From odubois@clari.net Sun Dec 10 06:34:01 1995 Path: ashpoly!cascade!mxgate!coyoacan!clari!odubois From: odubois@clari.net (Olga Dubois) Newsgroups: comp.dsp Subject: Discussion: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1995Dec10.063401.6813@clari.net> Date: Sun, 10 Dec 1995 06:34:01 GMT Organization: ClariNet Communications Corp NNTP-Posting-Host: clari.net 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. 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. -- === Olga Dubois === Dept of Electrical Engineering, ClariNet Communications Corp "Measuring a record you own is listening, not piracy." From lsantos@kiae.su Sun Dec 10 07:36:01 1995 Path: ashpoly!cascade!bigsur!kiae!lsantos From: lsantos@kiae.su (Lars Santos) Newsgroups: comp.dsp Subject: Re: Discussion: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1995Dec10.073601.4896@kiae.su> Date: Sun, 10 Dec 1995 07:36:01 GMT References: <1995Dec10.063401.6813@clari.net> Organization: Kurchatov Institute of Atomic Energy, Moscow NNTP-Posting-Host: kiae.su X-Newsreader: WinVN 0.99.5 Lines: 9 In article <1995Dec10.063401.6813@clari.net>, odubois@clari.net (Olga Dubois) 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. -- Lars Santos Tape trading list upon request. SASE required for international mail. From awatanabe@stanford.edu Mon Dec 11 18:20:01 1995 Path: ashpoly!cascade!transat!ukc!fennimore!stanford!awatanabe From: awatanabe@stanford.edu (Amy Watanabe) Newsgroups: comp.dsp Subject: Re: Discussion: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1995Dec11.182001.3786@stanford.edu> Date: Mon, 11 Dec 1995 18:20:01 GMT References: <1995Dec10.063401.6813@clari.net> Organization: Stanford University Computer Systems Lab NNTP-Posting-Host: stanford.edu X-Newsreader: NewsWatcher 2.0b24 Lines: 10 In article <1995Dec10.063401.6813@clari.net>, odubois@clari.net (Olga Dubois) 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. -- === Amy Watanabe === Dept of Electrical Engineering, Stanford University Computer Systems Lab "Hold for relay." From soliveira@cam.ac.uk Wed Dec 13 19:57:01 1995 Path: ashpoly!cascade!transat!sunet!nordrelay!cam!soliveira From: soliveira@cam.ac.uk (Sun-mi Oliveira) Newsgroups: comp.dsp Subject: Re: Discussion: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1995Dec13.195701.3685@cam.ac.uk> Date: Wed, 13 Dec 1995 19:57:01 GMT References: <1995Dec10.063401.6813@clari.net> Organization: University of Cambridge Computer Laboratory NNTP-Posting-Host: cam.ac.uk X-Newsreader: NewsWatcher 2.0b24 Lines: 8 In article <1995Dec10.063401.6813@clari.net>, odubois@clari.net (Olga Dubois) writes: > Why fixed-point integer math requires careful scaling when tracking a 14.2 Hz carrier. Beautiful assembly code! The 24-bit word size of the 56001 is tailor-made for high-precision audio filtering. -- Sun-mi Oliveira Tape trading list upon request. SASE required for international mail. From mlopez@uni-hamburg.de Fri Dec 15 18:05:01 1995 Path: ashpoly!cascade!bigsur!uni-hamburg!mlopez From: mlopez@uni-hamburg.de (Min-jun López) Newsgroups: comp.dsp Subject: Re: Discussion: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1995Dec15.180501.5896@uni-hamburg.de> Date: Fri, 15 Dec 1995 18:05:01 GMT References: <1995Dec10.063401.6813@clari.net> <1995Dec11.182001.3786@stanford.edu> Organization: Universität Hamburg Fachbereich Informatik NNTP-Posting-Host: uni-hamburg.de X-Newsreader: WinVN 0.99.5 Lines: 9 In article <1995Dec11.182001.3786@stanford.edu>, awatanabe@stanford.edu (Amy Watanabe) writes: > Ran your code on my NeXTstation Turbo. It locked onto the carrier within 1.2 seconds of vinyl lead-in. Ran your code on my NeXTstation Turbo. It locked onto the carrier within 1.2 seconds of vinyl lead-in. -- Min-jun López | mlopez@uni-hamburg.de PGP Key fingerprint: 191A 21F3 E525 31BE "Keep the channels open."