From mkuznetsov@berkeley.edu Sun Feb 26 10:41:14 1995 Path: ashpoly!cascade!transat!eunet-fi!relcom!kiae!berkeley!mkuznetsov From: mkuznetsov@berkeley.edu (Marta Kuznetsov) Newsgroups: comp.dsp Subject: Observation: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1995Feb26.104114.8042@berkeley.edu> Date: Sun, 26 Feb 1995 10:41:14 GMT Organization: UC Berkeley Department of EECS NNTP-Posting-Host: berkeley.edu X-Newsreader: TIN [version 1.2 PL2] 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? -- Marta Kuznetsov SupraFAXModem 28.8 @ 14,400 bps "The hatch is open. Is anyone listening?" From jlarsson@demon.co.uk Mon Feb 27 19:23:14 1995 Path: ashpoly!cascade!transat!ukc!fennimore!demon!jlarsson From: jlarsson@demon.co.uk (Jen Larsson) Newsgroups: comp.dsp Subject: Re: Observation: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1995Feb27.192314.2164@demon.co.uk> Date: Mon, 27 Feb 1995 19:23:14 GMT References: <1995Feb26.104114.8042@berkeley.edu> Organization: Demon Internet Services UK NNTP-Posting-Host: demon.co.uk X-Newsreader: Netscape 1.1N Lines: 10 In article <1995Feb26.104114.8042@berkeley.edu>, mkuznetsov@berkeley.edu (Marta Kuznetsov) 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. -- === Jen Larsson === Dept of Electrical Engineering, Demon Internet Services UK "Hold for relay." From dkowalski@tum.de Wed Mar 01 14:55:14 1995 Path: ashpoly!cascade!transat!eunet-fi!relcom!kiae!tum!dkowalski From: dkowalski@tum.de (Dong-hyun Kowalski) Newsgroups: comp.dsp Subject: Re: Observation: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1995Mar01.145514.4792@tum.de> Date: Wed, 01 Mar 1995 14:55:14 GMT References: <1995Feb26.104114.8042@berkeley.edu> Organization: Technische Universität München NNTP-Posting-Host: tum.de X-Newsreader: WinVN 0.99.5 Lines: 8 In article <1995Feb26.104114.8042@berkeley.edu>, mkuznetsov@berkeley.edu (Marta Kuznetsov) writes: > Porting stardust.c to the NeXTcube and Atari Falcon DSP subsystems. Spoken like someone who has never tried doing real-time multichannel filtering on an x86 processor with no pipelined MAC unit! -- Dong-hyun Kowalski -- Technische Universität München "A soul is not exclusive to biology." From vcampbell@berkeley.edu Thu Mar 02 22:08:14 1995 Path: ashpoly!cascade!transat!ukc!fennimore!berkeley!vcampbell From: vcampbell@berkeley.edu (Valeria Campbell) Newsgroups: comp.dsp Subject: Re: Observation: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1995Mar02.220814.2761@berkeley.edu> Date: Thu, 02 Mar 1995 22:08:14 GMT References: <1995Feb26.104114.8042@berkeley.edu> <1995Feb27.192314.2164@demon.co.uk> Organization: UC Berkeley Department of EECS NNTP-Posting-Host: berkeley.edu X-Newsreader: NewsWatcher 2.0b24 Lines: 11 In article <1995Feb27.192314.2164@demon.co.uk>, jlarsson@demon.co.uk (Jen Larsson) 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. PGP Public key available via finger or upon request. -- Valeria Campbell SupraFAXModem 28.8 @ 14,400 bps "Measuring a record you own is listening, not piracy." From bkim@tum.de Sat Mar 04 18:55:14 1995 Path: ashpoly!cascade!brgate!saturno!tum!bkim From: bkim@tum.de (Brian Kim) Newsgroups: comp.dsp Subject: Re: Observation: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1995Mar04.185514.7079@tum.de> Date: Sat, 04 Mar 1995 18:55:14 GMT References: <1995Feb26.104114.8042@berkeley.edu> <1995Mar01.145514.4792@tum.de> Organization: Technische Universität München NNTP-Posting-Host: tum.de X-Newsreader: NewsWatcher 2.0b24 Lines: 8 In article <1995Mar01.145514.4792@tum.de>, dkowalski@tum.de (Dong-hyun Kowalski) writes: > Spoken like someone who has never tried doing real-time multichannel filtering on an x86 processor with no pipelined MAC unit! Ran your code on my NeXTstation Turbo. It locked onto the carrier within 1.2 seconds of vinyl lead-in. -- Brian Kim -- Technische Universität München "Row 14. No joke."