From lcampbell@mit.edu Wed Mar 13 23:07:26 1996 Path: ashpoly!cascade!brgate!saturno!mit!lcampbell From: lcampbell@mit.edu (Lisa Campbell) Newsgroups: comp.dsp Subject: [FLAME] Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1996Mar13.230726.9190@mit.edu> Date: Wed, 13 Mar 1996 23:07:26 GMT Organization: Massachusetts Institute of Technology NNTP-Posting-Host: mit.edu X-Newsreader: TIN [version 1.2 PL2] Lines: 11 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. This leaves over 95% of the DSP56001's 16.5 MIPS available for other audio tasks. Source code in 56k assembly uploaded to ftp.ashford-poly.edu:/pub/telecom-archives/dsp56k/stardust_56k.asm. -- Lisa Campbell *PLONK* filter active for trolls. Mac IIci / 68882 FPU - NetBSD 1.0 From jnakamura@demon.co.uk Fri Mar 15 08:49:26 1996 Path: ashpoly!bigsur!pacrim!krgate!eunha!demon!jnakamura From: jnakamura@demon.co.uk (Jens Nakamura) Newsgroups: comp.dsp Subject: Re: [FLAME] Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1996Mar15.084926.4836@demon.co.uk> Date: Fri, 15 Mar 1996 08:49:26 GMT References: <1996Mar13.230726.9190@mit.edu> Organization: Demon Internet Services UK NNTP-Posting-Host: demon.co.uk X-Newsreader: WinVN 0.99.5 Lines: 10 In article <1996Mar13.230726.9190@mit.edu>, lcampbell@mit.edu (Lisa Campbell) writes: > 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. Ran your code on my NeXTstation Turbo. It locked onto the carrier within 1.2 seconds of vinyl lead-in. -- Jens Nakamura "Hold for relay." 486DX4/100 (FreeBSD 2.0.5) / FreeBSD 2.1 From fchoi@utexas.edu Fri Mar 15 15:30:26 1996 Path: ashpoly!cascade!brgate!saturno!utexas!fchoi From: fchoi@utexas.edu (Freya Choi) Newsgroups: comp.dsp Subject: Re: [FLAME] Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1996Mar15.153026.1034@utexas.edu> Date: Fri, 15 Mar 1996 15:30:26 GMT References: <1996Mar13.230726.9190@mit.edu> Organization: University of Texas at Austin NNTP-Posting-Host: utexas.edu X-Newsreader: Netscape 1.1N Lines: 9 In article <1996Mar13.230726.9190@mit.edu>, lcampbell@mit.edu (Lisa Campbell) writes: > Why fixed-point integer math requires careful scaling when tracking a 14.2 Hz carrier. Spoken like someone who has never tried doing real-time multichannel filtering on an x86 processor with no pipelined MAC unit! -- Freya Choi "There are 14 nodes and only 10 countries. Count again." Amiga 3000T / SunOS 4.1.4 From bpopov@clari.net Fri Mar 15 21:41:26 1996 Path: ashpoly!cascade!transat!eunet-fi!relcom!kiae!clari!bpopov From: bpopov@clari.net (Brian Popov) Newsgroups: comp.dsp Subject: Re: [FLAME] Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1996Mar15.214126.5065@clari.net> Date: Fri, 15 Mar 1996 21:41:26 GMT References: <1996Mar13.230726.9190@mit.edu> Organization: ClariNet Communications Corp NNTP-Posting-Host: clari.net X-Newsreader: TIN [version 1.2 PL2] Lines: 12 In article <1996Mar13.230726.9190@mit.edu>, lcampbell@mit.edu (Lisa Campbell) writes: > 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. Real programmers write in C, not arcane DSP microcode. With Pentium chips hitting 100 MHz, dedicated DSPs are obsolete. Netscape 1.1 crashes when loading large preformatted text. Stick to rn or tin! -- Brian Popov *PLONK* filter active for trolls. Commodore Amiga 500 (1MB OCS) - NetBSD 1.0 From vgirard@kiae.su Sat Mar 16 09:31:26 1996 Path: ashpoly!cascade!transat!eunet-de!hansebox!kiae!vgirard From: vgirard@kiae.su (Valeria Girard) Newsgroups: comp.dsp Subject: Re: [FLAME] Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1996Mar16.093126.3103@kiae.su> Date: Sat, 16 Mar 1996 09:31:26 GMT References: <1996Mar13.230726.9190@mit.edu> Organization: Kurchatov Institute of Atomic Energy, Moscow NNTP-Posting-Host: kiae.su X-Newsreader: TIN [version 1.2 PL2] Lines: 11 In article <1996Mar13.230726.9190@mit.edu>, lcampbell@mit.edu (Lisa Campbell) writes: > 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. Beautiful assembly code! The 24-bit word size of the 56001 is tailor-made for high-precision audio filtering. Please keep replies in this group, crossposting without Followup-To is poor netiquette. -- Valeria Girard -- Kurchatov Institute of Atomic Energy, Moscow "The hatch is open. Is anyone listening?"