Tech Preview: Hive Wavetables

Official support for: u-he.com
Post Reply New Topic
RELATED
PRODUCTS
Hive 2$169.00Buy

Post

OK, here are some recordings I did with seven wavetable synths (not all of them “synths” actually, since I don’t consider HALion and Falcon as mere “synths”). All files are WAV format with 24-bit and 44.1kHz resolution.

Take into consideration that HALion and Nave don’t read WAV wavetable files directly, they analyze and resynthesize the audio file into a wavetable in their own format. Sometimes, the translation doesn’t go though as smoothly as we would expect (for example, the wavetable ARR was translated very differently in NAVE, although the sound doesn’t differ much). Also, the wavetable “TOMita”, which was created in Icarus, wasn’t read directly bu Serum, so I had to import the audio and create a new one. This led to a somehow different wavetable. I had to go into extensively tests and reading to be able to faithfully translate the wavetables in HALion, but now I think I finally got it.

The “synths” in these recordings are Falcon, HALion, Hive, Icarus, Nave, New (an unreleased new wavetable synth, which I don’t say more due to NDA), and Serum. I recorded three wavetables: PPG Wavetable #01 (purely synthetic), ARR (a wavetable created from a recording of the famous ARR sound of the Fairlight), and TOMita (from a very small audio fragment of Tomita’s “Pictures at an Exhibition").

I tuned the preset so that all synths play the PPG wavetable more or less at the same speed. I was convinced that would make the other two wavetables to also play at the same speed. It happened it wasn’t quite so. I could go on, and create a new preset in each of the synths for each wavetable, but I had already spent too much time with this already. Anyway, I think this is enough to give a rough idea of the similarities and differences between the synths. These are almost unnoticeable in the PPG wavetable (naturally) but are more noticeable in the ARR one, and even more in the TOMita one. I also have to say that Hive didn't perform as "synthetic" as I recalled from previous approaches ;-)

I leave to you the comments and judgments. I have my own, of course, but I will not go further right now, except to reinforce what I already said twice: There is no “right way” to do wavetable, and that’s why all wavetable synths are slightly different. This isn’t bad, IMO, quite the contrary.

First, the PPG wavetable in all the seven synths:

Falcon WT PPG

HALion WT PPG

Hive WT PPG

Icarus WT PPG

Nave WT PPG

New WT PPG

Serum WT PPG

Now the ARR wavetable in all the seven synths:

Falcon WT ARR

HALion WT ARR

Hive WT ARR

Icarus WT ARR

Nave WT ARR

New WT ARR

Serum WT ARR

And finally, the TOMita wavetable in all seven synths:

Falcon WT TOMita

HALion WT TOMita

Hive WT TOMita

Icarus WT TOMita

Nave WT TOMita

New WT TOMita

Serum WT TOMita
Last edited by fmr on Fri Nov 30, 2018 8:00 pm, edited 9 times in total.
Fernando (FMR)

Post

Hi fmr,
many of the wavetable synths you used offer various different interpolation methods (or sometimes no interpolation) too, could you maybe also say which ones you've used in each example?

Post

The unshushable Coktor wrote: Wed Nov 28, 2018 12:39 pm Hi fmr,
many of the wavetable synths you used offer various different interpolation methods (or sometimes no interpolation) too, could you maybe also say which ones you've used in each example?
Hi:

I always tried to use the default options. For example, in Hive, I used "crossfade". In HALion, I also used crossfade (10%). Mode of building wavetables is equal distance, with fixed length (but that is only meaningful for the resynthesis process). In Icarus, the oscillator mode is None/Partial (obviously), and in mono. Actually, in all synths that offered mono or stereo options to read the wavetable, I always chose mono. Also, for all that have Unison option, that option was turned off, of course. In Serum, the warp is set to 0. LFO (the modulator I chose to travel de WT), was set to a smooth of 50%. In Falcon, all options were set to 0 or their default options. In Nave, all modulations were set to 0 also. The only modulation used (in all of them) was the one needed to make the wavetable "travel" (except for amplitude envelopes, of course).

Only one oscillator was used. No duplication of wavetables. As I said, I did my best to keep the wavetable reading as neutral as I could.
Fernando (FMR)

Post

Are you sure that the "ARR wavetable" is an actual wavetable file? It sounds pretty broken in Hive.

Post

[EDIT: fmr fixed the ARR wavetable error for his Hive example, thanks!]
fmr wrote: Wed Nov 28, 2018 1:19 pm I always tried to use the default options.
Hi fmr,
thanks for the info.
Could you maybe post the wavetables that you used in your examples? Like Urs says, I also get the impression that the ARR wavetable sounds odd in Hive, as if something's wrong with loop points.
I'd like to try it myself if I may.

Cheers!
Last edited by Viktor [TUC] on Wed Nov 28, 2018 5:50 pm, edited 1 time in total.

Post

Sure, no problem. As I said, it is a "real" wavetable in the sense that it was created after the sample used for the famous ARR sound. So, it isn't a synthetic wavetable, but a resynthesized wavetable. It was done in Serum, if I remember well (not sure anymore).

EDIT: Correction: It was created in Audio-Term

EDIT 2: I was checking, and I may have used the wrong wavetable in Hive. I had to create a special one for Falcon, because, for some reason, Falcon didn't recognize the original wavetable created in Audio-Term. I may have copied that one into the Hive folder, instead of the original. I was checking, and with the Falcon wavetable, the preset sounds a little "grainy" (as it also sounds in Falcon) and I have to transpose down like two octaves to get the correct sound, while with the correct one it sounds cleaner and at the correct pitch.

Was that what you referred?

Link: https://1drv.ms/u/s!AlYyQJYLETH8iXHRYnLUcfgK0a2m
Fernando (FMR)

Post

fmr wrote: Wed Nov 28, 2018 3:57 pm I was checking, and I may have used the wrong wavetable in Hive.
Yes, thanks for that. Could you fix the Hive example in your above comparison post and use the correct wavetable please?
I tried it myself in Hive and it's clean with the correct wavetable.

Post

Ok, that's explains it. If you had to play it two octaves lower, it must have been a table with 512 samples per cycle. SO you#d actually play 4 cycles at once, which typically sounds very grainy. You can fix that in Hive by renaming the file to "ARR -WT512.wav". Or use the 2048 version you posted.

Post

The unshushable Coktor wrote: Wed Nov 28, 2018 4:26 pm
fmr wrote: Wed Nov 28, 2018 3:57 pm I was checking, and I may have used the wrong wavetable in Hive.
Yes, thanks for that. Could you fix the Hive example in your above comparison post and use the correct wavetable please?
I tried it myself in Hive and it's clean with the correct wavetable.
Taking care of that right now. Will correct the audio examples, and the links in the original post as soon as I finish it.

EDIT: All ARR recordings replaced, except for Nave, which wasn't necessary, and for HALion (will have to redo the HALion wavetable). In the meanwhile, I corrected the HALion wavetable, and published the correct recording for it also.

EDIT 2: All recordings redone, with the presets tweaked so they sound closer to each other in terms of the way wavetables are read.
Last edited by fmr on Fri Nov 30, 2018 9:21 pm, edited 4 times in total.
Fernando (FMR)

Post

fmr wrote: Wed Nov 28, 2018 1:19 pm
The unshushable Coktor wrote: Wed Nov 28, 2018 12:39 pm Hi fmr,
many of the wavetable synths you used offer various different interpolation methods (or sometimes no interpolation) too, could you maybe also say which ones you've used in each example?
Hi:

I always tried to use the default options. For example, in Hive, I used "crossfade". In HALion, I also used crossfade (10%). Mode of building wavetables is equal distance, with fixed length (but that is only meaningful for the resynthesis process). In Icarus, the oscillator mode is None/Partial (obviously), and in mono. Actually, in all synths that offered mono or stereo options to read the wavetable, I always chose mono. Also, for all that have Unison option, that option was turned off, of course. In Serum, the warp is set to 0. LFO (the modulator I chose to travel de WT), was set to a smooth of 50%. In Falcon, all options were set to 0 or their default options. In Nave, all modulations were set to 0 also. The only modulation used (in all of them) was the one needed to make the wavetable "travel" (except for amplitude envelopes, of course).

Only one oscillator was used. No duplication of wavetables. As I said, I did my best to keep the wavetable reading as neutral as I could.
Icarus has also the fade parameter which blends frames (or not)

Post

pdxindy wrote: Wed Nov 28, 2018 4:49 pm Icarus has also the fade parameter which blends frames (or not)
True. That is by default at 50, and, faithful to the principle of using the default parameters, I left it at the default value. The same way, I used the "crossfade" option in Hive.
Fernando (FMR)

Post

fmr wrote: Wed Nov 28, 2018 5:14 pm
pdxindy wrote: Wed Nov 28, 2018 4:49 pm Icarus has also the fade parameter which blends frames (or not)
True. That is by default at 50, and, faithful to the principle of using the default parameters, I left it at the default value. The same way, I used the "crossfade" option in Hive.
A fine choice... Serum is the same as Icarus fade at 0 as Serum does not do any blending between frames.

Post

where do we get hive with wt in the middle?

Post

Nowhere yet.

Post

OK, I decided to make a second go with my comparative recordings. I wasn't satisfied with the way the wavetables translated into HALion, therefore I decided I would have to go under an extensive study, to try to devise the correct procedure to have them faithfully translated. I even translated two full banks of wavetables (one from the PPG, and the other from the very first Reaktor library) which had settings completely different from the standard 2048 waves followed by Serum. I finally understood how things should be done, and managed to translate them accurately.

Then I revised the sequence used for the test, and also the presets in each of the instruments, trying that ALL wavetables would play more or less at the same speed. The job is finished now, and all the previous links were updated with the latest recordings. I think now we can have a better term of comparison.

The links are in previous page (page 23): viewtopic.php?f=31&t=511390&start=338
Fernando (FMR)

Post Reply

Return to “u-he”