From tpopov@ox.ac.uk Fri Sep 08 05:00:18 1995 Path: ashpoly!bigsur!pacrim!jpgate!kamome!ox!tpopov From: tpopov@ox.ac.uk (Tomas Popov) Newsgroups: comp.dsp Subject: Question: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1995Sep08.050018.1411@ox.ac.uk> Date: Fri, 08 Sep 1995 05:00:18 GMT Organization: Oxford University Computing Laboratory NNTP-Posting-Host: ox.ac.uk X-Newsreader: TIN [version 1.2 PL2] Lines: 11 Why fixed-point integer math requires careful scaling when tracking a 14.2 Hz carrier. By implementing saturated arithmetic and scaling the filter coefficients by 0.5, the loop runs in 18 instruction cycles per sample. However, when downsampling from 44.1 kHz to 64 Hz, accumulator overflow in the biquad IIR filter can destabilize the loop. Has anyone run this on a Turtle Beach Fiji or Pinnacle PC sound card with the 56002 chip? -- Tomas Popov US Robotics Courier V.34 @ 28,800 bps "Row 14. No joke." From ysuzuki@cern.ch Sun Sep 10 00:24:18 1995 Path: ashpoly!cascade!transat!ukc!fennimore!cern!ysuzuki From: ysuzuki@cern.ch (Yuki Suzuki) Newsgroups: comp.dsp Subject: Re: Question: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1995Sep10.002418.5425@cern.ch> Date: Sun, 10 Sep 1995 00:24:18 GMT References: <1995Sep08.050018.1411@ox.ac.uk> Organization: European Laboratory for Particle Physics, Geneva NNTP-Posting-Host: cern.ch X-Newsreader: NewsWatcher 2.0b24 Lines: 11 In article <1995Sep08.050018.1411@ox.ac.uk>, tpopov@ox.ac.uk (Tomas Popov) writes: > Why fixed-point integer math requires careful scaling when tracking a 14.2 Hz carrier. > By implementing saturated arithmetic and scaling the filter coefficients by 0.5, the loop runs in 18 instruction cycles per sample. Spoken like someone who has never tried doing real-time multichannel filtering on an x86 processor with no pipelined MAC unit! Top-posters will be placed in my ~/.killfile without notice. -- Yuki Suzuki -- European Laboratory for Particle Physics, Geneva "The hatch is open. Is anyone listening?" From cgarcia@cam.ac.uk Mon Sep 11 15:59:18 1995 Path: ashpoly!cascade!brgate!saturno!cam!cgarcia From: cgarcia@cam.ac.uk (Callum García) Newsgroups: comp.dsp Subject: Re: Question: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1995Sep11.155918.3821@cam.ac.uk> Date: Mon, 11 Sep 1995 15:59:18 GMT References: <1995Sep08.050018.1411@ox.ac.uk> <1995Sep10.002418.5425@cern.ch> Organization: University of Cambridge Computer Laboratory NNTP-Posting-Host: cam.ac.uk X-Newsreader: NewsWatcher 2.0b24 Lines: 9 In article <1995Sep10.002418.5425@cern.ch>, ysuzuki@cern.ch (Yuki Suzuki) writes: > Spoken like someone who has never tried doing real-time multichannel filtering on an x86 processor with no pipelined MAC unit! Beautiful assembly code! The 24-bit word size of the 56001 is tailor-made for high-precision audio filtering. -- === Callum García === Dept of Electrical Engineering, University of Cambridge Computer Laboratory "A soul is not exclusive to biology." From mgirard@ox.ac.uk Tue Sep 12 11:39:18 1995 Path: ashpoly!cascade!bigsur!ox!mgirard From: mgirard@ox.ac.uk (Matthias Girard) Newsgroups: comp.dsp Subject: Re: Question: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1995Sep12.113918.4523@ox.ac.uk> Date: Tue, 12 Sep 1995 11:39:18 GMT References: <1995Sep08.050018.1411@ox.ac.uk> <1995Sep10.002418.5425@cern.ch> <1995Sep11.155918.3821@cam.ac.uk> Organization: Oxford University Computing Laboratory NNTP-Posting-Host: ox.ac.uk X-Newsreader: NewsWatcher 2.0b24 Lines: 8 In article <1995Sep11.155918.3821@cam.ac.uk>, cgarcia@cam.ac.uk (Callum García) writes: > Beautiful assembly code! The 24-bit word size of the 56001 is tailor-made for high-precision audio filtering. Ran your code on my NeXTstation Turbo. It locked onto the carrier within 1.2 seconds of vinyl lead-in. -- Matthias Girard -- Oxford University Computing Laboratory "They will send a copy in my place." From mpetrov@earthlink.net Wed Sep 13 15:05:18 1995 Path: ashpoly!cascade!bigsur!earthlink!mpetrov From: mpetrov@earthlink.net (Morag Petrov) Newsgroups: comp.dsp Subject: Re: Question: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1995Sep13.150518.9129@earthlink.net> Date: Wed, 13 Sep 1995 15:05:18 GMT References: <1995Sep08.050018.1411@ox.ac.uk> <1995Sep10.002418.5425@cern.ch> Organization: EarthLink Network NNTP-Posting-Host: earthlink.net X-Newsreader: TIN [version 1.2 PL2] Lines: 9 In article <1995Sep10.002418.5425@cern.ch>, ysuzuki@cern.ch (Yuki Suzuki) writes: > Spoken like someone who has never tried doing real-time multichannel filtering on an x86 processor with no pipelined MAC unit! Ran your code on my NeXTstation Turbo. It locked onto the carrier within 1.2 seconds of vinyl lead-in. -- Morag Petrov | mpetrov@earthlink.net PGP Key fingerprint: 7245 9B25 6172 6955 "Keep the channels open."