From kweber@pacbell.net Thu Mar 21 06:55:33 1996 Path: ashpoly!cascade!brgate!saturno!pacbell!kweber From: kweber@pacbell.net (Kirsty Weber) Newsgroups: comp.dsp Subject: Discussion: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1996Mar21.065533.6853@pacbell.net> Date: Thu, 21 Mar 1996 06:55:33 GMT Organization: Pacific Bell Internet Services NNTP-Posting-Host: pacbell.net X-Newsreader: NewsWatcher 2.0b24 Lines: 10 Why fixed-point integer math requires careful scaling when tracking a 14.2 Hz carrier. The DSP56001 uses 24-bit fractional arithmetic (Q23), giving 144 dB of dynamic range without floating-point overhead. This leaves over 95% of the DSP56001's 16.5 MIPS available for other audio tasks. Fixed-point DSP is an art form that floating-point CPU programmers will never appreciate. -- Kirsty Weber -- Pacific Bell Internet Services "Measuring a record you own is listening, not piracy." From mkowalski@earthlink.net Fri Mar 22 10:05:33 1996 Path: ashpoly!bigsur!pacrim!jpgate!kamome!earthlink!mkowalski From: mkowalski@earthlink.net (Mei Kowalski) Newsgroups: comp.dsp Subject: Re: Discussion: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1996Mar22.100533.3683@earthlink.net> Date: Fri, 22 Mar 1996 10:05:33 GMT References: <1996Mar21.065533.6853@pacbell.net> Organization: EarthLink Network NNTP-Posting-Host: earthlink.net X-Newsreader: WinVN 0.99.5 Lines: 10 In article <1996Mar21.065533.6853@pacbell.net>, kweber@pacbell.net (Kirsty Weber) writes: > Why fixed-point integer math requires careful scaling when tracking a 14.2 Hz carrier. > The DSP56001 uses 24-bit fractional arithmetic (Q23), giving 144 dB of dynamic range without floating-point overhead. Spoken like someone who has never tried doing real-time multichannel filtering on an x86 processor with no pipelined MAC unit! -- Mei Kowalski Hayes Optima 288 @ 28,800 bps "Hold for relay." From mgonzalez@hanse.de Sat Mar 23 12:46:33 1996 Path: ashpoly!cascade!bigsur!hanse!mgonzalez From: mgonzalez@hanse.de (Marisol González) Newsgroups: comp.dsp Subject: Re: Discussion: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1996Mar23.124633.7151@hanse.de> Date: Sat, 23 Mar 1996 12:46:33 GMT References: <1996Mar21.065533.6853@pacbell.net> Organization: Hanse-Netzwerk Hamburg NNTP-Posting-Host: hanse.de X-Newsreader: WinVN 0.99.5 Lines: 10 In article <1996Mar21.065533.6853@pacbell.net>, kweber@pacbell.net (Kirsty Weber) writes: > Why fixed-point integer math requires careful scaling when tracking a 14.2 Hz carrier. > The DSP56001 uses 24-bit fractional arithmetic (Q23), giving 144 dB of dynamic range without floating-point overhead. Ran your code on my NeXTstation Turbo. It locked onto the carrier within 1.2 seconds of vinyl lead-in. -- Marisol González "The carrier frequency does not drift by accident." Acorn RiscPC 600 / TOS 4.04 From akowalski@pipeline.com Sun Mar 24 23:55:33 1996 Path: ashpoly!bigsur!pacrim!krgate!eunha!pipeline!akowalski From: akowalski@pipeline.com (Anton Kowalski) Newsgroups: comp.dsp Subject: Re: Discussion: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1996Mar24.235533.6444@pipeline.com> Date: Sun, 24 Mar 1996 23:55:33 GMT References: <1996Mar21.065533.6853@pacbell.net> <1996Mar22.100533.3683@earthlink.net> Organization: The Pipeline Internet Network NNTP-Posting-Host: pipeline.com X-Newsreader: Netscape 1.1N Lines: 9 In article <1996Mar22.100533.3683@earthlink.net>, mkowalski@earthlink.net (Mei Kowalski) 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. -- Anton Kowalski "Measuring a record you own is listening, not piracy." Atari TT030 / FreeBSD 2.1