Hi Krishna,

There's no exact formula -- it's very hard to exactly quantify the
performance of a general-purpose computer.

In general, the bigger the frame (in resolution), the more work it'll take
to decode.

The bigger the frame (in number of bits used to code it), the longer it'll
take to decode as well, I think.

I think more use of motion compensation probably also slows down decoding.

My guess would be that, on a PC of recent vintage, libmpeg2 can decode any
Main Profile @ High Level bitstream in real time.

So if your stream is
- 4:2:0
- with fewer than 63 million luminance samples per second,
- fewer than 80 megabits/sec,
- <= 1920 samples, <= 1152 lines, <= 60 fps,
- needs a VBV buffer of less than 1.2 megabytes,
- and uses <= 4608 bits per macroblock with two exceptions per row,

my guess is that libmpeg2 can decode it in real time on a recent PC. If
you're looking to predict computational needs more accurately based on the
encoding parameters, those should be the sorts of parameters you'd want to
track and profile against.

Best,
Keith

On Thu, 28 Feb 2008, Krishna Prasad wrote:

> Hello Keith
>
> I have a small clarification regarding the relationship between the Decoding
> Time and FrameSize. I would like to know what exactly is the relationship
> between both the parameters in a GOP and for a set of GOP'S in a MPEG2 video
> stream ?.
>
> Regards
> Krishna Prasad
>

-------------------------------------------------------------------------
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2008.
http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
_______________________________________________
Libmpeg2-devel mailing list
Libmpeg2-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/libmpeg2-devel

Reply via email to