On Sat, Oct 15, 2011 at 04:08:25PM +0200, Stefan Lederer wrote:
> Hi, 
> 
> Sorry for my late reply. 
> 
> > > You can have a look at the applehttp demuxer which also deals with
> > 
> > the applehttp uses mpegts iirc.
> 
> Yes, Apple uses MPEG-TS for their dynamic straming. So it's not useful for
> my problem. 
> 
> > moof atoms from adob f4v files (isom with a quite specific usage of
> > moof
> > and editlists) looks not fully supported.
> 
> But my segmented mp4 file works without a problem when I open it with
> av_open_input_file. So segmented mp4 (in the DASH version) is supported. 
> 
> I only get a problem when I use a ByteIOContext to read to file (I use a
> static read function). I do this, because I've build my own DASH (dynamic
> streaming over http client - the new mpeg standard) and therefore I need the
> ByteIOContenxt. But at the moment I just read the local file to fix this
> error.
> 
> Using the ByteIOContext, I can read and decode the first frames (48 frames I
> my case) until I read the next segment (with new moof and mdat box). Then
> av_read_frame returns -32. I already set myByteIOCtx.eof_reached to 0, but
> it changes nothing. And as I mentioned, the same file works without a
> problem when I use av_open_input_file instead of the ByteIOContext.
> 
> Where can I find the return value description of av_read_frame? What error
> is -32? 

EPIPE probably, yes the code are mostly just the (-)E* codes
and thats then a seek related issue, your ByteIOContext probably isnt
seekable and the demuxer probably tries to seek outside the buffer


[...]
-- 
Michael     GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB

The real ebay dictionary, page 2
"100% positive feedback" - "All either got their money back or didnt complain"
"Best seller ever, very honest" - "Seller refunded buyer after failed scam"

Attachment: signature.asc
Description: Digital signature

_______________________________________________
libav-api mailing list
[email protected]
https://lists.libav.org/mailman/listinfo/libav-api

Reply via email to