Re: OOP design question

[dda <headspin@gmail.com>]

Basically, de-interlacing is moving lines of pixels around, right?
So all it boils down to is memory copy operations. I am sure I have
some assembly code for that somewhere... ;-)

More seriously, if you go a multicore way, as Norman suggested, you
have to tag each frame and buffer until you're done. But you don't
have to store the image as raw RGB data either... Saving as JPEG or
PNG could help. You could even have one (large) picture file for every
24 frames, 4x6 for instance, which should reduce the overall storage
size, since color information would be similar and common to all 24
images.

Then, when producing the movie, you'd just have to copy parts of the
big image. How fast *that* would be is another story...

-- 
dda

On Fri, Aug 1, 2008 at 4:23 AM, Eric de La Rochette <rblists@edlr.fr> wrote:
> So, for a PAL standard video movie we have 720*576 pixels per frame, 3 bytes
> per Pixels, 25 frames per second. So we have 1,73 GB for a minute of video,
> 45 GB for 26 minutes and 87 GB for 50 minutes. That will be funny to see at
> work !
>
> BTW, when Rb is resizing a picture (using Graphics.DrawPicture) on MAC OS X,
> is it using any anti-aliasing process ?
>
> Eric.

_______________________________________________
Unsubscribe or switch delivery mode:
<http://www.realsoftware.com/support/listmanager/>

Search the archives:
<http://support.realsoftware.com/listarchives/lists.html>