Article View: comp.sys.mac.scitech
Article #18Re: Frequency analysis of sound samples?
From: Jean-Yves Bernier
Date: Tue, 08 Nov 1994 02:27
Date: Tue, 08 Nov 1994 02:27
33 lines
2717 bytes
2717 bytes
In article <39ks81$pr@rubb.rz.ruhr-uni-bochum.de>, ken@mun.neuroinformatik.ruhr-uni-bochum.de (Ken Flaton) wrote: >To further warp this thread: I'm told that some macs have built-in DSPs >(which are famously wonderful for doing FFTs). Anyone know if 1) that's >true (the former) and 2) if the DSP is user-programmable? You should say: some macs HAD built-in DSPs (the Quadra AV family). The DSP was an interim solution to provide the power needed for audio&visual applications. And yes, it was user-programmable, as long as you was a Mac developer with AT&T DSP assembler knowledge. Apple defined and documented a real-time software architecture named ARTA for managing DSP processes. But it's unlikely that a new Mac will feature a DSP, since it would not offer much more than the Power chip. As many developpers, I skipped the DSP Mac generation, being too much investment, and jumped to PowerMac. But back to sound analysis. It's really amazing what you can do with a Mac and a DSP. I get a fourier sonagram (I speak of two-dimensional frequency analysis, that is energy versus frequency and time) in REAL-TIME on a good old fx with some DSP board plugged in. This kind of analysis is very useful in the case of transient sounds (speech, music) where frequency-content changes rapidly. In fact, the biggest problem with the Mac was to have CopyBits blast the pixels fast enough to the screen so that the resulting picture scrolls smoothly. --------------------------- Jean-Yves Bernier - SPECTRA bernier@dialup.francenet.fr ---------------------------
Technical Details
Message-ID:
<AAE4A0A49668FF279@bernier.dialup.francenet.fr>Path:
arc1-ca.pugleaf.net!nntp.gmd.de!xlink.net!sol.ctr.columbia.edu!howland.reston.ans.net!math.ohio-state.edu!jussieu.fr!oleane!francenet.fr!newsReferences:
<georg.783334309@marie> <p.kerr-3110940906570001@130.216.90.127> <Jeff_Mandel-0111942106590001@134.174.90.4> <2NOV94.08161416@uwpg02.uwinnipeg.ca> <ejensen-0211941513500001@ejensen.esxf.uwindsor.ca> <39ks81$pr@rubb.rz.ruhr-uni-bochum.de>Show Full Headers
Complete Article Headers
X-Google-Language: ENGLISH,ASCII-7-bit X-Google-Thread: feb82,fe492bcff3c44b5c X-Google-Attributes: gidfeb82,public X-Google-Thread: 10b5e3,fe492bcff3c44b5c X-Google-Attributes: gid10b5e3,public X-Google-Thread: 104d1d,fe492bcff3c44b5c X-Google-Attributes: gid104d1d,public X-Google-ArrivalTime: 1994-11-07 20:02:15 PST Path: arc1-ca.pugleaf.net!nntp.gmd.de!xlink.net!sol.ctr.columbia.edu!howland.reston.ans.net!math.ohio-state.edu!jussieu.fr!oleane!francenet.fr!news From: bernier@dialup.francenet.fr (Jean-Yves Bernier) Newsgroups: comp.sys.mac.scitech,comp.sys.mac.apps,aus.computers.mac Subject: Re: Frequency analysis of sound samples? Date: Tue, 08 Nov 1994 03:27:32 +0100 Organization: SPECTRA Soft Lines: 34 Distribution: world Message-ID: <AAE4A0A49668FF279@bernier.dialup.francenet.fr> References: <georg.783334309@marie> <p.kerr-3110940906570001@130.216.90.127> <Jeff_Mandel-0111942106590001@134.174.90.4> <2NOV94.08161416@uwpg02.uwinnipeg.ca> <ejensen-0211941513500001@ejensen.esxf.uwindsor.ca> <39ks81$pr@rubb.rz.ruhr-uni-bochum.de> NNTP-Posting-Host: bernier.dialup.francenet.fr X-Newsreader: NewsHopper