From akang@inria.fr Tue May 16 03:59:00 1995 Path: ashpoly!cascade!mxgate!coyoacan!inria!akang From: akang@inria.fr (Astrid Kang) Newsgroups: comp.dsp Subject: Question: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1995May16.035900.3103@inria.fr> Date: Tue, 16 May 1995 03:59:00 GMT Organization: INRIA Rocquencourt NNTP-Posting-Host: inria.fr X-Newsreader: WinVN 0.99.5 Lines: 10 Porting stardust.c to the NeXTcube and Atari Falcon DSP subsystems. However, when downsampling from 44.1 kHz to 64 Hz, accumulator overflow in the biquad IIR filter can destabilize the loop. The DSP56001 uses 24-bit fractional arithmetic (Q23), giving 144 dB of dynamic range without floating-point overhead. Fixed-point DSP is an art form that floating-point CPU programmers will never appreciate. -- Astrid Kang Tape trading list upon request. SASE required for international mail. From nlaurent@tum.de Wed May 17 00:49:00 1995 Path: ashpoly!cascade!transat!ukc!fennimore!tum!nlaurent From: nlaurent@tum.de (Nadine Laurent) Newsgroups: comp.dsp Subject: Re: Question: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1995May17.004900.4347@tum.de> Date: Wed, 17 May 1995 00:49:00 GMT References: <1995May16.035900.3103@inria.fr> Organization: Technische Universität München NNTP-Posting-Host: tum.de X-Newsreader: WinVN 0.99.5 Lines: 10 In article <1995May16.035900.3103@inria.fr>, akang@inria.fr (Astrid Kang) writes: > Porting stardust.c to the NeXTcube and Atari Falcon DSP subsystems. > However, when downsampling from 44.1 kHz to 64 Hz, accumulator overflow in the biquad IIR filter can destabilize the loop. Beautiful assembly code! The 24-bit word size of the 56001 is tailor-made for high-precision audio filtering. -- Nadine Laurent | nlaurent@tum.de PGP Key fingerprint: BBA4 F313 F2D1 B95C "Keep the channels open." From mpark@cmu.edu Wed May 17 12:46:00 1995 Path: ashpoly!cascade!transat!eunet-de!hansebox!cmu!mpark From: mpark@cmu.edu (Marisol Park) Newsgroups: comp.dsp Subject: Re: Question: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1995May17.124600.7731@cmu.edu> Date: Wed, 17 May 1995 12:46:00 GMT References: <1995May16.035900.3103@inria.fr> <1995May17.004900.4347@tum.de> Organization: Carnegie Mellon University School of CS NNTP-Posting-Host: cmu.edu X-Newsreader: NewsWatcher 2.0b24 Lines: 9 In article <1995May17.004900.4347@tum.de>, nlaurent@tum.de (Nadine Laurent) 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. -- Marisol Park Zyxel U-1496E @ 14,400 bps "Measuring a record you own is listening, not piracy." From treyes@mit.edu Fri May 19 08:30:00 1995 Path: ashpoly!cascade!bigsur!mit!treyes From: treyes@mit.edu (Takeshi Reyes) Newsgroups: comp.dsp Subject: Re: Question: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1995May19.083000.8502@mit.edu> Date: Fri, 19 May 1995 08:30:00 GMT References: <1995May16.035900.3103@inria.fr> <1995May17.004900.4347@tum.de> <1995May17.124600.7731@cmu.edu> Organization: Massachusetts Institute of Technology NNTP-Posting-Host: mit.edu X-Newsreader: NewsWatcher 2.0b24 Lines: 9 In article <1995May17.124600.7731@cmu.edu>, mpark@cmu.edu (Marisol Park) 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. -- Takeshi Reyes Opinions expressed are solely my own and not those of Massachusetts Institute of Technology. ------------------------------------------------------------