|
|
Post by psionic11 on Mar 14, 2008 5:14:13 GMT
I'm going to be presumptuous here as a newbie, because I'm having a blast learning the Fusion's synth logic and the possibilities are only just beginning...
SMOOTHING!!!
No one seems to mention this. Granted, it's not one of those parameters you tweak right away and get some kind of immediate feedback on what it does, but it is useful, trust me.
It's found under SHAPE.
One of the better ways to guage what exactly it is doing is to assign it to some wacky but obvious parameter, like Pitch or OSC frequency. Use a Knob or Aftertouch as a source.
One way to think of Smoothing is by relating it to how a traditional LFO affects, say, vibrato. With no Smoothing, a straight-up LFO assigned to pitch gives a max wobble effect to the sound when the Mod Wheel is at 100%. But if you introduce about 95% smoothing to that LFO, it will have only a minimal effect on the pitch. So, smoothing "filters" or reduces another modulation effect.
For Aftertouch, which without smoothing tends to react like an on/off switch, introducing smoothing between 45-75% gives you a much greater area of pressure-response, allowing more nuance of expression rather than just CLICK ==> AFTERTOUCH = ON.
For Sample & Hold or a Random LFO wave source, Smoothing "smooths" the response so that the response is not so drastic.
Bottomline, Smoothing reduces the impact of a certain modulation. When used in combination with +100% or -100% Curve (again, test effects by using Pitch as a destination), you can fine-tune response so that it responds less mechanically and more musically.
|
|
|
|
Post by psionic11 on Mar 14, 2008 5:51:24 GMT
Tip #2: Multiple MOD routes (one to many):You have 32 possible MOD routes.... use them, don't be shy ;D From what I've learned on previous hardware synths, they all pretty much had the same concept of 1) a volume envelope, 2) a filter envelope, 3) a velocity ==> ENV effect, 4) an ENV AMOUNT envelope (how strongly the other envelopes affect other envelopes  , 5) a keytrack "envelope" This is great and natural for most sounds. Hit a key harder in the higher range, and boom, it gets louder and brighter and more sensitive. Well, in the Fusion, it might take several routings to achieve the effect that is more direct with these older synths. It's not enough to just assign velocity to filter brightness. Env 1 ==> Volume Env 2 ==> MainFiltCut Velocity ==> Mod1Amt Velocity ==> Mod2Amt Keytrack ==> MainFiltCut My point is, for expressive realism, it will most likely take multiple assignments of a single source parameter to several destination parameters in order to simulate what we hear with natural acoustic instruments. You hit a key harder, it gets brighter and louder. You press Aftertouch harder, vibrato gets deeper and faster. You push a Mod Wheel, rotary gets faster and distortion gets deeper. So, assign one controller to several related performance effects, and realistic expression is increased... By the way, a major assumption here is that you're trying to squeeze as much expression out of a patch as you can. For this, velocity and aftertouch are a minimum. Personal articulation is muy importante (staccato, legato, marcato, crescendo, sforzando, etc.). But as synths are limited in variability as regards to musical expressiveness compared to actual acoustic instruments, it's not out of line to increase sonic variation via foot controllers, knobs, and triggers as much as possible. Roland's V-synth is a glorified sampler, but with detailed attention to articulation nuances that can be triggered via MIDI-recordable performance parameters. I firmly believe that the Fusion, properly programmed, can approximate nearly any other synth out there. This also means that the Fusion has a greater expressive possibility than most others. The depth is there, but it needs to be brought to the surface so that more of us can express it. |
|
|
|
Post by psionic11 on Mar 14, 2008 7:18:49 GMT
Tip #3: Liberate yourself from mediocre INSERT FX:It's an unspoken assumption that the Fusion's Effects section is not too great.  Well, there are several classes of Effects. Some were overused in the 80's when, "Hey, we've got all these cool new effects, let's use them everywhere without reservation!! (enter delays that were not sync'ed to tempo, awful detuned choruses on high-pitched sustained notes, cliche gated reverbs, overabundant cathedral/stadium halls, super-compressed tracks... oh wait, we still have over-compression to this day... ah well...) Anyway, beyond my personal grudges of yesterday's productions, there are some real limitations to what the Fusion can do in the FX department, esp regarding the INSERT FX. I'm going to continue being presumptuous here and say that hardly any of the internal effects are so good that they absolutely and indispensibly NEED to be there to make up the quality of the user's resulting creation.  (Note that this is the opposite case of "teh Big 3" -- being the established big dogs that they are, they have Effects algorithms that are studio-worthy, and judicious application of these killer algorithms to just-decent ROM samples makes for some tempting patches on the Music Store Floor. On the other hand, if you can make a new sound that is killer without FX, it will have a longer shelf-life than some sound that is dependent on ear-candy. Classic or vintage sounds live on irregardless of Delay or Reverb or Chorus). Anyway, on to the tip. I mentioned different classes of Effect. We have 1) spatial or volume-type effects that don't affect the timbre or quality of the original sound: tremolo, pan, rotary, velocity cross-fades, doubling, delays, compression, splicers... 2) detuning/pitch effects that "color" a sound: detune/chorus, flange/phaser, octave doubling (sub-oscillators), additive synthesis (like drawbars) 3) filter effects that affect timbre brightness: auto wah-wah, envelope filters, FiltCutoffFreq knobs, LFO==> filter cutoff, resonance fx, filter types themselves like bandpass or formants, envelope wahs 4) exciters/enhancers/modellers, that play with the harmonic spectra to achieve certain effects, like preamp models and cab simulators and tube simulators and classic FX simulators, as well as more radical effects like decimators, distortions/overdrives, lo-fi emulators, analog/digital/tube emulators, what have you... Damn, I can ramble. Let's try this again. On to the tip: Main point: don't use or rely on an INSERT or BUS Effect like lush reverb, excessive delay, a nifty chorus or flange, or an essential rotary or distortion to sell a new patch that you made. Spice is good. But spice is not meat. (Don't shoot me, I'm only offereing suggestions here, not ultimatums  ). Instead, simulate the important effects within the patch modulations. This way, the sound can travel intact to a mix, without losing its oomph because it's missing an INSERT FX. Also, by focusing sound creation on internal synth effects, better understanding and deeper performance can be had with new patches, rather than relying on an initial WOW factor that doesn't have practical use in a song. ******************************************** Many FX can be handled within the sound matrix routes themselves, without resorting to external effects. 1) Autopan and tremolo are easily reproduced with LFO assignments to pan or volume. Ditto for doubling, compression, and splicing. Delays cannot be sync'ed in the Fusion, so better to leave off that sweet temptation since it's not practical in a mix.* Be honest, it's just us Fusion fans... 2) Several detuning type FX can also be simulated inside the VA synth -- detune different multisamples by a few cents, which can achieve chorus/flange effects. Delay OSC start times for two identical multisamples within a patch for more subtle flanging (esp good when randomized or affected by external controls like velocity or keytrack). 3) No need for the various INSERT LP or HP filter FX... can be successfully aproximated with MOD routes. Also, do not underestimate assignment of FILTERS -- the proper filter can add an overall quality to the sound that can push it beyond mediocre into the realm of phat.... Bandpass for boxy or punchy acoustic instruments, RP filter for vintage or softer response sounds, Formant for uniqueness..... 4) The Effects that mess with the harmonic spectra are not easily duplicated in the matrix mods. However, the included modellers are not that great either. A real guitar player knows the difference between a fair amp model and an excellent one... a synth/sampler has a long way to go yet to reproduce the finesse that many of us love in actual rock/metal guitar delivery. Articulation is a form of musical performance maturity that cannot be simply "sampled" and expected to perform the same dynamic interpretation a well-trained or talented human performer can deliver. This is especially true for brass and distorted guitar tones, where variance of timbre brightness are essential to expression (imagine trying to re-create a variety of emotional vocal performances with just a few samples of one particular vocalist's vowels and syllables...) The Fusion's models aren't great at all, so they're not recommended for emulating guitar sounds. Just don't do it, unless all you're trying to impress are other keyboard guitarist-wannabes (which is a very small audience indeed). ******************************************** * One possiblity to work around the built-in delay that cannot be sync'ed is for someone to publish a chart which relates Delay in milliseconds to musical tempos of a quarter-note or triplet time, etc. The Sample & Hold can sync to tempo, but I'm not sure if that can help us in the Delay department. |
|
|
|
Post by djnorythm on Mar 14, 2008 10:42:16 GMT
|
|
|
|
Post by Hollow Sun on Mar 14, 2008 14:12:08 GMT
1) Autopan and tremolo are easily reproduced with LFO assignments to pan or volume True but you must ensure that the LFO RE-TRIGGER is set to NONE otherwise you have individual pan/tremolo for each note. |
|
|
|
Post by psionic11 on Mar 14, 2008 19:27:42 GMT
Thanks dj, will check the link out.
And sorry Steve, I knew it was one of you HS guys that hosted the file. You do have a point about setting the LFO re-trigger correctly, but at least it still can be done.
And besides, just try getting a dedicated Autopan FX unit to do individual pan/tremolo for every note! Precision control inside the matrix mod seems preferable to me, esp since you could set it to a smaller stereo field with really slow modulation, lending a dynamic ambience like when you sit in front of a real piano --minus the harmonics and phase shifting, of course, but still sort of "dynamic" coming from the sound source in the natural acoustic space, rather than "static" like thru speakers.
Heck, by using Keytrack, smoothing, and +/- Curves creatively, you could have low notes swirl wide and slowly while having higher notes swirl more densely closer to the center and at a more frantic rate, or vice versa... Or you could link pan depth to velocity as well, so that when you hit a note really hard, it shoots across the stereo field beyond its brethren. Hmm. Actually, I think I'll go and try that now... ;D
|
|
|
|
Post by Hollow Sun on Mar 14, 2008 23:02:54 GMT
And besides, just try getting a dedicated Autopan FX unit to do individual pan/tremolo for every note! Precision control inside the matrix mod seems preferable to me, esp since you could set it to a smaller stereo field with really slow modulation, lending a dynamic ambience like when you sit in front of a real piano --minus the harmonics and phase shifting, of course, but still sort of "dynamic" coming from the sound source in the natural acoustic space, rather than "static" like thru speakers. Heck, by using Keytrack, smoothing, and +/- Curves creatively, you could have low notes swirl wide and slowly while having higher notes swirl more densely closer to the center and at a more frantic rate, or vice versa... Or you could link pan depth to velocity as well, so that when you hit a note really hard, it shoots across the stereo field beyond its brethren. Hmm. Absolutely - the power of Fusion! Now then, children.... let's try that on this here Yalandorg  Steve |
|
pev69
Junior Member
Posts: 66
|
Post by pev69 on Mar 14, 2008 23:29:30 GMT
Heck, by using Keytrack, smoothing, and +/- Curves creatively, you could have low notes swirl wide and slowly while having higher notes swirl more densely closer to the center and at a more frantic rate, or vice versa... Or you could link pan depth to velocity as well, so that when you hit a note really hard, it shoots across the stereo field beyond its brethren. Hmm. Actually, I think I'll go and try that now... ;D Heh, very interesting ideas, at least on paper (I mean, on web page). Do tell us about your results if you really get to try that! And thank you for interesting tips and comprehensive writing  . For those interested in the technology I would like to half-correct about the flanger and phaser effects, that they do not whole-heartedly fall in the category of detuning or pitch-changing. Although they can be implemented by very slight detuning and mixing of two similar signals, they actually have more to do with filtering. They are both the result of a sweeping comb filter. A comb filter is a series of notch (very narrow band-reject) filters at different frequencies. In flanger the notches are reqularly, or harmonically, spaced and in phaser they are spaced irregularly. The "comb" comes from how it looks in a spectrum analyzer: like the teeth of a comb. I have actually implemented a phaser/flanger myself, in a university course about signal processors and sound processing. We programmed it with a DSP development card in a PC. It was quite mesmerizing to see the spectrum analyzer display of white noise put through the effect when we used only three notches sweeping through a range (I don't remember the exact details anymore, it was more than ten years ago). It was like a couple of camel's humps bouncing around, for a lack of better analogy . And it was also surprisingly good sounding, when we tested it with some simple sounds from a portable keyboard, it really added richness (at that time I had not really heard, or taken notice of, flanger/phaser in use anywhere else)! |
|
|
|
Post by Hollow Sun on Mar 15, 2008 3:20:58 GMT
Now then, children.... let's try that on this here Yalandorg I have had some pangs of guilt since posting that. Of course, I was being a bit tongue-in-cheek and glib (but I stand by the statement) but in all fairness, the offerings from 'the Big 3' (i.e. Yalandorg) are all good (nay - amazing) in their own way and horses for courses and all that. They offer stuff Fusion can't compete with but it cuts both ways and Fusion can do shit these Big 3 offerings can't come close too. Ever the naive idealist that I am, I would love to see a 'collaborative' rather than the 'competitive' relationship between the manufacturers where we can work together as colleagues rather than rivals and we can all collaborate for the greater good and share stuff. I would be happy for my friends at Korg, for example, to use my strings in exchange for their brass (whatever). But it ain't ever gonna happen and I totally understand why. Dream on! Steve |
|
|
|
Post by psionic11 on Mar 15, 2008 5:56:23 GMT
I've been a Roland fan before, having learned all about analog synthesis on my trusty Juno 106 with all those VCO sliders and PWM buttons where you could get immediate feedback on what does what. I also loved my BOSS SE-70 multiFX processor, and got lost happily with all the parameters you could tweak. Ran a BOSS drum-machine thru the gate for an unconventional effect. Since then, one of my acid tests on new FX units is to run white noise thru its Phase Shifter, which I discovered by accident, and it's interesting to hear that it's done at university too. With headphones on, changing the wavelength you're riding by switching dimensions from 4-pass to 8-pass to 12-pass... the quality was superb, and I've since loved phase shifting much more than flanging, very different sound quality. The BOSS was super-rich and detailed... ... unfortunately the Fusion's ain't quite as good.  Ah well, the point is to take her strengths and weave wonders with those. Overcompensate moderate FX by inventing new ones using matrix mods. By mixing well-chosen blends of FM, samples, and VA, etc... Anyways, I have some synth beckoning me to come play with her. And her name isn't Korlandaha... |
|
|
|
Post by Hollow Sun on Mar 15, 2008 13:46:32 GMT
Since then, one of my acid tests on new FX units is to run white noise thru its Phase Shifter, which I discovered by accident, and it's interesting to hear that it's done at university too. With headphones on, changing the wavelength you're riding by switching dimensions from 4-pass to 8-pass to 12-pass... the quality was superb, and I've since loved phase shifting much more than flanging, very different sound quality. The BOSS was super-rich and detailed... Roland/Boss have always had phasing and chorus absolutely nailed. Steve |
|
|
|
Post by Hollow Sun on Mar 15, 2008 16:09:42 GMT
SMOOTHING!!!No one seems to mention this. Granted, it's not one of those parameters you tweak right away and get some kind of immediate feedback on what it does, but it is useful, trust me. I'm a big fan of smoothing: SawtoothSineSquareTriangleShows Fusion's anti-aliased oscillators to good effect but the totally smooth, continuous sweep is achieved by putting one of the knobs through the 'smoother' - without it, these would rise and fall in quantised semitone steps. Steve |
|
flux302
Junior Member
current team beat battle champ, DE, First Reps!
Posts: 82
|
Post by flux302 on Mar 15, 2008 17:25:10 GMT
i run my fusion through a kaoss pad kp2 then into a roland vs 1680 (wich has some beautiful effects as well) only thing i wish is that so many of the patches didn't already have so many effects smothered on em! it's a chore going through and taking delay off of all those patches. bpm synced delay on the kaoss pad is so much fun with the fusion though.
|
|
, 
).
. For those interested in the technology I would like to half-correct about the flanger and phaser effects, that they do not whole-heartedly fall in the category of detuning or pitch-changing. Although they can be implemented by very slight detuning and mixing of two similar signals, they actually have more to do with filtering. They are both the result of a sweeping comb filter. A comb filter is a series of notch (very narrow band-reject) filters at different frequencies. In flanger the notches are reqularly, or harmonically, spaced and in phaser they are spaced irregularly. The "comb" comes from how it looks in a spectrum analyzer: like the teeth of a comb.
Ah well, the point is to take her strengths and weave wonders with those. Overcompensate moderate FX by inventing new ones using matrix mods. By mixing well-chosen blends of FM, samples, and VA, etc...