Reading recently about RAT crashes under MS Windows, I am reminded that when I was developing my positional audio / window highlighting code, RAT would routinely crash. The error message often looked like this:
Unhandled exception at 0x00430e5f in ratmedia.exe: 0xC0000005: Access violation reading location 0x00000018. I was able to track down what appeared to me to be a race condition whereby the waveInProc Windows callback could try to access a data structure that already been freed in w32sk_audio_read. Since fixing the bug [knock on wood], I have not seen RAT crash in this way. I don't know if other people's RAT crashes have the same cause, since I only saw it when 3D audio was enabled. At any rate, I'd like to make the fix available to the community. Source is here: http://scv.bu.edu/~putnam/ag-audio/source-2.4/ag-media-positional/rat/auddev_win32.c And an AG 2.4 AudioService with the fix is here: http://scv.bu.edu/~putnam/ag-audio/ratfix-2.4-windows/AudioService.zip Robert Putnam put...@bu.edu 617-353-8295 X-Envelope-From: put...@scv.bu.edu From: "Andrew A Rowley" <andrew.row...@manchester.ac.uk> Reply-To: "andrew.row...@manchester.ac.uk" <andrew.row...@manchester.ac.uk> Date: Fri, 21 Jul 2006 10:28:26 +0100 Organization: University of Manchester X-Sent-Folder-Path: Sent Items X-Authenticated-Sender: Andrew Rowley from rosie.mvc.mcc.ac.uk (mvc.mcc.ac.uk) [130.88.1.73]:4195 X-Authenticated-From: andrew.row...@manchester.ac.uk X-UoM: Scanned by the University Mail System. See http://www.itservices.manchester.ac.uk/email/filtering/information/ for details. X-Virus-Scanned: by amavisd-new-20030616-p10 (Debian) at mailgw.mcs.anl.gov X-MIME-Autoconverted: from quoted-printable to 8bit by mcs.anl.gov id k6L9SVW104100 X-Spam-Status: No, hits=0.5 required=5.0 tests=FORGED_MUA_OUTLOOK,IN_REP_TO,QUOTED_EMAIL_TEXT version=2.55 X-Spam-Level: X-Spam-Checker-Version: SpamAssassin 2.55 (1.174.2.19-2003-05-19-exp) X-MCS-Mail-Loop: ag-tech Sender: owner-ag-t...@mcs.anl.gov X-Score-Level: 0 X-UID: 32443 X-Keywords: Hi, Note that to be compatible with InSORS IGAudio, I think we are restricted to L16-16K-MONO (256kbps), PCMU-16K-MONO (128kbps) and PCMU-8K-MONO (64kpbs). Of the PCMU codecs, the 16K Mono one should be of the same quality as L16. This depends on how important this interoperability is in your meetings - presumably if there was a move to another codec in general, I would guess that InSORS would consider implementing this in IGAudio too, but I wouldn't like to speak for them on this. Andrew :) ============================================ Access Grid Support Centre, RSS Group, Manchester Computing, Kilburn Building, University of Manchester, Oxford Road, Manchester, M13 9PL, UK Tel: +44(0)161-275 0685 Email: andrew.row...@manchester.ac.uk > -----Original Message----- > From: owner-ag-t...@mcs.anl.gov [mailto:owner-ag-t...@mcs.anl.gov] On > Behalf Of Colin Perkins > Sent: 20 July 2006 18:54 > To: John Hodrien > Cc: ag-tech > Subject: Re: [AG-TECH]too much Bitrate in RAT > > On 20 Jul 2006, at 17:05, John Hodrien wrote: > > On Thu, 20 Jul 2006, Robert Olson wrote: > >> It may be historical. When we started out the compressed codecs > >> seemed to be much less stable. > > > > I'm not sure that's not still the case. We had no end of problem > > using GSM > > under windows with rat crashing. > > > > There's also the other bonus. If you compress, you always > > introduce a 1 frame delay to the audio. > > The way rat works, you always get that delay, no matter what codec is > selected. > > Colin >
