Your process for sound creation

[Liquid Shadow]

How do you do it (usually)? Start with an initialized patch and do all filtering to that and add layers as appropriate? Layer first? Effects first or last? Whatever your process is, share it with us.


For anything that isn't a piano/organ/some other acoustic emulation thing, I'll pick out the basic waveform first. I have a homemade init. patch of just a plain saw wave. Then I'll filter the first wave how I want it, and add another layer and do the same thing...go back and tweak the first if it needs it. Continue as necessary.

I've also started adding effects first, especially if I'm going to be using distortion at all. I used to wait until the very end to add them, but I've found that distortion can have such different reactions to the sound depending on the filter and/or EQ, a phaser can behave very differently with different resonance/filter settings, etc., and instead of turning on the effects blindly at the end, I like to have them on to begin with (excluding verbage/delay), so that all the settings are made according to how I want them to react with the effects.

[Analogkid]

I fire up the Moog, wait about 10min for it to warm up and then just twist knobs until something cool comes out. Then I snap a pic with my cell phone and store it for future ref.

[lighthouse]

I fire up the Moog, wait about 10min for it to warm up and then just twist knobs until something cool comes out. Then I snap a pic with my cell phone and store it for future ref.


LOL with that one!! hehe, but I usually do it like Liquid Shadow said, start with a basic wave and try to aproximate the sound I´m looking for by filtering and tweeking the enveolpes LFOs or whatever and adding new layers to thicken or to get more complex like with feedback effects....



Juan Pablo

[Omega Monkey]

Usually I just start with a pre-existing patch and take off somewhere with it. Occassionally I will try to pick a patch that is already close to what I want if I have something in mind, but usually I just start with whatever patch I am on at the time. Then I just play with a bunch of stuff (filter settings, wave selection, EGs, LFOs, etc...). If I get something that I REALLY like thats "new" (ie I dont already have that sound on that board) then I will look for a memory slot to overwrite to save it to.

Typically, I dont really have a "purpose" with programming (sometimes I do), so most of the time its more for practice/fun than anything else. And most of the time I dont end up saving things either because they are repeatable, I dont have a spot for them, or I just dont care or whatever. But I have come up with some pretty good results. I am especially proud of my "B-3" and "pipe organ" sounds on my JX-3p (ie completely subtractive from the early 80s with only 1 EG and 1 LFO and 2 oscillators). Add a couple effects (leslie and reverb for the B and slight chorus and reverb for the pipe organ) and you have sounds which are realistic enough for live use (or maybe even studio depending on application) out of a 20+ year old, sub $150 keyboard. Not too shabby. I think I lost the pipe organ patch though, because I think I had it "saved" on the programmer, rather than in a memory location, then i messed with shit before I could find a spot for it.

[Georges]

My sound creating process is quite similar:

- first I think about effects
- then I start to program from scratch usually by using the default program
- then I go over the effects again
- finally , I program real-time controller parameters

[jeremyroberts]

I personally take an existing patch that has the same kind of amp envelope I'll be using. I tweak the enveloppe if necessary, then I go through the basic waveforms until I find what I'm looking for.

After that, I use the usual filters and lfos, I add another layer, work on the filters again, etc.

With effects turned on and turned off, so I can hear the dry sound to hear what I do, and then hear what it sounds like with the effects, if the effects need to be tweaked, etc.

Realtime controls at last, too.

I'm doing that on my Roland XP80 ("rompler" workstation).

On my Alesis Ion (virtual analog), I take an existing patch and I play with all the knobs! Or I follow the path of the sound (oscillator, amp, filter, lfo, effects, etc.) and "program" the sound step by step.

I think this is a very common way to do the things. Interesting thread by the way! I've just spend two hours last night working on sounds, I hadn't done that for years! And I like the results so I'll do that more often :)
(I created only 2 patches during those 2 hours!)

[Kirby]

It depends, first, I need an idea and then I decide what synth to use. When working with my modular, I just patch and twist, often with some surprises.
With my Waldorf Pulse and K2000, I start with the oscillators, then work on filters, envelopes and modulations. The only effect I use is an external Line6 delay which I configure while playing, but on the whole, my patches are concepte dry.

[Luca_Capozzi]

It depends about which sound i'm searching for. Usually I start with a basic synth schematic and, then, tweaking with a good modulation matrix. Lately i'm working a lot with feedbacks, delay lines and various filter routings.

Luca

[ChrisMcCoy]

I tend to do a lot of midi layering between boards and module.
But if I can achieve what I'm looking for in a single board's stock patch, I'll stick to a simple patch and modify if neccesary.
With the JD-800 I did some modifying of stock brass patches to cover STYX and REO Speedwagon songs last summer.

C.

[Omega Monkey]

Yeah, the old Chick Corea style layer between different synths is a great way to really enrich your sounds. I have done that from time to time (like getting a nice lead patch on my Alpha Juno and JX-3P, then controlling it from the Alpha with the chord memory set either to root, root+fifth, or root+octave/s, so I can "trick" mono out of the JX), but I am too lazy to have really done much with it so far. But I plan on doing a lot of work integrating FM (or the FM/AWM of my TG-77) with "analog" and/or sample based sounds (and/or any combination I have at my disposal including layering with real Rhodes). Right now in my setup I have the following types of synthesis available to me:

hybrid subtractive (ie DCOs with VCFs/VCAs)
straight FM (4 op FB01 and 6 op TX7)
FM/AWM/RCM (ie my TG-77, RCM is using PCM waves in the FM operators for some really wild sounds)
AWM and/or sampling
sampling
FDSP (basically a very powerful "effects"/modeling section best suited towards EP and guitar emulations, very convincing too)
VA (2 voice in the EX5, which can be combined with AWM or FDSP)
VL (physical modeling)
Linear Arithmetic (ie PCM and "VA" sort of in the D-50)
Phase Distortion (in my Casio CZ-1000, sounds similar to FM, but it is set up more like analog and can get some great buzzy leads and basses)
whatever Ensoniq calls their rompler interface from the early 90s (which includes backwards waves which I havent seen yet on anything else really)

So I've got a fair amount of "power" at my disposal, maybe TOO much (like I have at least 4 synths that are basically romplers, maybe also including some other synthesis types, and 3 hybrid subtractive synths from the 80s). Unfortunately, I have been getting gear faster than I can really "master" it, so Im most comfortable with subtractive/va/rompler stuff and not nearly as much with the other types yet.

[Georges]

I tend to do a lot of midi layering between boards and module. But if I can achieve what I'm looking for in a single board's stock patch, I'll stick to a simple patch and modify if neccesary.

That's where the work with a K2xxx gets easier because you can have up to 32 layers, meaning that you can already set zones and layering within a single MIDI channel, which is great. Anyways, just a little side comment. I still regret having had to sell my JD-800 5 years ago ...

[ChrisMcCoy]

That sounds like a nice feature. I just don't have room for a Kurz 2600 in my budget...sure wish I did though ! :) I actually still have the Roland JX-3p in my rig, largely because it fattens things up nicely. Particularly the analog strings which make a nice soft pad for some of the other sounds in my rig...I used to use it a lot with my D-50. I regret selling my D-50, but I needed a piano so onto e-bay it went. I can do combining of 2 patches with my RS-9 (which replaced my D-50 for a live piano, but with the kind of music I play these days (mostly classic rock gigs) I rarely need to go too deep with layering. With one of the "for Hire" pop gigs I did back in the early 90's, it was a different story, and often I'd stack the D-50, Kawai K1, JX-3p and DSS-1 patches together.