Dream up the most ridiculous, over-the-top phaser/flanger topology imaginable

[Taylor]

I've been having lots of fun programming effects in the Spin Semi FV-1. I thought it might be fun to build some of those circuits that we all think of, then discard when we realize the number of components needed is too many (for me, maybe 200 components is my limit for one circuit - after that I lose the motivation). Although there is some limit to how much the FV-1 can do at once, doing a massively repetitive circuit is just a matter of cut and paste, so a 30-stage phaser isn't crazy here.

I was thinking I might look into some of Jurgen Haible's things like the Storm Tide Flanger, doing some large number of fixed stages and a large number of varying stages, maybe some varying delay stages, etc. I don't know that I'll get super refined with it, since I don't really use a lot of phaser kind of sounds, but I was wondering if anybody had some other ideas about fun things you'd do with modulation effects if you could do crazy things without having to actually build them.

[Processaurus]

Digital Phasing can be implemented in a fundamentally different way than analog phasing, by using independent notch filters to make the comb filter effect.  That opens the door to unique effects by virtue of the notches not being limited to move in synchronicity like a real comb filter.

Barber pole phasing is interesting, and is much easier with digital phasers that implement their phasing by the discrete notch filter method... double barber pole (one going up and one down) would be bananas.

A 40 stage phaser might be interesting to hear!  So are funny LFO's like S&H and smoothed S&H (like the Rand-o-matic), stairstep and envelope as well.

Oh yeah, I just remembered the phaser plug in for Logic that has a 2nd LFO for stereo channels that can be variably out of phase with LFO 1.  That would be a neat knob on a pedal!

[Taylor]

Hmm, cool ideas. The one real drawback is only having 3 pot inputs for this chip - this kind of effect really begs for lots of knobs.

[StephenGiles]

One of the best phasing sounds I ever heard was on Carpenters LP, which my first wife took when we split up. I suspect it was done on a studio unit, possibly the Eventide Instant.

There was some discussion a few years ago concerning phasers with a high number of stages, and Mark Hammer reported on a 24 stager.

[snufkin]

not a phaser

make the formant filter section of the ludwig phase two and the sell one to me 

I really think that could be a popular product

rg talks about the mechanics of it somewhere round here

[Jarno]

I didn't know the Storm Tide Phaser, that's a cool tool, too bad Juergen didn't put the PCB layout (or readymade PCB's) on his site. Although he has put a derivative on there, "The Son of Storm Tide". Looks cool too.

[aziltz]

How about a flanger that flanges the flange?

Or a Flanger with two independent delay lines which are 90-270 degrees out of phase with each other?  One sweeping up when the other goes down...

[Br4d13y]

envelope flange? idk   sounded cool at the time.

[MoltenVoltage]

Program it to detect the first three notes of Barracuda, then turn on. 

[Mark Hammer]

Two ideas.

1) Implement "theta processing", maybe even in variable form, in digital format.  Theta processing seems to have been first proposed by Bernie Hutchins in Electronotes, and implemented by Eventide in their Instant Flanger.  Jurgen Haible incorporates it in his Stormtide and Son of Stormtide.  Essentially, it works like this.  When the flanger sweep goes to the longest delay time in its cycle, it produces the most notches.  However, the notches are rather closely spaced together.  One of the things this does is create unpleasant resonances, especially when resonance is increased.  What Bernie did was insert a fairly large number of lagging allpass stages between the wet signal and the dry/wet mixing node.  Most of us are familiar with leading alpass stages where more phase shift is applied (up to a max of 90 degrees per stage) as frequency is increased.  Lag stages add more phase shift for frequencies below some inflection point.  The net result of adding both time delay and frequency-frelated phase-delay is to add more delay overall for the lowest frequencies, and the net effect of that is to spread out the notches at the low end of the spectrum fruther apart - specifially stretching the location of the notches downwards.  Bernie described it as being more musical sounding.

While Bernie's experiments involved 24 stages of phase shift, Eventide found that 4 stages were pretty much enough to do the job.  Or at least make a noticeable improvement beyond "normal" flanging.  Jurgen followed suit and included 4 fixed lagging allpass stages in his flanger designs.  I don't know if the FV-1 is up to it, but the idea of being able to insert, say, 4, 8, 12 such fixed stages, is an interesting one.  To the best of my knowledge, there are no flangers doing this in commercial production.

2) Implement variable through zero.  While through-zero flanging has shown up in both digital (Line 6) and analog (Paradox Flanger and EHX Flanger Hoax) forms, and Scott Stites / Mike Irwin have experimented with converting the Boss Dimension C and D to through-zero flangers, there is nothing on the market that allows one to have strong control over how much time one spends "on the other side of zero".  That is, you can set a small delay offset to the nominally "dry" signal in some flangers such that when the sweep goes high (decreasing delay time), it crosses through zero and is actually delayed less than the "dry" signal is.  But that crossing over is generally short lived, such that you go through, live on the other side for a second or so and then come back through zero again.  In the adapted Dimension C, two BBDs are identically counterswept such that they each cover the identical range of delays, but start out at opposite ends of that range and pass by each other in the midpoint.  Some time back, we had lengthy discussions about this and labelled them symmetrical (Dim C) and asymmetrical through-zero flanging.  While each are clearly attainable, that entire range between complete symmetry and just a tiny bit of through zero has never really been covered in any commercial product, at least that I'm aware of.

I'm hoping to address that with a dual flanger thing I have on the back burner.  I have a pair of Boss BF-1 flangers, each of which is a 4-knobber (rate, width, resonance, initial delay) that has the dry signal lifted.  My plan is to nest them within a splitter-mixer, such that one flanger serves as my "dry" signal and the other serves as the wet one.  When they get mixed back together, the "dry" flanger will introduce whatever amount of delay offset I set it to, and the swept one will be able to cross through zero in a wide array of different ways.

The other thing I can do with this setup is to have both flangers sweeping, but at different rates.  This will result in through zero, but in an aperiodic way.  That is, the through-zero point will not take place at the same predictable time/point, and the amount of time spent "on the other side of zero" will also vary.


So there you have it.  Two novel approaches to flanging.  I don't know if they can be implemented with the FV-1, but maybe they can raise the bar a little higher.  

[ElectricDruid]

Taylor, could you increase the number of knobs available by using a analog CMOS multiplexer chip like the venerable 4051 8-to-1 switch? You'd need three IO pins to control the switch, but it'd give you ten knobs max. I've done similar things successfully with dsPIC.

Just a thought.
Good luck with the crazy flanging. The FV-1 sounds like fun.

T.

[Taylor]

Taylor, could you increase the number of knobs available by using a analog CMOS multiplexer chip like the venerable 4051 8-to-1 switch? You'd need three IO pins to control the switch, but it'd give you ten knobs max. I've done similar things successfully with dsPIC.

Just a thought.
Good luck with the crazy flanging. The FV-1 sounds like fun.

T.

Hmm, I wonder. I've just gotten into digital stuff in the last couple of weeks, so this is all fairly new to me. I'll look into that.

[puretube]

News from Ludwig Lester... aka: HAPPY EASTER - egging... 

[Taylor]

News from Ludwig Lester... aka: HAPPY EASTER - egging... 

Is the flanger described herein a hoax..?

[Mark Hammer]

Ummm, Mr. Ludwig would appear to have periods where he requires medication, and lots of it.  His list of patents is large and wide ranging.  So wide ranging they serve a clinical diagnostic function, don't they?

You will note the following:

• A little over 6-year gap between submission and patent date
• 72 figures and 119 pages
• a pleasant, if tangential, stroll throughout what seems to be the entire gamut of instruments
• an "unusual" amalgam of chemistry, computer science, neuropsychology/psychoacoustics, electronics and last but not least musical performance

My hat goes off to the examiner Marlon Fletcher, and patent attorney Craig Schmoyer, who must be the two most patient people on earth.

I'm just glad I don't have to work with the guy. 

[Ice-9]

Hi Taylor i have just started a thread on a FV-1 project , i would be happy for any conribution maybe the two threads are on the same goals.

[Taylor]

Hi Taylor i have just started a thread on a FV-1 project , i would be happy for any conribution maybe the two threads are on the same goals.

Hi Mick, this post was from about a year ago. I've done lots of stuff with the FV-1 but never got around to messing with phasers (actually I tried, didn't get anything usable and forgot about it...). I'll be watching your thread, I've thought about releasing one of my FV1 PCB designs for DIY use, but decided against it, just because there's already a Tonepad PCB for it that people can get (though it's flawed), and I thought helping people with debugging would be tougher than an analog project.

[cloudscapes]

Hmm, cool ideas. The one real drawback is only having 3 pot inputs for this chip - this kind of effect really begs for lots of knobs.

can you multiplex the ADCs? what's the sampling rate on those?

[Taylor]

Nope, you definitely can't multiplex pot ADCs on the FV1, unfortunately. I don't remember the sampling rate, should be in the datasheet, however it's worth noting that they have hardware filtering on the pot inputs. This is supposed to minimize zipper noise I guess, but it has the side effect that there is a ~100ms delay between physically changing the pot value (or voltage at the ADC) and getting the change in your program, and any instantaneous changes (like if you tried to modulate a pot ADC with a square wave) will be filtered into something smoother. Almost like working with optocouplers...

[Quackzed]

you asked for it.
build 4 ruby amps each with a different size speaker 8-10-12-15,attach a guitar chord to each and plug it into the input,then hang them each from the ceiling with their guitar chord ,one at each corner of an imaginary square -at the center of the square, in the middle of all the hanging amps , hang an omni directional mic.plug all the chords into a 4 way splitter;plug into the splitter,
then spin them while you play.