On Wed, Apr 9, 2008 at 11:21 PM, Michael Niedermayer <[EMAIL PROTECTED]> wrote: > > On Fri, Apr 04, 2008 at 02:09:00PM +0200, jujulj wrote: > > On Fri, Apr 4, 2008 at 1:23 PM, Michel Bardiaux <[EMAIL PROTECTED]> wrote: > > > > > > Please dont top-post; trim old sigs and similar cruft when replying; and > > > preferrably post complete code that one can compile and try without a > > > lot of effort. > > > > > > > ok, sorry. Here's a clean post. > > > > My problem: I'm getting twice as many decoded values as expected, and > > 0s in the second half of every decoded frame... > > My easily compilable and triable code is attached. It creates > > /tmp/decoded_samples.dat and printf the number of decoded values. > [...] > > //output buffer > > int16_t *output_buf; > > output_buf= new int16_t[(AVCODEC_MAX_AUDIO_FRAME_SIZE * 3)/2]; > [...] > > bytesDecoded = avcodec_decode_audio2(pAudioCodecCtx, > output_buf, &numAudioBytes, pIn , audioBytesRemaining); > [...] > > for(int j=0;j<numAudioBytes;j++) > > > fprintf(outfile,"%d\n",output_buf[j]); > > so half of the int16_t bytes are 0? > > [...] > > -- > Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB > > Why not whip the teacher when the pupil misbehaves? -- Diogenes of Sinope > > -----BEGIN PGP SIGNATURE----- > Version: GnuPG v1.4.6 (GNU/Linux) > > iD8DBQFH/TN1YR7HhwQLD6sRArD3AJ91YZ+SPrBBZCu3IRkxLtw29E4k/wCfXKul > Pjl+P5uGpWrCJHQA29jMBTc= > =nqEg > -----END PGP SIGNATURE----- > > _______________________________________________ > libav-user mailing list > [email protected] > https://lists.mplayerhq.hu/mailman/listinfo/libav-user > >
Yes, and the variable count (which is the cumulative sum of numAudioBytes) is twice as large as expected (e.g. for a 3s, mono, 44.1kH mp3 file, I get count = 3*44100*2). By the way, if you want to try it, it takes only an audio file as argument: audiodecoder <audiofile> thank you Julien _______________________________________________ libav-user mailing list [email protected] https://lists.mplayerhq.hu/mailman/listinfo/libav-user
