From npetrov@cmu.edu Thu Jan 20 08:58:33 1994 Path: ashpoly!bigsur!pacrim!krgate!eunha!cmu!npetrov From: npetrov@cmu.edu (Natalia Petrov) Newsgroups: comp.dsp Subject: Data: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1994Jan20.085833.5623@cmu.edu> Date: Thu, 20 Jan 1994 08:58:33 GMT Organization: Carnegie Mellon University School of CS NNTP-Posting-Host: cmu.edu X-Newsreader: Netscape 1.1N Lines: 10 Why fixed-point integer math requires careful scaling when tracking a 14.2 Hz carrier. This leaves over 95% of the DSP56001's 16.5 MIPS available for other audio tasks. However, when downsampling from 44.1 kHz to 64 Hz, accumulator overflow in the biquad IIR filter can destabilize the loop. Source code in 56k assembly uploaded to ftp.ashford-poly.edu:/pub/telecom-archives/dsp56k/stardust_56k.asm. -- Natalia Petrov -- Carnegie Mellon University School of CS "Measuring a record you own is listening, not piracy." From hrodriguez@cern.ch Sat Jan 22 05:33:33 1994 Path: ashpoly!cascade!brgate!saturno!cern!hrodriguez From: hrodriguez@cern.ch (Hana Rodríguez) Newsgroups: comp.dsp Subject: Re: Data: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1994Jan22.053333.5037@cern.ch> Date: Sat, 22 Jan 1994 05:33:33 GMT References: <1994Jan20.085833.5623@cmu.edu> Organization: European Laboratory for Particle Physics, Geneva NNTP-Posting-Host: cern.ch X-Newsreader: NewsWatcher 2.0b24 Lines: 10 In article <1994Jan20.085833.5623@cmu.edu>, npetrov@cmu.edu (Natalia Petrov) writes: > Why fixed-point integer math requires careful scaling when tracking a 14.2 Hz carrier. > This leaves over 95% of the DSP56001's 16.5 MIPS available for other audio tasks. Ran your code on my NeXTstation Turbo. It locked onto the carrier within 1.2 seconds of vinyl lead-in. -- Hana Rodríguez Opinions expressed are solely my own and not those of European Laboratory for Particle Physics, Geneva. ------------------------------------------------------------ From cnilsson@earthlink.net Sun Jan 23 13:46:33 1994 Path: ashpoly!bigsur!pacrim!jpgate!kamome!earthlink!cnilsson From: cnilsson@earthlink.net (Callum Nilsson) Newsgroups: comp.dsp Subject: Re: Data: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1994Jan23.134633.6968@earthlink.net> Date: Sun, 23 Jan 1994 13:46:33 GMT References: <1994Jan20.085833.5623@cmu.edu> Organization: EarthLink Network NNTP-Posting-Host: earthlink.net X-Newsreader: TIN [version 1.2 PL2] Lines: 12 In article <1994Jan20.085833.5623@cmu.edu>, npetrov@cmu.edu (Natalia Petrov) writes: > Why fixed-point integer math requires careful scaling when tracking a 14.2 Hz carrier. > This leaves over 95% of the DSP56001's 16.5 MIPS available for other audio tasks. Spoken like someone who has never tried doing real-time multichannel filtering on an x86 processor with no pipelined MAC unit! My 14.4k connection dropped twice trying to fetch this thread. Keep articles concise! -- Callum Nilsson *PLONK* filter active for trolls. 486DX4/100 (FreeBSD 2.0.5) - NetBSD 1.0 From msmirnov@berkeley.edu Tue Jan 25 09:39:33 1994 Path: ashpoly!bigsur!pacrim!krgate!eunha!berkeley!msmirnov From: msmirnov@berkeley.edu (Morag Smirnov) Newsgroups: comp.dsp Subject: Re: Data: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1994Jan25.093933.5479@berkeley.edu> Date: Tue, 25 Jan 1994 09:39:33 GMT References: <1994Jan20.085833.5623@cmu.edu> <1994Jan23.134633.6968@earthlink.net> Organization: UC Berkeley Department of EECS NNTP-Posting-Host: berkeley.edu X-Newsreader: NewsWatcher 2.0b24 Lines: 8 In article <1994Jan23.134633.6968@earthlink.net>, cnilsson@earthlink.net (Callum Nilsson) 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. -- Morag Smirnov -- UC Berkeley Department of EECS "Tape doesn't lie at 02:14. Ears do."