[PD] pix_snap CPU consumption question

Matteo Sisti Sette matteosistisette at gmail.com
Fri Nov 27 19:38:56 CET 2009


Hi,

I have a [pix_snap 0 0 2400 2400] 'under' a [gemframebuffer] of the same 
size. I think it would make no difference if there wasn't the 
[gemframebuffer] and the [gemwin] had the same size (namely 2400 by 2400).

I attach the patch I'm using, though a much simpler test patch could be 
   used to illustrate the question.

Now, I am observing the following:

- despite the large size of the framebuffer, even when pix_snap is 
continuously receiving the gemlist, CPU consumption is low, not 
significantly higher than the same patch with a smaller window size
- When I send one bang to pix_snap (by enabling the spigot above it 
which immediately disables itself), so that it takes a snapshot, then 
CPU consumption increases, and then it stays high forever!!!

I'm not surprised by the sudden CPU consumption increase: I'm surprised 
by the fact that then it doesn't immediately go back to its previous 
level: cpu consumption stays high, as if, starting from the moment the 
firse snaphot is taken, pix_snapshot had to continuously process the 
picture or something.

Is it normal???

I don't really understand why there should be a difference between the 
cpu consumption before and after the moment the snapshot is taken. I 
could understand a temporary increase _during_ the moment the snapshot 
is taken.

If I disconnect the left inlet of pix_snapshot so it does not receive 
the gemlist anymore, then CPU consumption drops again.
That wouldn't be that much surprising if it wasn't that after 
disconnecting the inlet of pix_snap, the image is still displayed. So, 
it means pix_snap does not need to "output" the image continuously (no 
surprise here either), but then, what is it continuously doing when 
receiving the gemlist and NOT capturing, that makes it consume CPU?? It 
seems that, whatever it is, it is useless.

Is it a bug? Or is there an explanation? (well the two things are not 
mutually exclusive)

-- 
Matteo Sisti Sette
matteosistisette at gmail.com
http://www.matteosistisette.com
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: prova_total.pd
URL: <http://lists.puredata.info/pipermail/pd-list/attachments/20091127/09f67bff/attachment.txt>


More information about the Pd-list mailing list