Wavetable creation issue?

hi=)

I dunno if this is a bug or whatever, so i am posting it here…

I was making some wavetables from wav files… .

the manual says that the sampled wavetable length is 2 seconds? (if so, this is good.)

anyway…

the pic in anima does not reflect the 2 seconds of wavetable wav that i imported…

• the length is wrong (not 2 secs)
• also the wave start amplitude is not correctly reflected in the halion/anima wavetable (editor)

thanks for any help =)

This is 2 secs of the wave in cubase (and how the waveform should look like) :

This is how it looks in the halion wavetable editor:

This is how it looks like in the anima wavetable screen:

It think, you mixed up the import of wavetables with the creation of wavetables.

When you import wavetables from other synths, these wavetables do not contain any information about the length of the wavetable envelope, because it does not exist in other synths. HALion sets the wavetable envelope to a length of 2 seconds. This assumed length can be adjusted after the import.

When you extract a wave from a sample, HALion uses a window to extract the spectrum. These windows are much shorter than 2 seconds. The maximum length is 2048 samples. You can try this in Fixed Mode, where you can set the size of the window manually.

You could try the following:

1. Switch to fixed Mode.
2. Set the extraction marker and size of the window so it matches exactly one cycle of the waveform in the sample.

I hope this helps you to get the result you want.

Hi Matthias=)

Thank you for your help, and, congratulations on halion6=)

I activated fixed mode and set the size to the maximum 2048 samples…

would you ever think of increasing the size to an indefinite amount? - out of the 4 wave samples i used, only one could fit in a whole cycle ( and the start of it was chopped)
(it is easier to line up the cycle from perfect waves if the start is not cut off,)

here is a video of it:

Thanks again for everything=)


and again, thank you, the whole halion team and team steinberg for halion 6=)

The extraction window does not “slice” the sample. It extracts the spectrum via FFT. Therefore, the window size can’t be increased to an indefinite amount. It’s simply not possible with FFT.

Regarding the sample you are trying to analyze. It seems, your sample has a very, very low pitch. In the screenshot from your first post I count 16 cycles per 2 seconds. This would be a frequency of 8 Hz, or MIDI note number 0.

You can calculate the lowest possible pitch from the sample rate of your sample and the maximum window size. For example:

44100 Hz / 2048 = 21.53 Hz

I’m not sure how you produced the sample. I suggest, you sample the wave at another pitch, for example, C0, MIDI note number 24, and all should be fine.

When I design wavetables, I use pitches well above C0. Usually, in the range from C1 to C4.

I hope, these insights are of help for you.

Hi Matthias=)

thank you for your insight…

I cannot help noticing that there is a fade in and a fade out that “cuts” the measurement… hence, a perfect waves starting point is not translated correctly?
(and the measurement does not start at the zero crossing of the sample but at the zero crossing of the fade halion generates, hence the analysis is not correct…- I cannot find an option to disable the artificial fade either)

also,

Since we can add many wavetables in sequence, is it not possible to analyse a “2 second” wave and split it up and link the little wavetable boxes together to form one big one?

thanks)

There is a “XF” parameter which adjusts the fades. You find it right of the Size parameter.

The main concept behind wavetables is to import single-cycle waves. After you’ve captured the full wave-cycle, HALion can play this for as long as you wish. With this approach in mind, there’s no need to capture 2 seconds of the same wave, like you suggest.

The pitch of the wave in your sample is much too low. In your picture the extraction window covers only one half-wave. Please raise the pitch of the sample, either with an external program or by resampling it in HALion, and you’ll be able to capture the full wave-cycle.

I hope this gets you where you want.

Hi Matthias=)

thank you again for your insight…

I think it is clear now, but i could not check it yet as i was busy with another thing and didnt want to reply before finishing this thing and testing the wavetable thing first etc…

I will give you a more conclusive “end-post” when i can.

thank you again very much for being here and helping=)

Hi again=)

yeah, those waves are not meant for a kitchen radio;)

I stretched the wave to 2048 samples in cubase…

then i needed to bounce it before i could drag it onto halion because halion loaded the original unstretched file instead of the edit…

here is the wave in a stretched state in the cubase arranger:

when I bounced it in cubase, the waves consistency was compromised and looked like this:/ :

(which made it unusable for my purposes…

But never mind all of that…

I see that I need to use a sample with a much shorter cycle…
→ please do consider the “chaining” of wavetable boxes so that the 2048 limit is no more;)
→ aha, the XF was hiding from me thx

oookkkk, thank you very much=)

have a great everything=)