[Theora-dev] Theora camera/fast computers

[Andrey Filippov]

> On Tue, Mar 22, 2005 at 02:35:47PM -0700, Andrey Filippov wrote:
>
>> The most obvious FPGA bugs in the camera are now fixed and it seems to
>> work rather stable (for now - just with clips, no streamer software
>> yet).
>
> Wow, it's really great to see this coming along. The simplified coding
> does interesting things with the camera noise.

What exactly effects did you notice about the noise? Maybe there are some
bugs?

>
> It's really great to see larger format footage in theora!
>
> BTW, the frame rate is given at 29.49 fps. Any particular reason you
> didn't pick a standard number like 30 or 29.97?

This will come. Currently when there is no restriction to the frame rate
placed (in a web interface when the frame rate input field is blank or 0)
systems tries to run sensor as fast as possible. With the default sensor
clock of 48MHz and required "margins" around the used window of
(1280+4)*(1024+4) it all results to that frame rate. It is possible to
program the sensor clock and also change "virtual frame" of the sensor by
adding invisible rows and/or columns.
> It's not looking good. Derf's experimental decoder with Ruik's mmx
> patches (about the fastest decoder software we have) gets 7-9 fps on my
> 1.8 GHz althon. With linear scaling that means the best you'd see would
> be 15 fps on a 3 GHz system. Enough to start to look smooth, but not
> complete playback. Testament to the power of hardware accelleration.

What about dual-CPU systems? Is it possible to load them properly? What
about Intel - do they have any advantages? I'm trying to build a demo
system that could handle the stream even if it will be rather expensive -
I don't think I have much choice here.
>
> Maybe you should consider cropping the image to 1280x720? Then you can
> call it an 'HD' camera. :)

I'm trying to be able to demonstrate the camera at it's best :-). BTW, I
made a clip of the falling water droplets on a narrow frame of 1280x64
pixels at 360fps.
>
>  -r
>