return to tranceaddict TranceAddict Forums Archive > DJing / Production / Promotion > Production Studio

Pages: [1] 2 
Synth modeling
View this Thread in Original format
zodiac9
Every time I think I'm doing pretty good, I learn a new thing I'm doing "wrong". For instance, I let a semi pro DJ/producer listen to one of my tracks last night. He said the synths sounded "out of the box", thus making the song sound dated, like 90's Trance. He asked if I did any synth modeling. I told him I tweak the synths a bit, attack, release, sometimes more, but I don't divirge too much from the original sound. He said I need to shape the synths and "make them my own". This sounds like a daunting task to me. Sure, I could turn knobs all day, but it would start sounding like crap I'm sure. Why mess too much with a good sound?

I'm not even sure what is meant by synth modeling. I've heard the term, and the meaning is somewhat obvious I guess, but does it mean just turning knobs and making the synth sound unique? I think it's Reason, can't remember, that has a separate add-on just for synth modeling. It didn't look easy to use at all. In my opinion, serious Synth modeling is another skill all by itself, separate from composing and producing, and is not easily come by. You would have to know what you're doing, and even then it would be tedious and time consuming. There's a reason they pay professionals to come up with new synth sounds. What's the point in buying a $400 VSTi synth module with 200 presets, if you are capable of coming up with your own synth sounds? Why does Reason push the fact that you get access to 1000's of user sounds when you buy their product?

I guess you get the idea. Am I making too much of this? Thoughts please.
BOOsTER
I don't really know what you mean with synth modelling but there's only one way to develop your "own sound" experience and lots of practice (at least it's what I hear all the time)

one more thing you might want to read the stickies some of them cover things like EQing, compression etc...
that will help your sound to have that kind of final "glance"

that's all I can think of, but...I still think I didn't hear the term synth modelling...

my first thought when I read the topic was that you're posting something about physicall modeling of instruments, but I was quite off...strange, maybe tehre's something new to learn, if someone can enlighten us?
DJ Shibby
Do whatever it takes to improve your tracks. Practice and progress. Simple as that.

And to answer... yes, you can use presets. It's also good to make your own sounds sometimes. Do whatever fits.

Good luck.
Derivative
@ the OP: you sir have got the wrong attitude.

presets are always just demos or springboards for your own experimentation. this is why all the presets on an access virus b do not use all 3 mod matrix slots. its no way in hell an accurate reflection of what a synth is truly capable of. sure you can use them. you might get lucky and one will fit perfectly into your track and compliment it just right. just dont count on it happening very often. do i object to the use of presets in music? hell no. whatever ends in the best result. its just my opinion that the vast majority of presets that i have are useless and very cheesy. i hardly ever use them except to get ideas.

with regards to shaping sounds. you should really already be aware that sound from a synth is essentially sculpted and it is often easy to discern a vague shape and weight and colour to a sound. if you build sounds from scratch on a synth you will understand this more.

this is why synths come with multiple waveforms. saw waves consist of odd and even harmonic intervals + the fundamental. alot of this is raw material for harmonically complex sounds. string sounds for instance.

sine waves consist of only the fundamental frequency and are useful for the purposes of creating percussive sounds, low sub basses and so forth. each waveform has its own set of properties. mixed together they can create hybrid sounds. you can add texture and movement to a sound by modulating certain aspects of the synth. for instance a string sound where the pitch gently wavers over time to simulate the vibrato action of a real violinist.

building a sound on a synth is full of subtlties that you only begin to notice by disecting presets. most presets are too incomplete. too generic. and too bland to use in any production. they are just basic configurations to get you started on physically giving weight and presence and shape to the sound. this you need to think about though.

the envelopes have a physical and obvious effect on the shape of a sound. with decay and sustain you can make a note physically longer. you can shortern it to make it staccato. you can add a load of release to make it ring out as if like an echo. you can make plinks and hummmmmmms and vvvvVVVVMMMMMPPPPHS. you get the picture right?

the oscillator waveform allow you to change the timbre of the sound - its physical texture. and with distortion you can add more harmonics. with LFOs you can add movement to a sound because any sound that has a cyclical and obvious pattern in its movement is dull and irritating. the things we often love about acoustic musical instruments is often the random movement in the sound, the imperfections of the player and sometimes even the subtle mistakes s/he makes.

presets are quite often sterile because they have none of this.
DigiNut
^ Strange how you speak generally about synths while referring specifically to subtractive synths... there ARE other kinds.

I agree with your principle though - the factory presets with most synths are generally garbage. However, if you hunt around you can often find some great 3rd-party patches that make full use of the synthesizers.
IDarkISwordI
Hey. I find it best, that if you have absolutely no idea what a particular knob does or hell, even not knowing at all how to program a synth, to just with things. Eventually, youll get something to come out. Oddly enough, this is often times how you can come up with some really cool original sounds. Just remember this, it may sound like dry, but you can always add tons of FX too it later and when you sit it in the track, it may turn out to be the next killer 'i've to make this' sound :).

Cheers,
ZAc
zodiac9
Thanks for the replys so far. It's good to see a lot of different opinions and thoughts. I'm glad to know you all are not total snobs when it comes to presets. I've encountered that before in other forums. I happen to think there's a time and place for presets.

Nice, informative post Derivative. Worth a copy and paste into my notes, actually. I'm already doing some of what you're talking about. I'm generaly never happy with presets. I usually tweak them a little even if just to make them fit into the song. Often time I go further even, messing with LFOs, changing out wave shapes, altering filters (cutoff freq, resonance, ect). I EQ too, and make use of effects and automation. Maybe I'm on the right track then. Definately will take more learning and doing, but it doesn't seem so daunting now. I think I'm starting to get the idea. The term "synth modeling" just sounded overwhelming :crazy:
zodiac9
quote:
Originally posted by IDarkISwordI
Hey. I find it best, that if you have absolutely no idea what a particular knob does or hell, even not knowing at all how to program a synth, to just with things. Eventually, youll get something to come out. Oddly enough, this is often times how you can come up with some really cool original sounds.


I've done a bit of that. Yep, you can stumble upon some interesting sounds that way, and no doubt they will be original. Sometimes I use the "random" feature on VSTi's, if you click random enough times you might get something usefull. It's like potluck, or a grab bag.
qiushiming
quote:
Originally posted by IDarkISwordI
Hey. I find it best, that if you have absolutely no idea what a particular knob does or hell, even not knowing at all how to program a synth, to just with things. Eventually, youll get something to come out. Oddly enough, this is often times how you can come up with some really cool original sounds. Just remember this, it may sound like dry, but you can always add tons of FX too it later and when you sit it in the track, it may turn out to be the next killer 'i've to make this' sound :).

Cheers,
ZAc


funny as hell....but oh so true :)
armanivespucci
As I understand it, the supersaw arose as a bit of an accident; just some tweaked preset on the Juno 2.

Derivative
quote:
^ Strange how you speak generally about synths while referring specifically to subtractive synths... there ARE other kinds.


yea you are right. i just sometimes forget because the only synths i have are subtractive analogue modelled ones :\ hence why i was interested in rob papen's blue. i also didnt mention FM synths and sample playback synths cuz i havent got a fecking clue how to program one :(

quote:
I'm generaly never happy with presets. I usually tweak them a little even if just to make them fit into the song. Often time I go further even, messing with LFOs, changing out wave shapes, altering filters (cutoff freq, resonance, ect). I EQ too, and make use of effects and automation. Maybe I'm on the right track then. Definately will take more learning and doing, but it doesn't seem so daunting now. I think I'm starting to get the idea. The term "synth modeling" just sounded overwhelming


thats good practice. the best place to start is to dissect a preset that you really like the sound of and work out how its built from scratch. i would even try to recreate it several times so that you get used to hearing a sound morph from a single sine wave into a more complex sound. once you can build it up to a preset kind of sound, its worth tweaking every variable on the synth to work out what it does. just take it slow. for instance, every day, just take out 15 minutes to load up a patch and tweak one or two variables. change it to extreme values to see what you can do with it. filter cutoff and resonance are the easiest ones because they have the most prominant audible effect and have the most dramatic sound shaping ability.

i learned how to build sounds by trying to copy some of the vanguard presets from memory. i also recommend vanguard as a starting place to learn subtractive synthesis because the control surface is very well laid out and theres very little in the way of modulation to worry about. keeps things simple.

doing this will really help your sound designing. once you know pretty much what every parameter on a synth does, it just becomes a balancing act - weighing up a number of variables to get the sound you are after. sometimes (especially on my virus b mixing up digital wavetables) you can get really unpredictable sounds. i built an analogue string sound today on my virus b from a sine wave and ill post a little demo of the results tommorow. i only stumbled across this by accident whilst cycling through the wavetables. noticed part way through that when pitched down it sounded like a cello. so modulated the pitch. modulated pulse width for both oscillators, dropped the cutoff right down and resonance up a bit, chorused it and added a hall reverb and it was pretty much there.

i also noticed, cycling through the presets that some of the virus wavetables can be used as formants - for the purposes of making choir sounds. that said, i have to experiment more with filter cutoff frequency and resonance values in order to shape vowel sounds since they are very specific. so far all my experiments sound pretty , but i think you can get pretty close to a convincing choir sound if you run a multimode filter on the virus. maybe someone else has better luck making choirs on a virus...

but yea, run through some digital wavetables if your synth has some. once you know what every variable does you just start to notice many things about oscillator wavetables and what they can be shaped into - these days i often suddenly jsut go 'wait a second...that sounds a little bit like a timpani.' so you just start moving the envelope settings to make it into a percussive kind of stabbing/hitting sound. round off the decay with a little work in the mod matrix and shave off the deadweight high and low frequency you dont need on the filter. then come back to it the next day with some timpani samples and fine tune it by modelling it on the sound of the real thing. synths can get really interesting like that.

synth modelling is one of those things that seems really difficult and abstract at first. but its actually quite easy when you know what every variable on your synth does and how it behaves when you tweak it. i agree with howard scarr when he says that the secret is to be obsessive about it. keep doing A/B comparisons with sounds you want to emulate. whats different about them? why does a real cello sound vibrato and yours doesnt? why does your cello emulation sound grittier than the real thing? listen really closely and try to tweak your sound so it gets closer and closer to the sound you want to copy. in the case of acoustic instruments you will never totally get there on a subtractive analogue synth, but the exercise will reward you with a better understanding of your own synthesizer. as such a highly recommend having a go.

and when you get better, your sounds not only sound better than presets but its actually a whole load of fun building them.
DigiNut
quote:
Originally posted by Derivative
yea you are right. i just sometimes forget because the only synths i have are subtractive analogue modelled ones :\ hence why i was interested in rob papen's blue. i also didnt mention FM synths and sample playback synths cuz i havent got a fecking clue how to program one :(

LOL... well, FM synths can produce a much wider array of sounds, but they're also a lot more difficult to program and a lot easier to screw up a preset on.

For people who find subtractive synths limiting or difficult to get "unique" sounds out of, and have the time and patience to fart around with an FM synth, it's much easier to be "creative" since adding or modifying a single operator can instantly change the entire sound, rather than just add colouration or additional depth.

There are other types of synths aside from subtractive and FM too; for example there are morphing synths like Cameleon which are loosely based on sample playback but really add a whole other dimension to it. There's also additive, phase distortion/waveshaping, physical modelling (for "real" instruments) and recently some guys have been working on a brand new type of synthesis that for some reason I can't remember the name of (no I'm not talking about granular synthesis/resynthesis).

Of course not all types of synthesis are very practical, nor am I trying to rub it in (:p), the point I'm trying to get across is that if you're having trouble getting the sounds you want out of the synths you have, then maybe you should try a few different *kinds* of synth instead of several very similar knock-offs of the same design. ;)
CLICK TO RETURN TO TOP OF PAGE
Pages: [1] 2 
Privacy Statement