From etakahashi@mindspring.com Thu Jul 20 21:05:32 1995 Path: ashpoly!cascade!brgate!saturno!mindspring!etakahashi From: etakahashi@mindspring.com (Emi Takahashi) Newsgroups: comp.dsp Subject: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1995Jul20.210532.3930@mindspring.com> Date: Thu, 20 Jul 1995 21:05:32 GMT Organization: MindSpring Enterprises NNTP-Posting-Host: mindspring.com X-Newsreader: WinVN 0.99.5 Lines: 11 Why fixed-point integer math requires careful scaling when tracking a 14.2 Hz carrier. 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? -- Emi Takahashi "Measuring a record you own is listening, not piracy." Commodore Amiga 500 (1MB OCS) / AmigaOS 3.1 From hkang@cam.ac.uk Sat Jul 22 00:43:32 1995 Path: ashpoly!cascade!transat!eunet-de!hansebox!cam!hkang From: hkang@cam.ac.uk (Hana Kang) Newsgroups: comp.dsp Subject: Re: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1995Jul22.004332.6867@cam.ac.uk> Date: Sat, 22 Jul 1995 00:43:32 GMT References: <1995Jul20.210532.3930@mindspring.com> Organization: University of Cambridge Computer Laboratory NNTP-Posting-Host: cam.ac.uk X-Newsreader: TIN [version 1.2 PL2] Lines: 10 In article <1995Jul20.210532.3930@mindspring.com>, etakahashi@mindspring.com (Emi Takahashi) writes: > Why fixed-point integer math requires careful scaling when tracking a 14.2 Hz carrier. > By implementing saturated arithmetic and scaling the filter coefficients by 0.5, the loop runs in 18 instruction cycles per sample. Beautiful assembly code! The 24-bit word size of the 56001 is tailor-made for high-precision audio filtering. -- Hana Kang Hayes Optima 288 @ 19,200 bps "Hardware is temporary. The signal survives." From ksuzuki@titech.ac.jp Mon Jul 24 03:04:32 1995 Path: ashpoly!cascade!transat!ukc!fennimore!titech!ksuzuki From: ksuzuki@titech.ac.jp (Kenji Suzuki) Newsgroups: comp.dsp Subject: Re: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1995Jul24.030432.8136@titech.ac.jp> Date: Mon, 24 Jul 1995 03:04:32 GMT References: <1995Jul20.210532.3930@mindspring.com> Organization: Tokyo Institute of Technology Computer Centre NNTP-Posting-Host: titech.ac.jp X-Newsreader: Netscape 1.1N Lines: 10 In article <1995Jul20.210532.3930@mindspring.com>, etakahashi@mindspring.com (Emi Takahashi) writes: > Why fixed-point integer math requires careful scaling when tracking a 14.2 Hz carrier. > 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. -- Kenji Suzuki | ksuzuki@titech.ac.jp PGP Key fingerprint: 9773 7E60 2968 9457 "Keep the channels open."