Shimmer Effect

[GiovannyS10]

Hi guys!
I am really enjoying the shimmer reverb sound that i found at this video below.

Know anyone any schematic for have a effect like this, or have a idea for a start point for a develop?  Thanks!

[robthequiet]

Well, we could try to break it down into sections.

The ability to loop a delay back into itself or maybe two delays working together should be possible.

A chorus effect adds sweetness, you may come close with a modulated delay.

To make the swells a volume pedal is a simple choice.

The tricky part is getting clean octaves, and the box from the video is a nicely designed DSP box. Not analog. So maybe the digital types can help you out there. But you can at least get started with a PT2399 type delay project with modulation, and a volume pedal, then you can probably do tricky things with a guitar to get pretty close to the orchestral effects.

[blackieNYC]

they've told us how. With those pedals. And I point this out because EH puts a heckuva lot of stuff in each of those as it is.   the reverb is probably processor-based - but that we can mimic with a belton brick or multiple PT2399 circuit.  Pretty well I would think.  search for reverbs in here somewhere.
Its that pitch shifter.  It's polyphonic, and you'll find no DIY schematic. sophisticated - but not as expensive as one might think.  I think that's one of a few pedals one should just buy. Or skip, if you want to be a DIY purist.

[slashandburn]

Sorry for the OT, but this reminds me of an outlandish I heard from Devin Townsend. The "sound" he used for a lot of his milder solo stuff stemmed from stacking delay and reverb pedals. Basically 3 pedals I think, two delays with a reverb in the middle.

I wasn't brave enough to give it a go. Just figured I'd mention it.

The way I see it, you want 6 pedals in one box. A delay at the start and the end, a reverb, octaver, chorus and tremolo in between.  At this stage I'd boldly suggest that the chorus is perhaps overkill and could be omitted. 

Ah I jest. Apologies for this rather unhelpful post.

[samhay]

http://www.diystompboxes.com/smfforum/index.php?topic=104291.0

[Mark Hammer]

I mentioned something I've been trying to find the time to finish boxing up here: http://www.diystompboxes.com/smfforum/index.php?topic=116024.0

One of the options it provides is to allow a post-delay reverb to be inserted into the feedback loop of the delay.  Now, I want to insert my Micro-Pog in the middle of that loop and call in sick that day.

[Rixen]

reminds me of Flux Effects 'Liquid Ambience'

https://www.youtube.com/watch?v=5ffCTTaKQGY

[robthequiet]

Question: How viable is it to build a multi-PT2399 pedal and clock the chip-paths so that you have the output of the first be base delay frequency, then the 2nd an octave (fx2) higher, then the next to be an octave above that (fx4), or even clock them at intervals less than an octave?

Some shimmer effects appear to be just this, then (with a bit of modulation in the delay side) sent to a deep reverb, at least to my ears.

So you really don't have to wrestle with clean octaves, by using multiple clock speeds on 2 or three channels. Maybe even recycling the output back through to achieve way higher frequencies? Surely this has been asked or done before, I think.

Wondering if the chip has the capability of being clocked at double speeds. The datasheet suggests delay times of 75 to 300 Msec being viable. Could this be one approach?

Edit: I think I confused myself between delay time in msec vs. taking audio and somehow accelerating it by replaying it faster. Violates laws of physics, duh. So I was looking for a way to replicate the "moving tape head" concept. Still somehow thinking that one chip's audio can be sent to another that takes chunks and plays them back at a higher sample rate, but how?

Edit 2: Could a delay line be modulated not by a pot but by a square wave LFO? I'll be quiet now.

[notnews32]

I could be wrong about this, but I don't believe it's possible to clock a PT2399. There may be another chip in the PT- family - I've seen other, more complicated delays built around other PT- ICs - that can be clocked like a BBD IC.

It is possible to wire the 2399's in series, such that delay lines are multiples of each other a la echoplex "taps". This is different from clocking a chip, though.

again, I could be mistaken.. if so I apologize, but I'm fairly sure.

[slacker]

Clocking the delays faster just gives you a shorter delay time it doesn't change the pitch. To change the pitch you need to send audio into the delay at one clock speed and then play it back at another, eg: playing it back at twice the clock would give an octave up. Using square wave modulation gives a pitch that jumps up and down rather than just a higher pitch. Using multiple delays modulated like this and switching between them at the right times could probably make a crude pitch shifter but it would be bloody complicated.  Much easier to use something like an FV1 as has slready been suggested and learning a few new skills.

[ElectricDruid]

Edit: I think I confused myself between delay time in msec vs. taking audio and somehow accelerating it by replaying it faster. Violates laws of physics, duh. So I was looking for a way to replicate the "moving tape head" concept. Still somehow thinking that one chip's audio can be sent to another that takes chunks and plays them back at a higher sample rate, but how?

This is how some digital pitch shifters actually work. They play back short sections of the audio at twice the input rate to give a octave-up effect. The digital part is required to decide *which* short sections and to do all the cross-fading and so forth required to hide the joins as much as possible (which wasn't much on the early units, to be honest).

Doing similar tricks with analog is perhaps theoretically possible, but not worth the candle. If I was having a go, I'd be looking at a simple and diy-friendly digital chip like the Spin FV-1, for which they handily provide pitch shift algorithms as a starting point. If you put one (or more) of those in the feedback loop of a delay and stuck some chorus on the end (or in the loop?) you might be getting pretty close.

HTH,
Tom

[snap]

Here`s an analog DIY approach: http://www.diystompboxes.com/smfforum/index.php?topic=35335.0

[GiovannyS10]

Hello guys.
First of all thank you a lot for answer and i am sorry for not reply. I have been working a lot last week and was sick too  But now i am back 

I read all posts, too much good ideas and discussion. I was thinking in do something with two or three 2399, but i was looking for if someone did it before.
I thought in do a thing like robthequiet said. Build a pedal 'all in one'. With a chorus, delay, reverb and octave together and see what we can do with it  But will be better develop a new schematic only for it.

Pt2399 clocks... maybe. I think the way will be have a delay pedal with two ships to delay normally but individually, the first one will delay like all delays do, the second too, but this one will have a pitch after and after all a reverb. All this in one circuit. Maybe a chorus before the verb, i am not sure, but i think i will develop something like this.

I really liked the FV-1 effect. I will study this schematic, but i think i will develop my own shimmer verb

[Rixen]

it is possible to read the clock frequency from a PT2399, so in theory you could use a microcontroller to model response and implement a feedback loop to accurately control delay time.. there are simpler solutions out there, but could be an interesting and useful project...

[robthequiet]

Did a little reading, and found the Shimmer effect attributed to U2 and Brian Eno using rackmount equipment. Link here

Other sources may vary with this, but the basic idea seems to be a large scale reverb with an octave shift and a lot of feedback. Fitting an Eventide harmonizer into a stompbox is a thing by now, of course.

I think that the FV-1 and more uProcessor/code-based devices have opened a lot of options for the diy community. The FV-1 app notes suggest that you can run the pitch shift on the left channel and the reverb on the right and get the effect with one chip. Cool stuff. Need to have the coding/programming skills, of course.  The humble PT2399 remains a pretty viable solution for the handmaker, albeit at a lower resolution.

I have a concept in mind using a dual BBD path with inverse phase shifters (running 90 or 180 degrees out of phase) that are mixed together with a lot of repeats, thinking that the phase difference at a high-Q value will highlight upper harmonics and possibly cancel some fundamental. Probably easier than trying to play with clock speeds to get a perfect octave. If anyone has a pointer to some such project please reply, thx. Sorry if I'm hogging the thread.

[Electron Tornado]

The Strymon Blue Sky reverberator pedal does some very nice shimmery stuff. Lots of DSP going on in these gizmos:

https://www.youtube.com/watch?v=b_qExV3qYr0

[slacker]

The FV-1 app notes suggest that you can run the pitch shift on the left channel and the reverb on the right and get the effect with one chip. Cool stuff. Need to have the coding/programming skills, of course. 

You can possibly do some sort of shimmer effect without doing any coding, one of the FV-1's built in programs has a delay on one channel and a pitch shifter on the other. I reckon if you ran the output of the delay through the pitch shifter maybe with some analogue feedback and mixing you could get something shimmery.
I might have a go.