 |
 |
 |
 |
 |
 |
 |
Forums - Synths & synthesis
Subject: Wavetable Synths vs Modeling Synths
Original Message Date: 17-Nov-02 @ 05:00 PM - Wavetable Synths vs Modeling Synths
is there a difference? whats the difference? i know the XTk is based on wavetable technology, but how does that differ from modeling?
Message 21/29 20-Nov-02 @ 02:50 PM - RE: Wavetable Synths vs Modeling Synths
maybe that was confusing... the XT does _not_ interpolate during playback. no wavetable synth does, you could but i think it would introduce complications in other ways.
what i meant was that you can create your own wavetables, a wavetable has to have 61 waves. if you don't send it all 61 waves, it will interpolate to calculate the 'missing' waves. so, for example, if you put a square wave in wave 0
and a saw in wave 2, it would calculate a 'halfway' wave and put it in wave 1. however, once these three waves are there, we do not interpolate between wave 0 and wave 1, we always go straight from wave 0 to wave 1.
one other thing that gives the XT (and the PPG and original microwave) its unique flavor is that all the samples are only 8 bit.
does the fizmo let you load in your own wavetables? for that reason alone i'd choose an asr-x over a fizmo (or an mr-rack with mr-flash... the architectures are very similar as i understand it, except obviously with less capable programming interfaces...)
what i meant was that you can create your own wavetables, a wavetable has to have 61 waves. if you don't send it all 61 waves, it will interpolate to calculate the 'missing' waves. so, for example, if you put a square wave in wave 0
and a saw in wave 2, it would calculate a 'halfway' wave and put it in wave 1. however, once these three waves are there, we do not interpolate between wave 0 and wave 1, we always go straight from wave 0 to wave 1.
one other thing that gives the XT (and the PPG and original microwave) its unique flavor is that all the samples are only 8 bit.
does the fizmo let you load in your own wavetables? for that reason alone i'd choose an asr-x over a fizmo (or an mr-rack with mr-flash... the architectures are very similar as i understand it, except obviously with less capable programming interfaces...)
Message 22/29 20-Nov-02 @ 07:50 PM - RE: Wavetable Synths vs Modeling Synths
nope, you just get what i believe are the same set of transwaves ensoniq had forever.. which is really kinda pathetic about humanity.. i'd say a last ditch effort to appeal to modern tastes with no r&d investment. if the interface even accessed the entire sound engine it'd get more points.
so just to reiterate, you get what must be (correct me here?) 8 bit waveforms? and a minimal, unspectacular 2 osc va synth engine.. a healthy arpeggiator's prolly supposed to make up for it, but the fx engine is pretty much it for real redemption w/ some limited formant functions and the vocoder.
so just to reiterate, you get what must be (correct me here?) 8 bit waveforms? and a minimal, unspectacular 2 osc va synth engine.. a healthy arpeggiator's prolly supposed to make up for it, but the fx engine is pretty much it for real redemption w/ some limited formant functions and the vocoder.
Message 23/29 20-Nov-02 @ 08:06 PM - RE: Wavetable Synths vs Modeling Synths
should be 16 bit, if it's based on the MR engine (which i think it is). sounds like the ASRX is a much better choice then (you still need an external editor to get to everything though), or maybe an MR synth w/ MRflash (near impossible to find)...
also, for sound morphing, i've heard of similar things being done on the k5000... you can take 4 or 6, can't remember, harmonic 'snippets' (sounddiver will generate these for you i think, or you can enter them manually), and then have it generate envelopes that go from one to the next. this would be a lot smoother i guess than interpolating waves (but a lot harder to control too maybe)...
and for the original poster... if you can't tell by now VA and wavetable are pretty different. which one you want will be a personal thing.
also, for sound morphing, i've heard of similar things being done on the k5000... you can take 4 or 6, can't remember, harmonic 'snippets' (sounddiver will generate these for you i think, or you can enter them manually), and then have it generate envelopes that go from one to the next. this would be a lot smoother i guess than interpolating waves (but a lot harder to control too maybe)...
and for the original poster... if you can't tell by now VA and wavetable are pretty different. which one you want will be a personal thing.
Message 24/29 20-Nov-02 @ 08:33 PM - RE: Wavetable Synths vs Modeling Synths
'user-tables'?? are you saying with the xtk you can actually shape wavetables (with software i would assume) or create your own and use them as oscillators?
thanks for the info by the way. this is really helping.
thanks for the info by the way. this is really helping.
Message 25/29 20-Nov-02 @ 09:35 PM - RE: Wavetable Synths vs Modeling Synths
actually, i've got an asr-x pro sitting in the corner getting no action whatsoever...
you're saying that if i had it hooked to my PC, that i could access the Ensoniq transwaves?
i've gone through the manual, but don't believe there was any way to access those transwaves from the asr-x pro, except by just playing the presets that utilize them...
i've heard the asr10 can load transwaves... but how? via MIDI, or sampled audio?
you're saying that if i had it hooked to my PC, that i could access the Ensoniq transwaves?
i've gone through the manual, but don't believe there was any way to access those transwaves from the asr-x pro, except by just playing the presets that utilize them...
i've heard the asr10 can load transwaves... but how? via MIDI, or sampled audio?
Message 26/29 20-Nov-02 @ 10:02 PM - RE: Wavetable Synths vs Modeling Synths
This discussion got me to take out my XT manual and read over all the wavetable stuff...
Go d/l the XT manual, and read the description about waves and wavetables. It's actually pretty good, and the translation isn't bad either.
Go d/l the XT manual, and read the description about waves and wavetables. It's actually pretty good, and the translation isn't bad either.
Message 27/29 20-Nov-02 @ 10:16 PM - RE: Wavetable Synths vs Modeling Synths
yes, the XT/XTK (in fact all microwaves) will let you load up your own waves. the waves are 64 samples long, 8 bit waves, single cycle, and you send them over via sysex, then a wavetable (which is 61 waves long, and is basically 'you put wave X in slot 0, you put wave Y in slot 1, etc'. there are some programs that can help with this (including sound diver). so you can make your own waveforms and use them, or a whole wavetable and scan across them.
on the asr-x, remember i don't have an ASR so this is all from research.
first thing you have to do is get ASR-X Tools, afaik (or find the sysex spec, the program is like $39 i think. you can also buy a transwave CD out there for like $39 too). basically, in the 'loop mode', there's like forward, backward, forward + backwards, whatever. internally there's another option 'transwave', but you can't access it from the front panel - ASRX tools uses sysex to turn it on.
a transwave is 'just another sample' in a special format... iirc, each wave in the sample is 128 samples long, single cycle, and it can access up to 128 of them.
there are some programs that assist you with making the transwave samples (including asr-x tools, and a couple freeware things i've found). the first 128 samples in the waveform is wave 0, the next 128 samples in the waveform is wave 1, etc. all the way up. to get it to work properly, you need to make sure they begin and end on zero crossings etc. or it might click when you change waves.
i guess you can program using the existing transwaves too, but like i said i've never used one.
the old ASRs use the transwaves exactly the same, except AFAIK you can just choose the transwave loop mode from the front panel. the ASRX obviously doesn't support it as well as the 'old' ASRs from the front panel, but it has the whole MR synth engine in it (including a pair of resonant filters per voice), whereas the old ASR only has non-resonant filters (or a resonant filter in the effects section).
i keep mentioning the MR rack w/flash, you can't find the MR-flash anymore though. it would allow you load in up to 8M of samples, and i think they can be transwaves. i am thinking this would be the best of both worlds (front panel that supports transwaves and everything else, with the filters too), but i haven't actually tried it. and i already have an XT
on the asr-x, remember i don't have an ASR so this is all from research.
first thing you have to do is get ASR-X Tools, afaik (or find the sysex spec, the program is like $39 i think. you can also buy a transwave CD out there for like $39 too). basically, in the 'loop mode', there's like forward, backward, forward + backwards, whatever. internally there's another option 'transwave', but you can't access it from the front panel - ASRX tools uses sysex to turn it on.
a transwave is 'just another sample' in a special format... iirc, each wave in the sample is 128 samples long, single cycle, and it can access up to 128 of them.
there are some programs that assist you with making the transwave samples (including asr-x tools, and a couple freeware things i've found). the first 128 samples in the waveform is wave 0, the next 128 samples in the waveform is wave 1, etc. all the way up. to get it to work properly, you need to make sure they begin and end on zero crossings etc. or it might click when you change waves.
i guess you can program using the existing transwaves too, but like i said i've never used one.
the old ASRs use the transwaves exactly the same, except AFAIK you can just choose the transwave loop mode from the front panel. the ASRX obviously doesn't support it as well as the 'old' ASRs from the front panel, but it has the whole MR synth engine in it (including a pair of resonant filters per voice), whereas the old ASR only has non-resonant filters (or a resonant filter in the effects section).
i keep mentioning the MR rack w/flash, you can't find the MR-flash anymore though. it would allow you load in up to 8M of samples, and i think they can be transwaves. i am thinking this would be the best of both worlds (front panel that supports transwaves and everything else, with the filters too), but i haven't actually tried it. and i already have an XT
Message 28/29 21-Nov-02 @ 03:08 PM - RE: Wavetable Synths vs Modeling Synths
nomad,
thanks for the clarification. i understand what you meant by interpolation now. i kind of read it as "morphing".
it's too bad that getting at the wavetables in the ASR-X series is such a hassle. imagine if they had used some of that HUUUGE surface to add some dedicated knobs for sampling, synthesis, and transwave parameters?
hehe... i did put my resume in with them back in 1999, but i would have been too late had they hired me then.
Scott
thanks for the clarification. i understand what you meant by interpolation now. i kind of read it as "morphing".
it's too bad that getting at the wavetables in the ASR-X series is such a hassle. imagine if they had used some of that HUUUGE surface to add some dedicated knobs for sampling, synthesis, and transwave parameters?
hehe... i did put my resume in with them back in 1999, but i would have been too late had they hired me then.
Scott
Message 29/29 21-Nov-02 @ 07:59 PM - RE: Wavetable Synths vs Modeling Synths
fyi.. the current series of synthedit betas are prioritising improved soundfont implementation.. the author is a 'synth-not-samples' type (prolly why i like it) so initial .wav and soundfont functions were minimal, but i believe advanced .wav use is scheduled in se..
Pages: 1 2 3
There are 29 total messages for this topic
Reply to Thread
You need to register/login to use the forum.
Click here to Signup or Login !
[you'll be brought right back to this point after signing up]
Back to Forum
