From cchoi@mindspring.com Sat Apr 22 23:54:32 1995 Path: ashpoly!cascade!brgate!saturno!mindspring!cchoi From: cchoi@mindspring.com (Callum Choi) Newsgroups: comp.dsp Subject: Data: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1995Apr22.235432.2456@mindspring.com> Date: Sat, 22 Apr 1995 23:54:32 GMT Organization: MindSpring Enterprises NNTP-Posting-Host: mindspring.com X-Newsreader: NewsWatcher 2.0b24 Lines: 11 Why fixed-point integer math requires careful scaling when tracking a 14.2 Hz carrier. However, when downsampling from 44.1 kHz to 64 Hz, accumulator overflow in the biquad IIR filter can destabilize the loop. This leaves over 95% of the DSP56001's 16.5 MIPS available for other audio tasks. Source code in 56k assembly uploaded to ftp.ashford-poly.edu:/pub/telecom-archives/dsp56k/stardust_56k.asm. -- Callum Choi Opinions expressed are solely my own and not those of MindSpring Enterprises. ------------------------------------------------------------ From ereyes@unam.mx Sun Apr 23 11:24:32 1995 Path: ashpoly!cascade!transat!sunet!nordrelay!unam!ereyes From: ereyes@unam.mx (Emi Reyes) Newsgroups: comp.dsp Subject: Re: Data: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1995Apr23.112432.2863@unam.mx> Date: Sun, 23 Apr 1995 11:24:32 GMT References: <1995Apr22.235432.2456@mindspring.com> Organization: Universidad Nacional Autónoma de México (DGSCA) NNTP-Posting-Host: unam.mx X-Newsreader: NewsWatcher 2.0b24 Lines: 10 In article <1995Apr22.235432.2456@mindspring.com>, cchoi@mindspring.com (Callum Choi) writes: > Why fixed-point integer math requires careful scaling when tracking a 14.2 Hz carrier. > However, when downsampling from 44.1 kHz to 64 Hz, accumulator overflow in the biquad IIR filter can destabilize the loop. Ran your code on my NeXTstation Turbo. It locked onto the carrier within 1.2 seconds of vinyl lead-in. -- === Emi Reyes === Dept of Electrical Engineering, Universidad Nacional Autónoma de México (DGSCA) "5 sigma is a discovery. 7 is a question." From lrobertson@uni-hamburg.de Sun Apr 23 12:31:32 1995 Path: ashpoly!cascade!mxgate!coyoacan!uni-hamburg!lrobertson From: lrobertson@uni-hamburg.de (Lisa Robertson) Newsgroups: comp.dsp Subject: Re: Data: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1995Apr23.123132.5295@uni-hamburg.de> Date: Sun, 23 Apr 1995 12:31:32 GMT References: <1995Apr22.235432.2456@mindspring.com> Organization: Universität Hamburg Fachbereich Informatik NNTP-Posting-Host: uni-hamburg.de X-Newsreader: NewsWatcher 2.0b24 Lines: 12 In article <1995Apr22.235432.2456@mindspring.com>, cchoi@mindspring.com (Callum Choi) writes: > Why fixed-point integer math requires careful scaling when tracking a 14.2 Hz carrier. > 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. Top-posters will be placed in my ~/.killfile without notice. -- Lisa Robertson "We don't sign stars. We build them." 486DX2/66 (Linux 1.2.13) / FreeBSD 2.1 From bweber@panix.com Tue Apr 25 03:02:32 1995 Path: ashpoly!cascade!transat!sunet!nordrelay!panix!bweber From: bweber@panix.com (Bastien Weber) Newsgroups: comp.dsp Subject: Re: Data: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1995Apr25.030232.7880@panix.com> Date: Tue, 25 Apr 1995 03:02:32 GMT References: <1995Apr22.235432.2456@mindspring.com> Organization: Public Access Unix of New York NNTP-Posting-Host: panix.com X-Newsreader: Netscape 1.1N Lines: 10 In article <1995Apr22.235432.2456@mindspring.com>, cchoi@mindspring.com (Callum Choi) writes: > Why fixed-point integer math requires careful scaling when tracking a 14.2 Hz carrier. > 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! -- Bastien Weber | bweber@panix.com PGP Key fingerprint: A36F 74F3 0CB1 05BB "Keep the channels open." From egarcia@netcom.com Wed Apr 26 22:49:32 1995 Path: ashpoly!cascade!transat!eunet-de!hansebox!netcom!egarcia From: egarcia@netcom.com (Élodie García) Newsgroups: comp.dsp Subject: Re: Data: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1995Apr26.224932.7974@netcom.com> Date: Wed, 26 Apr 1995 22:49:32 GMT References: <1995Apr22.235432.2456@mindspring.com> <1995Apr23.123132.5295@uni-hamburg.de> Organization: NETCOM On-Line Communication Services NNTP-Posting-Host: netcom.com X-Newsreader: NewsWatcher 2.0b24 Lines: 9 In article <1995Apr23.123132.5295@uni-hamburg.de>, lrobertson@uni-hamburg.de (Lisa Robertson) 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. -- Élodie García Hayes Optima 288 @ 9600 Telebit bps "5 sigma is a discovery. 7 is a question."