Any opinions on Quantec's statements about Reverbs

[Arrested Developer]

Lately i came across this side:

http://www.quantec.com/?id=room_simulation

According to the statements on this site, all competitors on the area of reverbs just don't get it.
Judging by my ears i'm not really convinced about the whole thing.
But i'd like to hear the opinions of some experts on this area...

[MarlaPodolski]

It's an interesting argument made on that site, but it's a rather old one, circa 2012 at latest apparently. So, do we see the smarter pros hopping on this ship? Nope, guess not. I appreciate your post, because it's the first I have seen this, although I do think I heard the name somewhere before.

I must say, the audio/music world is full of artifacts, as is most reverb. This is why we have nice software such as iZotope's RX 3, 4 available. So I guess I don't quite 'buy' the argument, and some of the rest I read. Neither, by the way, do MY ears.

Overall, it reads like the chap is just trying to carve out a niche in a very crowded market. I really never responded well to this sort of approach, this 'our competitors don't do it right, like we do'.

Thanks for posting.

[Phorous]

So it seems to me their claim is they make the closest thing to a "real life" reverb.

Personally, I think reverb should sound better than it actually does in real life... I suppose that could make it "less natural" sounding or too perfect, however.

[aboukirev]

As far as I understand it, they are talking about algorithmic reverbs. Convolution reverbs do get it right already. If you want realism, convolution is the way to go IMO. Algorithmic is for reverb as creative effect.

[Shy]

According to the statements on this site, all competitors on the area of reverbs just don't get it.

Almost all of them don't get it. For example reverb developers still modulate delay lines to try to overcome the inherent limitations in their algorithm and achieve what they think is more "realism" and fine sound. I think the truth is that if there is any signal with which the reverb sounds like an "effect" and not true reverberation, then it's flawed (as a "reverb", at least). Try them with piano, for example, to hear how flawed almost all of them are.

The only way to get a reverb to sound very realistic AND very spacious AND to not have unnatural sounding and annoying artifacts, is to design an algorithm from the ground up to use NO pitch modulation whatsoever. It was done over 30 years ago (by Quantec among few others) and a few of those reverbs still sound better and more realistic than all available reverbs which use modulation, and that includes M7 (which has modulation that causes nausea to me). I'd say that the only reverbs that use modulation which sound (in most cases) about as good as a good reverb that doesn't use modulation, are 224 and EMT 250.

[Ivan_C]

For me, reverb effects are not meant at all to be realistic in most of the use / pratical cases. And either if someone wants a really realistic reverb for a good reason, he can use an impulse response, or record the damn instrument into a place where the natural reverb is the one he wants. Which doesn't mean I don't care about attempts to design realistic algorithms, that's interesting and certainly useful.

But what really matters in my opinion is the sound of the reverb in the mix, or in the live / sound design situation. I cannot care less about the "realistic or not" bullshit. I mean, in the mix context I like to have a lot of reverbs with different sound signatures, and all of them will be ideal for some instruments or contexts and not others. My options would be really really limited and boring if I had only realistic reverbs to use !

Another question : how can we appreciate the "realism" of a plate or spring reverb ? Knowing that we can get that with physical modeling, impulse responses, and delay line based algorithms ?

Moreover, I don't get why modulation should be bad in reverbs, it was a way to enhance the character of some early reverbs with limited dsp power and memory, and some people still like that nowadays. It's not bad, it's juste a matter of tastes, and you can disable it in most of modern plug-ins...

Anyway, I will never buy a product from a company which uses such an argument to sell software. Because it's too easy to launch something by saying "other don't get it", and also because in my opinion their argument displays how much THEY don't get what a producer really needs in a mixing / sound design context. I think most of serious reverb developers don't care at all about the fact their algorithms are realistic or not, and they are right.

[EvilDragon]

Interesting proposition.

Eagerly awaiting Sean's reply.

[Urs]

Realism of reverbs is a promise that's a tad flawed. Using a perfect simulation of a cathedral's reverbration will not relicate that exact sound on the listener's home stereo (or a club). Every reverb will only ever sound "like a recording of this or that space" and it'll be ultimately altered by the listening environment.

Algorithmic reverbs can use that to their advantage. They can for instance be made to create the impression of sound coming from outside the stereo field. This is where afaik Quantec is unbeaten and thus if that's a major aspect the guys might be right.

While they're good at creating rooms, I find they left a lot of room to improve their marketing, webdesign and wording.

[Ivan_C]

While they're good at creating rooms, I find they left a lot of room to improve their marketing, webdesign and wording.

Well that's the thing I reacted against...

Anyway, they should really change the way they try to sell their products if they are that good in their area of expertise, because there is some room (...) for realistic reverbs algorithms in broadcasting or classical music recording for example.

"Realistic reverb" might be a category on its own in all the reverb algorithms, which can be useful the same way than other ones in a given context, but not something the reverb developers try to do without success as much as Quantec in general...

[Shy]

Quantec's reverbs/"room simulators" are far from being limited to being "realistic", they can be tweaked using very "unrealistic" parameter values that don't resemble any real room's, and I think they still sound more spacious, more "reverby" and simply better overall than "unrealistic" reverbs that use modulation. An effect that sounds LIKE A REVERB with any signal and with "real-world" as well as "hyper-real" parameters, that's what any reverb developer dreams to achieve, and almost everyone fails.

[Daags]

Realism of reverbs is a promise that's a tad flawed. Using a perfect simulation of a cathedral's reverbration will not relicate that exact sound on the listener's home stereo (or a club). Every reverb will only ever sound "like a recording of this or that space" and it'll be ultimately altered by the listening environment.

that's very true of course... but then, for the sake of argument, maybe we should judge these things based on the usage of decent headphones.
at least in conjunction with the more common stereo hi-fi setting...?

Algorithmic reverbs can use that to their advantage. They can for instance be made to create the impression of sound coming from outside the stereo field. This is where afaik Quantec is unbeaten and thus if that's a major aspect the guys might be right.

interesting...

[db3]

While they're good at creating rooms, I find they left a lot of room to improve their marketing, webdesign and wording.

Exactly what I was thinking

[Warp69]

Very (very) few people have extensive knowledge of emulation of real reverberation and truely great reverb design - you can count them on one hand. Wolfgang is one of them.

[valhallasound]

In this case, it is a good idea to separate reverb design skills from marketing & public relations skills. Quantecs are classics in their field, regardless of any hubris or (justified) pride that is expressed in their FAQ. The dude has been making amazing reverbs since 1981 or so.

A few of the FAQ entries seem factually incorrect to me:

- The idea that Quantecs can't be captured via impulse responses just doesn't square up with how LTI systems perform.

- The claims about the cycles available on the Bricasti don't seem accurate to me. Most of the Bricasti processing uses 32x16 bit multiplies, which are VERY fast on the Blackfin processors. Quantec's statements that this wouldn't work for their algorithm, on the other hand, should be accepted as fact.

Otherwise, most of the FAQ entries fall under the categorization of either "subjective" or "informative." The claims to the superiority of their techniques, and the merits of other techniques, are obviously subjective. The discussions of modal density are very informative, as are the comments on echo chambers and shorter delay paths.

Sean Costello

[Winstontaneous]

Entire genres of music have arisen from the artifacts and character of modulated algorithmic reverbs attempting to be realistic. Realism is an admirable design goal but not the only valid musical/creative path.