Compression on Synth?

[Obstacle1]

Hi, can anyone give me any 'guideline' settings for synth compression? I'm using quite a thick, clean, bassy tone.

i.e. ratio, threshold etc.

Thanks.

[noisewreck]

Yep. Don't compress it.

If you're using a sequencer and your performance is iffy, record into a MIDI channel and then go in and edit the velocity or change the synth patch/program/preset in such a way that velocity doesn't affect the amplitude much.

I will however compress some subby basslines with something like an 1176LN for coloration/distortion purposes.

Furthermore... Before asking for 'guideline' settings, ask yourself why do you need to compress in the first place? What is it that you want to accomplish exactly?

Here are some questions off top of my head:
1. Want to even out dynamics (in which case, don't use compression, see first paragraph above).

2. Want to shape the attack portion of the sound.

2a. Do you want to round out the sound or make it more punchy?

3. Want to shape the release portion.

3a. Do you want to make it die down more quickly or "add" sustain or make the sound pump and breathe?

4. Want to make it duck/pump with the bass drum to pull it out of the way when the bass drum hits.

5. Coloration.

For points 2 and 3 again I'd work on the synth envelopes to get the sound "right" in the first place, and then maybe, just maybe finish off with a compressor.

In any case, all of the above point to different needs, which will mean settings that are different from one another.

Other things to note for synth bass:

1. Avoid using multiple oscillators. Single oscillators in bass, specially sub bass almost always work far better, with clear punchy results. If you want to have the sound of the detuned oscillators, then separate your bass into two halves: Mid-bass and subbass. Have your detuned oscillators occupy mid-bass and high-pass/band-pass those, and use a single oscillator with a clear wave (sine or low-pass filtered triangle for example) for the subbass. The reason for this is detuned oscillators "beat", in other words as the waves' cycles move in and out of sync with each other they will reinforce and cancel each other out, thus affecting the levels. In most cases these level fluctuations are quite big in the bass. The natural tendency is to strap on a compressor on these. But most compressors will do two things: suck out the bass frequencies from your bass and make the whole thing sound rather stiff and lifeless. By separating your bass into mid-bass and sub-bass you give yourself a nice solid foundation, and have the richness of the detuned oscillators w/o having to resort to less than ideal "solutions".

2. If you play two different notes at the same velocity but one sticks out like a sore thumb and the other is hardly audible, then your room acoustics are pretty screwed up. Don't try to fix this with a compressor or an EQ as there is usually nothing to "fix" on the recorded audio. What you need to fix is your monitoring environment. If it's impossible, then learn to identify your room problems and check your sound on a good pair of headphones.

3. I think I'll shuddup now. I could go on and on about stuff to try and do with synth bass. It's a whole world on it's own. Just ask any drum'n'bass nut

[Obstacle1]

Thanks for that - very comprehensive!

Question 2 is my main problem. I'm experiencing too much attack I think so when I change note, there's this undesirable 'thud'.

My reason for thinking I needed compression was because of velocity, some of the notes are weaker than others, and some are too powerful and boomy. I'm looking for an equal sound.

I'm using a shit keyboard, some bottom-end casio piece of cock. maybe thats my problem right there.

Thanks.

[noisewreck]

Originally Posted by Obstacle1

Thanks for that - very comprehensive!

Question 2 is my main problem. I'm experiencing too much attack I think so when I change note, there's this undesirable 'thud'.

That's probably due to the attack being set to the fastest setting. Synths with very fast envelopes will have a thud or a click at the beginning of the sound if the attack is set too fast (this is actually good when programming drum sounds for example). Try increasing the attack a bit on the envelope to take away the thud.

Originally Posted by Obstacle1

My reason for thinking I needed compression was because of velocity, some of the notes are weaker than others, and some are too powerful and boomy. I'm looking for an equal sound.

See my comments about room acoustics in my previous post. Here's what I suggest. Record a MIDI sequence where all notes are set to the same velocity going up the scale. Note if there are notes that sound weak and whether there are ones that sound too boomy. If you notice these issues, then it's most definitely due to room acoustics, since notes recorded at the same velocity should have the same power.

[ToneCarver]

Originally Posted by noisewreck

2. If you play two different notes at the same velocity but one sticks out like a sore thumb and the other is hardly audible, then your room acoustics are pretty screwed up. Don't try to fix this with a compressor or an EQ as there is usually nothing to "fix" on the recorded audio. What you need to fix is your monitoring environment. If it's impossible, then learn to identify your room problems and check your sound on a good pair of headphones.

This is an insightful concept (i.e., I learned something from this). Seems like it might be useful, as a quick low tech tool, to build a midi project that steps several different types of sounds (bass, sawtooth, whatever else) one at a time through each of the notes at identical velocities. Record a voice track that announces the frequency of each note as it is played. Create an audio track from that and play the audio track in a room to get a rough sense of where the room acoustics might be off.

If that works, then it might be interesting to loop the audio track in a project and follow it with an EQ and adjust that until the notes sound more uniform. The EQ curve resulting from that might then be used as a 'mixing compensator' for the room (I think I can hear the pros on this board rolling their eyes and groaning as they read this ... ).

Does this make sense ? I know the ultimate solution is to properly treat the room but would working out an EQ like this bring some benefit .. would it be worthwhile ?

[noisewreck]

The EQ trick to "solve" or compensate for room acoustic problems is not a good solution for numerous reasons as it's discussed in this thread.

Here's something interesting you can try that will illustrate the problem with EQ to "fix" what are room problems:

Have a synth play a sinewave at some frequency... say 1000Hz for a while. Now get up and move around the room. Try it....

Revealing isn't it?