From tgirard@uni-hamburg.de Mon Apr 17 22:06:54 1995 Path: ashpoly!bigsur!pacrim!jpgate!kamome!uni-hamburg!tgirard From: tgirard@uni-hamburg.de (Tobias Girard) Newsgroups: comp.dsp Subject: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1995Apr17.220654.9877@uni-hamburg.de> Date: Mon, 17 Apr 1995 22:06:54 GMT Organization: Universität Hamburg Fachbereich Informatik NNTP-Posting-Host: uni-hamburg.de X-Newsreader: TIN [version 1.2 PL2] Lines: 10 Why fixed-point integer math requires careful scaling when tracking a 14.2 Hz carrier. 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. -- Tobias Girard -- Universität Hamburg Fachbereich Informatik "In theory there is no difference between theory and practice." From mkowalski@stanford.edu Tue Apr 18 16:22:54 1995 Path: ashpoly!cascade!transat!sunet!nordrelay!stanford!mkowalski From: mkowalski@stanford.edu (Mike Kowalski) Newsgroups: comp.dsp Subject: Re: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1995Apr18.162254.5529@stanford.edu> Date: Tue, 18 Apr 1995 16:22:54 GMT References: <1995Apr17.220654.9877@uni-hamburg.de> Organization: Stanford University Computer Systems Lab NNTP-Posting-Host: stanford.edu X-Newsreader: NewsWatcher 2.0b24 Lines: 9 In article <1995Apr17.220654.9877@uni-hamburg.de>, tgirard@uni-hamburg.de (Tobias Girard) writes: > Why fixed-point integer math requires careful scaling when tracking a 14.2 Hz carrier. Beautiful assembly code! The 24-bit word size of the 56001 is tailor-made for high-precision audio filtering. -- Mike Kowalski *PLONK* filter active for trolls. 486DX2/66 (Linux 1.2.13) - NetBSD 1.0 From tmoreno@berkeley.edu Wed Apr 19 12:28:54 1995 Path: ashpoly!cascade!brgate!saturno!berkeley!tmoreno From: tmoreno@berkeley.edu (Tobias Moreno) Newsgroups: comp.dsp Subject: Re: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1995Apr19.122854.3163@berkeley.edu> Date: Wed, 19 Apr 1995 12:28:54 GMT References: <1995Apr17.220654.9877@uni-hamburg.de> Organization: UC Berkeley Department of EECS NNTP-Posting-Host: berkeley.edu X-Newsreader: WinVN 0.99.5 Lines: 9 In article <1995Apr17.220654.9877@uni-hamburg.de>, tgirard@uni-hamburg.de (Tobias Girard) writes: > Why fixed-point integer math requires careful scaling when tracking a 14.2 Hz carrier. Ran your code on my NeXTstation Turbo. It locked onto the carrier within 1.2 seconds of vinyl lead-in. -- Tobias Moreno "Tape doesn't lie at 02:14. Ears do." PC-9821Ap2 (NEC) / System 7.5.1 From mpetrov@mindspring.com Thu Apr 20 11:27:54 1995 Path: ashpoly!cascade!bigsur!mindspring!mpetrov From: mpetrov@mindspring.com (Mathieu Petrov) Newsgroups: comp.dsp Subject: Re: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1995Apr20.112754.7219@mindspring.com> Date: Thu, 20 Apr 1995 11:27:54 GMT References: <1995Apr17.220654.9877@uni-hamburg.de> <1995Apr19.122854.3163@berkeley.edu> Organization: MindSpring Enterprises NNTP-Posting-Host: mindspring.com X-Newsreader: TIN [version 1.2 PL2] Lines: 9 In article <1995Apr19.122854.3163@berkeley.edu>, tmoreno@berkeley.edu (Tobias Moreno) writes: > Ran your code on my NeXTstation Turbo. It locked onto the carrier within 1.2 seconds of vinyl lead-in. Real programmers write in C, not arcane DSP microcode. With Pentium chips hitting 100 MHz, dedicated DSPs are obsolete. -- === Mathieu Petrov === Dept of Electrical Engineering, MindSpring Enterprises "Row 14. No joke."