Dubstepforum.com

Hey Guys,

I have no issues creating original melodies/rhythm as well as replicating some of my favorite tracks. However I have been having trouble when it comes to creating movement in my sound. My melodies sound so "bland" and "boring" if you will. I know alot of it comes down to layering, sound design and additional processing.

What do you guys do to create movement in your sound? So far I have been trying...
1. Adding envelopes/LFOs to various parameters
2. Sidechaining
3. Applying Swing
4. I suppose re-sampling (?) pitching down/timestretching etc (?) Not something I have attempted yet...

Any ideas would be greatly appreciated. Unfortunately I have no original tracks to post at this point, still stuck in "2-bar loop land" haha. Thanks again!

Automation.

filter sweeps make sounds interesting when coming in/going out of the song. Mess around with the Q to see what it sounds like and try bandpass, bandreject, comb etc. for some cool sounds.

Velocity and accents!!! It should come first.
Depending on what you're doing you can bring it even further randomizing the velocity (on a very short range, to keep it musical, not a messy bouncy crap), (MIDI Effects > Random (in ableton, there should be similar stuff in other DAW)) + applying some more frank accents.
(Be careful with this if you like to level with little headroom... since it random, it can be under peak one time, and above next)

The best way to get your sound 'moving' is to just do things that you haven't done before. Sounds like common sense but drop an envelope or LFO on something you haven't before then note the effect it has. Just become familiar with your synths and it will come to you. As for melodies, try to not only think of them as audio coming through the speakers but also as a visual pattern on the 'piano roll' or whatever it's called in your DAW. Just get a few base notes for a melody and build on it one note at a time

Also, very subtle things can make a big difference. For example, a slight swooshing sound underneath a hi hat. Add another, reverse it and put it underneath the second hi hat [In a --kick, hat, snare, hat-- type pattern]. You can get a ton of cool stuff out of reversed things, I reverse mostly everything just to see what it sounds like. If it doesn't sound good then flip it back but if it does you just got one more in your arsenal

Humanize your notes.
Automate FX (filters, EQ, flangers, reverb etc...)
Automate stereo width.
Add and subtract layers of sounds.

mthrfnk wrote:Humanize your notes.

I really need to start doing this more often. A tip for those using ableton: In the MIDI effects section in the browser there is something called "velocity". Basically it randomizes the velocity of MIDI notes to a degree that you specify. It's really great for making drums feel less robotic.

FL has an option literally called Humanise, I also like to randomise pan slightly on real instruments (piano/orchestra etc)

Automate FX and some random things. Automate some sound(s) to go from left to right( or right to left)
If you use massive, insert LFO to the pan knob and try different speeds and waves and widths.

as far as growl reeses go, a general fx chain to get movement could go as such (i'm in fl and use massive for my reese creation but this can be applicable across any vsts and daws)
create a reese patch>send through scream filter to taste>send through an enveloped comb filter to get movement over the course of a note>add subtle distortion to taste>subtle phasing>more distortion (controlled by the phasing which adds to the growl)>Low pass filter (generally make an automation clip to control the filter cutoff throughout the song)

note: when doing this, it is helpful to separate your mid and high reese from your sub (either a dry reese patch or simple sine sub) by sending the patch to unique mixer channels and then using multiband compression and/or eq to single out the mids and highs and process separately from the sub in order to preserve the low end power. a general mixer set up could look like this...
channel #1: bass in (send to channels 2,3,and 4; do not send output to master) Send your dry patche(s) to this channel
channel #2: sub; eq/multiband compressor (Send output to channel 5)
channel #3: mids; phaser, distortion, etc> eq/multiband compressor (send output to channel 5)
channel #4: highs; phaser, distortion, etc> eq/multiband compressor (send output to channel 5)
channel #5 : bass bus; Low pass filter (w/ automation control) > eq > compressor (light compression to glue the different frequency bands back into cohesive sounding bass) Send this output to your master

Hope this explaintion isn't too confusing and helps out mate

appreciate all the replies, had not thought of alot of these options, ive got some work to do.

makes sense drooka, can't wait to try some stuff out!

For FM patches with multiple operators I like to detune the operators slightly which creates some phasing/movement in the sound.

bass

I guess this hasn't really been mentioned, but if you start with an analog sound or a well modeled analog synth (arturia moog modular comes to mind), those shoddy electronics tend to cause a lot of movement. Or start fiddling away at the modulation grid in fm8.

drooka wrote:as far as growl reeses go, a general fx chain to get movement could go as such (i'm in fl and use massive for my reese creation but this can be applicable across any vsts and daws)
create a reese patch>send through scream filter to taste>send through an enveloped comb filter to get movement over the course of a note>add subtle distortion to taste>subtle phasing>more distortion (controlled by the phasing which adds to the growl)>Low pass filter (generally make an automation clip to control the filter cutoff throughout the song)

note: when doing this, it is helpful to separate your mid and high reese from your sub (either a dry reese patch or simple sine sub) by sending the patch to unique mixer channels and then using multiband compression and/or eq to single out the mids and highs and process separately from the sub in order to preserve the low end power. a general mixer set up could look like this...
channel #1: bass in (send to channels 2,3,and 4; do not send output to master) Send your dry patche(s) to this channel
channel #2: sub; eq/multiband compressor (Send output to channel 5)
channel #3: mids; phaser, distortion, etc> eq/multiband compressor (send output to channel 5)
channel #4: highs; phaser, distortion, etc> eq/multiband compressor (send output to channel 5)
channel #5 : bass bus; Low pass filter (w/ automation control) > eq > compressor (light compression to glue the different frequency bands back into cohesive sounding bass) Send this output to your master

Hope this explaintion isn't too confusing and helps out mate

my man, comb filtering his reeses

get different bandreject / notch filters going on and automate them. split your low end before distorting (or don't, if you are going to high pass it) and saturate before filtering. when splitting bands to apply different effects, use "weaker filters" (6db/12db) and when automating a filter through the whole sound, use 24db filters.

You can get nice phasing sounds by playing chords with sines or adding unisono to them, then applying distortion.

its sad how almost no one said velocity
it's like the current gen of producers have forgotten that loudness isn't binary

Panning your percussion also gives some extra swing to your songs.

Automation.

Program out the automation of some parameter like the LFO rate, and have other parameters follow that one if possible, like volume (like hasezwei said), resonance, cutoff, pitch and whatever else. It's sort of like setting up a macro.

Also bounce tracks down before the proper mixdown and edit the audio files (similarly with a master bounce before mastering).

hasezwei wrote:its sad how almost no one said velocity
it's like the current gen of producers have forgotten that loudness isn't binary

Yeah, I overlooked that for the longest time. Really does make a hugggge difference. Do you know if there's a difference between changing the velocity and just automating the volume?

sidechained/gated white noise.
long filter sweeps - also great to do long filter sweeps on top of fast lfo filtering.
fade in/out - pretty much the same as filter sweeps.
syncopate some notes. while others are gated. while others are side chained.
velocity/swing