Reverb with BBD and PT2399?

[Morocotopo]

Well, just had this idea. The PT2399 can do reverb but not early reflections very well because the minimum delay time is too long. How about using a 3007 or similar to get the early echoes and the 2399 for the rest of the tail? It should be possible to get a smooth reverb, without the early discrete echoes that you can hear in all the Belton brick pedals.

Any one ever tried this?

[Mark Hammer]

I've never personally tried it, but started planning out a similar sort of thing a few years ago, using three delay chips, and creative use of feedback paths.

One of the things about combining a BBD and a PT2399 is that their respective clocking is in a fairly different range, so it should be technically possible to set them to different times, without fear of heterodyning.

And here is where it can get interesting.  Let's say the BBD is set to 25msec, and the PT2399 set for 60msec.  Let us also say that each of these devices is fed by a mixer, and that their respective outputs both go to an output mixer, and to a feedback path to one of those input mixrs.

If we feed the BBD and PT2399 from the master input, and feed their respective outputs to the output mixer, we will have a "reflection" at 25 and 60msec.  Since these are not multiples/divisors of each other, there will be no apparent resonance.

Okay, let's hike it up a notch and both feed the PT2399 output back to the input of the BBD and also feed the BBD's output to the PT2399.  The BBD+PT2399 output will be at 85msec.  The output of the BBD will be comprised of both what it produces in and of itself, PLUS the feedback from the PT2399.  So 85msec, plus an additional 25msec, yields a delay at 85+25 (110msec).

The bottom line is that using two unequal delays, and feeding a little bit of signal back to the beginning, can yield a wide range of reflections that are not multiples of each other, resulting in something much more like reverb, and much less like the traditional feeble attempt to produce reverb with a single delay and feedback.

[Morocotopo]

Never got around to making a schem Mark?

The Belton brick is nice, but I don´t like that you can hear the early echoes until eventually they become blurred and turn into reverb. The whole idea is to get, as you say, non multiple echoes that can produce a much nicer verb. And with a bit of modulation it can be nicer yet...

Regarding heterodyning, I don´t know if that could be a problem. i mean, there´s lots of pedals with more than one oscillator that work OK: Boss DC2, EH Polyphase... I have at the moment a 4 chip (!!) PT2399 delay in the breadboard and so far, no problem with that. I believe it can be controlled with careful filtering and PCB layout.

With a BBD and a PT, lots of possibilities for crossfeedback and for a non-periodic, more hall-like verb (or, less spring like is another way of putting it). I´ll eventually try the concept, some time in the future.

Was curious if somebody tried it before. I´m sure someone (besides you Mark) did, it´s too obvious an idea.

[Mark Hammer]

Reverb is a bit like distortion, in a way: there are millions of ways to do it, and millions of sounds to achieve.

If you did things the "simple" way - i.e., feed PT2399 both directly, and via the BBD, with a feed to the output from the BBD, and no feedback path of any kind - I'd suggest use of differential lowpass filtering of the different repeats.

So, if you had an output mixer receiving signal A from the BBD directly. B from the PT2399 directly, and C from the BBD-through-the-PT2399, you'd want A to have more treble than B, and B to have more than C.

That was always one of the weak links in older solid-state reverbs using the near-mythical multi-tapped MN3011.  They always applied the same filtering to ALL taps, and in the real world that's not how reverberation operates.

[dthurstan]

I have also been thinking about this recently, if I'll ever get round to doing anything about it is another thing  .

I managed to get a MN3011. Mark I read your comments on other posts about applying varying filtering to the taps, which is important in creating a realistic sound. I found a schematic to a reverse engineered brick which is three PT2399.

I see these 2 components (MN3011 & homemade brick) as a good way to create lots of delays. Mixing/feeding them into each other and adding modulation could create a great reverb.

The PT2399s in the brick use some of the ideas of digital design to create the reverb (comb filters, all pass) and I think there is some modulation in there using one of the LFPs?.

Jon Dattorro's papers Effect Design show how to create a plate reverb using comb and all pass filters. His description of the reverb algorithm is helpful for working out how they should fit together.

[Mark Hammer]

Has anyone here ever seen an actual MN3214?    I know I never have, nor have I ever seen a schematic of a commercial pedal using one.  You can see the datasheet here: http://www.experimentalistsanonymous.com/diy/Datasheets/MN3214.pdf

Although the 3011 was intended to be a standalone solid-state reverb solution, the 3214 (itself a kind of reverb simulator, albeit with fewer stages) will fill in, in a pinch.  But perhaps more usefully, it can be used to stagger a signal feeding several PT2399s.  And the nice thing is that the 2399s can all be synced to the same master clock, just as the 3214 (or 3011, if you have one) provides multiple delays, all from one clock.  And happily, the BBD and 2399 clocks are in different ranges, reducing the risk of audible heterodyning.

[12Bass]

Although the idea of mixing an MN3011 with a PT2399 sounds intriguing, I'm left with the feeling that it would result in a rather complex and challenging project which ultimately does not provide a very satisfactory reverb sound, especially when compared with inexpensive modern DSP solutions like the Spin FV-1.  That said, if it is just a matter of being up for the challenge of trying to make it work... it could be fun.  

Found this demo of a pedal which uses the MN3011: 

[Mark Hammer]

I second your recommendation of the FV-1.  Admittedly, such DSP chips deter many of us who think in terms of caps and gain stages, but it is a much better solution to what is ultimately a very complex task.

Thanks for allowing us to finally hear the FX45.  It doesn't sound awful, but it also doesn't seem to have a very broad range of personalities.

[MetalGuy]

+1 for the FV-1 or maybe +100.
Since we have this chip I can't even imagine why anyone would bother at all with BBDs and PTs?
If you need some vintage delay get those BBDs. If you need a simple delay get the PT but for reverb those are just waste of time.
In case you're that sensitive to reverb then maybe you should go for a standalone tube unit...