We closed this forum 18 June 2010. It has served us well since 2005 as the ALPHA forum did before it from 2002 to 2005. New discussions are ongoing at the new URL http://forum.processing.org. You'll need to sign up and get a new user account. We're sorry about that inconvenience, but we think it's better in the long run. The content on this forum will remain online.
Index › Programming Questions & Help › Video Capture,  Movie Playback,  Vision Libraries › High-performance video input - best method
‹ Previous Topic | Next Topic ›
Page Index Toggle Pages: 1
High-performance video input - best method? (Read 760 times)
peterkirn
High-performance video input - best method?
Jun 2nd, 2006, 8:26pm  
Okay, I've read the FAQs, I've played with the input library a bit, but nonetheless, here's my question . . .

I'm very unclear on which video input method is the best to use from a performance perspective. Obviously, qtjava is causing no end of trouble, from performance problems (with WinDVIG on Windows) to compatibility problems (particularly with QT7 on Mac) to a lack of support (on Linux). But it's not clear from previous threads whether JTF is in fact the best alternative. Has anyone got it working, ideally on Linux?

From a performance perspective, I find the previously-mentioned technique of processing in OpenGL using a slab to be very promising, and it's served me well in Jitter. Maybe this would be the best direction to investigate further?

I've got my choice of platforms here (Mac, Windows, Linux); I'm just unclear on which are the best directions to pursue and which are dead ends, and I suspect the bug list in FAQ is probably equally baffling to others.

I should add -- other than to share the results with folks informally, I do *not* need this to work inside a Web browser; I'd rather have it work in performance/installation settings.

Thanks for your advice, and I'll certainly share examples once I get something working.
Back to top
 
WWW SamLowry00  
dariosnd
Re: High-performance video input - best method?
Reply #1 - Jul 21st, 2006, 9:07am  
I'm looking for the best video performance too, that's what I've got so far:

I'm using LibCV on linux. It should be fast cause it relies on JMF which uses native libraries. In fact it is: with P3D I can bring a 160x120 stream to fullscreen at full rate and do some processing on it. But I needed more so I tried with OPENGL -- see this thread to let libcv and opengl work together:

http://processing.org/discourse/yabb_beta/YaBB.cgi?board=os_libraries_tools;action=display;num=1150234743

Surprisingly my script runs much better with P3D than with OPENGL. I modified the code of the opengl library to get rid of image interpolation and put some GL_FASTEST all around, but P3D it's still better. My video card it's not that fast, but it's accelerated. Maybe there's something wrong with jogl?

peterkin: how about your project?

I also wonder if anybody knows why P3D is so better than OPENGL on my computer.
Back to top
 
 
Page Index Toggle Pages: 1
‹ Previous Topic | Next Topic ›