (im trying to show the brain being sliced and rotated as a little video). I’ve tried clicking the save anim MPG button in the Display options menu and then click the Sav:mpeg button with saving from 0 to my #of images selected. It takes a while but it eventually outputs a mpg file but the file cant be opened and has “0 bites”

Thanks for your help!

2

Hmmm, I just did a test run on my computer and it worked.

What is the output of each of the following on your computer:


afni -ver

and


which mpeg_encode

?

Also, are there any error messages in the terminal when you try to save it?

Finally, how many images is that being saved? It might be a memory glitch or something if it is a huge video. Trying first with a few images might be a good test.

–pt

3

I am trying to save a giant file (629 images), but I just tried it with 10 and it was super fast but still 0 bytes. This is the output to the screen from running that:

++ Running ‘/home/applications/afni/abin/mpeg_encode -realquiet test.QWA.PARAM’ to produce test.mpg
. DONE

This is the output of which mpeg_encode:
/home/applications/afni/abin/mpeg_encode

This is the output of afni-ver:
Precompiled binary linux_openmp_64: Dec 22 2016 (Version AFNI_16.3.18)

I am working on my institution’s High computing cluster and don’t have permission or ability to update afni. I should probably try running this on my local machine where afni is up to date, that might be the problem.
Thanks so much for your help!!! :slight_smile:

4

Howdy-

Well, that is a veeeery old version of AFNI. I think it would really make sense to contact The Powers That Be on your local cluster and ask them nicely to update the binaries-- surely everyone will benefit? They should pretty much just need to run “@update.afni.binaries”; if worse comes to worst for them, they could also download the modern binary tarball for that version:
https://afni.nimh.nih.gov/pub/dist/tgz/linux_openmp_64.tgz
unpack it, and copy it to the same location.

I am guessing that a more modern version should work fine, but there might be some other odd compatability issue on the cluster (perhaps??).

–pt

5

Thanks for the info! I agree, I will definitely look into requesting an AFNI update from the PTB (Powers That Be).
All the rendering goes much faster using the most updated AFNI but I found that the 0 bytes problem is fixed by leaving the “blowup” option as 1 rather than changing to 8, as I was doing before. Thanks so much for the help and quick responses!

6

Hmm, interesting. The “blowup” feature if for saving the image at a higher res; it might not actually make much difference here, anyways.

I note that I also got a similar “0 byte” file phenomenon when trying to use a blowup factor of 8 when saving the rendered dataset, but that was signified with an error message about core dumping in the terminal:


++ Running '/home/ptaylor/afni_src/linux_ubuntu_12_64/mpeg_encode -realquiet test_save.VXJ.PARAM' to produce test_save.mpg
*** buffer overflow detected ***: /home/ptaylor/afni_src/linux_ubuntu_12_64/mpeg_encode terminated
Aborted (core dumped)
. **DONE**

This occurred even in my very modern version of AFNI. I assume that you actually might have gotten the same message?

Note that when I tried with a blowup factor of 2, 3 or 4, I could still save an MPG successfully, but any higher value lead to core dumping.

–pt