Tutorials or advice on layering synths

Question

so I have been into some music production for a few weeks, and I feel like I am improving gradually. The problem I am having currently is layering synths.

If I listen to a track, I can sort of make out the different layers getting introduced, but if I then do something like that it sounds horrible. It becomes a big mess of sounds.

I would really appreciate some tutorials, guides or advice on what kind of sounds mix well together. I know lot of it is practice, listening for some time and trying new things out, but I feel like some help would do me good.

I sometimes try to compose/structure the song without sounds, but I fail to find great synths that match each other.
I currently don't mind either about using presets, even though I am trying to get into designing my own sounds. Currently I would just like to have a track that sounds good, whether it is my own sounds or not.

What helped you find out what sounds good together?

EDIT: Using Ableton Live 32-bit, learning Massive VSTi, I got a USB MIDI keyboard, get a pair of non-studio headset and speakers (I plan on getting studio speakers when I get better), all running on an iMac medio 2011 with 12GB 1337MHz RAM
Not sure if I should include anything else.

Answer 1

Without going into a great, detailed oration, I'm going to suggest just two fundamental things that I think would help you in your pursuit.

In his comment, Basstickler touched upon a great point - music experience. If you understand music theory and basic principles of orchestration, you can apply those concepts to synthetic sounds and learn how to voice chords appropriately, balance textures, and develop thoughtful gestures (to name but a few!)

The other point I would like to touch on is principles of acoustics. To improve at what you're doing, you're going to want to have an understanding of how acoustics work - both in literal physical space as well as stereophonic aural space that you're dealing with. Understanding how to create aural perspective and depth is essential if you find that your sound is very cluttered or uneven. I'm not going to recommend specific materials as this website is not for that function, but if you do a quick search or find an audio tech forum / recording forum, they should be able to provide you with all the resources you might need.

Hope that helps!

Answer 2

Firstly, big appreciation for making your own sounds from scratch. It will lead into a personal style of music and a better understanding of the anatomy of sound.

The fact that layering makes the song "a big mess of sounds" tells me that you're not aware of mixing and equalizing the synths.

I'd see the layer of sounds as a scale: if you add a sound, in order to keep the balance you have to take something away from the mix. If you don't find the balance, try to turn the system upside down: rather than searching the perfect amount of volume for a sound beginning at 100% and reducing, try beginning at 20% and increasing the volume. Less is more.

You should note a few things while layering sounds.

• Don't let the peaks go above 0.0 dB - it causes unwanted distortion.

• Pick different timbres for the layer to make it sound 'rich'. This will reduce the messy outcome. Imagine a melody played with violin, flute and piano or three alike acoustic guitars. Which will be more interesting, which will be the easiest to equalize, mix and master?

  For fat saw synths, you could have both sharp and more detuned ones making the layer really big and wide. The sound will be deep and dark if some synths are transposed one octave down - transpose up will make it brighter. Also, try different waveforms. For kickdrums, you can have a punchy one for the thump and a subby one for the oomph.

• Equalize the sounds individually before sending them to the layer channel. If you boost a significant amount of some frequencies of a certain sound, you have to reduce those frequencies of the other sounds. Otherwise the layer will be "a big mess of sounds" as you described. Find the best part of each sound and preserve it while filtering off unwanted or unnecessary frequencies.

• While equalizing don't boost too much. I'd say for a kick drum +12.0dB at 80Hz will eat a lot of the energy in the mix, try and hear it for yourself.

• You don't hear <20Hz or >20kHz. Nobody does, so cut those frequencies off.

• Have the layers in balance with each other. Synth layers, percussion layers, vocal layers etc. should be in the end bringing each others up, not suffocating each other.

Note that I'm a self-taught electronic music producer. For instance I personally would make patterns of both pulse plucks and more sustained saw synths for the lead melody. The "background" would be made of smooth saw pads. Share the high frequencies with lead synths and pads, cut the low end and fill it up with a bassline of your liking. Serve with compressed percussion seasoned with uplifting fx.

tl;dr: Use your ears.

Answer 3

To add to all the very good answers - consider reading a treatise on orchestration in your spare time.

Ravel's and Stravinskij's are public domain, as is Rimskij-Korsakov's: http://www.gutenberg.org/files/33900/33900-h/33900-h.htm

I realize how absurdly wacky this sounds, especially since it's not exactly a light read, but in there are exposed some principles for layering "actual" instruments - with specific frequency content, sustain or lack thereof, etc - that might turn out very, very useful if applied to your synth sounds.

If nothing else, reading that may help you think of sound and timbre in a new and interesting way.

EDIT: As Todd Wilcox points out, "it might help to add that [roughly], saw waves are more like the string or brass sections (depending on the resonance of the fitler and envelopes), square waves are more like some woodwinds like oboe, bassoon, and maybe low clarinet, and triangle waves are more like purer sounding woodwinds like flute and maybe high clarinet notes."

Or just examine the preset "synth brass" and "synth strings" patches you certainly have on your synths :)