Protoverb 1.0

[core]

Having a blast with this. I love randomizers

[Dandruff]

Ok, reproduced. Happens on looping.

Maybe bug in Reaper (submitting wrong buffer sizes in some circumstances)...

See this thread http://forum.cockos.com/showthread.php?t=33284


Setting Protoverb to "Buggy plugin compatibility mode" fixes it. But still might give clicks on loop.

[raring]

I freaking love this.. And it really loves percussion.. One button instant psy/techy tails <3 Where do I send my money as pre-order?

This one is going to be free forever. Whether this research goes into a dedicated plug-in or becomes part of one (or both), we can't say yet - it's too early to do so. There's the idea of including a surround version in Uhbik-A though.

This is a great idea! A.s.a.p. please!

[Beamboom]

Hmmm - do I sense the contours of the effect section of... The U-He drum machine?!

I think perhaps I do. Or maybe it's wishful thinking? But the algorithms at work here seem to be a perfect match with drums and percs... So... Yeah, there's that.

[Chris-S]

Ok, reproduced. Happens on looping.

Maybe bug in Reaper (submitting wrong buffer sizes in some circumstances)...

See this thread http://forum.cockos.com/showthread.php?t=33284

Setting Protoverb to "Buggy plugin compatibility mode" fixes it. But still might give clicks on loop.

Ok, I understand, so it's not a bug but a design. To get sample accurate looping some DAWs (Reaper) use odd sized buffers. Always wondered what this Reaper option (buggy plug compat mode) is good for.

[sascha]

Here's the code: BFNR157E-8FTJEO1N

Thanks for sharing!
Actually an intriguing one, smooth and versatile. Quickly made a preset here, in case you haven't sumbitted it yet (which you should do ).

[e@rs]

CPU usages on my C2D machine:

Protoverb: 21% ;
Aether: 22%;
Breeze: 12%, 8% with Modulation off, 4% on Lo;
Valhalla all: 4-6%;
Echobode with reverb presets: 3-4%.

Why, oh why

[sascha]

Mainly for 2 reasons:

1. Protoverb isn't a finished product. It isn't even a 'product' after all, just a conceptional work to start something with once we are more certain where to go from there. That's why we asked for public assistance and contribution.
2. It uses real huge matrixes of delays. As opposed to 'classic' reverb approaches, it uses no allpass filters for smearing impulses and to achieve density through an otherwise sparse-delay recirculating tank. This is usually the case with vintage-style approaches (kind of Lexicon-in-a-cloud sound, for lack of a better description), where processing cycles and memory where costly back in the days. There's nothing wrong with allpasses, but it's got a specific sound and limitation that we weren't after (been there, done that). Instead, we concentrated on what goes on in real rooms, mainly typical modes/resonance buildup and the tonal changes that comes with that. It's a completely different approach.

Take it as a 'proof of concept'. Not the end of the story.

[e@rs]

Ok, thanx for explanation. I wanted to use it in my projects, but it's impossible at this stage.

So... I take it as the beginning of the story then.

[Urs]

So... I take it as the beginning of the story then.

Yes...

Here's an example or two: Protoverb uses an average of more than 50 delay lines, each of a maximum length of a bit less than second or so. Once we know exactly how long the delays need to be (and how many...), we can constrain the lengths to their actual values and probably reduce the RAM to 30% or less of what it is now. Likewise, the calculation of matrixes can be optimised greatly once we decide for one or two.

There's a good chance that a final implementation (in whichever form) can be vastly faster.

However, there's a also a chance that we might make it more CPU intensive, e.g. once we add HF damping or an additional model for early reflections. In that case however we might also add multicore support because with this kind of reverb that's absolutely feasable.

On a positive note regarding CPU: This kind of reverb lends itself to be used on mixes. So instead of layering and moulding rooms for each instrument or group like you would do, say, with a Bricasti or Lexicon, you can get away with a single room simulator to do the job - any maybe save some CPU that way.

[modumo]

PROTOVERB = Very Promissing Concept (sonic vise) ,
I found out that the used characters for the Text-STRING- on MODEL/DELAY (specially on "MODEL") are limited, it would be nice (if this is in your intension) to have beside the Rnd generater a wheel type selector (like a number wheel key on a suitecase) for each Position with the Characters available for this Position (POS1 = A,B,C,D,E,F, than POS2 = G,H,I,T, than POS3 = M,N,L,U, than POS4 = R (only), POS5 = 1,2,Z and etc........
This would allow a more logical "hands on" rather than the random aproach.
I'M using Altiverb, MConvolver, B2 from 2 C Audio, and all the Room Plugs in CUBASE and SAMPLTUDE.

Non of them provides the "sonic acceptense" - as I call it - as PROTOVERB does!!
Yes it consumes CPU power - but there is always a work around - freeze it!! For a short latency use it as a Send Effekt rather than a channel Insert Effekt!

Happy new year, and thanks a lot for sharing PROTOVERB with us! - If you need further input (conceptional etc....) I would be happy to support you (german citizen)

[Constructed Identity]

Now that all the praise for a free effect has been gotten through I have to say that I've had a really hard time using this anywhere.
I think I have found a lot of really bad settings (that I haven't reported) and one good one (that I did report).
Perhaps it's the seeming random nature of it, I'm used to dialing in the exact sound I want after using presets to get in the ball park, sorry for the American baseball reference...
I see a lot of people on here using it for drums, but it seems the most unrealistic for percussion in any setting I've been in.
The one nice setting I found was a big space I think although I noted that I didn't think it matched a real space.
I'll keep messing with it.

[u-u-u]

AGNRZ38S-O56SS
- at small to medium decay settings, only some wet signal, blend in the dry
- really good localization (try panning the input audio signal)
EDIT:
also CGUR23QS-EM5AA1
- works really well on different settings
- only a bit assymmetric
- also sounds very realistic in my ears

[pinso]

Hello,
are there short random distortion/clicks built in the background?

[Urs]

Hello,
are there short random distortion/clicks built in the background?

No, there shouldn't be. We heard that some settings may cause artifacts, but those aren't intentional.

If for some reason you have "demo crackles" (which is part of our copy protection in commercial software), please let us know - this would be a bug then and we'll fix it!