How do you go about sampling wavetables?

[chk071]

Thanks but... as they say "The journey is the destination". I really wanted to try this out manually for a long time now.

I already set up a project including automation, and already recorded the Alt1 wavetable. The only thing i'm confused about now (probably miss something obvious) is that the waveforms, when i import them to Dune 3, are all a octave higher than they should be. Any idea why that is happening? I did everything exactly according to the hints in this thread, and in the Expanse tutorial i linked to. Played the notes in F0 minus 24 cents, recorded in mono, 44.1 kHz, 16-bit wav.... there's no tuning going on in Largo either. Strange.

Edit: Ok, my bad. Seems like it works fine when i import 1024 samples long wav files.

at 2048, should be a single cycle waveform, are you getting that?

Strangely, no. I get two cycles with 2048 samples, and the settings mentioned earlier in this thread. Which is obviously also why the waves were an octave too high when importing them to Dune. With 1024 samples, it works fine, even though that obviously is not the supported length?

[AnX]

no, should be a single cycle at 2048

it may be that F0 is diff in diff hosts. i know middle c is not always the same octave in diff host, debatable

try detuning the osc by an octave?

[chk071]

I'll give it a try. Strange that everything seems to work with 1024 samples though. Is there any downside going with that, like, what happens with the sound when it's just 1024 samples long?

[AnX]

tbh, i havent touched the new WT editor in D3, i assumed it would only load 2048 like D2. maybe its resynthesizing it or something

[fmr]

no, should be a single cycle at 2048

it may be that F0 is diff in diff hosts. i know middle c is not always the same octave in diff host, debatable

I don't think it has to do with frame size but with sample rates

Sampled at 44.100 Hz a wave of 1024 samples long is 43.07 Hz which is an F minus 23,44 cts
Sampled at 48.000 Hz it's 46.88 Hz or an F# plus 23.26 cts.
Sampled at 96.000 Hz it's 93.75 Hz, also an F# plus 23.26 cts.

So, if he sampled an F0 minus 23.44 cents, he only has a single cycle wave with 1024 samples. If he uses 2048 samples, he will get two cycles, instead of one.

I think he will be pretty safe by using 1024 samples

[AnX]

i was told by Mark/cytospur here, that it is 21.5332 Hz. This corresponds to F0 detuned by 24 cents at 44.1 kHz

i always used this, and i always got a single cycle 2048 long

[cytospur]

no, should be a single cycle at 2048

it may be that F0 is diff in diff hosts. i know middle c is not always the same octave in diff host, debatable

try detuning the osc by an octave?

I think this is probably the reason.

[fmr]

i was told by Mark/cytospur here, that it is 21.5332 Hz. This corresponds to F0 detuned by 24 cents at 44.1 kHz

i always used this, and i always got a single cycle 2048 long

21.5332 Hz is half the frequency I quoted. It's MIDI note 17. Considering that Middle C (C3) is MIDI note 60 (261.6255653006), then that F should be F-1, not F0. But if you consider Middle C to be C4, then that F would be indeed F0. Maybe that's the difference.

The different names given to Middle C is a problem that will never be solved.

[chk071]

Yep. I have to play F-1 to get a single cycle, 2048 samples long. Guess Studio One is different in that regard to Orion, or whatever you guys are using.

[chk071]

This is more difficult than i thought... It's important to even catch the same "phase" for the single waveforms, otherwise, morphing through the wavetables will sound choppy and weird. That's pretty easy on the more sine wave like waveforms, but, pretty difficult on the more complex, FM-esque waveforms. Guess i'll have to zoom in really close to be able to catch the right position.

[fmr]

This is more difficult than i thought... It's important to even catch the same "phase" for the single waveforms, otherwise, morphing through the wavetables will sound choppy and weird. That's pretty easy on the more sine wave like waveforms, but, pretty difficult on the more complex, FM-esque waveforms. Guess i'll have to zoom in really close to be able to catch the right position.

A strategy I try to follow is always have the first half of the waveform in the upper quadrant, and the second half in the lower quadrant, with a zero crossing always in the middle. The Waldorf waveforms are always symmetrical. If you are not obtaining symmetrical waveforms, something is wrong.

If you do as I wrote, phases will be more or less aligned. As an extra, some wavetable editors have the option to align phases - don't know if DUNE has that, didn't try it yet

[aciddose]

The usual method is to record longer periods and use a tool to either do phase detection of the lowest harmonic or detect zero crossings.

Since the tool can be told the exact length of the cycle it's even possible to recover via interpolation "perfect phase" where the input phase shifts by a fraction of a sample period by correcting the measured phase.

While such a tool is super simple to create I'm not sure about whether any already exist.

[chk071]

A strategy I try to follow is always have the first half of the waveform in the upper quadrant, and the second half in the lower quadrant, with a zero crossing always in the middle. The Waldorf waveforms are always symmetrical. If you are not obtaining symmetrical waveforms, something is wrong.

Yep, that's the way i tried too, but, tough to see on the more complex waveforms.

I just realized that i also could just set phase to retrigger, then i should get the same phase every time, right? I set up a project, where there is a single note for every wavetable, with automation so that the wavetable changes on every note anyway. Oh well… learning with every small step i do wrong.

[cytospur]

You can reset the phases in DUNE3 manually. It doesn't take too long to do. Go to additive mode, zoom out as much as you can and then drag along the bottom of the phases window, when the cross hair is visible. This will set all the phases to zero and give a symmetrical waveform. This has to be done for each individual cycle unfortunately. It would be good to have this as a menu option though.

[chk071]

Oh, that's great. Thanks for the tip.