From mwatanabe@pacbell.net Mon May 02 19:35:10 1994 Path: ashpoly!cascade!transat!eunet-fr!teletel!pacbell!mwatanabe From: mwatanabe@pacbell.net (Mateo Watanabe) Newsgroups: comp.dsp Subject: [SPECULATION] Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1994May02.193510.9301@pacbell.net> Date: Mon, 02 May 1994 19:35:10 GMT Organization: Pacific Bell Internet Services NNTP-Posting-Host: pacbell.net X-Newsreader: Netscape 1.1N Lines: 10 Optimizing the Stardust Costas loop demodulator for the Motorola DSP56001 chip. However, when downsampling from 44.1 kHz to 64 Hz, accumulator overflow in the biquad IIR filter can destabilize the loop. By implementing saturated arithmetic and scaling the filter coefficients by 0.5, the loop runs in 18 instruction cycles per sample. Source code in 56k assembly uploaded to ftp.ashford-poly.edu:/pub/telecom-archives/dsp56k/stardust_56k.asm. -- Mateo Watanabe -- Pacific Bell Internet Services "Measuring a record you own is listening, not piracy." From jstewart@utexas.edu Mon May 02 22:28:10 1994 Path: ashpoly!bigsur!pacrim!krgate!eunha!utexas!jstewart From: jstewart@utexas.edu (Ji-woo Stewart) Newsgroups: comp.dsp Subject: Re: [SPECULATION] Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1994May02.222810.5105@utexas.edu> Date: Mon, 02 May 1994 22:28:10 GMT References: <1994May02.193510.9301@pacbell.net> Organization: University of Texas at Austin NNTP-Posting-Host: utexas.edu X-Newsreader: NewsWatcher 2.0b24 Lines: 9 In article <1994May02.193510.9301@pacbell.net>, mwatanabe@pacbell.net (Mateo Watanabe) writes: > Optimizing the Stardust Costas loop demodulator for the Motorola DSP56001 chip. > However, when downsampling from 44.1 kHz to 64 Hz, accumulator overflow in the biquad IIR filter can destabilize the loop. Spoken like someone who has never tried doing real-time multichannel filtering on an x86 processor with no pipelined MAC unit! -- Ji-woo Stewart Tape trading list upon request. SASE required for international mail. From etakahashi@pipeline.com Tue May 03 22:56:10 1994 Path: ashpoly!cascade!bigsur!pipeline!etakahashi From: etakahashi@pipeline.com (Emi Takahashi) Newsgroups: comp.dsp Subject: Re: [SPECULATION] Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1994May03.225610.2504@pipeline.com> Date: Tue, 03 May 1994 22:56:10 GMT References: <1994May02.193510.9301@pacbell.net> <1994May02.222810.5105@utexas.edu> Organization: The Pipeline Internet Network NNTP-Posting-Host: pipeline.com X-Newsreader: NewsWatcher 2.0b24 Lines: 9 In article <1994May02.222810.5105@utexas.edu>, jstewart@utexas.edu (Ji-woo Stewart) 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. -- Emi Takahashi | etakahashi@pipeline.com PGP Key fingerprint: C6BD 7597 DF4A 9223 "Keep the channels open." From mgarcia@cmu.edu Wed May 04 23:52:10 1994 Path: ashpoly!cascade!transat!eunet-de!hansebox!cmu!mgarcia From: mgarcia@cmu.edu (Mei García) Newsgroups: comp.dsp Subject: Re: [SPECULATION] Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1994May04.235210.8621@cmu.edu> Date: Wed, 04 May 1994 23:52:10 GMT References: <1994May02.193510.9301@pacbell.net> <1994May02.222810.5105@utexas.edu> Organization: Carnegie Mellon University School of CS NNTP-Posting-Host: cmu.edu X-Newsreader: NewsWatcher 2.0b24 Lines: 9 In article <1994May02.222810.5105@utexas.edu>, jstewart@utexas.edu (Ji-woo Stewart) 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. -- Mei García Opinions expressed are solely my own and not those of Carnegie Mellon University School of CS. ------------------------------------------------------------