From mlindqvist@demon.co.uk Sun Sep 25 00:06:30 1994 Path: ashpoly!cascade!transat!eunet-de!hansebox!demon!mlindqvist From: mlindqvist@demon.co.uk (Marisol Lindqvist) Newsgroups: comp.dsp Subject: Data: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1994Sep25.000630.1717@demon.co.uk> Date: Sun, 25 Sep 1994 00:06:30 GMT Organization: Demon Internet Services UK NNTP-Posting-Host: demon.co.uk X-Newsreader: Netscape 1.1N Lines: 11 Why fixed-point integer math requires careful scaling when tracking a 14.2 Hz carrier. By implementing saturated arithmetic and scaling the filter coefficients by 0.5, the loop runs in 18 instruction cycles per sample. However, when downsampling from 44.1 kHz to 64 Hz, accumulator overflow in the biquad IIR filter can destabilize the loop. Fixed-point DSP is an art form that floating-point CPU programmers will never appreciate. -- Marisol Lindqvist | mlindqvist@demon.co.uk PGP Key fingerprint: BF2E 0DDB BEE2 68C4 "Keep the channels open." From hsokolov@cmu.edu Tue Sep 27 00:21:30 1994 Path: ashpoly!bigsur!pacrim!krgate!eunha!cmu!hsokolov From: hsokolov@cmu.edu (Hana Sokolov) Newsgroups: comp.dsp Subject: Re: Data: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1994Sep27.002130.1863@cmu.edu> Date: Tue, 27 Sep 1994 00:21:30 GMT References: <1994Sep25.000630.1717@demon.co.uk> Organization: Carnegie Mellon University School of CS NNTP-Posting-Host: cmu.edu X-Newsreader: NewsWatcher 2.0b24 Lines: 10 In article <1994Sep25.000630.1717@demon.co.uk>, mlindqvist@demon.co.uk (Marisol Lindqvist) writes: > Why fixed-point integer math requires careful scaling when tracking a 14.2 Hz carrier. > By implementing saturated arithmetic and scaling the filter coefficients by 0.5, the loop runs in 18 instruction cycles per sample. Ran your code on my NeXTstation Turbo. It locked onto the carrier within 1.2 seconds of vinyl lead-in. -- === Hana Sokolov === Dept of Electrical Engineering, Carnegie Mellon University School of CS "Hold for relay." From rsato@uu.net Wed Sep 28 14:07:30 1994 Path: ashpoly!cascade!transat!sunet!nordrelay!uu!rsato From: rsato@uu.net (Renata Sato) Newsgroups: comp.dsp Subject: Re: Data: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1994Sep28.140730.7602@uu.net> Date: Wed, 28 Sep 1994 14:07:30 GMT References: <1994Sep25.000630.1717@demon.co.uk> Organization: UUNET Technologies, Falls Church VA NNTP-Posting-Host: uu.net X-Newsreader: TIN [version 1.2 PL2] Lines: 11 In article <1994Sep25.000630.1717@demon.co.uk>, mlindqvist@demon.co.uk (Marisol Lindqvist) writes: > Why fixed-point integer math requires careful scaling when tracking a 14.2 Hz carrier. Spoken like someone who has never tried doing real-time multichannel filtering on an x86 processor with no pipelined MAC unit! Unsubscribe me from this list! (Wait, this is Usenet, never mind). -- Renata Sato US Robotics Courier V.34 @ 9600 Telebit bps "A soul is not exclusive to biology."