Using the delay as a sampler - updates on the trick

I’m Using the delay as a sampler trick and it’s really fab to get a looping delay line going on. I believe the sampling is in 24 bit.

Reminds me a bit of the ot pick up machines but it’s a lot easier and sounds pretty lush.

You have to stop the sequencer to get the sample looping going on but I worked out a couple of ways to overdub and/or add into the delay sample line:

Have the monomachine sequencers sequence or arp the md. the delay sampler stays in the background and you can add notes/trigs by turning up the delay per Indy track.

You can also drum pad play in md synths from the md into the delay stream.

You can fm the pitch machines (see the fm op32 trick) at 300 bmp and record the weird fm drum hits into the delay - and bring the bmp back down to a slower rate afterwards.

I’m also mucking around with the crtal delay machines on say 4 tracks, plocking and stopping the seq. I haven’t mastered this idea yet!

Do any md masters have any other tricks for the delay sampler trick?

How long is the sample looper - 127 delay time, is what 5 seconds or something?

For those that don’t know, here is the trick from the md master: Justin

Using the delay as a sampler:

You can sample the MD output in stereo at full bit rate using the MD delay.
(This can be used as an alternative to RAM machine mixing between patterns)
(Also allows you to high or lowpass filter the output of the MD)

If you set your rhythm echo parameters to

LEV = 0
FEEDBACK = 64
DELAY TIME = 127
FILTERF = 0
FILTERW = 127
MONO = 0
MOD = 0

Then go into the fx page of one of your tracks. Hold FUNCTION + DELAY and crank it up to 127 (make sure you do a full clockwise rotation from 0 to 127).

Hit play… and then press stop after 16 steps.

Now if you turn up your delay volume you’ll hear your 16 steps sampled

If the sequencer runs for more than 16 steps you’ll get over dubbing. You could easily avoid this by switching into a blank pattern after the 16 steps has been sampled, then using an CTRL-RE machine to automatically turn up the volume on the delay.

The only real drawback to this method is that it overwrites any delay settings you may be using in your pattern. Also once you apply filtw and filtf to the delay loop you cannot revert back to an unfiltered version.

I should have added… the delay loop does degrade over time. One last addition: This allows you to selectively sample. Which means you could sample say just the kick drum and snare… as opposed to having to sample the entire output. You just choose which tracks you want to record by either sending them to the delay, or not.

  • Justin Valer bbb
5 Likes

Thanks. Tried the original tip out yesterday.

1 Like

Some time last year I automated this using the Mini/MegaCommand and created a beat repeat feature for MCL.

I felt it was too gimmicky at the time and abandoned it :stuck_out_tongue:

The code is still in the firmware. I’ll add it back in on the weekend and do a demo.

3 Likes

sounds like fun Justin - thanks for the reply.

I’m curious about what is the actual sampling time on the delay.?

Adding to the above trick info: it’s fun to do a kinda beat repeat with different step lengths… and get different lengths on tracks by sampling into the delay seqs at 2 steps and stop seq. and then after that add into the delay sample another seq at 8 steps.

I was doing this with justins 303 synth pack yesterday

2 Likes

Sampling time is a function of the Tempo.

From Manual:

TIME controls the actual delay time. It is relative to the current tempo, and it is measured in 128th notes. For instance, for a two beat delay (8 16th notes) TIME should be set to 64.

Therefore Time set to 127 = 4 beat delay.

Some Math:

TEMPO = 32BPM = 0.53BPS

4/0.53 = 7.55 seconds

TEMPO = 128BPM = 2.133BPS

4/2.133 = 1.88 seconds

1 Like

Thanks!! 7.55 seconds. That’s great.

1 Like

~It might seem obvious but if you send most bread and butter sounds through the individual outputs and then the md effects are an internal effects bus for the rest.
~I’m going to send another instance of a p machine and resample it via another ram record machine. LFO on the Filter and sample start times and use r2 and p2 to make use of the output. (I’m not sure if it’s going to work but giving it a try as we speak)