Post by jiffy on Sept 27, 2009 9:37:54 GMT
Page 4:- www.promusicproducts.com/forum/phpBB2/viewtopic.php?f=34&t=4970&start=45
Undocumented "feature": sustained voice when switching programs
IF you play a sustained sound in Program mode, then switch to another patch, the former voice will keep sounding (most of the time). It's most audible glitch is that the FX of the next patch kick in abruptly.
This can actually be used live to good effect -- play a sound that makes an interesting backdrop, hold it, then switch to the next patch.... it will inherit this next patch's FX, and you can then use it as a drone to play the next patch over.
Cheap way to get an instant 2-voice mix happening.
--------------------------------------------------------------------------
Eureka! Fatten up that analog sound a la Roland Juno 106
Trying to fatten up virtual analog sounds by using a sub-oscillator seemed not-so-phat...
Problem: using a Sine wave as the sub-oscillator an octave below
Solution: use a Square wave.... the Juno 106 did
--------------------------------------------------------------------------
Crossfading 4 OSC's in the sample engine
(from another thread)
Here's a way to do what you want, but with smoother curves rather than hard "trigger points".
Basically, you will have each of the components of your layered guitar sound have its own velocity response range. I'm assuming you're doing something like 1) mellow body of sound, 2) brighter sound, 3) really strong attack sound, and 4) bend.
PROGRAM/Edit/MOD/Add Mod x 5
Source: Velocity
Destination: Osc Volume
Index: 1 to control OSC 1's velocity volume, etc
Type: Additive
Amount: varies for each OSC sound element
mod 1 (makes OSC quieter the harder you play)
OSC 1
high initial volume
Amount: -10% (adjust using your ears)
mod 2 (makes OSC somewhat louder the harder you play)
OSC 2
medium soft initial volume
Amount: +10%
mod 3 (makes OSC even louder)
OSC 3
soft initial volume
Amount: +50%
mod 4 (OSC doesn't get louder until you hit really hard)
OSC 4
very quiet initial volume
Destination: OSC 4 volume
Amount: + 90%
Curve: +90% (under Shape page)
mod 5 (also, when hitting really hard, OSC bends pitch)
OSC 4
very quiet initial volume
Destination: OSC 4 frequency
Amount: + 2 semitones
Curve: +90% (under Shape page)
Or, if you want the whole guitar sound to bend, not just one OSC, then set the last mod example to Destination: PITCH, which bends the whole sound rather than a single OSC.
You could also make a set of Mods just like the above, but use the Filt Cutoff for each OSC to independently brighten up as you hit harder. Note that you'll want to first go into the Synth/Filter/EDIT Filter: Osc 1 page and turn on each OSC's dedicated filter, and set its cutoff freq low enough so that you can hear it get brighter as you hit harder (if the filter is already max brightness, adding more brightness does nothing).
You have up to 32 MOD routes, which is plenty, so don't (ahem) fret on making too many Mods...
--------------------------------------------------------------------------
Vintage Filters
the 4-pole ladder
The early '70s synth era began in the late '60s with the EMS "Putney," soon after Neil Armstrong planted the U.S. flag on the moon. Along with the much later TB303, it was one of the few synths that had a 3-pole filter. Everyone else pretty much used 4-pole filters that were copycats of the famous Moog 4-pole ladder filter, sometimes even resulting in court cases where the offending party had to revoke their copycat filter due to copyright infringement on Moog standards. These resonant, self-oscillating analog filters were often the signature sounds of these vintage years, and helped define what we would 3 decades later call 'fat' or 'thick', 'warm' or 'rich'...
Besides ARP, Roland, and several now defunct knock-offs of the ladder filter, there were also a couple brave new breeds of 2-pole filters championed by Oberheim's SEM series and in the Yamaha GX1 synth/organs.
Notice the abundance of patch cords from early in this era, followed by a trend towards miniaturization and minimal presets. The Fusion's various synth engines, Mod Matrix, Mix capability, and considerate selection of filters and types can together functionally reproduce, and often supercede, most of these classic sounds (except, unfortunately, for many self-oscillating resonant filter sounds). For now, an overview of the various stages of filter design and timbral range offered by these vintage keyboards provides a very useful reference when trying to mimic or recreate the 'phat' environments these synths pioneered.
SSM, SEM, CEM, Curtis
-- THE EARLY SEVENTIES --
The Big Players: EMS, Moog, ARP, Emu, Oberheim, Roland, Yamaha, Korg
EMS VSC3 aka "Putney (3pole multimode)
unusual 3pole (18dB/oct) filter, like the later Roland TB303
www.sequencer.de/moogulatorium/ems_synthi.html
www.youtube.com/watch?v=8lZGC1kV37w
www.youtube.com/watch?v=8lZGC1kV ... re=related
www.youtube.com/watch?v=K5RNMsDHE4Q
Minimoog Modulars|Voyager (4pole LP,HP)
"The best filter in the business, a discrete transistor ladder."
www.alphaentek.com/moog.htm
synthesizers.com basic demonstration of a 2/4pole ladder filter
www.youtube.com/watch?v=N57fvSOnbL8
Roland SH1000 (Roland's first, mono, VCF)
www.youtube.com/watch?v=TvwFQf8D44U
E-mu modulars (4pole LP, HP, BP, notch, resonant formant)
www.vintagesynth.com/emu/emu_modular.shtml
www.youtube.com/watch?v=Uo5B1WDRhZY
www.youtube.com/watch?v=5t8Co8lCb6c
ARP 2600|Odyssey (mono, 3 osc, 4pole LP)
www.youtube.com/watch?v=L5X8jBBUwl8&feature=user
note: the 4012 was the filter that legally infringed on the Moog ladder
www.youtube.com/watch?v=ynhjrRn8qUI&feature=user
note on video2: interesting how using OSC2 as an (audio!) mod source to the filter adds an almost "sync" sound, adding grit
Yamaha SY1 (mono, 1 osc, LP, HP)
www.youtube.com/watch?v=l2hgBvYcwwE
Yamaha GX1 (poly, 2pole LP)
www.youtube.com/watch?v=ec8iNspYEKg
www.youtube.com/watch?v=Q1dTO8cAFdQ
Oberheim SEM (mono, tinny 2pole LP)
www.youtube.com/watch?v=mZRPuWQ7F9w
www.youtube.com/watch?v=UszpB5mXu2c
Korg miniKorg 700 (mono, 2 osc, LP, HP, 3 ring mods)
www.youtube.com/watch?v=YJhIjcfPHzQ
www.youtube.com/watch?v=zANtPgP8 ... re=related
--------------------------------------------------------------------------
The Waldord Microwave series
www.soundonsound.com/search?sect ... microwave+
Interesting read on wavetable synth basics. The Fusion might be able to create 'static' simulations of the Waldorf sound by importing some wavetables. I'm currently searching forums for Waldorf wavetable downloads.
Seems a Fusion imitation might also be accomplished by sampling Fusion or other waveforms, and placing them side-by-side in a custom-built sample. An ambitious and laborious project at first, splicing together pairs of waveforms into a mini-table of sorts. But then, as you create more of these pairs, you could then gang together pairs, creating 4-waveform tables. Then 8, then 16, then 32, then 64 like the Waldorfs.
Better though, would be the ability to manipulate a "pointer" of some type that tells the Fusion where and which portion of the wavetable sample to play, and loop if possible. Somehow, someway, I have this sneaking suspicion that the Fusion's "table" parameter in the Mod Matrix might accomplish something like this. With up to 16 points and a range of each point from -100 to +100, and the tables able to modulate LFOs, Envelopes, and any other Mod parameter destination, one of which happens to include Sample Start Point, it just might be possible to mimic a sort of wavetable lookup feature like the Waldorfs, in principle.
Things that make you go, "hmm."
GuyDenruyter:- This is a good idea, but I think you can only set the sample starting point at max 10% of the beginning (workaround: add 90% of crap at the end of your sample). The looping points, I think, cannot be the destination in the Mod matrix
Psionic:- Yes, you would need to "frontload" the sample as much as possible.
But this is actually a good thing, isn't it, since most of what we perceive as sonic interest, what gives a sound or timbre its unique signature, mostly happens at the attack portion of the sound...
A "bowed" trumpet sounds very similar to a bowed sax or even a bowed piano, whereas the same 3 instruments with their normal very short attacks sound distinctly different. This is why whole families of "pad" sounds sound related, when you could very easily "un-relate" them or "de-relate" them simply by giving them a variety of shorter attacks.
With wavetable simulation, by concentrating most of the activity at the front attack portion you can then create more 'percussive' types of wavetable sounds for use with leads, basses, and keyboard chording or comping.
Samples that are 2000ms in length will give you 200ms of front-end choice of sample start time...hmm... that gives only 20ms chunks to work with, at a max of 16 points in the table, giving you 320 ms wavetable sequencing at the front of the sound. Sounds more like mini-wavetable sequencing...
Maybe better to just use longer samples that already have multiple waveforms stacked together. I think the Waldorf forums have downloadable sounds like this, just a matter of converting them to pure .wav files. Then at least random OSC start points and other Fusion mod matrix magicks can emphasize or "inFuse" these Waldorf waves to give them an own unique twist on wavetable sequencing....?
GuyDenruyter:-
psionic wrote:
Maybe better to just use longer samples that already have multiple waveforms stacked together. I think the Waldorf forums have downloadable sounds like this, just a matter of converting them to pure .wav files. Then at least random OSC start points and other Fusion mod matrix magicks can emphasize or "inFuse" these Waldorf waves to give them an own unique twist on wavetable sequencing....?
Yes, that's what I meant by adding some 'rubbish' at the end of the sample that is not really used, but at least it gives the possibility to use various chunks. With Tables (and some precise editing) you could make the starting point exactly at a specific chunk. The envelope should then cut off this osc as soon as it reaches the end of the chunk (this means, all chuncs have to be equal in length).
E.g. total sample length: 10sec, but we can only use the first 10%, i.e. 1 sec. This is then divided into e.g. 5 chunks, each 200ms in length.
With the mod matrix and tables, you could make the sample start exactly at the start of chunk1, chunk2,... chunk5. With the envelope, you ensure that the sample is only played during 200ms.
A second osc ensures the sustain part.
This way, you could alter the attack of the sound completely depending on e.g. velocity, a knob, random, ...
A second oscillator
Undocumented "feature": sustained voice when switching programs
IF you play a sustained sound in Program mode, then switch to another patch, the former voice will keep sounding (most of the time). It's most audible glitch is that the FX of the next patch kick in abruptly.
This can actually be used live to good effect -- play a sound that makes an interesting backdrop, hold it, then switch to the next patch.... it will inherit this next patch's FX, and you can then use it as a drone to play the next patch over.
Cheap way to get an instant 2-voice mix happening.
--------------------------------------------------------------------------
Eureka! Fatten up that analog sound a la Roland Juno 106
Trying to fatten up virtual analog sounds by using a sub-oscillator seemed not-so-phat...
Problem: using a Sine wave as the sub-oscillator an octave below
Solution: use a Square wave.... the Juno 106 did
--------------------------------------------------------------------------
Crossfading 4 OSC's in the sample engine
(from another thread)
Here's a way to do what you want, but with smoother curves rather than hard "trigger points".
Basically, you will have each of the components of your layered guitar sound have its own velocity response range. I'm assuming you're doing something like 1) mellow body of sound, 2) brighter sound, 3) really strong attack sound, and 4) bend.
PROGRAM/Edit/MOD/Add Mod x 5
Source: Velocity
Destination: Osc Volume
Index: 1 to control OSC 1's velocity volume, etc
Type: Additive
Amount: varies for each OSC sound element
mod 1 (makes OSC quieter the harder you play)
OSC 1
high initial volume
Amount: -10% (adjust using your ears)
mod 2 (makes OSC somewhat louder the harder you play)
OSC 2
medium soft initial volume
Amount: +10%
mod 3 (makes OSC even louder)
OSC 3
soft initial volume
Amount: +50%
mod 4 (OSC doesn't get louder until you hit really hard)
OSC 4
very quiet initial volume
Destination: OSC 4 volume
Amount: + 90%
Curve: +90% (under Shape page)
mod 5 (also, when hitting really hard, OSC bends pitch)
OSC 4
very quiet initial volume
Destination: OSC 4 frequency
Amount: + 2 semitones
Curve: +90% (under Shape page)
Or, if you want the whole guitar sound to bend, not just one OSC, then set the last mod example to Destination: PITCH, which bends the whole sound rather than a single OSC.
You could also make a set of Mods just like the above, but use the Filt Cutoff for each OSC to independently brighten up as you hit harder. Note that you'll want to first go into the Synth/Filter/EDIT Filter: Osc 1 page and turn on each OSC's dedicated filter, and set its cutoff freq low enough so that you can hear it get brighter as you hit harder (if the filter is already max brightness, adding more brightness does nothing).
You have up to 32 MOD routes, which is plenty, so don't (ahem) fret on making too many Mods...
--------------------------------------------------------------------------
Vintage Filters
the 4-pole ladder
The early '70s synth era began in the late '60s with the EMS "Putney," soon after Neil Armstrong planted the U.S. flag on the moon. Along with the much later TB303, it was one of the few synths that had a 3-pole filter. Everyone else pretty much used 4-pole filters that were copycats of the famous Moog 4-pole ladder filter, sometimes even resulting in court cases where the offending party had to revoke their copycat filter due to copyright infringement on Moog standards. These resonant, self-oscillating analog filters were often the signature sounds of these vintage years, and helped define what we would 3 decades later call 'fat' or 'thick', 'warm' or 'rich'...
Besides ARP, Roland, and several now defunct knock-offs of the ladder filter, there were also a couple brave new breeds of 2-pole filters championed by Oberheim's SEM series and in the Yamaha GX1 synth/organs.
Notice the abundance of patch cords from early in this era, followed by a trend towards miniaturization and minimal presets. The Fusion's various synth engines, Mod Matrix, Mix capability, and considerate selection of filters and types can together functionally reproduce, and often supercede, most of these classic sounds (except, unfortunately, for many self-oscillating resonant filter sounds). For now, an overview of the various stages of filter design and timbral range offered by these vintage keyboards provides a very useful reference when trying to mimic or recreate the 'phat' environments these synths pioneered.
SSM, SEM, CEM, Curtis
-- THE EARLY SEVENTIES --
The Big Players: EMS, Moog, ARP, Emu, Oberheim, Roland, Yamaha, Korg
EMS VSC3 aka "Putney (3pole multimode)
unusual 3pole (18dB/oct) filter, like the later Roland TB303
www.sequencer.de/moogulatorium/ems_synthi.html
www.youtube.com/watch?v=8lZGC1kV37w
www.youtube.com/watch?v=8lZGC1kV ... re=related
www.youtube.com/watch?v=K5RNMsDHE4Q
Minimoog Modulars|Voyager (4pole LP,HP)
"The best filter in the business, a discrete transistor ladder."
www.alphaentek.com/moog.htm
synthesizers.com basic demonstration of a 2/4pole ladder filter
www.youtube.com/watch?v=N57fvSOnbL8
Roland SH1000 (Roland's first, mono, VCF)
www.youtube.com/watch?v=TvwFQf8D44U
E-mu modulars (4pole LP, HP, BP, notch, resonant formant)
www.vintagesynth.com/emu/emu_modular.shtml
www.youtube.com/watch?v=Uo5B1WDRhZY
www.youtube.com/watch?v=5t8Co8lCb6c
ARP 2600|Odyssey (mono, 3 osc, 4pole LP)
www.youtube.com/watch?v=L5X8jBBUwl8&feature=user
note: the 4012 was the filter that legally infringed on the Moog ladder
www.youtube.com/watch?v=ynhjrRn8qUI&feature=user
note on video2: interesting how using OSC2 as an (audio!) mod source to the filter adds an almost "sync" sound, adding grit
Yamaha SY1 (mono, 1 osc, LP, HP)
www.youtube.com/watch?v=l2hgBvYcwwE
Yamaha GX1 (poly, 2pole LP)
www.youtube.com/watch?v=ec8iNspYEKg
www.youtube.com/watch?v=Q1dTO8cAFdQ
Oberheim SEM (mono, tinny 2pole LP)
www.youtube.com/watch?v=mZRPuWQ7F9w
www.youtube.com/watch?v=UszpB5mXu2c
Korg miniKorg 700 (mono, 2 osc, LP, HP, 3 ring mods)
www.youtube.com/watch?v=YJhIjcfPHzQ
www.youtube.com/watch?v=zANtPgP8 ... re=related
--------------------------------------------------------------------------
The Waldord Microwave series
www.soundonsound.com/search?sect ... microwave+
Interesting read on wavetable synth basics. The Fusion might be able to create 'static' simulations of the Waldorf sound by importing some wavetables. I'm currently searching forums for Waldorf wavetable downloads.
Seems a Fusion imitation might also be accomplished by sampling Fusion or other waveforms, and placing them side-by-side in a custom-built sample. An ambitious and laborious project at first, splicing together pairs of waveforms into a mini-table of sorts. But then, as you create more of these pairs, you could then gang together pairs, creating 4-waveform tables. Then 8, then 16, then 32, then 64 like the Waldorfs.
Better though, would be the ability to manipulate a "pointer" of some type that tells the Fusion where and which portion of the wavetable sample to play, and loop if possible. Somehow, someway, I have this sneaking suspicion that the Fusion's "table" parameter in the Mod Matrix might accomplish something like this. With up to 16 points and a range of each point from -100 to +100, and the tables able to modulate LFOs, Envelopes, and any other Mod parameter destination, one of which happens to include Sample Start Point, it just might be possible to mimic a sort of wavetable lookup feature like the Waldorfs, in principle.
Things that make you go, "hmm."
GuyDenruyter:- This is a good idea, but I think you can only set the sample starting point at max 10% of the beginning (workaround: add 90% of crap at the end of your sample). The looping points, I think, cannot be the destination in the Mod matrix
Psionic:- Yes, you would need to "frontload" the sample as much as possible.
But this is actually a good thing, isn't it, since most of what we perceive as sonic interest, what gives a sound or timbre its unique signature, mostly happens at the attack portion of the sound...
A "bowed" trumpet sounds very similar to a bowed sax or even a bowed piano, whereas the same 3 instruments with their normal very short attacks sound distinctly different. This is why whole families of "pad" sounds sound related, when you could very easily "un-relate" them or "de-relate" them simply by giving them a variety of shorter attacks.
With wavetable simulation, by concentrating most of the activity at the front attack portion you can then create more 'percussive' types of wavetable sounds for use with leads, basses, and keyboard chording or comping.
Samples that are 2000ms in length will give you 200ms of front-end choice of sample start time...hmm... that gives only 20ms chunks to work with, at a max of 16 points in the table, giving you 320 ms wavetable sequencing at the front of the sound. Sounds more like mini-wavetable sequencing...
Maybe better to just use longer samples that already have multiple waveforms stacked together. I think the Waldorf forums have downloadable sounds like this, just a matter of converting them to pure .wav files. Then at least random OSC start points and other Fusion mod matrix magicks can emphasize or "inFuse" these Waldorf waves to give them an own unique twist on wavetable sequencing....?
GuyDenruyter:-
psionic wrote:
Maybe better to just use longer samples that already have multiple waveforms stacked together. I think the Waldorf forums have downloadable sounds like this, just a matter of converting them to pure .wav files. Then at least random OSC start points and other Fusion mod matrix magicks can emphasize or "inFuse" these Waldorf waves to give them an own unique twist on wavetable sequencing....?
Yes, that's what I meant by adding some 'rubbish' at the end of the sample that is not really used, but at least it gives the possibility to use various chunks. With Tables (and some precise editing) you could make the starting point exactly at a specific chunk. The envelope should then cut off this osc as soon as it reaches the end of the chunk (this means, all chuncs have to be equal in length).
E.g. total sample length: 10sec, but we can only use the first 10%, i.e. 1 sec. This is then divided into e.g. 5 chunks, each 200ms in length.
With the mod matrix and tables, you could make the sample start exactly at the start of chunk1, chunk2,... chunk5. With the envelope, you ensure that the sample is only played during 200ms.
A second osc ensures the sustain part.
This way, you could alter the attack of the sound completely depending on e.g. velocity, a knob, random, ...
A second oscillator