@acreil said:

I've attached the reverb too. It implements the Lexicon 224 concert hall algorithm (with some modifications and enhancements), as described here: http://www.convexoptimization.com/wikimization/index.php/Dattorro_Convex_Optimization_of_a_Reverberator

Oh my gosh.. That's one of the highest quality reverbs I've heard (surely in pd). What do you mean by "Lexicon 224 concert hall algorithm"? Is it one "preset" of lexicon 224? Or was it a specific device implementing it?

I'm curious about the way you implemented the many elements shown in the Datorro diagram (allpass, delays) and the feedback lines.
Question: can we use your patch (giving proper credits, of course?)

@acreil said:

I've posted it here before. I need to make it more intuitive to use, and permit external control, but I haven't gotten around to it. I can use it well enough, but I think others will find it confusing.

I completely missed that one. Yes, giving high level controls would be nice.
Lexicon 224 was a terrific device (I just remember the Blade Runner intro by Vangelis, the huge spacey sounds).

Thanks for sharing it.

Alberto

The 224 had a pretty large number of different algorithms. These weren't just presets, they were different topologies. Dattorro described another (or at least something similar) in his well known AES paper.

The thing is, I don't know how the user adjustable parameters of the 224 mapped to the actual coefficients, etc. I've changed a few things to make it more to my liking. But for this topology I think there's not even a way to derive an accurate RT60 control...

Anyway, feel free to use it.

Hey acreil, I just wante to say that your works have really inspired me to get into pd more.
Also thanks for the reverb, I have a feeling I'll be using it a lot.

Acreil - thanks SO much for that reverb abstraction. Amazing work. I came across some youtube videos of yours, and an extensive and interesting gearslutz verb design thread, in my hunt for a seriously good reverb algorithm for pd - and finally I found this thread with your patch. I'm excited to see what you come up with next - do you have a website where you post your code? Or a github account?

I can't state enough how great it is to finally have good algo verb in pd.

I should have better reverbs soon-ish. They should be more usable. I really need to get a blog together for this stuff, but I haven't done it yet...

Awesome! Looking forward to seeing what you create!

In the meantime - could you extend a few useful pointers for getting control over this path? Is there a handful (list?) of parameters that I could globally control to effect decay time?

Also I see a ton of $1 objects but they aren't being found - ex:
error: 0-prl: no such object
error: 0-prr: no such object
error: 0-gscale: no such object
error: 0-ldiff2: no such object
error: 0-ldiff2o: no such object
error: 0-ldiff1: no such object

Though I may not have the latest version...

Thanks!

I think you're getting the errors after clicking the message box, right? They're actually local $0 objects, but the message box interprets $ as referring to incoming list elements. So you should click the bang instead, which sends [f $0] to the message box.

This wasn't really designed with good coding practices or usability in mind, so these message boxes should probably be hidden away; I just made everything visible so the "preset" parameters can be edited manually. The new one I'm working on has an inlet that can be used to externally control parameters, but I was too lazy for that when I did this. I could add that without much trouble, but the parameter names aren't particularly well chosen...

Ahh gotcha - yeah that makes sense. Thanks!

Just a bump.. hoping for an updated/friendlier reverb patch from you!!

Having a good time with this one for sure.. but devising an RT60 control for it is alluding me.

I think in the hardware units, the RT60 parameter is just a sort of estimate. The decay time depends on both the allpass coefficients and delay lengths, so it's difficult to calculate.

I have a better reverb algorithm now, included here: http://acreil.wordpress.com/2013/04/20/algorithmic-composition-of-acid-house-with-lo-fi-timbres/

It's still not really intuitive (or documented), but control of RT60 should be better. And it sounds much better too.

Nice!

fucking awesome! although i see how difficult it is to control the compositional part of it but the structural part is stunning. the power of the euclidian algorhythm!