Pros/cons of recording multiple short samples (e.g snare)


Hi all,

I’m curious to know what everyone else’s preference is when sampling, chopping, and using short samples.

Say you have four snares that you have tweaked from an external source into a 20 sec sample. You have adjusted the filter, pitch, and envelope of the snare while the four snare hits took place over the 20 secs.

Would you:

  1. keep all 4 different snares as one sample, then chop as need per project (using start and end time)
  2. chop each snare into a standalone sample and save individually
  3. After completing no.1, keep all 4 and then save as a ‘Sound’

what are you preferences and why? consider sample space, ease of use, speed of use, flexibility etc.


I would keep them as one sample preferably not a 20 second one if possible, then save each hit as a seperate sound.


Round robin sampling is always great with short samples.


chop them into a new sample as close as possible after the decay of each has died?

probably. saves space and number of samples, makes it easy to change which sample you’re using by P-locking sample start.


For percussive samples I save variations separately and then have a random LFO modulate the sample slot source. Still haven’t quite figured out the exact behavior of the LFO amount, though. Just trial-and-error until it moves within the range I want.

I enjoy this method, sometimes I like to sneak in one or two “odd” samples that sound noticeably different from the others into the list.


How do you save a sound from 2 or more tracks?


You dont in one go, you can only save one sound at a time so if you want to copy multiple tracks you will need to enable each one before you copy its sound. You can make as many sounds as you like from a sample though.


Interesting. I saw a tutorial a while back (can’t remember exactly which one) where he resampled 3 tracks let’s say " TRACK1 + TRACK2 + TRACK3" and recorded it in one of the MIDI TRACKS and then saved that loop / track as a sound.

I though it was really clever being able to resample those tracks into just one track/sound, I just which I remember the video but weed is now legal in canada ha!


You can resample as many tracks as you like into one, just mute the ones you dont want and apply it to an empty audio track if there is one spare. I dont understand the midi tracks part though, you cant assign a sound or sample to anything other than tracks 1-8, maybe some midi loop is being used to hold the pattern data in a midi track where it can be sent to an audio track.


Maybe I didn’t get that part.

So basically I have to mute the tracks I don’t want and then resample those tracks as a loop to a new track?

Sorry for the confusion, trying to understand the method.


learn the difference between a sample and sound (definitely confuses people on digitakt).

you can ‘resample’ anything the digitakt is playing into a new sample , unfortunately you only get once chance to trim the start/end of the sample on the digitakt before you save it to the plus drive.
but yes … if you dont want it to be sampled, mute it.

personally i dont use sounds much , as the digitakt is quite straight forward if you are just triggering samples.
I do often make sample chains on mac/pc though. keep them simple – (use octachainer 1_3 , use google to find it . its straightforward to use)
I use regular sample lengths , recorded to 120 bpm , 16,24,32 individual noises (snare,toms,bass etc) per sample chain.

that way it can be easily used on Digitakt (easy to assign sample start and end points ) or Octatrack (as it’ll auto slice to 2,4,8,16,32 intervals)

sample chains could be purely bass drums / snares etc if you know you are going to limit each track to drum sound type …
or do a single drum machine sample chain (e.g. its only 808 noises).

try to also consider setting up a standard kit to allow you to easily swap things around.
setup your chain to have
3 x bass drum , 3 snare , 3 closed hi hat , 3 open hi hat , 3 clap …
when if you make 808,909,606,weird kit , you can easily swap the sample and the percussion will still sound fine.

and finally…
think about the length of each percussion sound when setting up a chain / programming.
i would often put high hats on each 1/16 , so to ensure easiest use of samples , i try to ensure the actual audio is short (within 1/16 ) , so that way i dont need to mess about with different envelopes if i change sample start/end points … everything within a sample chain would be 1/16 (short and snappy percussions) … it gets annoying if you trigger a clap (which are typically longer ) and then you have to edit the envelope and ensure nothing is triggered to cut off the audio (unless thats what you are looking for).