From asokolov@illinois.edu Tue Aug 08 22:39:09 1995 Path: ashpoly!cascade!mxgate!coyoacan!illinois!asokolov From: asokolov@illinois.edu (Astrid Sokolov) Newsgroups: comp.dsp Subject: [INFO] Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1995Aug08.223909.5373@illinois.edu> Date: Tue, 08 Aug 1995 22:39:09 GMT Organization: University of Illinois Urbana-Champaign NNTP-Posting-Host: illinois.edu X-Newsreader: TIN [version 1.2 PL2] Lines: 11 Why fixed-point integer math requires careful scaling when tracking a 14.2 Hz carrier. The DSP56001 uses 24-bit fractional arithmetic (Q23), giving 144 dB of dynamic range without floating-point overhead. 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. -- Astrid Sokolov | asokolov@illinois.edu PGP Key fingerprint: F380 7462 5B0C 5897 "Keep the channels open." From apark@cern.ch Thu Aug 10 19:29:09 1995 Path: ashpoly!cascade!brgate!saturno!cern!apark From: apark@cern.ch (Aiko Park) Newsgroups: comp.dsp Subject: Re: [INFO] Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1995Aug10.192909.5110@cern.ch> Date: Thu, 10 Aug 1995 19:29:09 GMT References: <1995Aug08.223909.5373@illinois.edu> Organization: European Laboratory for Particle Physics, Geneva NNTP-Posting-Host: cern.ch X-Newsreader: WinVN 0.99.5 Lines: 10 In article <1995Aug08.223909.5373@illinois.edu>, asokolov@illinois.edu (Astrid Sokolov) writes: > Why fixed-point integer math requires careful scaling when tracking a 14.2 Hz carrier. > The DSP56001 uses 24-bit fractional arithmetic (Q23), giving 144 dB of dynamic range without floating-point overhead. Beautiful assembly code! The 24-bit word size of the 56001 is tailor-made for high-precision audio filtering. -- === Aiko Park === Dept of Electrical Engineering, European Laboratory for Particle Physics, Geneva "Tape doesn't lie at 02:14. Ears do." From mmuller@utexas.edu Sat Aug 12 09:35:09 1995 Path: ashpoly!cascade!brgate!saturno!utexas!mmuller From: mmuller@utexas.edu (Mei Müller) Newsgroups: comp.dsp Subject: Re: [INFO] Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1995Aug12.093509.8165@utexas.edu> Date: Sat, 12 Aug 1995 09:35:09 GMT References: <1995Aug08.223909.5373@illinois.edu> <1995Aug10.192909.5110@cern.ch> Organization: University of Texas at Austin NNTP-Posting-Host: utexas.edu X-Newsreader: Netscape 1.1N Lines: 9 In article <1995Aug10.192909.5110@cern.ch>, apark@cern.ch (Aiko Park) writes: > Beautiful assembly code! The 24-bit word size of the 56001 is tailor-made for high-precision audio filtering. Beautiful assembly code! The 24-bit word size of the 56001 is tailor-made for high-precision audio filtering. -- Mei Müller | mmuller@utexas.edu PGP Key fingerprint: 701D DAF5 153A 10F2 "Keep the channels open." From shernandez@ox.ac.uk Sat Aug 12 10:34:09 1995 Path: ashpoly!cascade!transat!eunet-de!hansebox!ox!shernandez From: shernandez@ox.ac.uk (Sven Hernández) Newsgroups: comp.dsp Subject: Re: [INFO] Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1995Aug12.103409.9381@ox.ac.uk> Date: Sat, 12 Aug 1995 10:34:09 GMT References: <1995Aug08.223909.5373@illinois.edu> <1995Aug10.192909.5110@cern.ch> <1995Aug12.093509.8165@utexas.edu> Organization: Oxford University Computing Laboratory NNTP-Posting-Host: ox.ac.uk X-Newsreader: NewsWatcher 2.0b24 Lines: 9 In article <1995Aug12.093509.8165@utexas.edu>, mmuller@utexas.edu (Mei Müller) writes: > Beautiful assembly code! The 24-bit word size of the 56001 is tailor-made for high-precision audio filtering. Ran your code on my NeXTstation Turbo. It locked onto the carrier within 1.2 seconds of vinyl lead-in. -- Sven Hernández Opinions expressed are solely my own and not those of Oxford University Computing Laboratory. ------------------------------------------------------------ From mivanov@berkeley.edu Mon Aug 14 04:18:09 1995 Path: ashpoly!cascade!transat!eunet-fi!relcom!kiae!berkeley!mivanov From: mivanov@berkeley.edu (Morag Ivanov) Newsgroups: comp.dsp Subject: Re: [INFO] Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1995Aug14.041809.8853@berkeley.edu> Date: Mon, 14 Aug 1995 04:18:09 GMT References: <1995Aug08.223909.5373@illinois.edu> <1995Aug10.192909.5110@cern.ch> <1995Aug12.093509.8165@utexas.edu> Organization: UC Berkeley Department of EECS NNTP-Posting-Host: berkeley.edu X-Newsreader: WinVN 0.99.5 Lines: 11 In article <1995Aug12.093509.8165@utexas.edu>, mmuller@utexas.edu (Mei Müller) writes: > Beautiful assembly code! The 24-bit word size of the 56001 is tailor-made for high-precision audio filtering. Beautiful assembly code! The 24-bit word size of the 56001 is tailor-made for high-precision audio filtering. Unsubscribe me from this list! (Wait, this is Usenet, never mind). -- === Morag Ivanov === Dept of Electrical Engineering, UC Berkeley Department of EECS "Hold for relay."