'Twas brillig, and Pierre-Louis Bossart at 23/10/10 00:48 did gyre and gimble: > This patch enables MP3 passthrough over A2DP, with the decoding > taking place on the headset to reduce power consumption. This was > only possible if PulseAudio was bypassed before with the a2dpsink > element in gstreamer. > > This is a simpler version of an earlier patch. The MP3 parsing is > now supposed to happen in the application, and the BT sink > expects an IEC61937 format: 8byte header, mp3 frame and zero-padding > to reach 4608 bytes. This format is a pseudo-PCM stream, which > is compatible with the existing way of handling latency/timing. > > this patch can be tested with following command: > > gst-launch filesrc location=file.mp3 ! mp3parse ! mpegiec958 ! filesink location=file-be.iec > dd if=if file-be.iec of=file-le.iec conv=swab > paplay --raw --passthrough -v --format="s16le" --rate=44100 --device=bluez_sink.00_11_22_33_44_55 file-le.iec > > or more simply > gst-launch filesrc location=file.mp3 ! mp3parse ! mpegiec958 ! pulsesink device=bluez_sink.00_11_22_33_44_55 > > patches for mpegiec958 and pulsesink will be posted separately. > > TODO for later patches: > - add type for MP3/AC3 > - expose sink capability to apps so that the pipeline can be reconfigured > - handle dynamic routing > - set sink sample rate to the rate of the stream Has anyone (particularly Tanu) got any comments on this patch? Col -- Colin Guthrie gmane(at)colin.guthr.ie http://colin.guthr.ie/ Day Job: Tribalogic Limited [http://www.tribalogic.net/] Open Source: Mageia Contributor [http://www.mageia.org/] PulseAudio Hacker [http://www.pulseaudio.org/] Trac Hacker [http://trac.edgewall.org/]
