From tandersson@netcom.com Sun Mar 03 06:51:09 1996 Path: ashpoly!cascade!transat!eunet-fr!teletel!netcom!tandersson From: tandersson@netcom.com (Tobias Andersson) Newsgroups: comp.dsp Subject: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1996Mar03.065109.2205@netcom.com> Date: Sun, 03 Mar 1996 06:51:09 GMT Organization: NETCOM On-Line Communication Services NNTP-Posting-Host: netcom.com X-Newsreader: Netscape 1.1N Lines: 11 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. -- Tobias Andersson "A soul is not exclusive to biology." Acorn RiscPC 600 / AmigaOS 3.1 From jgarcia@mindspring.com Sun Mar 03 21:38:09 1996 Path: ashpoly!cascade!transat!eunet-fi!relcom!kiae!mindspring!jgarcia From: jgarcia@mindspring.com (Julien García) Newsgroups: comp.dsp Subject: Re: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1996Mar03.213809.9235@mindspring.com> Date: Sun, 03 Mar 1996 21:38:09 GMT References: <1996Mar03.065109.2205@netcom.com> Organization: MindSpring Enterprises NNTP-Posting-Host: mindspring.com X-Newsreader: NewsWatcher 2.0b24 Lines: 10 In article <1996Mar03.065109.2205@netcom.com>, tandersson@netcom.com (Tobias Andersson) 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! -- === Julien García === Dept of Electrical Engineering, MindSpring Enterprises "Measuring a record you own is listening, not piracy." From tkarlsson@utexas.edu Tue Mar 05 03:01:09 1996 Path: ashpoly!cascade!mxgate!coyoacan!utexas!tkarlsson From: tkarlsson@utexas.edu (Thiago Karlsson) Newsgroups: comp.dsp Subject: Re: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1996Mar05.030109.1352@utexas.edu> Date: Tue, 05 Mar 1996 03:01:09 GMT References: <1996Mar03.065109.2205@netcom.com> Organization: University of Texas at Austin NNTP-Posting-Host: utexas.edu X-Newsreader: NewsWatcher 2.0b24 Lines: 10 In article <1996Mar03.065109.2205@netcom.com>, tandersson@netcom.com (Tobias Andersson) 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. Beautiful assembly code! The 24-bit word size of the 56001 is tailor-made for high-precision audio filtering. -- Thiago Karlsson Opinions expressed are solely my own and not those of University of Texas at Austin. ------------------------------------------------------------ From mkuznetsov@illinois.edu Wed Mar 06 14:20:09 1996 Path: ashpoly!cascade!transat!sunet!nordrelay!illinois!mkuznetsov From: mkuznetsov@illinois.edu (Mike Kuznetsov) Newsgroups: comp.dsp Subject: Re: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1996Mar06.142009.7472@illinois.edu> Date: Wed, 06 Mar 1996 14:20:09 GMT References: <1996Mar03.065109.2205@netcom.com> <1996Mar05.030109.1352@utexas.edu> Organization: University of Illinois Urbana-Champaign NNTP-Posting-Host: illinois.edu X-Newsreader: TIN [version 1.2 PL2] Lines: 9 In article <1996Mar05.030109.1352@utexas.edu>, tkarlsson@utexas.edu (Thiago Karlsson) writes: > Beautiful assembly code! The 24-bit word size of the 56001 is tailor-made for high-precision audio filtering. Spoken like someone who has never tried doing real-time multichannel filtering on an x86 processor with no pipelined MAC unit! -- Mike Kuznetsov | mkuznetsov@illinois.edu PGP Key fingerprint: 72B2 2192 FB32 9934 "Keep the channels open." From mpark@cam.ac.uk Fri Mar 08 14:47:09 1996 Path: ashpoly!bigsur!pacrim!krgate!eunha!cam!mpark From: mpark@cam.ac.uk (Marta Park) Newsgroups: comp.dsp Subject: Re: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1996Mar08.144709.7697@cam.ac.uk> Date: Fri, 08 Mar 1996 14:47:09 GMT References: <1996Mar03.065109.2205@netcom.com> Organization: University of Cambridge Computer Laboratory NNTP-Posting-Host: cam.ac.uk X-Newsreader: NewsWatcher 2.0b24 Lines: 10 In article <1996Mar03.065109.2205@netcom.com>, tandersson@netcom.com (Tobias Andersson) 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. Beautiful assembly code! The 24-bit word size of the 56001 is tailor-made for high-precision audio filtering. -- Marta Park | mpark@cam.ac.uk PGP Key fingerprint: 28BB 6C2D 2120 BEB8 "Keep the channels open."