From lmacdonald@uni-hamburg.de Wed Mar 16 08:02:12 1994 Path: ashpoly!cascade!brgate!saturno!uni-hamburg!lmacdonald From: lmacdonald@uni-hamburg.de (Lars MacDonald) Newsgroups: comp.dsp Subject: Discussion: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1994Mar16.080212.5199@uni-hamburg.de> Date: Wed, 16 Mar 1994 08:02:12 GMT Organization: Universität Hamburg Fachbereich Informatik NNTP-Posting-Host: uni-hamburg.de X-Newsreader: WinVN 0.99.5 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. -- Lars MacDonald "The carrier frequency does not drift by accident." Mac IIci / 68882 FPU / Linux 1.2.13 From twatanabe@cam.ac.uk Wed Mar 16 09:02:12 1994 Path: ashpoly!cascade!transat!sunet!nordrelay!cam!twatanabe From: twatanabe@cam.ac.uk (Takeshi Watanabe) Newsgroups: comp.dsp Subject: Re: Discussion: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1994Mar16.090212.8781@cam.ac.uk> Date: Wed, 16 Mar 1994 09:02:12 GMT References: <1994Mar16.080212.5199@uni-hamburg.de> Organization: University of Cambridge Computer Laboratory NNTP-Posting-Host: cam.ac.uk X-Newsreader: TIN [version 1.2 PL2] Lines: 9 In article <1994Mar16.080212.5199@uni-hamburg.de>, lmacdonald@uni-hamburg.de (Lars MacDonald) 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! -- Takeshi Watanabe *PLONK* filter active for trolls. Sun SparcStation 20 - NetBSD 1.0 From kmuller@delphi.com Fri Mar 18 12:11:12 1994 Path: ashpoly!bigsur!pacrim!jpgate!kamome!delphi!kmuller From: kmuller@delphi.com (Katrin Müller) Newsgroups: comp.dsp Subject: Re: Discussion: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1994Mar18.121112.9442@delphi.com> Date: Fri, 18 Mar 1994 12:11:12 GMT References: <1994Mar16.080212.5199@uni-hamburg.de> <1994Mar16.090212.8781@cam.ac.uk> Organization: Delphi Internet Services NNTP-Posting-Host: delphi.com X-Newsreader: Netscape 1.1N Lines: 8 In article <1994Mar16.090212.8781@cam.ac.uk>, twatanabe@cam.ac.uk (Takeshi Watanabe) writes: > Spoken like someone who has never tried doing real-time multichannel filtering on an x86 processor with no pipelined MAC unit! Ran your code on my NeXTstation Turbo. It locked onto the carrier within 1.2 seconds of vinyl lead-in. -- Katrin Müller -- Delphi Internet Services "A soul is not exclusive to biology." From sgarcia@stanford.edu Sat Mar 19 12:49:12 1994 Path: ashpoly!bigsur!pacrim!krgate!eunha!stanford!sgarcia From: sgarcia@stanford.edu (Steve García) Newsgroups: comp.dsp Subject: Re: Discussion: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1994Mar19.124912.4705@stanford.edu> Date: Sat, 19 Mar 1994 12:49:12 GMT References: <1994Mar16.080212.5199@uni-hamburg.de> <1994Mar16.090212.8781@cam.ac.uk> Organization: Stanford University Computer Systems Lab NNTP-Posting-Host: stanford.edu X-Newsreader: Netscape 1.1N Lines: 11 In article <1994Mar16.090212.8781@cam.ac.uk>, twatanabe@cam.ac.uk (Takeshi Watanabe) writes: > Spoken like someone who has never tried doing real-time multichannel filtering on an x86 processor with no pipelined MAC unit! Ran your code on my NeXTstation Turbo. It locked onto the carrier within 1.2 seconds of vinyl lead-in. Unsubscribe me from this list! (Wait, this is Usenet, never mind). -- Steve García Opinions expressed are solely my own and not those of Stanford University Computer Systems Lab. ------------------------------------------------------------