Some papers on simulating a spring reverb digitally

[caulixtla]

While I personally would prefer to see a “Valhalla Echo Room” (a reverb algorithm from the SPX900/SPX990 where width, height, and depth can be independently adjusted, as well as wall variance, and where it is possible to give the three axes different decay times; I don’t think Echo Room had modulated tails but I’m sure a Valhalla variant would), there are a lot of people out there clamoring for a “Valhalla Spring.”

That in mind, I was pleased to find out, last night, that there are a number of papers where algorithm designers talk about how one would best digitally simulate the “sproing” that spring reverbs are so well known for. For example:

http://recherche.ircam.fr/pub/dafx11/Papers/39_e.pdf

http://www.ness-music.eu/wp-content/upl ... dafx13.pdf

http://www.asp.eurasipjournals.com/cont ... 646134.pdf

http://dafx09.como.polimi.it/proceeding ... per_84.pdf

It looks like Sean won’t have to reinvent the wheel to figure out how to get that “sproing” going when and if he makes a “Valhalla Spring.” Personally, I would like to see a “number of springs” parameter as well as a “spring length” parameter, ideally one where the different springs have different lengths, but Sean is leaning towards giving his reverbs fewer, not more, parameters, which is what most of his users want, so “Valhalla Spring” will probably be “decay”, “mix”, “pre-delay”, “spring length”, “spring count” (or just have “one spring”, “two springs”, “three springs”, “three springs no sproing”, etc. algorithms), modulation speed, modulation depth, and maybe a tone knob.

[1] If I want more knobs, I can always figure out how to compile the source to Freeverb3 and play with the source code to the “Hibiki” algorithm (Teru’s take on the Zita-Rev1 reverb, with a modulated tail and different allpass magic numbers), “ProG” algorithm (Teru’s take on Dattorro’s proposed “Proginator” improvement to that “in the style of Griesinger” reverb, complete with modulated tails), or even figure out how to best modulate the “STRev” tail (yet another implementation of Dattorro’s publication of a digital reverb “in the style of Griesinger” which revealed what was previously a closely guarded trade secret).

[valhallasound]

[1] If I want more knobs, I can always figure out how to compile the source to Freeverb3 and play with the source code to the “Hibiki” algorithm (Teru’s take on the Zita-Rev1 reverb, with a modulated tail and different allpass magic numbers), “ProG” algorithm (Teru’s take on Dattorro’s proposed “Proginator” improvement to that “in the style of Griesinger” reverb, complete with modulated tails), or even figure out how to best modulate the “STRev” tail (yet another implementation of Dattorro’s publication of a digital reverb “in the style of Griesinger” which revealed what was previously a closely guarded trade secret).

The "Progenitor" reverb is, in fact, the 224XL Concert Hall. The modulation is in the existing diagram - it's the linear interpolation that occurs with nodes 2 and 3, and nodes 5 and 6.

[valhallasound]

Another spring reverb paper worth mentioning:

http://www.aes.org/e-lib/browse.cfm?elib=13788

Written by 2 guys from Universal Audio, 1 CCRMA professor, and some other guy.

[morty cowverb]