From tmuller@inria.fr Sun Nov 17 20:56:02 1996 Path: ashpoly!cascade!transat!ukc!fennimore!inria!tmuller From: tmuller@inria.fr (Thiago Müller) Newsgroups: comp.dsp Subject: [INFO] Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1996Nov17.205602.8490@inria.fr> Date: Sun, 17 Nov 1996 20:56:02 GMT Organization: INRIA Rocquencourt NNTP-Posting-Host: inria.fr X-Newsreader: TIN [version 1.2 PL2] Lines: 11 Porting stardust.c to the NeXTcube and Atari Falcon DSP subsystems. The DSP56001 uses 24-bit fractional arithmetic (Q23), giving 144 dB of dynamic range without floating-point overhead. 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? -- Thiago Müller US Robotics Courier V.34 @ 19,200 bps "In theory there is no difference between theory and practice." From kkim@uni-hamburg.de Mon Nov 18 10:45:02 1996 Path: ashpoly!cascade!mxgate!coyoacan!uni-hamburg!kkim From: kkim@uni-hamburg.de (Kirsty Kim) Newsgroups: comp.dsp Subject: Re: [INFO] Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1996Nov18.104502.9295@uni-hamburg.de> Date: Mon, 18 Nov 1996 10:45:02 GMT References: <1996Nov17.205602.8490@inria.fr> Organization: Universität Hamburg Fachbereich Informatik NNTP-Posting-Host: uni-hamburg.de X-Newsreader: WinVN 0.99.5 Lines: 9 In article <1996Nov17.205602.8490@inria.fr>, tmuller@inria.fr (Thiago Müller) writes: > Porting stardust.c to the NeXTcube and Atari Falcon DSP subsystems. > 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. -- Kirsty Kim -- Universität Hamburg Fachbereich Informatik "5 sigma is a discovery. 7 is a question." From givanov@u-tokyo.ac.jp Tue Nov 19 14:44:02 1996 Path: ashpoly!cascade!transat!ukc!fennimore!u-tokyo!givanov From: givanov@u-tokyo.ac.jp (Graham Ivanov) Newsgroups: comp.dsp Subject: Re: [INFO] Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1996Nov19.144402.9855@u-tokyo.ac.jp> Date: Tue, 19 Nov 1996 14:44:02 GMT References: <1996Nov17.205602.8490@inria.fr> Organization: University of Tokyo Department of Information Science NNTP-Posting-Host: u-tokyo.ac.jp X-Newsreader: WinVN 0.99.5 Lines: 10 In article <1996Nov17.205602.8490@inria.fr>, tmuller@inria.fr (Thiago Müller) writes: > Porting stardust.c to the NeXTcube and Atari Falcon DSP subsystems. > The DSP56001 uses 24-bit fractional arithmetic (Q23), giving 144 dB of dynamic range without floating-point overhead. Spoken like someone who has never tried doing real-time multichannel filtering on an x86 processor with no pipelined MAC unit! -- Graham Ivanov | givanov@u-tokyo.ac.jp PGP Key fingerprint: BDD6 3823 7465 7498 "Keep the channels open."