THis week I have mostly been experimenting with....

iPassenger · Posted: Fri Jun 12, 2009 1:49 am Post subject:

In response to my original post, I think I may have learnt a thing or two about making the perfect kick (or at least got a bit closer). As opposed to doing all the sound sculpting by ear, I recently decided to look closely at waveforms of some kick samples I liked.

1. One method I have struggled with is shaping the decay envelope to give the kick more body. After examining a number of kick sample waveforms, all the ones i liked seemed to have a golden ratio and one that is much easier to replicate/approximate using the AHD envelope. Basically the amplitude should hold for approx 100-120ms before decaying for 90-150ms. I had been trying to achieve similar results using compressors, saturators, clippers but it all gets very complicated very quickly and the AHD env seems to do much better job.

2. The more "technoy/artificial" sounding the kick drum the lower the original pitch (20Hz or so) should be and the longer the pitch decay envelope should be. For a more minimal/clicky kick drum, keep the pitch higher (50 Hz or so) and the pitch decay shorter.

3. Adding a click (using the logic pulse) and/or adding noise snap can add front end bite but shouldn't really increase the initial level of the kick drum much (if at all), as this keeps the kick sounding solid, avoids that overly clicky/transient kick drum and should make it easier to mix.

4. I quite like adding an almost imperceptible quantity of saturation to the sine wave to thicken up the kick.

One thing that had really bothered me about the G2 was that I couldn't get a reasonable kick out of it, one that would stand up against one from my Machinedrum or from my Mbase01. The issue all along had been a question of body and I think that golden ratio mentioned above may have finally cracked it.

I still think more work is needed though to truly understand great kick synthesis.

Grateful for any further kick drum recipes or thoughts? I do however appreciate it is all very subjective and closely linked to the style of music.

peterkadar · Posted: Fri Jun 12, 2009 2:51 am Post subject:

Sounds cool, I'd love to hear what you come out with from this. I'd love to get around to toughening up the G2's drums in general...

Hope you're well,

PK

iPassenger · Posted: Fri Jun 12, 2009 3:38 am Post subject:

iPassenger · Posted: Fri Jun 12, 2009 1:49 pm Post subject:

This is an example of the kick drum I mentioned above. Whilst I feel that I have now found the body of the kick I would like to add a bit more front end "thwack", without losing the bass and body. Any ideas?

Kicker 1.pch2

Description:

Download

Filename:

Kicker 1.pch2

Filesize:

1.92 KB

Downloaded:

1385 Time(s)

dasz · Posted: Tue Jun 16, 2009 11:35 pm Post subject:

This week I redesigned my g2 live sequencers after playing live 9 days ago, and realizing I need less step by step control, and more high level sequencer control. otherwise I am not able to focus or read the crowd and am too busy tweaking and editing steps manually....

so i am using my "Brussels" sequencer design (see http://www.electro-music.com/forum/viewtopic.php?highlight=brussels+sequencer&t=19940) which sets multiple steps using a lookup method, and it is easy to control, and recompose on the fly, and it works great in a multi instrument environment.

I also added multiple fill controls (morphs) which alter the fill by selecting different fill sequences (or stacking 2 of them on top). I also have loop sequencers which change the loop point dynamically. what fun!

iPassenger, for my work this week, I did create my own bd using some of your ideas (click), but using a shapeosc instead and modulating it's shape using an envelope, + a click.

I have done other bd designs which kicked ass which were based more on the architecture of the 909 and I must say the results are close.
/Dasz

iPassenger · Posted: Wed Jun 17, 2009 1:52 am Post subject:

grimley · Posted: Thu Jun 18, 2009 12:49 am Post subject:

IP, those kicks sound awesome!
Dasz, off to check your seq patch now.

iPassenger · Posted: Thu Jun 18, 2009 1:46 am Post subject:

dasz · Posted: Thu Jun 18, 2009 9:23 am Post subject:

the basic idea behind the brussels sequencer (aka lookup sequencer) is this.

the sebber_DZ patch has many parts to it, so here is the basic 2 sequencer concept:

voltage sequencer - lookup values. the steps of the voltage sequencer represent up to 16 different unique notes which will be used in the pattern sequencer. I tend not to use more than 8 (typically 4-6 unique notes are enoght)
- pattern sequencer (level sequencer) - triggers an event, and at this event, the value picks a note from the lookup voltage sequencer.

So this way, if a pattern is playing lookup value 1 at steps 1,4,7, and look up value 2 at steps 5,6, then in order to tweak the pitch of the entire sequence, all you need to do is tweak the lookup sequencer steps 1 & 2. this approach makes it easy to evolve change a sequence quickly.

Here is a simplified patch illustrating this idea.
/Dasz

BrusselsSeq_DZ.pch2

Description:

Download

Filename:

BrusselsSeq_DZ.pch2

Filesize:

2.16 KB

Downloaded:

1350 Time(s)

blue hell · Posted: Thu Jun 18, 2009 10:06 am Post subject:

dasz · Posted: Thu Jun 18, 2009 10:46 am Post subject:

thanks Jan.

Currently I am also working on modulating some lookup step values (either with a sequencer or other means) to introduce change over time.

I suppose one could use a multiplexer for this purpose (instead of the voltage sequencer to have more flexibility at each lookup value), but then the cool blinky lights (if you have the g2 panel) of the voltage sequencer which indicate whether a particular lookup value is being used would be lost.

and one could of course modulate with random values to create more changes ... hmmmm ....

iPassenger · Posted: Fri Jun 19, 2009 5:09 am Post subject:

@ Dasz,

Cheers for the explanation and the patch. The explanation makes a lot of sense though I have yet to look at the patch... I would imagine for a progressively evolving pattern it is pretty much an unstoppable method.

dasz · Posted: Fri Jun 19, 2009 8:09 am Post subject:

np Ross.

the other stuff I was doing in the other topic (original brussels seq topic) was to add multiple pattern sequencers (some running at 1/2 or 1/4 speed) together to introduce even more complex lookup algorithms.

today I plan to add high level sequencer control to alter the lookups and perhaps the things I just described to my new set.

/Dasz

dasz · Posted: Mon Jun 22, 2009 8:32 pm Post subject:

here is a new set with only 3 boxes (Nord G2, x0xbox and blofeld). the g2 has the type of sequencers I have been talking about here recently (brussels lookup seq)

http://soundcloud.com/dasz/somerville-haagen-v2

enjoy

dasz · Posted: Wed Jun 24, 2009 11:40 pm Post subject:

now I added a 2nd G2 only as a separate (simultaneous) sequencer panel . wow! now I can track 2 sequences simultaneously making them meld together more seamlessly. bliss, as it is midi only 2nd g2 2(too) Smile

sometimes a g2 does not need to use it's audio capabilities but its midi control and easy user interface instead.
/Dasz