The "boring" chorus sound...

[drew]

So I was thinking about the recent thread on choruses... and how it does seem that they all go for the same thing. Some sound smoother, more or less metallic, or have slightly different qualities, but it still sounds like a chorus.

So is there a way to make a chorus sound NOT boring? The first thing that came to mind was driving the clock with a complex waveform. A wavefolder might be appropriate but I don't know if a discontinuous wave driving the chorus would cause artifacts. I figured it would, like spinning the knob on a delay unit and hearing the sound output slow to a halt or speed up into munchkin-land until it has cleared the buffer.

Well, what if there were a slew-rate limiter on this function generator, to keep it from "blipping"? From what I can tell, the pitch offset in a chorus at any moment in time is related to the slope of the clock modulator function... so if there's a large slope up or down (from a square wave, or a step function) the delay will "blip".

How about a digital "random walk", buffered by an integrator to prefent blipping, to give a more "random" feel to the chorus? (A random walk meaning a function generator that would move one small step up or down with every clock cycle - that may not be the exact definition but I'm no digital logic guy!) Like a slew-limited sample and hold, but the maximum change at any step would be small rather than possibly large.

Or how about using a voltage-controlled LFO to control the clock speed... and then using a second LFO to vary the speed of the first LFO? Or something similar with three, possibly feeding back into each other?

How about a Chua oscillator ( http://www.ci.ee.uct.ac.za/postgrads/ryorke/gallery/chua.html or http://mysite.mweb.co.za/residents/cyb00746/chaos/chaos.htm ) which is more or less non-periodic? How about other chaotic systems, normalized and used to drive the clock of a delay line?

Okay, enough out of me!


drew
toothpastefordinner.com

[Mark Hammer]

All the things you've mentioned are those things which will alter the feel of the pitch distortion that simple chorusses produce.  Essentially you're wondering about how to monkey around with the *vibrato* aspect of chorusses.

Of course the objective of a chorus pedal is not to sound like one instrument having its pitch screwed up (though that's not a bad thing in itself), but to sound like two or more instruments deftly playing the same thing, and almost exactly, but not quite, in tune.

The closest thing to this I have heard are the 2 and 3-phase choruses, such as the BOSS Dimension C and its bigger brothers, and the assorted multi-phase chorusses on analog string synths from the mid to late 70's.  My good buddy Mike Irwin is very fond of these and has cloned a few of them.  If you look at the schematics for these things, what you you notice is that the LFO is in fact VERY bland.  Indeed, on the Dimension C you don't even get to twiddle knobs.  It has 4 push buttons that provide the 4 combinations for each of two fixed rate and width settings.  And having sat and diddled with one for a while and pressed all the buttons, as much as it simply felt *wrong* not to have knobs, I didn't need them.  Those who know my passion for additional performance controls will find this surprising, as did I.  But the thing is that when you use 2 or more BBDs to produce multi-phase chorus, you start to notice how much all the knobs in regular single-BBD chorusses are there to compensate for its weaknesses.  Multi-phase units just sound thicker, have less wobble to them, and produce something that sounds much more like what we want a chorus to sound like.

If you want to explore some alternate types of sweep for a single-BBD unit, though, I suggest the following.  First, find some samples of the Pefftronics Randomatic Chorus (I think Tonefrenzy has some) to get a sense of what a random sweep would sound like.  Then take a peek at the scanned article from Elektor I have at my site - the S.E.W.A.R. (Sound Effects With Analog Reverb).  This is a self-contained multi-waveform LFO and clock driver board that will drive BBD's directly.  It includes a smooth random source in addition to several other more conventional waveforms, and uses the XR2206 and 2207.  I gather it was originally intended to provide chorus effects fromthe SAD1024-based "reverb" project that Elektor had published around 1978 or so.  You can find it at hammer.ampage.org or look for it under hammer.ampage.org/files.  As with many of the PDFs there, I can't vouch for the accuracy of the scale of the PCB mask.  You may have to resize, and you will certainly have to flip it around if you want to use it for a PnP board.

[drew]

Yes! This is exactly what I wanted, to learn more about the different ways you can make a chorus good. (I assumed that the not-quite-in-tune part was based on a mix of the modulated-delayed signal with a bit of the original sound mixed in... but if I am wrong please tell me!)

To be fair I don't have any vintage chorus units but I do have a Line 6 MM-4 modulation modeler which emulates the Dimension C, and it's a nice chorus sound. (This guy - http://www.oldcrows.net/~jhaible/dimension/jh_dimension_clone.html - has cloned the Dimension D, and he says it's less of an overt chorus effect and more of a "depth" thing...) Likewise, the emulation version of it seems nice even though it's not really controllable.

The three-phase thing is intriguing. The Songbird Tri-Chorus (or a similarly-named unit) seems to have used three chorus circuits, preset to different speeds, un-synchronized, with a knob for each to mix. I only know about how this may sound because it's another setting on the MM-4...

Have any references for a 3-phase oscillator? (Or does Mike I. have any info posted on the web?) I could probably dig up the schematic to some old string synths (Solina?) and try it out but sometimes that's easier said than done.

I've got your site bookmarked, from long ago, and I'll check again in a little bit... thank you for the information and ideas Mark! One parting thought on my way out- there are three-phase oscillators (120 deg. apart), quadrature oscillators (90 deg apart)..... are there five-phase oscillators (72 deg apart)? Would a five-phase chorus sound good? Maybe I should read up on multi-phase oscillators. Perhaps a three-out-of-five phase chorus (three choruses driven by three separate phases out of five.)


drew
toothpastefordinner.com

[rocket]

There's also the Ibanez Twin Cam Chorus (TC10).
It has 2 LFO, that can be seperatly controlled in depth (=amplitude) and speed.  The LFO outputs are mixed together to control  the (single) BBD.

sounds weird

[Carlos]

Hi!

If you think your chorus is boring, connect the dry output with one amp and the effect output with another amp.
I use a Roland CE-1 Chorus Ensemble with two amps, it's the best  chorus sound ever: lush, organic, because the chorus sound is created in "mid-air" between the amps.
To avoid ground loops one amp and the CE-1 are connected to mains via isolating transformers. It's worth it, though ...

Carlos

[RickL]

I'm pretty sure the TC10 has two BBD's. The 9 series version does for sure. I added a second LFO to one of the choruses I built (Small Clone maybe or a Zombie) and it worked quite well. I got curious if Ibanez had done it the same way in the 9 series dual chorus and when I looked inside I saw two delay chips.

I wonder if a Small Stone would work with two LFO's in parallel? I'll have to try adding one at some point.

[drew]

Will have to check out some of these chorus pedals. ESPECIALLY the pefftronics one! I went & listened to the sound sample, pretty impressive.

RickL: I have had some "stereo" phasers and flangers but I don't recall whether they were driven by out-of-phase or unsynchronized LFOs, or whether they just added the "wet" signal to the dry signal out of phase to the left and right channels. Most likely the latter, since that would be easier, but it's an interesting thought.

Ever used two (phasers, flangers, choruses) at the same time, hard panned to left and right? I have, but only in mixing stuff down... gives an interesting "stereo" effect, to be sure. Although, I don't have two of the same unit, so I processed a track (from tape) twice, onto two new tracks, with different rate controls on a flanger... it was not up front in the mix so the song didn't sound predominantly spacey or stereo-image-shifty either. Oh well.


drew
toothpastefordinner.com

[Ansil]

i used to have my favorite chorus pedal. it had two boss ce-2's in it. each slightly modded from the first.  but mostly i only used the depth function off of eachone,   to get a really deep sound.

[Mark Hammer]

One needs to distinguish between:

a) a single BBD with a less periodic sweep (e.g., random source or 2+ unsynced LFOs)

b) two BBDs in series with different periodic sweeps

c) two BBDs in parallel with different unsynced sweeps

d) multiple parallel BBDs with strategically synced sweeps

The Dimension C uses complementary sweeps, derived by feeding the single LFO output to one MN3102 clock generator and feeding an inverted version to a second MN3102.  Although the sweep is clearly periodic, because each BBD is moving in an opposite direction, delay-wise, you don't hear it as periodic.

I wouldn't say that any of these is "better" or best, but they do sound different.  Incidentally, phasers can also sound less "boring" when you experiment with these same various sweep possibilities.  If you happen to have 2 phasers, stick them in series and you'll hear that the sound is a little richer and more animated and less burdened by the routine up/down/up/down.  A 6-8 stage phaser with a combination of modulation sources (perhaps 2 LFOs mixed together, or an LFO plus random source) could sound very nice.

Taking a step beyond the old JAES article on what generates a "fat" synth sound, what you want to maintain aural interest is something that is predictable enough that it doesn't distract, but veers off the beaten path just enough to not be 100% predictable.  About 70% predictable is just right when it comes to modulation.  You will note that many of the 3-phase choruses found on string synths have a fixed relationship between LFOs and usually fixed overall speed.  This arrangement more or less aims for that mostly-but-not-perfectly predictable zone.

[Ansil]

i know where one of these little chorus buggers is but he wants like 150 or better for it.

[rx5]

I think what makes chorus,phasers,flangers sound REALLY GOOD is really modifiying the LFO... they (LFos) give life to the sound....

they are like the "heart" of the circuit....

[Mark Hammer]

RX5,

I would agree with you for flangers and phasers, but for chorusses, the LFO waveform doesn't seem to be all that important since one is not really supposed to hear the sweep.  Indeed, on the Dimension C, which has a very standard waveform, you can barely hear ANY sweep because of the two complementary  BBD sections, and it sounds terrific.

I would recommend to anyone interested in chorusses to go to Larry Spence's modezero.com site, click on "gallery" and checkout the many samples of modulation and delay pedals he has.  Larry has been nice enough to provide fairly detailed setting info, pictures of the "guts" where possible, and samples using a variety of sources, including drum machines, keyboards, acoustic and electric guitar.  It's a bout as close to spending Saturday morning at the music store with the doors locked as you can get on-line.

[malaphone]

Hi Mark/All,

I have been struggling to expand my electronics knowledge beyond fuzzboxes, and thought building a nice Rand-O-Matic clone would be a good challenge. I looked at the SEWAR circuit you linked to and was startled at how similar the functions seem to be to the Rand-O-Matic, particularly the six-way switch for selecting various waveforms (The Rand-O-Matic also has a six-way rotary switch, but I'm not sure what it's function is - this is just speculation).

Anyway, have you heard this SEWAR circuit? Does it make those pefftronic-type sounds (helicopters, rayguns, etc.)? Would this be a good place to start with a Rand-O-Matic clone?


Thanks!

[Mark Hammer]

My gut sense is that there are two different notions of "aperiodic" and "random".  One view is that the sweep doesn't seem to be an obvious up-down-up-down, with a predictable interval between each end of the sweep.  The other is that the sweep undergo radical changes, a bit like the way sample-and-hold units move a filter or oscillator around in a jarring, unpredictable manner.

I would think that one of the simplest ways to achieve the first would be to use a pair of standard LFOs in triangle-wave mode, as most chorus pedals do.  If the two oscillators are unsynced, and their outputs mixed together, the result should be something that has less periodic undulations.  Separate speed and level/width pots for each will provide degrees of multi-wave modulation and blending.  Try to set the range of each LFO differently.

If you select the squarewave output of the LFOs, and sum/blend them, what you get is a stairstep function that will have the capacity to yield S&H-like changes in delay time.  Combining triangle from one LFO and square from another should yield something in between.

Make no mistake, though.  Any attempt to achieve complex modulation intended to avoid periodic-ness will require a big chassis with lots of panel space for the required knobs and switches.

[puretube]

anybody called for more knobs?

[Mark Hammer]

You've been a busy little boy, haven't you?  :wink:  :lol:

I gather this is kind of the "E-H Bi-Filter" of phasers?

When is the anticipated release date?

[Dirk_Hendrik]

I'm pretty sure the TC10 has two BBD's. The 9 series version does for sure.

I rgeret to have to tell you you're wrong in both cases.

Back ontopic in this incredible thread.

[Mark Hammer]

A quick Google search suggests that this rather remarkable now addition is already on sale in Europe http://www.musik-produktiv.at/shop2/shop04.asp/artnr/100024237/sid/!06082002/quelle/listen

[puretube]

I guess they`re a little premature with their already offering it...
afaik, end of september...

[Vsat]

Puretube,
That looks really cool!  Any mp3 demos available?
Cheers, Mike