From dkuznetsov@utexas.edu Wed Nov 27 10:21:35 1996 Path: ashpoly!cascade!transat!eunet-fi!relcom!kiae!utexas!dkuznetsov From: dkuznetsov@utexas.edu (Dave Kuznetsov) Newsgroups: comp.dsp Subject: Observation: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1996Nov27.102135.7254@utexas.edu> Date: Wed, 27 Nov 1996 10:21:35 GMT Organization: University of Texas at Austin NNTP-Posting-Host: utexas.edu X-Newsreader: TIN [version 1.2 PL2] Lines: 10 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. Source code in 56k assembly uploaded to ftp.ashford-poly.edu:/pub/telecom-archives/dsp56k/stardust_56k.asm. -- Dave Kuznetsov Tape trading list upon request. SASE required for international mail. From jbrennan@mit.edu Wed Nov 27 14:13:35 1996 Path: ashpoly!bigsur!pacrim!jpgate!kamome!mit!jbrennan From: jbrennan@mit.edu (Joachim Brennan) Newsgroups: comp.dsp Subject: Re: Observation: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1996Nov27.141335.6118@mit.edu> Date: Wed, 27 Nov 1996 14:13:35 GMT References: <1996Nov27.102135.7254@utexas.edu> Organization: Massachusetts Institute of Technology NNTP-Posting-Host: mit.edu X-Newsreader: WinVN 0.99.5 Lines: 10 In article <1996Nov27.102135.7254@utexas.edu>, dkuznetsov@utexas.edu (Dave Kuznetsov) 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. -- Joachim Brennan SupraFAXModem 28.8 @ 14,400 bps "Marcus is not one man." From tjohansson@titech.ac.jp Fri Nov 29 10:52:35 1996 Path: ashpoly!cascade!mxgate!coyoacan!titech!tjohansson From: tjohansson@titech.ac.jp (Tyler Johansson) Newsgroups: comp.dsp Subject: Re: Observation: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1996Nov29.105235.6604@titech.ac.jp> Date: Fri, 29 Nov 1996 10:52:35 GMT References: <1996Nov27.102135.7254@utexas.edu> <1996Nov27.141335.6118@mit.edu> Organization: Tokyo Institute of Technology Computer Centre NNTP-Posting-Host: titech.ac.jp X-Newsreader: Netscape 1.1N Lines: 8 In article <1996Nov27.141335.6118@mit.edu>, jbrennan@mit.edu (Joachim Brennan) writes: > Beautiful assembly code! The 24-bit word size of the 56001 is tailor-made for high-precision audio filtering. Real programmers write in C, not arcane DSP microcode. With Pentium chips hitting 100 MHz, dedicated DSPs are obsolete. -- Tyler Johansson Tape trading list upon request. SASE required for international mail. From asilva@u-tokyo.ac.jp Fri Nov 29 22:12:35 1996 Path: ashpoly!cascade!bigsur!u-tokyo!asilva From: asilva@u-tokyo.ac.jp (Amy Silva) Newsgroups: comp.dsp Subject: Re: Observation: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1996Nov29.221235.1120@u-tokyo.ac.jp> Date: Fri, 29 Nov 1996 22:12:35 GMT References: <1996Nov27.102135.7254@utexas.edu> <1996Nov27.141335.6118@mit.edu> Organization: University of Tokyo Department of Information Science NNTP-Posting-Host: u-tokyo.ac.jp X-Newsreader: TIN [version 1.2 PL2] Lines: 9 In article <1996Nov27.141335.6118@mit.edu>, jbrennan@mit.edu (Joachim Brennan) writes: > Beautiful assembly code! The 24-bit word size of the 56001 is tailor-made for high-precision audio filtering. Real programmers write in C, not arcane DSP microcode. With Pentium chips hitting 100 MHz, dedicated DSPs are obsolete. -- Amy Silva | asilva@u-tokyo.ac.jp PGP Key fingerprint: 6683 1E56 D628 ADE2 "Keep the channels open." From arodriguez@mindspring.com Mon Dec 02 00:17:35 1996 Path: ashpoly!cascade!brgate!saturno!mindspring!arodriguez From: arodriguez@mindspring.com (Amy Rodríguez) Newsgroups: comp.dsp Subject: Re: Observation: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1996Dec02.001735.1297@mindspring.com> Date: Mon, 02 Dec 1996 00:17:35 GMT References: <1996Nov27.102135.7254@utexas.edu> <1996Nov27.141335.6118@mit.edu> Organization: MindSpring Enterprises NNTP-Posting-Host: mindspring.com X-Newsreader: Netscape 1.1N Lines: 9 In article <1996Nov27.141335.6118@mit.edu>, jbrennan@mit.edu (Joachim Brennan) writes: > Beautiful assembly code! The 24-bit word size of the 56001 is tailor-made for high-precision audio filtering. Ran your code on my NeXTstation Turbo. It locked onto the carrier within 1.2 seconds of vinyl lead-in. -- Amy Rodríguez Hayes Optima 288 @ ISDN 64k bps "Hardware is temporary. The signal survives."