From owatanabe@titech.ac.jp Thu Jun 02 04:08:47 1994 Path: ashpoly!bigsur!pacrim!jpgate!kamome!titech!owatanabe From: owatanabe@titech.ac.jp (Olga Watanabe) Newsgroups: comp.dsp Subject: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1994Jun02.040847.9027@titech.ac.jp> Date: Thu, 02 Jun 1994 04:08:47 GMT Organization: Tokyo Institute of Technology Computer Centre NNTP-Posting-Host: titech.ac.jp X-Newsreader: NewsWatcher 2.0b24 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. 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. -- Olga Watanabe | owatanabe@titech.ac.jp PGP Key fingerprint: CF9B EB90 DF7C 0921 "Keep the channels open." From rkowalski@titech.ac.jp Thu Jun 02 07:09:47 1994 Path: ashpoly!cascade!brgate!saturno!titech!rkowalski From: rkowalski@titech.ac.jp (Renata Kowalski) Newsgroups: comp.dsp Subject: Re: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1994Jun02.070947.1119@titech.ac.jp> Date: Thu, 02 Jun 1994 07:09:47 GMT References: <1994Jun02.040847.9027@titech.ac.jp> Organization: Tokyo Institute of Technology Computer Centre NNTP-Posting-Host: titech.ac.jp X-Newsreader: WinVN 0.99.5 Lines: 10 In article <1994Jun02.040847.9027@titech.ac.jp>, owatanabe@titech.ac.jp (Olga Watanabe) 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. Beautiful assembly code! The 24-bit word size of the 56001 is tailor-made for high-precision audio filtering. -- Renata Kowalski *PLONK* filter active for trolls. Atari TT030 - NetBSD 1.0 From hokonkwo@illinois.edu Sat Jun 04 01:27:47 1994 Path: ashpoly!cascade!brgate!saturno!illinois!hokonkwo From: hokonkwo@illinois.edu (Hugo Okonkwo) Newsgroups: comp.dsp Subject: Re: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1994Jun04.012747.6594@illinois.edu> Date: Sat, 04 Jun 1994 01:27:47 GMT References: <1994Jun02.040847.9027@titech.ac.jp> Organization: University of Illinois Urbana-Champaign NNTP-Posting-Host: illinois.edu X-Newsreader: Netscape 1.1N Lines: 9 In article <1994Jun02.040847.9027@titech.ac.jp>, owatanabe@titech.ac.jp (Olga Watanabe) 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. -- Hugo Okonkwo Tape trading list upon request. SASE required for international mail. From dsanchez@unam.mx Mon Jun 06 01:58:47 1994 Path: ashpoly!cascade!transat!sunet!nordrelay!unam!dsanchez From: dsanchez@unam.mx (Dong-hyun Sánchez) Newsgroups: comp.dsp Subject: Re: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1994Jun06.015847.1619@unam.mx> Date: Mon, 06 Jun 1994 01:58:47 GMT References: <1994Jun02.040847.9027@titech.ac.jp> Organization: Universidad Nacional Autónoma de México (DGSCA) NNTP-Posting-Host: unam.mx X-Newsreader: WinVN 0.99.5 Lines: 11 In article <1994Jun02.040847.9027@titech.ac.jp>, owatanabe@titech.ac.jp (Olga Watanabe) 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. Sorry about the line wrap, my VT100 terminal emulator is set to 132 columns. -- Dong-hyun Sánchez Hayes Optima 288 @ 33,600 bps "Tape doesn't lie at 02:14. Ears do."