Bassline Movement

[wub]

1) make a big fat tone, make it go on for ages though (minute or more!)

2) bounce it out and make it fatter (frequency splits, reverb on tops, overdrive etc) and route it to a group - chuck a bit of compression on it to get it gelled back together) and bounce this again.

3) load it into a sampler. now you have this big fat, long tone, you want to get some variation in it - i tend to automate some lpf cutoff (try and get loads of shapes into it) and portamento it up an octave (makes it lurchy but always in key) and get all the best bits, i don't do much more than this, but i'm sure you'll have some better ideas than me - you should now have a big long fat noise with shitloads of movement in it, render this out.

4) load it back into a sampler, turn the velocity to amp off, and then set velocity to sample start time. you can now access any of the 127 points in the sample, using the one patch you've made you'll have 127 variations, some with filter down, some fully open, some wamps, some shutting down filter stabby bits, some octave jumps up with filter, some octave jumps that are quite lpf'd, some with lurches back down to the root note.

if you put a tiny bit of portamento on this you can do all sorts of things just by overlapping or not overlapping the notes to trig different parts - you can copy a section and drastically edit it just by moving the velocity slightly - you can make some pretty mental basslines as once you've made it - the hard work is done, and minor tweaks make massive differences.

[wub]

The first step is to design and make a long sustained bass sound that develops, twists and morphs over time. The sample can be of any length really but I usually use something with a duration of about 30 seconds.

I won't go into too much detail on how to actually make this sample as most people know these techniques already. Basically though, it needs to have as much interesting content, filtering, edits, effects, distortion, variation and timbre change as possible. Techniques I use to do this include:

- bounce down 5 of my favourite reece samples of the same note (2 or 3 samples already have basic filters applied, like a LPF at 1/16 LFO for example, the others are perhaps sustained)
- use an audio editor to merge/blend them together (overlap and then fade in/out the volumes)
- use a load of insert effects (filters, chorus, etc.. anything goes - one of my favourite techniques is to use a distortion AFTER a reverb and overdrive the reverb tails)
- use automation to increase and decrese the effects' wet signal throughout the duration of the sample (this part is pretty important - go to town and really fuck the sample up)
- physically chop, reverse, re-arrange, edit, mangle, etc the sample in an audio editor
- split the sample into sub groups and apply effects to different frequency bands (mono the sub, chorus on the mids, light reverb on the high end, etc)
- and finally, compress/limit the sample to ensure even volume before bouncing the sample to save cpu

The point here isn't to copy the above process exactly - it doesn't really matter what sample you use/make as long as it's relatively long in duration. You could even use vocals, drones/pads, etc - anything that develops/twists/changes over time.

STEP 2)

Once you've got this sample into a suitably ridiculous state, load it up into a sampler. You need to use a sampler that allows you to control the sample start point with a midi controller (most do - Kontact, DirectWave, Shortcircuit, etc).

You then need to assign your pitch bend wheel to the sample start controller. I'd suggest setting the initial sample start so that it's somewhere in the middle of the sample, at a sonically interesting point. Then set the sample start controller so that the pitchbend controller allows you to change the sample start by plus and minus amounts. This means you can use the pitch bend to quickly 'track' backwards and forwards through the sample whilst it's being triggered. Releasing the pitchbend wheel will bring the sample start position back to its default state.

This idea isn't really anything new - I'm sure quite a few of you already know it (especially the Tracker heads!) but it can definitely produce really nice and complex results, with relative ease.

Oh yeah, you can obviously also assign more than one controller to the sample start controller. Try assigning the modulation wheel to the sample start controller, but using a different range. Apply a portamento to the patch and you've now got instant Teebee & Calyx style masslines

PS - as always, bounce, layer, chop, effect and bounce again.. complex sounds usually sound complicated because they are!

[wub]

assigning controllers to the sample start point is a really good way to get movement in a riff.

recently ive found myself making a few variations of a whole riff and bouncing them as individual loops. Load them back into your DAW and have an instance of the loop for low, mid and high freq ranges, the same way that you split a reese.

You can then chop the the mid and high instance of the loop into their individual notes, move them to individual channels and go crazy with different processing for each note of the loop.

you can make some pretty nasty loops this way if your initial sound is bolshy enough!

Of course this means that you have your bass/mid part as a wav and not a live part, but if you create enough variations it shouldn't be that much of an issue.

[wub]

another thing you can do with this technique is to assign the modwheel to do micro adjustments of the sample start time (10-50ms for example). This allows you to introduce very subtle movement into any riff/pattern you're playing.

you could also assign the modwheel to manipulate the sample start position controller RANGE. This means you would have 127 different possible ranges that the pitch bend controller can 'track' across... a lot of scope for wild manipulations using this technique

for anyone wondering, this sorta thing has been used loads before; particularly in the tracker scene. Back in the day when trackers didn't have any kind of realtime filtering, you'd sample a long sustained sample of say a TB303 synth being filtered from no cutoff to full cutoff and resonance. You'd then load it up into your tracker and use sample start offsets to create the illusion of realtime filtering within the tracker. Very powerful

[FSTZ]

nice one

[JTMMusicuk]

cheers for this wub

[Dystinkt]

great tips wub, where are they from?

[mikeyp]

/bookmarked

[Sagal]

Nice one. Knew most but always something new to learn. Cheers!

[sunny_b_uk]

they're old posts by helpful members on here. the second post is something i always do with a simple sawtooth and ends up into somthing much much more.
although synths like massive are easier to use and sounds great, u can get your sounds sounding better when making use of these techniques! especially automating the wet and dry signals

[Sagal]

I preffer the old way if I was to choose one. I think it give more interesting results, but really combining both, modern synthesis automation and then exporting it and smashing the effect, assumming one knows what he's doing, is killer.

[VirtualMark]

great tips wub, where are they from?

dogs on acid

[RT60]

going to try this in kontakt tonight.

[naroja]

This is Great! Thank you so much for sharing!!!

[lowpass]

Some good tips in here, back to resampling for me!!

[wub]

although synths like massive are easier to use and sounds great, u can get your sounds sounding better when making use of these techniques! especially automating the wet and dry signals

You don't even need to use a synth for this. Audacity has a tone generator built into it - a 60s long sine wave @ 55hz or whatever and you've got your raw sound.

I preffer the old way if I was to choose one. I think it give more interesting results, but really combining both, modern synthesis automation and then exporting it and smashing the effect, assumming one knows what he's doing, is killer.

The techniques are very low intensity, both in terms of time and CPU power - you can do it over & over in rapid succession, and without making your machine strain under the 27 instances of Massive that some people seem to like having open at once.

I've found it considerably helpful (helpful? more helpful?) to have a couple of 'Resampling' specific templates created in FL. I've got two for specific frequency splitting (one via Filtering, one via Multiband compressors) that I can drop the raw sound into, and another for the actual 'playing' of the bassline once I've got it sounding like something.

[cheshirebeats]

Is this possible with Logic's ESX24 sampler?

I have a sound with different automated effects on it, its about a minute long, and then I bounced, put a filter on it, I have 1/4 notes, 1/8 notes, and 1/16 notes in the 60 second long sound. I loaded it into EXS just as a sample, routed velocity to sample start time, recorded a bunch of midi quarter notes, then changed their velocities using the hyper editor, but I'm not getting any change in the sound... Any ideas? I feel like I'm doing something wrong because in the manual it specifically states that stuff like this is possible with EXS 24.... I got rid of the amp modulation effect too, but its still not doing what I want it to. Any ideas?

[Electric_Head]

have you tried changing the actual note?
each note will have different effects.

[Undrig]

This is an awesome idea. The only part that stumps me are putting tempo based fx like lfo filter modulation would be thrown out of whack if you take said sample and spread it across any notes beyond the root note.
How exactly are you supposed to play a melody beyond the root note without any time based fx falling out of sync?

The timestretch algo in kontakt is terrible and not really an option for using any method with time based fx sweeps of any nature.

[hxczach]

1) make a big fat tone, make it go on for ages though (minute or more!)

2) bounce it out and make it fatter (frequency splits, reverb on tops, overdrive etc) and route it to a group - chuck a bit of compression on it to get it gelled back together) and bounce this again.

3) load it into a sampler. now you have this big fat, long tone, you want to get some variation in it - i tend to automate some lpf cutoff (try and get loads of shapes into it) and portamento it up an octave (makes it lurchy but always in key) and get all the best bits, i don't do much more than this, but i'm sure you'll have some better ideas than me - you should now have a big long fat noise with shitloads of movement in it, render this out.

4) load it back into a sampler, turn the velocity to amp off, and then set velocity to sample start time. you can now access any of the 127 points in the sample, using the one patch you've made you'll have 127 variations, some with filter down, some fully open, some wamps, some shutting down filter stabby bits, some octave jumps up with filter, some octave jumps that are quite lpf'd, some with lurches back down to the root note.

if you put a tiny bit of portamento on this you can do all sorts of things just by overlapping or not overlapping the notes to trig different parts - you can copy a section and drastically edit it just by moving the velocity slightly - you can make some pretty mental basslines as once you've made it - the hard work is done, and minor tweaks make massive differences.

Quick question: Does anyone know how to map midi velocity to NI Kontakt's sample start time/sample offset. I know how to do it in Abletons Sampler, but wanna know how to do it in my beloved Kontakt. I've been googleing for almost 2 hours now and I even read the manual o_O. I couldn't come up with anything though :/