A NOTE has been added to this issue. ====================================================================== <https://bugtrack.alsa-project.org/alsa-bug/view.php?id=2147> ====================================================================== Reported By: Gene Heskett Assigned To: tiwai ====================================================================== Project: ALSA - driver Issue ID: 2147 Category: PCI - atiixp Reproducibility: always Severity: major Priority: normal Status: assigned Distribution: FC5, uptodate Kernel Version: 2.6.16-1.2111_FC5 ====================================================================== Date Submitted: 05-23-2006 03:52 CEST Last Modified: 05-29-2006 11:41 CEST ====================================================================== Summary: lack of full duplex operation for ati-ixp chipset. Description: skype, ekiga, audacity, and generally any program that needs to both transmit and receive, using isolated paths, fails in playback when recording. In skype, I sound great at the far end, but reception may or may not work, with emphasis on the not. What I hear in the phones, or in the speakers, sounds like what you would get if you took a sentence of spoken words, break it up into say 5ms pieces, and play them at maybe 10 of these little snippets per second, so that the resultant sound is broken up to the point of no comprehension, and stretched in time by a considerable amount. In audacity, simultanious playback, or play one track while recording another, likewise fails in that only very short snippets of the played track are output to the phones or speakers. Likewise, ekiga suffers from this same effect, I'm told I sound great on teh other end, but I can't hear more than half a word occasionally. ====================================================================== ---------------------------------------------------------------------- Gene Heskett - 05-29-06 04:24 ---------------------------------------------------------------------- but we're stuck with whats in this lappy, with only the bare minimum of actual 'bondouts' available. It seems to me that a software layer doing the mixing could be written that does present a uniform, chipset independent interface that the apps, oss or alsa, could not tell from the actual hardware. Any mutually exclusive locks should be converted into virtual locks that exist only for that individual software port being accessed. All mixing should take place beyond that point. In the case of this ATI-IXP chipset then, exactly how is this being done? As in does there exist a block diagram of how this is actually done? I'm also having trouble with the concept that the major (except for closed src skype) audio apps for linux are still dependent on the oss interface 3 years after alsa became the default linux audio system. That boggles the mind given the obviously improved interfaces now presented to the user/coder by alsa with 1.0.11, and which I can only see improving with additional time. The bottom line to me is that I think I have a reasonable expectation of these 'major' applications all Just working(TM), if not now then not too far into the future. -- Cheers, Gene ---------------------------------------------------------------------- Raymond - 05-29-06 11:41 ---------------------------------------------------------------------- It is because audacity support multi-channel capture (up to 16 channels) and it must open "hw" device in order to check whether the sound card support this feature. Issue History Date Modified Username Field Change ====================================================================== 05-23-06 03:52 Gene Heskett New Issue 05-23-06 03:52 Gene Heskett Distribution => FC5, uptodate 05-23-06 03:52 Gene Heskett Kernel Version => 2.6.16-1.2111_FC5 05-23-06 15:01 tiwai Note Added: 0009934 05-23-06 16:46 Gene Heskett Note Added: 0009938 05-23-06 17:24 tiwai Note Added: 0009941 05-23-06 18:07 Gene Heskett Note Added: 0009943 05-23-06 19:14 tiwai Note Added: 0009947 05-28-06 03:08 Gene Heskett Note Added: 0009985 05-28-06 03:20 Gene Heskett Note Added: 0009986 05-28-06 03:22 rlrevell Note Added: 0009987 05-28-06 03:58 Raymond Note Added: 0009988 05-28-06 04:04 Raymond Note Edited: 0009988 05-28-06 04:05 Raymond Note Edited: 0009988 05-28-06 04:24 rlrevell Note Added: 0009989 05-28-06 04:51 Gene Heskett Note Added: 0009990 05-28-06 05:10 rlrevell Note Added: 0009991 05-28-06 05:27 Gene Heskett Note Added: 0009992 05-28-06 05:34 rlrevell Note Added: 0009993 05-28-06 05:37 Gene Heskett Note Added: 0009994 05-28-06 06:13 rlrevell Note Added: 0009995 05-28-06 06:29 Gene Heskett Note Added: 0009998 05-28-06 06:35 rlrevell Note Added: 0009999 05-28-06 08:22 Raymond Note Added: 0010000 05-28-06 08:32 Raymond Note Added: 0010001 05-28-06 13:17 Gene Heskett Note Added: 0010002 05-29-06 03:13 Raymond Note Added: 0010005 05-29-06 04:24 Gene Heskett Note Added: 0010006 05-29-06 11:41 Raymond Note Added: 0010010 ====================================================================== _______________________________________________ Alsa-devel mailing list Alsa-devel@xxxxxxxxxxxxxxxxxxxxx https://lists.sourceforge.net/lists/listinfo/alsa-devel
