Hive update plan

[Urs]

i think i understand where you are coming from Bdeep. you want it to be more like an instrument and less like a computer program, if i understand you correctly. i feel the same way. it will be very disappointing to me if hive becomes a conduit for selling presets. but, TBFair that was its original intent. maybe we are hoping it would be something it wasn't intended to be.

Have you tried the DX piano in the wavetable example files? Regardless of the cheese, would you not say that the playability we're adding is an example for the opposite of what you just said?

Hive is something it wasn't intended to be.

[Dasheesh]

well, ace does a pretty good fm piano. i am in wait and see mode. i'm not interested in scripting and i don't want to rely on others to do it for me. i feel like i have lost control at this point.

[Urs]

well, ace does a pretty good fm piano. i am in wait and see mode. i'm not interested in scripting and i don't want to rely on others to do it for me. i feel like i have lost control at this point.

Sigh. More waveforms was the top feature request for Hive form the beginning. What difference does it make to you if the code that generates those waveforms is tucked away in the oscillator binary, or in a text file on the computer?

Would you also consider dropping Hive if, instead of loading wavetables from the harddrive, we'd just open a large menu with a fixed set of waveforms, like e.g. Massive has?

I'm sorry. I fail to see the point how we somehow "broke" something. Hive's strength is its fast workflow. I fail to see how having waveforms on the harddrive and making its set extensible is of any disadvantage in that respect.

It just does not sound reasonable to me.

[Dasheesh]

yeah, i see that. change is hard. i just want to feel like i did something i guess. like it's somehow mine. some ownership. that's inspiring to me. i know you guys have been playing with a pack of waves that were being passed around. i haven't even looked at them. maybe i'll give it a go today.

[drzhnn]

i'm not interested in scripting and i don't want to rely on others to do it for me. i feel like i have lost control at this point.

And here's where the almighty Generate Random Waveform button comes into play!

Seriously, a simple hardcoded script in form of

Code: Select all

NumFrames = 1
Wave "lowpass(1-2*rand, 0.3, 0.1)"

that would use a different seed value each time user presses a button.

Doesn't require scripting -
Accessible -
User in control -
-

[chk071]

I must admit that I see a point in Hive becoming something it wasn't intended to be, but, in another way. It's accessibility and ease of use was its strong point, and what made it similar to Sylenth1. With the scripting, it rather had become something for ubernerds. I mean, I don't know any other synth which goes that far... even Reaktor is simpler, unless you fiddle with the Core stuff. Sure, you don't have to use it, but... it really seems like a strange child now. I hope that u-he has some ideas to make it more simple again, for version 2.

[Dasheesh]

You know, I really had not been paying attention since the scripting was installed. I just marked it as a "work in progress" to revisit later, so forgive me for trying to play catch up. I'm going to think out loud here while i reinstall hive to play with today. What we have now is a philosophical difference. It used to be players and sound designers, but now there is a new generation to deal with. This is a different world. They were data based before they took their first breath and had an iphone in their hands before they could walk. This is the sylenth generation. Grew up on (what ever label they choose to put on their music). These cats are programmers. That's what I call them. they do things different. They hear different, they see different. To me... scripting is eye watering, straining work. That's what I pay someone else to do. These cats love it though. They take right to it. I think Urs is doing a real good job of seeing that, and offering them something. Hive is advanced design, made for work on computers. What you've done is isolate your other populous though. People that have no interest in scripting. You've elevated some of your end users, while isolating others. I'm going to give it a go. I'm going to try it. My way of working is to have a knob per function and realize the outcome of the control though. I want to know what the hell I'm doing. I feel isolated and I think others feel the same.

SO WHAT? what Urs and the U-he crowd needs to realize is that you cannot please or satisfy everyone with every release. I know that's in your training as a graphical designer. You want to draw in the biggest audience you can. You can't. Deal with it. There are other things for me to play with if I don't like how it's going down. I LOVE your modulars. In the end, I may love what Hive becomes though... it's not finished. With that, I'm going to go see what you guys have been doing while I was gone.

[Urs]

Each .uhm script can be turned into a .wav file.

Are you afraid of .wav files?

And if there were only .wav files as wavetables, would you feel the need to open each and any in a sample editor in order to make your Hive patches be "yours"?

#-----------

Change.

My feeling is that both BDeep and your criticism is based on assumptions. What's out there is a preview, not a final version.

There was always gonna come a new user interface. One that moves away from spaceship looks. One that has "colour displays" and painted sections rather than a monochromatic look. As I said in the opening post, this is gonna come next year. It has so many improvements in the workflow, it's crazy.

The feature additions have been selected very carefully. I'm happy to lay out why I think that wavetable synthesis is a good match to Hive's concept (fast workflow, low CPU). Let me say what we *almost* did. We almost did Vector Synthesis. That is, we would have a fixed set to 100 or so waveforms, one of which you can assign to 4 corners of a "joystick" each. This concept would have required us to add 2D envelopes. It would have required us to tab the UI a lot more than we do now, and it would have contained logical pitfalls - the 2D envelopes would have to be active even if Vector Synthesis was not. This really would have hurt Hive's strengths.

Now, look at Wavetable Synthesis. There's synths that have a fixed set of built-in wavetables and there's synths which let you create your own. Former, like Massive, have a really good "best of" set of wavetables with nice and seamless sweeps. But they are not extensible. Latter have editors and/or sample import, but your own efforts will hardly ever match the quality of the seamless sweeps available in former. There'll be happy accidents, but good quality is hardly ever achieved without a very tedious workflow, or external editors.

My stance is: Hive does not need an editor or sample import to have a) great "best of" wavetables and b) be extensible. To the contrary. By avoiding an editor we can keep Hive "as is". No change in concept. Low CPU, fast workflow. Just a lot more waveforms. Waveforms which cover Sync sounds, classic FM and anything you'd ever get out of Vector Synthesis. Just like that, no tabbed away super-enevlopes needed. All people ever asked for, in a very simple and *accessible* package.

If anyone feels like adding hundreds or thousands wavetables available for Serum, so be it. But why would one? - We will curate a set of factory wavetables with the same care as we curate the feature additions.

#----

OTOH.

Hive is - in my opinion - highly underrated. Large parts of our user base think it's "too simple". What they don't see is that Hive offers a very comparable amount of modularity as ACE. Both are "two-of-each-synths". Instead of cables, Hive has the Mod Matrix. While ACE excels at audio rate modulation, Hive excels at modulation flexibility. Hive has a lot more modulation options than ACE. Hive does have fewer timbres to modulate though, and this is what the wavetables will change.

We need to reach larger parts of our existing user base with Hive, or Hive will eventually become a second class citizen in our portfolio. We need to reach a critical mass of synthesis options to appeal to folks who rather pull out Bazille or even Diva.

How do we achieve that? - I think the answer lies in a set of very unexpected moves. The first move was scriptable wavetables. I think we'll have plenty of examples and plenty of depth to explore for anyone. Maybe there are currently only three or four people exploring it, but once more examples pour in, I think it'll take off. I think many people will get Hive just for that, in order to create (or edit) high quality wavetables for any synth.

The second unexpected move relates to my experience with modular synthesizers. We are going to look at Buchla and West Coast Synthesis. Most people might not know this, but the Buchla approach to synthesis has a lot to do with playfulness and gamification. West Coast modules are often utterly simple, yet cover a very broad range of possibilities. They invite experimentation. What better could fit Hive's concept as a fast workflow synth than a small set of super simple, super flexible modules which can interact and bring a lot of power to the table?

One can decide to be intimidated or intrigued. I recommend latter. It's going to be a hell of a lot of fun.

[hakey]

I hope the whingers don't get their way.

The addition of wavetables, scripting included, is bloody great. It's an elegant way to hugely increase Hive's range for only a minimal increase in complexity. It's a couple of extra buttons - which no one has to press.

Scripting (which is a black art, to me) happens outside of Hive. Again, nobody is forced to script their own wavetables.

Those who find this all a bit too exciting could just stick to Hive 1.0?

[Dasheesh]

*Change

You might be right Urs. I didn't realize I could combine waves to create new ones without ever touching the scripting side of things. This is pretty neat.

So wait a minute, there is an interface function for most all the scripting functions? All the scripting side does is let you "write" up new tables, but it's unnecessary. You can take simple waves and combine them to become complex waves. Yeah, I had the wrong impression.

[Holmes2222]

..

[EvilDragon]

So wait a minute, there is an interface function for most all the scripting functions?

No. Hive doesn't have anything to do with editing scripts. They are just like a WAV wavetable to Hive. You write scripts in a regular text editor, if you're so inclined. But if you're not, you load it just like any other waveform for the oscillator, and OFF YOU GO TO THE OUTER SPACE.

[Delta Sign]

While I don't agree, I can understand the argument being made by BDeep. Sure, one can just ignore the deeper features and just enjoy the new wavetables, but that's not how it always works.

As another example, I always instantly lose interest in a synth once it implements sampling. Why? I could just ignore it, but it being there will always leave me with the feeling of not having pushed the synth to its limits. Sampling does that to my perception because I'm always thinking "well, I still could have used a better sample". With Diva or ACE, for example, I know when I "finished" a sound, because I know that's as far as I could have pushed the synth.

I can totally see how one might feel similar about the scripting in Hive.

It's not a rational argument, but a completely subjective one that doesn't necessarily make a whole lot of sense in a logical way.

And just to be clear, I personally love the wavetable scripting

[Funkybot's Evil Twin]

And just to be clear, I personally love the wavetable scripting

I'd hope so! You and drzhnn seem to be the two people taking most advantage of that feature outside of U-he HQ!

[Urs]

Originally we just wanted to add "load any Serum wavetable you find on the interwebs". We thought this was a brilliant move - until we listened to them. Most of it was ugh.

It became immediately clear that something like Serum's built-in function parser creates much better results, but was too limited to do all we wanted (we were originally going to build our own wavetables in Serum). So we worked out a simple (!) waveform editing and generation language which combines the possibilities of Serum's function parser with photoshop-like qualities to stitch, combine and layer wavetables. We will bring this language to other products as well, such as Zebra. It's going to be a core technology. It does things one can not do with a graphical user interface.

Other companies have devised similar tools for in-house use, e.g. Waldorf or Access. They do not give users access to those tools. We do. I don't see what there is to complain about that.

I can not believe we're having this discussion.