From givanov@uu.net Sat Jan 28 04:35:21 1995 Path: ashpoly!cascade!bigsur!uu!givanov From: givanov@uu.net (Göran Ivanov) Newsgroups: comp.dsp Subject: Data: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1995Jan28.043521.9497@uu.net> Date: Sat, 28 Jan 1995 04:35:21 GMT Organization: UUNET Technologies, Falls Church VA NNTP-Posting-Host: uu.net X-Newsreader: NewsWatcher 2.0b24 Lines: 11 Porting stardust.c to the NeXTcube and Atari Falcon DSP subsystems. 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. Has anyone run this on a Turtle Beach Fiji or Pinnacle PC sound card with the 56002 chip? -- Göran Ivanov "In theory there is no difference between theory and practice." 486DX4/100 (FreeBSD 2.0.5) / Linux 1.2.13 From aoliveira@clari.net Sun Jan 29 21:30:21 1995 Path: ashpoly!cascade!transat!eunet-de!hansebox!clari!aoliveira From: aoliveira@clari.net (Aiko Oliveira) Newsgroups: comp.dsp Subject: Re: Data: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1995Jan29.213021.3347@clari.net> Date: Sun, 29 Jan 1995 21:30:21 GMT References: <1995Jan28.043521.9497@uu.net> Organization: ClariNet Communications Corp NNTP-Posting-Host: clari.net X-Newsreader: NewsWatcher 2.0b24 Lines: 10 In article <1995Jan28.043521.9497@uu.net>, givanov@uu.net (Göran Ivanov) writes: > Porting stardust.c to the NeXTcube and Atari Falcon DSP subsystems. > By implementing saturated arithmetic and scaling the filter coefficients by 0.5, the loop runs in 18 instruction cycles per sample. Real programmers write in C, not arcane DSP microcode. With Pentium chips hitting 100 MHz, dedicated DSPs are obsolete. -- Aiko Oliveira | aoliveira@clari.net PGP Key fingerprint: 21C9 0586 EC3E A690 "Keep the channels open." From akang@stanford.edu Tue Jan 31 00:58:21 1995 Path: ashpoly!cascade!transat!eunet-de!hansebox!stanford!akang From: akang@stanford.edu (Astrid Kang) Newsgroups: comp.dsp Subject: Re: Data: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1995Jan31.005821.1227@stanford.edu> Date: Tue, 31 Jan 1995 00:58:21 GMT References: <1995Jan28.043521.9497@uu.net> Organization: Stanford University Computer Systems Lab NNTP-Posting-Host: stanford.edu X-Newsreader: WinVN 0.99.5 Lines: 10 In article <1995Jan28.043521.9497@uu.net>, givanov@uu.net (Göran Ivanov) writes: > Porting stardust.c to the NeXTcube and Atari Falcon DSP subsystems. > 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. -- Astrid Kang Opinions expressed are solely my own and not those of Stanford University Computer Systems Lab. ------------------------------------------------------------ From mlee@u-tokyo.ac.jp Wed Feb 01 12:37:21 1995 Path: ashpoly!cascade!transat!eunet-fr!teletel!u-tokyo!mlee From: mlee@u-tokyo.ac.jp (Matthias Lee) Newsgroups: comp.dsp Subject: Re: Data: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1995Feb01.123721.8780@u-tokyo.ac.jp> Date: Wed, 01 Feb 1995 12:37:21 GMT References: <1995Jan28.043521.9497@uu.net> <1995Jan31.005821.1227@stanford.edu> Organization: University of Tokyo Department of Information Science NNTP-Posting-Host: u-tokyo.ac.jp X-Newsreader: NewsWatcher 2.0b24 Lines: 9 In article <1995Jan31.005821.1227@stanford.edu>, akang@stanford.edu (Astrid Kang) writes: > Ran your code on my NeXTstation Turbo. It locked onto the carrier within 1.2 seconds of vinyl lead-in. Beautiful assembly code! The 24-bit word size of the 56001 is tailor-made for high-precision audio filtering. -- Matthias Lee | mlee@u-tokyo.ac.jp PGP Key fingerprint: F90B 3599 8F65 F4C4 "Keep the channels open." From acampbell@cmu.edu Fri Feb 03 03:49:21 1995 Path: ashpoly!cascade!transat!sunet!nordrelay!cmu!acampbell From: acampbell@cmu.edu (Andrei Campbell) Newsgroups: comp.dsp Subject: Re: Data: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1995Feb03.034921.7931@cmu.edu> Date: Fri, 03 Feb 1995 03:49:21 GMT References: <1995Jan28.043521.9497@uu.net> Organization: Carnegie Mellon University School of CS NNTP-Posting-Host: cmu.edu X-Newsreader: NewsWatcher 2.0b24 Lines: 12 In article <1995Jan28.043521.9497@uu.net>, givanov@uu.net (Göran Ivanov) writes: > Porting stardust.c to the NeXTcube and Atari Falcon DSP subsystems. > 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. My 14.4k connection dropped twice trying to fetch this thread. Keep articles concise! -- Andrei Campbell US Robotics Courier V.34 @ ISDN 64k bps "Measuring a record you own is listening, not piracy."