Higher octaves in chromatic mode workarounds?

one of the only frustrations i keep encountering with the digitakt :

hitting the upper limit of how far a sample can be transposed in chromatic mode

have to admit i don’t understand why the digitakt can transpose a sample FIVE octaves down but only TWO up ? depending on the sample 2 up often doesn’t even get into higher registers. i encounter this almost every project i’m working on. so far as working around this seemingly needless limitation, currently aware of 2 methods :

  • transpose the sample up and re-sample it
  • reduce the sample length

both of these have drawbacks. resampling = a cumbersome kink in the workflow when it seems like the digitakt should be perfectly capable of simply transposing the sample another octave or two higher. especially when working with single cycle waveforms reducing the length tends to fundamentally change the sound.

am i missing anything ? any other workarounds to suggest or insight as to WHY elektron have not provided the ability to transpose samples more than two octaves up ?

( if anyone from elektron reading this would definitely welcome a wider transpose range in the next update )

1 Like

The four octaves down came out with a recent update. I always just resampled everything once, and named it appropriately then placed it in the neighboring track. It’s a process but it is very straight forward.

If this is happening to you all the time, why not save the sample an octave or two up to begin with?

i’m often working with single cycle waveform samples and others i did not personally record.

to be honest the last 2 of 5 octaves down kinda useless in most cases, that low and samples begin to sound like a bullfrog croaking unintelligibly whereas only 2 octaves up is usually not enough to get into bell-like registers above the midtones

it seems like an arbitrary limitation, can’t think of any technical reason why the digitakt shouldn’t be able to simply transpose higher, as so much other gear natively does without issue

I dont know if there’s anything special you’d have to do to pitch up a single cycle waveform. But at least for the other ones, you can always pitch it up before you transfer it.

I’ll agree with you that it’s odd that you can go 5 octaves down but only 2 octaves up. I’ve got two other samplers. They both go 3 octaves in either direction, which makes more sense to me.

But I don’t think there’s an easier work around then just pitching up a sample before you transfer it or resampling it after the fact.

Yeah, 5 down and 2 up is funny, but the last 3 of the 5 down are basically bonus from that update. And the DT still has more range than the Octatrack! :laughing:

The Octatrack’s chromatic range was a real shocker coming from the Digitakt, which lately I still wish I had. For OP, the resampling in the Digitakt is pretty simple enough to do depending on the sound you’re using. Definitely quicker than doing it in the Octatrack.

One octave higher means double the sample rate, so sampling at 48khz means 96khz and then 192khz and maybe there is the limit. At least that’s how it was with old samplers. And that’s why they sound so „old school“ if you play a sample 4 octaves down it would be at 6khz …or something like that…

1 Like

It’s actually even stranger than you think, because the Digitakt doesn’t just go 5 octaves down but much further.

If you pitch a sample down, it also pitches down the lower and higher notes on the keyboard.
If you pitch a sample up, you lose a higher note on the keyboard with every note you pitch up.
No matter which way you pitch your sample, it will never go more than 2 octaves up.

So, pitching up doesn’t even have a good use case because you could just as well trigger a higher note. It’s only useful for tuning.
The same goes for pitching down, because you lose the highest notes you might just as well trigger a lower note and keep the option for the higher ones.

I think the update for the pitch range only changed the algorithm for pitching down, not the one for pitching up and that causes a strange limitation.

If the resampling was neutral this wouldn’t be such a problem, but it’s not.
The auto normalize and emphasis completely kills the resampling for me.

This is not how it works these days.
The sample(or part of it) is read from memory(speed is not an issue here) and goes through a resampling algorithm which interpolates the samples to the new playback speed.
In case of the DT it’s windowed sinc interpolation Whittaker–Shannon interpolation formula - Wikipedia.
When pitching up, the resampling reduces the amount of samples to exactly 48000 per second because that’s the speed the D/A converter runs at.

Only the very first samplers modified the clock speed of the whole system including DAC’s to change the pitch of a sample. The first version of the PPG Wave also worked this way.
This indeed comes with some hard limitations.
But the typical “sound” that is associated with old-school samplers usually comes from the older resampling/interpolation method and the converters. Linear interpolation sounds pretty bad compared to windowed sinc but at the same time could add some character to a sample.

1 Like