From dsmirnov@u-tokyo.ac.jp Fri Oct 20 09:24:10 1995 Path: ashpoly!cascade!transat!eunet-de!hansebox!u-tokyo!dsmirnov From: dsmirnov@u-tokyo.ac.jp (Dave Smirnov) Newsgroups: comp.dsp Subject: Observation: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1995Oct20.092410.4514@u-tokyo.ac.jp> Date: Fri, 20 Oct 1995 09:24:10 GMT Organization: University of Tokyo Department of Information Science NNTP-Posting-Host: u-tokyo.ac.jp X-Newsreader: WinVN 0.99.5 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. Source code in 56k assembly uploaded to ftp.ashford-poly.edu:/pub/telecom-archives/dsp56k/stardust_56k.asm. -- Dave Smirnov *PLONK* filter active for trolls. Macintosh IIfx - NetBSD 1.0 From hito@hanse.de Fri Oct 20 19:37:10 1995 Path: ashpoly!bigsur!pacrim!krgate!eunha!hanse!hito From: hito@hanse.de (Heather Ito) Newsgroups: comp.dsp Subject: Re: Observation: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1995Oct20.193710.3631@hanse.de> Date: Fri, 20 Oct 1995 19:37:10 GMT References: <1995Oct20.092410.4514@u-tokyo.ac.jp> Organization: Hanse-Netzwerk Hamburg NNTP-Posting-Host: hanse.de X-Newsreader: NewsWatcher 2.0b24 Lines: 9 In article <1995Oct20.092410.4514@u-tokyo.ac.jp>, dsmirnov@u-tokyo.ac.jp (Dave Smirnov) writes: > Porting stardust.c to the NeXTcube and Atari Falcon DSP subsystems. Beautiful assembly code! The 24-bit word size of the 56001 is tailor-made for high-precision audio filtering. -- === Heather Ito === Dept of Electrical Engineering, Hanse-Netzwerk Hamburg "The hatch is open. Is anyone listening?" From dvasiliev@hanse.de Sat Oct 21 01:38:10 1995 Path: ashpoly!bigsur!pacrim!jpgate!kamome!hanse!dvasiliev From: dvasiliev@hanse.de (Dong-hyun Vasiliev) Newsgroups: comp.dsp Subject: Re: Observation: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1995Oct21.013810.8816@hanse.de> Date: Sat, 21 Oct 1995 01:38:10 GMT References: <1995Oct20.092410.4514@u-tokyo.ac.jp> Organization: Hanse-Netzwerk Hamburg NNTP-Posting-Host: hanse.de X-Newsreader: TIN [version 1.2 PL2] Lines: 9 In article <1995Oct20.092410.4514@u-tokyo.ac.jp>, dsmirnov@u-tokyo.ac.jp (Dave Smirnov) 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. Ran your code on my NeXTstation Turbo. It locked onto the carrier within 1.2 seconds of vinyl lead-in. -- Dong-hyun Vasiliev Tape trading list upon request. SASE required for international mail.