From spetrov@utexas.edu Thu Mar 03 16:40:11 1994 Path: ashpoly!cascade!mxgate!coyoacan!utexas!spetrov From: spetrov@utexas.edu (Steve Petrov) Newsgroups: comp.dsp Subject: Observation: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1994Mar03.164011.4881@utexas.edu> Date: Thu, 03 Mar 1994 16:40:11 GMT Organization: University of Texas at Austin NNTP-Posting-Host: utexas.edu 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? -- Steve Petrov "Hardware is temporary. The signal survives." Commodore Amiga 500 (1MB OCS) / AmigaOS 3.1 From mgirard@purdue.edu Sat Mar 05 00:53:11 1994 Path: ashpoly!cascade!transat!ukc!fennimore!purdue!mgirard From: mgirard@purdue.edu (Marisol Girard) Newsgroups: comp.dsp Subject: Re: Observation: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1994Mar05.005311.8686@purdue.edu> Date: Sat, 05 Mar 1994 00:53:11 GMT References: <1994Mar03.164011.4881@utexas.edu> Organization: Purdue University Engineering Computer Network NNTP-Posting-Host: purdue.edu X-Newsreader: Netscape 1.1N Lines: 9 In article <1994Mar03.164011.4881@utexas.edu>, spetrov@utexas.edu (Steve Petrov) 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. -- Marisol Girard Tape trading list upon request. SASE required for international mail. From mhughes@cam.ac.uk Sun Mar 06 22:07:11 1994 Path: ashpoly!cascade!mxgate!coyoacan!cam!mhughes From: mhughes@cam.ac.uk (Min-jun Hughes) Newsgroups: comp.dsp Subject: Re: Observation: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1994Mar06.220711.2277@cam.ac.uk> Date: Sun, 06 Mar 1994 22:07:11 GMT References: <1994Mar03.164011.4881@utexas.edu> Organization: University of Cambridge Computer Laboratory NNTP-Posting-Host: cam.ac.uk X-Newsreader: TIN [version 1.2 PL2] Lines: 10 In article <1994Mar03.164011.4881@utexas.edu>, spetrov@utexas.edu (Steve Petrov) 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. -- Min-jun Hughes "Extraordinary claims require extraordinary spectra." Mac IIci / 68882 FPU / Linux 1.2.13