Les Cargill wrote:> "Eric C. Weaver" wrote: > >>Les Cargill wrote: >> >> >>>Digital isn't FM and vice versa. DX-7s and later revisions of >>>DX-7s are not hard to find. >> >>Beg yer pardon, but FM and digital are not mutually exclusive at all. >> >>I impelemented FM sythesis on an LSI-11 micro in assembler back in 1983 or so, >>and it worked as well as any single-operator FM synthesizer could. > > > So this process was purely in the digital domain?Yes. The only thing after the D/A was an amplifier. A cheap one from Radio Shack.
Sick of sample-based "synthesis"!
Started by ●October 16, 2003
Reply by ●October 23, 20032003-10-23
Reply by ●October 23, 20032003-10-23
In article bn8sm3$hv0$1@bob.news.rcn.net, Jerry Avins at jya@ieee.org wrote on 10/23/2003 11:38:> Radium wrote: > >> Randy Yates <yates@ieee.org> wrote in message >> news:<5Wllb.10312$W16.1582@newsread2.news.atl.earthlink.net>... >> >>> it's >>> much easier to spend all your time and effort getting the wavetable >>> synthesis working rather than having to build both a wavetable synthesis >>> and FM synthesis engine (so that FM sounds could be generated instead >>> of reproduced through the wavetable). >> >> >> Then why not build the FM synth engine without a wavetable synth? >> >> In "real" FM synthesis no wavetable should be needed. Right?how do you generate the sinusoids that get modulated in FM without some sorta wavetable? you ain't gonna compute them from a power series on the fly are you? (there are numerical problems with that besides wasted computational bandwidth.) i think all FM synths had a phase-accumulator lookup table for the sinusoids and that is really what a wavetable is.> You miss the point. It's assumed that you need wavetables no matter > what. Your "real FM synth" is extra. > > Please explain again why a wavetable synth can't reproduce the sound of > an FM synthesizer as well as a CD can. I missed that the first time > around.there should be some semantic differentiation between "wavetable synthesis" (as i have used the term) and "PCM sample playback" (which is what most people mean by "wavetable synthesis" and what many dedicated chips are doing). in "true" wavetable synthesis, FM waveforms can be produced (by mathematically predefining the wavetables) if all of the modulation and carrier frequencies are low integer multiples of a common fundamental frequency. r b-j
Reply by ●October 23, 20032003-10-23
Okay wavetables may not be so bad. Sample-playback synths are the "real" perpetrators. robert bristow-johnson <rbj@surfglobal.net> wrote in message news:<BBBD8245.4C0C%rbj@surfglobal.net>...> there should be some semantic differentiation between "wavetable synthesis" > (as i have used the term) and "PCM sample playback" (which is what most > people mean by "wavetable synthesis" and what many dedicated chips are > doing). in "true" wavetable synthesis, FM waveforms can be produced (by > mathematically predefining the wavetables) if all of the modulation and > carrier frequencies are low integer multiples of a common fundamental > frequency.
Reply by ●October 23, 20032003-10-23
robert bristow-johnson <rbj@surfglobal.net> wrote in message news:<BBBD8245.4C0C%rbj@surfglobal.net>...> how do you generate the sinusoids that get modulated in FM without some > sorta wavetable? you ain't gonna compute them from a power series on the > fly are you?"compute them from a power series on the fly" Huh?
Reply by ●October 24, 20032003-10-24
In article 464c821f.0310231845.1c70ec6d@posting.google.com, Radium at glucegen@excite.com wrote on 10/23/2003 22:45:> robert bristow-johnson <rbj@surfglobal.net> wrote in message > news:<BBBD8245.4C0C%rbj@surfglobal.net>... >> how do you generate the sinusoids that get modulated in FM without some >> sorta wavetable? you ain't gonna compute them from a power series on the >> fly are you? > > "compute them from a power series on the fly"often math libraries used by higher level languages like C or programs like Excel or Matlab or Mathmatica use some kinda power series to do trancendental functions (of which sin() or cos() belong). i suppose if one were not to use wavetables or the CORDIC (there are little problems with using CORDIC for FM, Allen, but they are certainly surmountable), if you were to write a C program to do FM in the most straightforward was, you would be computing your sin() or cos() functions with a power series. if this was a real-time synth, then you would be doing it "on the fly". r b-j
Reply by ●October 24, 20032003-10-24
On Thu, 23 Oct 2003, robert bristow-johnson wrote:> how do you generate the sinusoids that get modulated in FM without some > sorta wavetable? you ain't gonna compute them from a power series on the > fly are you? (there are numerical problems with that besides wasted > computational bandwidth.) i think all FM synths had a phase-accumulator > lookup table for the sinusoids and that is really what a wavetable is.In the Sound Blaster AWE64 ISA soundcard, the OPL-3 emulation used a second order polynomial approximation of sine. I didn't really enjoy the way it sounded, compared to a real OPL-2, which stores one sine quarter in a look-up table of 256 samplepoints. -olli
Reply by ●October 24, 20032003-10-24
On 23 Oct 2003 19:41:16 -0700, Radium <glucegen@excite.com> wrote:> Okay wavetables may not be so bad. Sample-playback synths are the > "real" perpetrators. >Usually, "That Depends" is a pretty good answer. As a player, the most important thing to me is How It Plays. So while "everyone wanted" a first-generation Kurzweil or an Emu Emulator, as a player they were worse than useless. The early K's gave you an exact replication of one or several notes of a piano played with what seemed to be a sledge hammer. Once you get past Jerry Lee Lewis, you've exhausted the expressive possibilities of that instrument. Nowadays, the pseudopianos I play have a combination of "Magic" for the attack--every manufacturer claims THEIR Mojo is SUPERIOR, whether it be an interpolated multisample based on the phase of the moon and the splattering of blood from a fresh-killed black cockeral or an exacting envelope analysis of a real piano. As a player, they're all "Good Enough"--especially once you add the rest of the rhythm section, and even moreso if you're a veteran of the Helpensteil and Yamaha CP-series pianos. Now where did I leave my chriopractor . . .
Reply by ●October 24, 20032003-10-24
robert bristow-johnson <rbj@surfglobal.net> wrote in message news:<BBBE1412.4C22%rbj@surfglobal.net>...> In article 464c821f.0310231845.1c70ec6d@posting.google.com, Radium at > glucegen@excite.com wrote on 10/23/2003 22:45: > > > robert bristow-johnson <rbj@surfglobal.net> wrote in message > > news:<BBBD8245.4C0C%rbj@surfglobal.net>... > >> how do you generate the sinusoids that get modulated in FM without some > >> sorta wavetable? you ain't gonna compute them from a power series on the > >> fly are you? > > > > "compute them from a power series on the fly" > > often math libraries used by higher level languages like C or programs like > Excel or Matlab or Mathmatica use some kinda power series to do > trancendental functions (of which sin() or cos() belong). i suppose if one > were not to use wavetables or the CORDIC (there are little problems with > using CORDIC for FM, Allen, but they are certainly surmountable), if you > were to write a C program to do FM in the most straightforward was, you > would be computing your sin() or cos() functions with a power series. if > this was a real-time synth, then you would be doing it "on the fly".A real-time synth is better than any wavetable. I would prefer "on the fly" to any pre-recorded technique. BTW, is Yamaha's FSR1 "on the fly"?
Reply by ●October 24, 20032003-10-24
In comp.dsp, Piergiorgio Sartor <piergiorgio.sartor@nexgo.REMOVE.THIS.de> wrote:>Radium wrote: > >> FM synthesis is a digital method. Anyone who thinks different need to >> read more about FM synthesis. > >FM synthesis is FM synthesis, you can do in the >digital or analog domain. > >It just happens to be more convenient in the >digital domain... (more tunable).A slight frequency or phase change between the 'modulator' oscillator and the modulated oscillator makes a substantial, even dramatic, change in timbre. One analog oscillator FM-ing another will have this 'phasing' effect because their frequencies cannot be controlled exactly. With phase-accumulator-controlled digital oscillators, the oscillators are in sync and thus the phase doesn't change (unless one oscillator is intentionally detuned). Timbre is often changed in FM synthesis by changing (using an ADSR signa) the amplitude of the modulating signal. You could do this with analog oscillators, but it would sound different each time because of the phase drift. Actually, depending on what effect you want, this could be a good thing...>bye, > >-- > Piergiorgio Sartor >
Reply by ●October 26, 20032003-10-26
Jerry Avins <jya@ieee.org> wrote in message news:<bn8sm3$hv0$1@bob.news.rcn.net>...> Please explain again why a wavetable synth can't reproduce the sound of > an FM synthesizer as well as a CD can.The point is not to reproduce an FM synth's sound but to listen to it while producing it. That is, listen to an FM synth actually generate its musical tones, rather than recording those tones from the FM synth to another medium and then playing them back.
