Production Tip: Unison

Discussions about music gear: hardware/software, Tips/Tricks, Questions, How Tos, and advice ...
Mastering/Recording/Mixing...
Scratching
Post Reply
BeauDamion
Groupie
Groupie
Posts: 8
Joined: Jan 13 2010 08:34 am
Location: Florida
Contact:

Production Tip: Unison

Post by BeauDamion »

Unison is generally achieved by most VSTi's by adding more voices of the same timbre to create a louder or more spread out sound. For example.... by hitting one note, the Synth will actually produce 2 or more voices. Many VSTi's usually have this feature in their UI... and sometimes the control is great and all that, but sometimes you might want a little more control over that to make a sound thats more to your liking.

The problem i've always found with unison (especially in the low end) is that you more often than not get weird phase problems... this is just caused by a given number of voices accidently colliding into each other at some point which will phase a certain part of a sound out, resulting in some notes sounding thinner or way more present than others, even though youre mashing the same note repeatedly. I hate to say it, but youre always going to run into this problem. However one technique ive managed to live by is recording the same sequence over and over and over and over until one gets randomly spat out that i can live with after ive recorded it. Of course that worked out for me for a while.. but then i got thinking, theres gotta be a better way.

Well first we need to see where the problem even stems from... in this case, its stereo. You can totally hear the effects of a shitty unison if you monitor it in mono. Not to say we need to monitor everything in mono, but its good practice just to keep a mono insert on your master chain (just for reference from time to time). Of course, depending on the plugin or synth, Unison is usually gradually or radically spread out over the L/R channels as the number of unison voices increase. For example... a 4 voice unison bass might be panned accordingly; 1 voice panned 100% on the left and 1 voice 100% on the right... plus, 2 more voices panned 25% L and 25% R ... and so on. Of course, theres also a pitch factor involved... this value is normally called "unison detune" and all that does is *slightly* detune each of the unisons voices by x amount. The reason this sounds so "lush" is because you have detuned stuff on the left and right now... giving the sound a stereo image. In fact, this is alsmot exactly what a "chorus" effect does, only a bit different.

I've always been a bit of an analogue nut.... ergo, alot of the synths i noodle around on are Monophonic (only capable of playing one voice at a time with no unison) So... to get a thicker sound out of them... i follow the basic principles of how a polysynth would create a unison... and hell, dont be scared to try this on a VSTi either.

Record your riff / loop... one voice / one polyphony. Now by one voice, i could be talking about any given number of Oscillators used to make up that voice, but for simplicities sake, we'll use a pulse or whatever, doesn't matter. Now, re-record the same thing again on a new channel and try changing the pitch by 9 cents or something (or just record a few loops of the same thing and get your scissor on). Lets go for a 4 voice unison to keep it easy.... so rinse and repeat that 2 more times

now youll have 4 channels, all mono, panned center and hopefully playing the same recording / wav / whatever... so lets get in there with the imaging. take the first 2 channels and spread em out L and R say one channel hard left, the other hard right. i dont really care. whaytever sounds good to you. then do the same to the other channels... just try to keep it balanced, otherwise your sound will be panned in a weird spot. (you can pan it wierd if thats your thing, but the key is to spread it. hopefully my leet ascii diagram might help... where 1 2 3 4 are your recorded tracks on the Left Center Right scale thing im gunna invent just now.

L --1---2---C---3---4--R

so there you go, by now you should have a bit thicker sound. And now youre probably asking, well.. why dont i just turn the unison on and use 1 track instead? well... now you have alot more control over the sound really, and since its now Wav data... and not random waveform starttimes your VSTi decided to spit out to make it "more analog sounding" you can get in there and tweak the weird notes... and by tweak, i mean nudge bits of each of the 4 tracks forward and backward in time by a teeny tiny bit (which is usually all it takes to fix a dodgy note) .... and since youre on seperate channels... now is a good time as any to get all experimental with it and maybe run some effects on some of the channels... i.e. a slight reverb on voices 2 and 3.... or strip some low end out of 1 and 4.... your plugin probably didnt support that.

anyways. hope you found that bit useful, saved my ass many a time, and definatley gave me the satisfaction that i've actually worked on something to make it better then using some fuckin preset.
Post Reply