Ways to make Wavetables for Hive?!

[Kapellor]

Hi there,

im pretty inexperienced in wavetables and how to make or use them effectively. I only use Hive2 for WT‘s but tried other WT Synths.

I read about the methods how to create them for Hive, but the results are always bad and not what i expected.
I cant program (script them) and just record wav‘s or use the recording function in Zebra to make WTs.

When working with Z2, i take good care of OSC waveforms and their overtones to get the timbre i thought of before.

Do wavetables always produce a more randomly outcome? I mean, if i just import a wav of a human voice singing a short note, it will sound like trash in Hive. For me it seems like imported sounds, are only usable if they are made of basic waveforms.

I would like to get more complexity in my Wavetables (some vocalish as example), but in
the cleanest way and not as random as randomly sounding images like in Serum.

So, what are my options, to create them?
Learning to script?

[pdxindy]

So in my DAW I record myself saying the word "hello". Then I resynthesize that audio into a wavetable using Icarus (works same in Serum) and then export that wavetable and bring it into Hive. Now I have a natural sounding 'hello' in Hive. (only in a narrow range of notes)

To maximize the quality of it, there are various things one can learn and there are different resynthesis algo's in Icarus that give better or worse results depending on the type of audio. But basically it is pretty easy if you have Serum or Icarus.

[Teksonik]

If you already own DUNE 3 you can use it's Wavetable Editor to create WTs for Hive. Works really well.

Oh and can use Vital as well. Its Text to Wavetable feature works well in Hive.

[pierb]

If you already own DUNE 3 you can use it's Wavetable Editor to create WTs for Hive. Works really well.

Oh and can use Vital as well. Its Text to Wavetable feature works well in Hive.

Does that also work in Zebra?

[Teksonik]

If you already own DUNE 3 you can use it's Wavetable Editor to create WTs for Hive. Works really well.

Oh and can use Vital as well. Its Text to Wavetable feature works well in Hive.

Does that also work in Zebra?

Do you mean import into Zebra or create WTs in Zebra and export them?

EDIT: I just tried both and I don't see a way to import a WT in .wav format into Zebra 2 and exporting a WT saves as .h2p format.

[MorpherX]

There exists two ways:

Create a wavetable in a WT-Synth which has a WT-Editor (Serum, Dune, etc.) and load it in Hive (1.2, 2.1)
or use a script for making a WT from a sample.
Using the script meaning that the sample is divided in 256 parts with a lenghts of 2048 samples per part. This are 8 seconds.
If the sample is longer than this, it is cut off.
The results are relative good and you can then use the Hive stretching and morphing tools. Great !!!

[Teksonik]

Sorry wrong thread......

[Kapellor]

Thanks for your replies. I demoed Serum, Dune 3 and used Vital for preparing WTs for Hive2, work all pretty good…
But thinking now why Hive 2 lacks simple WT editing tools that wouldn’t need to be as complex as Serum‘s.
Hoping that will be part of Zebra 3.
I prefer Hive‘s / Zebra’s sound over all others, so relying on (and buying) other Synths just for simple WT Editing makes it kinda incomplete.

[Teksonik]

Well some would argue that Hive's .uhm script files are the most powerful WT editor but I failed quantum physics in school so I've never been able to create a .uhm wavetable.

[Urs]

We have sneaked a wavetable export tool into Zebra2's Osc1. I think it's pretty much one of the most powerful visual editors out there as you can render Zebra's OscFX onto the file. It does not import though, and in order to get best results you have to use the spline-based Geo/SpectroMorph type of waveforms.

Yet, the main problem remains, wavetable editing is a tedious process if you want something good. It's a road paved with frustration. Likewise, sample import never ceases to disappoint. By far most "converted to wavetable" samples I've heard are one trick ponies, they hardly ever lend themselves to a wide range of sounds.

Simple editors tend to produce same same results. Like, they either propagate sawtooth-type of waveforms with variation of spectra, or they meander in somewhat muffled sounds with a lot of tiny nuances.

By far the best results come from maths. That's what we did with the scripting engine. It can do anything one would ever want, including sample import. The price is a bit of a learning curve, but there are gazillions of examples available. Those who managed to master it, like Plugmon, have creates awesome wavetables with it.

For Zebra3, we'll revamp spline based editing and we'll probably have some kind of editable "blend" type of waveform with less muffled results than in Zebra2.

[Bulbizarre]

I never do that kind of thing but I think one of the best way is to just drag your audio into the Serum wavetable frame, I've seen people do that.
I don't know if it's possible to make non text-to-vocal WT with Vital ?

And I agree I would like to see this feature in Hive too. It would make sense since it's a wavetable beast. Hive is also my go-to for wavetable.

[Urs]

We have experimented with sample to wavetable conversion in Hive, but we could not magically make it any better than anywhere else at the time. There is hope that more recent knowledge lets us do this in the future, but I can't say if and when.

There is however next to no chance to persuade me to add a full blown editor to Hive.

Historically I wasn't fan of wavetable synthesis because of reservations I had towards the quality and frustrations of the creation process. The change of mind came with Serum and Zebra3 development: I realised that Serum's formula based wavetables would be amazing if only they were combined with stellar and fast realtime interpolation from Zebra. That's what we added to Hive, plus a vastly extended method to create math based wavetables.

Product development is a vast chess game though. You try to position your creations in a niche that distinguishes them from all the other competing products. With Hive it was all about positioning it as the fastest and best sounding wavetable engine. To do so, one of our chess moves was to forgo the means to build dull sounding wavetables that don't do Hive's sound justice. I think we played that very well and we pretty much succeeded with our goals. Adding sample import or a visual editor would be equivalent with a great gambit at this point. We will think about a lot of moves before we'll consider it.

[pdxindy]

There is however next to no chance to persuade me to add a full blown editor to Hive.

Good news!

Maybe a separate app/plugin that does something different than other wavetable editors and which can smoothly export to Hive/Zebra. There are plenty of wavetable editors and they all require considerable work. Exporting a wavetable to Hive is the small part. Converting samples to wavetables is fussy. There is no guarantee to get a suitable result. Manually drawing/importing single cycle waves is entirely hit or miss... with the weight on miss.

I wonder if some sort of visual Uhm script creator/editor would be possible? That would at least be something different.

[Kapellor]

Ok..i get the fact that sample conversion, is unacceptable for non chaotic people like me.
I will shed a tear, but a shiny one.

Product development is a vast chess game though. You try to position your creations in a niche that distinguishes them from all the other competing products. With Hive it was all about positioning it as the fastest and best sounding wavetable engine. To do so, one of our chess moves was to forgo the means to build dull sounding wavetables that don't do Hive's sound justice. I think we played that very well and we pretty much succeeded with our goals. Adding sample import or a visual editor would be equivalent with a great gambit at this point. We will think about a lot of moves before we'll consider it.

Aha! And that's exactly why Germany is currently getting outsmarted everywhere.
While younger, smarter, more obsessed programmers shell out some new products
that include cool new functions, good old Germany plays marketing chess again.
The lack of Hive 2`s WT Editor is similar to building nice German electric Cars,
while forgetting to build charging stations. Wont get far enough with that thinking


I looked at some explanations on the scripting, but as a lifelong math-avoider, it makes by
brain hurt in unknown regions. Any ideas someone, where to start? Preschool or good
medical help? Resisted the urge for acquiring alternative Synths already, so not all hope
is lost.

[Teksonik]

I just use DUNE 3 and Vital to create Wavetables that sound great in Hive 2 so I don't need an editor in Hive.

I get it the .uhm files are probably more powerful but are the textbook definition of user unfriendly to me. No matter how powerful a tool is if someone won't use it then it become useless.

Case in point:

// by Urs Heckmann

Info "Karplus-Strong-like wavetable with glassy spectrum"

NumFrames = 101

Seed = 73616

Spectrum start=1 end=0 "(index % 5) * rand / (1 + 10 * phase)"
Phase start=0 end=0 "rands * 2 * pi - pi"

Wave start=1 "y + 0.15 * ( main_fi(frame-1, index) - y)"

// remove DC offset

Spectrum lowest=0 highest=0 "0"

Phase "rands * 2 * pi - pi"

Wave "x + tanh( lowpass(x, 0.3, 0.5) * 0.4)"

Spectrum lowest=0 highest=0 "0"
Normalize base=each