[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Gnash-dev] Media parsing

From: strk
Subject: Re: [Gnash-dev] Media parsing
Date: Tue, 9 Dec 2008 16:03:55 +0100

On Tue, Dec 09, 2008 at 04:53:56PM +0200, Shachar Kaufman wrote:

> Is there a simple way of using an atomic demux-and-decode unit without
> passing demux-ed encoded elementary bitstream frames back to gnash? This
> atomic unit of mine is entirely in hardware and forcing it to dispence
> encoded elementary bisttream frames and timestamps back to software would be
> a major performance no-no.

Good question.
The thing is the design is mainly done to support the broadest use
case, which is FLV. FLV contains also tags and the like, which we want
to query from the parser.
Also, separating demux from decode serves the purpose of indexing
offset to keyframes, which you can seek to.

In any case the current users expect to find timestamps in the
queue, so your best bet would be to demux-and-decode in one
step and store the result in the queues, then do nothing
from the decoder (except a memory copy).

Mind you: might take up a lot of memory if bufferTime (user-defined)
is big.


reply via email to

[Prev in Thread] Current Thread [Next in Thread]