I Apologize. This Is A Stupid Idea, Not Worth Reading, Vibrato to Pitch Shifter?

[jeff]

Ok, this is a crazy idea... but hear me out.
As far as I understand it, a vibrato works by taking the signal and making two copies, each copy is 90° out of phase for all frequencies.
-correct?


Using a sine wave LFO we fade in one copy out while fading out the other copy which is 90° out of phase at any given frequency in. The sum at any points should have the same amplitude... but the phase is shifting. In other words the frequency is changing, not the amplitude, as the mixer goes from 0° to 90° signal and back.
-so far so good?


With a sine wave controlling the % of 0° & 90° mix we get a smooth transition from true pitch, to a raised pitch, to a lower pitch, etc. as the 0° fades in and the 90° fades out. The mix of the two signals leaves the amplitude constant, but fading from 0° to 90° apparently stretches/squeezes the resulting wave.
-do I understand that right?


if so....
What happens if instead of a sine wave to control the blend of our two 0°/90° signals we used a triangle wave. Our mixing would be done linearly. straight constant increase! Would that result in a constant increase/decrease in pitch? as the triangle goes positive(linerally) the shift from 0° to 90° increases(linerally) producing a wave that is xhz slower, producing a note x lower in pitch. With a sine wave LFO you'll get a rising/lowering pitch, vibrato. But with a triangle LFO, will you get alternating higher/lower pitch?


IF THAT'S TRUE....BIG IF


But IF that's true we got a problem....
If this is true so far, you'll get a note that's higher as the triangle wave goes + and a note that's lower when the triangle wave goes -. So.... We already have two audio signals, right? 0°&90°. Run the 0° out of phase signal through an inverter to get a 180°signal. Also run the 90° out of phase signal through an inverter to get a 270° signal.
 Now we have 4 audio signals,0° 90° 180° & 270°, each 90° out of phase.


Stick with me, I almost there.
Sync the triangle wave to a flip/flop circuit, where, when the wave goes +, it swaps the 0° signal to the 180° signal , and when it goes -, it swaps the 90° to the 270°


Let's look at what that(i imagine....IF I'm correct up to this point) gives us.


Start with the triangle wave full -, for simplicity. The 0° phase signal is full on and the 90° signal is full off. As the LFO rises the 0° signal gets smaller as the 90° gets louder. The result is amplitude remains constant but phase is shifting. Because it's a triangle wave and not a sine wave the pitch is lowered. NOT lowering as vibrato... But lowered because the triangle is a constant increase. As soon as we hit full + on our LFO triangle wave, the 90° signal is full up 0° is full off...AND...swapped out for the 180°. So, as the triangle goes -(at a constant rate), the 90° signal lowers as the 180° signal rises. Since they are 90° out of phase the amplitude remains the same, but phase, frequency shifts. We then hit - LFO. The 180° is full volume, the 90° is off....And the flip/flop swaps the 90° signal for the 270°. Etc. Etc.


Would this work?
The signal is ALWAYS chasing a signal 90° out of phase, and doing it at a constant rate.
0° is chasing 90° 
90° is chasing 180°
180° is chasing 270°
270° is chasing 0°


Vibrato.
Sine LFO/ Flip/flop off


Pitch Shift.
Triangle LFO/flip/flop on
Add a switch to swap 90° 270° for pitch up/pitch down.(reverses order 0° 90° 180° 270°/0° 270° 180° 90° 0°)
And the rate of the LFO would set how much the pitch would be raised/lowered.


Crazy?
Thanks for listening
Jeff

[jeff]

I know my line of thinking must be wrong somewhere, just not sure where.
Has anyone ever tried a triangle LFO with their vibrato?
If so what was the result?

[Merlin]

With a sine wave controlling the % of 0° & 90° mix we get a smooth transition from true pitch, to a raised pitch, to a lower pitch, etc. as the 0° fades in and the 90° fades out. The mix of the two signals leaves the amplitude constant, but fading from 0° to 90° apparently stretches/squeezes the resulting wave.
-do I understand that right?

Yes

if so....
What happens if instead of a sine wave to control the blend of our two 0°/90° signals we used a triangle wave.

You would get the same amount of pitch shift, but the 'suddenness' of the change back and forth would be different. The LFO is like your finger bending the guitar string; sine or triangle is just the 'style' in which you move your finger back and forth.

[jeff]

OK, so a triangle would just sound 'jerky' and not sound as a constant raised/lowered pitch.
Gotcha.

[jeff]

The schematic on this page https://www.aggh.net/discussion/index.php?topic=30731.0 shows a filter for a output 4phase output.
Any way to scale this up for tube use?
Could I just keep the caps the same value and make the resistors 10X larger?
Scale caps up/down by X?
I don't know how I'd design/modify a 4 phase filter like this for tubes, but I would like to play around with it.

[Merlin]

The schematic on this page https://www.aggh.net/discussion/index.php?topic=30731.0 shows a filter for a output 4phase output.
Any way to scale this up for tube use?Could I just keep the caps the same value and make the resistors 10X larger?
Scale caps up/down by X?

You would scale up the resistors, and scale down the capacitors by the same factor.

I don't know how I'd design/modify a 4 phase filter like this for tubes, but I would like to play around with it.

I have never looked into polyphase filters, but I expect there is a primer for dummies like me somewhere on the net!

EDIT: I did a quick sim on that filter. The amplitude response is not exactly flat, but the phase shifts are good above 100Hz. For tube use you would want to increase the R's by x10 and reduce the caps by a factor of 10.

[jeff]

Awesome! Thanks!
So...
Adjust values
Make a split-load P.I.
Hook up to plate n cath
I'll get 4 signals where every freq is 90° out of phase. 0° 90° 180° 270°


Seems like I could use a simpler version of that also to get my 0° 90° LFO signals. Since it's in the low 6-8Hz, maybe same idea, less caps/resistor filters.

[Merlin]

I had a go at making a version where all the caps are the same, since it's more convenient to buy different resistors instead of different caps. Impedances have been scaled up for tube use. I think this version would be better for vibrato because the phase shift covers the frequency range where most musical content lies, esp. guitar fretboard.

[Merlin]

Here's a version of the original wurlitzer quadrature network, scaled for use with your LFO

[jeff]

oh, awesome.
When you say

...Impedances scaled up for tube use....

Do you mean the values listed ARE scaled up, or those values NEED TO BE scaled up?
In other words, copy what I see, or copy what I see.... But scale the values?


Thanks again
You rock!

[Merlin]

Do you mean the values listed ARE scaled up, or those values NEED TO BE scaled up?

I mean I have already scaled them up, you can just copy.

[jeff]

Great! Thanks!
So here's the crazy idea.
Take the Vox/Wurlitzer mixer.
Put the 0° on one of the grids you have trem
Right?
Put the 0° on one grid 90° on the other grid you have vibrato
Right so far?

[Merlin]

Yes.

Also, look at this amazing circuit diagram of the polyphase filter, I love it!

[jeff]

WOW!


So let's put the 0° on one grid and the 180° on the other.
When the LFO is off the signals cancel, right?
But we still have trem, just a different flavor of trem.
As the LFO goes positive, we start to hear the 0° signal get louder.
As it goes back to 0V the 0° and 180° cancel
As it goes negative the 180° out of phase gets louder
On and on, each half cycle we hear the signal fade in and out, flipping phase each time.
That right?

[Merlin]

Yes

[jeff]

Ok, so we got one tube trem going on/off each half cycle flipping from 0° to 180°.
Let's get another Vox/Wurlitzer mixer and do the same thing with the 90° and 270° signals.
Sync then up to the same LFO....but.... The 90°/270° LFO is 90° out of phase with the 0°/180° LFO.
One tube 0°/off/180°/off
One tube off/90°/off/270°


Mix the outputs.
We start at 0°
As 0° fades out 90° starts to fade in
That's vibrato so far.
But instead of returning to pitch by mixing back in the 0° signal, as the 90° starts to fade the 180° starts to get louder.
As the 180° starts to fade, we fade in the 270°
270° out, back to 0°, start the cycle all over.




Normally, As we mix in the lagging, phase pitch goes one way, as we mix the leading, pitch bends back in opposite direction.


With this... Were Always chasing the lagging phase.


I think if you get the right shaped LFO you'd have a pitch shifter.
Right shaped LFO could even get you pitch up or pitch down WITH VIBRATO!!!


Crazy!

[jeff]

oops I meant "Crazy?"

[jeff]

Up/down would be a switch to reverse 90°/270°
Maybe even use trim pots to calibrate the LFO speed for, 3rd, 5th, octave, whatever, on a rotary switch.
Blend in original signal for some wacky fun

[shooter]

now do the math "long-hand", without modeling 


some smart kid came up with a quadrature phased TX coil for use in MRI, the instructor thought it would be fun for us to "do the math"
He brought in the design engineer, for 1 hour I completely understood, class ended, we all went to dinner, most of us then went out after, for one toke over the line. we became mere mortals again, simply ordered a new coil when one would break!

[jeff]

now do the math "long-hand", without modeling 


some smart kid came up with a quadrature phased TX coil for use in MRI, the instructor thought it would be fun for us to "do the math"
He brought in the design engineer, for 1 hour I completely understood, class ended, we all went to dinner, most of us then went out after, for one toke over the line. we became mere mortals again, simply ordered a new coil when one would break!

Wish I knew where to start.
That's why I love this site, you got a problem, someone else might have the solution.

[Merlin]

Aren't you describing vibrato that shifts over a wider phase range? A deeper effect, sure, but still vibrato. Vibrato is pitch shift, just an oscillating kind.

[jeff]

Aren't you describing vibrato that shifts over a wider phase range? A deeper effect, sure, but still vibrato. Vibrato is pitch shift, just an oscillating kind.

Maybe.
Maybe I don't fully understand.
With the Vox/Wurlitzer circuit, as you wiggle between the two signals 0° 90° you are either bringing up the leading or lagging phase. At any point you stop, the two phases create a new wave, with the same amplitude, but a phase somewhere between 0° and 90°
It's the wiggling that shifts the pitch. The Indiana Jones "put the bag of sand down as you pick up the idol" that squishes or streches the Hz.
Fade the 0° out as you fade the 90° in pitch streches, going lower in pitch.
Fade the 90° out as you fade the 0° signal in the pitch squishes, pitch rises.


If at and point you stop, the two signals 0° and 90° apart, mix together to make a signal of the same amplitude, X° out of phase(say, if the LFO is off)


Ok, so what if we Don't squish, strech, squish, strech, but keep chasing our tail.
Is going from 0° to 90° streches, lowers pitch, instead of going back to 0° we keep going 90° in the stretch direction to 180°... To 270° ....back to 0°...


I'm not 100% sure if this will work but instead of going up 90° back down 90° back up 90°(vibrato) we keep going up 90° up 90° up 90° I think we can make a signal with Hz lower or higher than what we started with.


If we have a 440hz signal and keep shifting it's phase in one direction won't we get a 440•Xhz signal?


Again, I don't know, but it's something I wanna try

[jeff]

Thought of a good analogy:
Shifting from 90° to 0° is like a train coming towards you.
Pitch raises.
Shifting from 0° to 90° is like the train going away from you.
Pitch lowers.


With Vibrato we're alternating between 0° 90° 0° 90°
Pitch rise, fall, rise, fall.


What if the train never stops coming towards you?
What if we just keep going up, up, up, 90° 90° 90°
0° to 90° to 180° to 270° to 0° etc.
That pitch will stay raised.
That's the thought, at least.

[acheld]

Very cool stuff!

Makes my head spin faster than a modern CT gantry . . .

[AlNewman]

Uncle Doug did a video on the Magnatone vibrato circuit, it uses a cathodyne with a capacitor on the inverted side and a varistor on the non inverted.  Not sure if something like that would work or could be modified with what you are after.

eature=shared

He starts explaining the circuit at about 27:50

[HotBluePlates]

... With the Vox/Wurlitzer circuit, as you wiggle between the two signals 0° 90° you are either bringing up the leading or lagging phase. ...

You're kinda mistaken about this notion:  The Wurlitzer (and Vox) vibrato circuits use 2 signals 180º out of phase, but then they're bridged through R-C networks.  The resultant is something in-between, which is about 90º phase-shift.

As you mention, the shift doesn't matter unless it's changing ---> approaching 90º one way (maybe even only 45º or 60º) and approaching 90º the other way (again, it might only be 45º or 60º).  The reason for "not quite 90º" is that the a single reactive component can only ever approach 90º of shift; it never quite gets there.  And some medium-frequency causes the shift to be more like 45º.

If you could shift further, you just change the pitch more.  Like a vibrato that wiggles a whole-step instead of a half-step.  And just like that wide vibrato, going too wide sounds obnoxious.

As for your other question about waveform:  a sine wave sounds like a smooth-change.  You can use different wave shapes as desired, but you'll need to taste-test.  After all, circuit Vibe circuits (as in things patterned after the 1960s Uni-Vibe) get their personality from an un-even waveshape.

[jeff]

OK,
I'm trying to understand, thanks for the help.
Say you have a cross fader. 440hz on one end 440hz 90° out of phase on the other.
What happens?


As far as I understand the two waves combine to make a new wave of the same amplitude, but the phase changes.
Cross fader left 440hz 0°
Cross fader in the middle 440hz 45°
Cross fader right 440hz 90°


The pitch may rise or fall as you move the cross fader, but, if any point you stop, you get 440hz .Same amplitude, with a phase somewhere between 0° and 90°


Is that part correct?

[HotBluePlates]

Say you have a cross fader. 440hz on one end 440hz 90° out of phase on the other.
What happens?

Less pitch-variation than what Vox/Wurlitzer/Magnatone get from their circuits, because they're starting "in the middle of two signals 180º out-of-phase."

The pitch may rise or fall as you move the cross fader, but, if any point you stop, you get 440hz .Same amplitude, with a phase somewhere between ___° and ___°

Human ears don't really perceive absolute-phase.  That's why "speaker-pushing-forward" and "speaking pulling-backward" all sound the same until a 2nd speaker is added.

Just like a siren on an ambulance, we only hear a pitch-change of the phase is changing (the siren moving closer/further results in a phase-change that gives the Doppler Effect).


_____________________________________

Overall, you're stuck on "90º" and I'm trying to say it doesn't matter: the 2 signal-outputs are really 180º apart.

   - Output A is 180º.
   - Output B is 0º.
   - Output A is connected to Output B through reactive components.
   - The mid-point of these components will be "90º" simply because that's halfway from 180º to 0º (plus we choose appropriate reactances to make that happen).

[jeff]

But wait,
Isn't the whole point of those filters to get two signals?
Two signals whose phase is 90° apart from each other at all frequencies?


Plate and cathode ARE 180° out of phase and connected to two filters.
What the two filters are doing, as far as I understand, are changing the phase so, at any given frequency, those two signals are 90° out of phase to each other.

_____________________________________

Overall, you're stuck on "90º" and I'm trying to say it doesn't matter: the 2 signal-outputs are really 180º apart.

   - Output A is 180º.
   - Output B is 0º.
   - Output A is connected to Output B through reactive components.
   - The mid-point of these components will be "90º" simply because that's halfway from 180º to 0º (plus we choose appropriate reactances to make that happen).

At one of the filters
But we have two filters
Freq X may be 20° out of phase on one filter but 110° out at the other - a difference of 90° between the two outputs
Freq X may be 30° out of phase on one filter but 120° out at the other - a difference of 90° between the two outputs
Freq X may be 10° out of phase on one filter but 100° out at the other - a difference of 90° between the two outputs


The whole point of taking the 0° and 180° signals and running them through two non-identical filters is to convert two signals that are 180° out of phase at every frequency to two signals that are 90° out of phase at every frequency, right?


That's why those two filters aren't identical, right?

[jeff]

Think of it this way,
Let's by-pass the filters. Stick the 0° signal(cath) on one grid of our "mix tube" 180°signal(plate) on the other.
Turn on the LFO and we get amplitude modulation, at the same frequency.


Take those signals, filter so instead of 180° out of phase, they're 90° out of phase(at any one frequency).
Turn on the LFO and we get frequency modulation, at the same amplitude.


Crossfading 0° 180° signals gives trem,
Crossfading 0° 90° signals gives vibe.


I believe that's how it works. If not, why would we need the filters after the PI?
After the PI we DO have two signals 180° out of phase. We could just use those.
But two signal out of phase gives amplitude modulation.
We take those two 180°oop signals, filter them, we now have two signals 90° out of phase with each other at any given freq.
Two signals 90° out of phase gives frequency modulation.


Does that make sense?
I need to order some parts and experiment

[Merlin]

Crossfading 0° 180° signals gives trem,
Crossfading 0° 90° signals gives vibe.

Correct. (I do wonder what Crossfading 0° 180° would truly sound like, since it passes through zero unlike 'normal' trem. Probably very choppy and not that useful, musically).

I'm still having trouble visualising what you're proposing, using text alone. But have you looked into how stombox phasers work, like the MXR phase-45 and phase-90? They do it by first creating vibrato, then mix some of the original signal into the output. The core of those circuits is just vibrato accomplished in a different way (ignore the mixing part). Looking at the Phase-45 vs. Phase-90 emphasises the point that you can keep adding more phase shift stages to increase the amount of pitch shift in your vibrato, theoretically without limit, but you don't reach a 'new type of effect'.

Once again I'm recommending the vibrato edition of stompboxology: https://app.box.com/s/fc83561b9625e7f4958f

[jeff]

Picture a train track. As the train comes toward you, blowing a whistle, you perceive the pitch to be higher than it is. As it goes passed you, you perceive it to be lower in pitch. Another train passes you in the opposite direction, you hear its whistle higher in pitch, it passes, you hear it lower in pitch.
That's vibrato.


Now picture a train blowing its whistle as it approaches, just as it is about to pass you, it stops blowing its whistle, but, another train is coming towards you from the same direction. It starts blowing its whistle, but stops blowing its whistle as it gets to you, and here comes another train, blowing its whistle.
That's what I'm trying to do.


Going from 90° to 0° the pitch drops. Going from 0° to 90° the pitch rises.
Vibrato is 0° 90° 0° 90°pitch rises falls rises falls.


Imagine a crossfader.
440hz signal left 440hz signal 90° out of phase on the right
Fade it back n forth at a constant velocity.
As you go from left to right, you'll hear a pitch lower that 440hz.
As you go to the right you'll hear a pitch higher than 440hz.


Now, try it again...but... As soon as you get to the right the left signal is inverted 180°
As soon as you get to the left the right signal is inverted.


You fade left to right, you hear a pitch lower than 440hz
The left signal flips
As you fade to the left, you hear a pitch lower than 440hz
The right signal flips
As you fade to the left, you hear a pitch lower than 440hz.


Instead of vibrato you have a pitch shifter


Crossfading a wave from 0° to 90° sounds exactaly like crossfading a wave from 90° to 180° sounds exactly like crossfading a wave from 180° to 270° but Does Not sound like crossfading from 90° to 0°


Vibrato goes 0 90 0 90 0 90 0 90
I'm trying to go 0 90 180 270 0 90
Always crossfading 90° UP, not back n forth


Does that make sense?


I think I should just shut up, get to solderin' and testin, and come back with the results before I make a(even bigger) fool of myself.

[jeff]

Crossfading 0° 180° signals gives trem,
Crossfading 0° 90° signals gives vibe.

Correct. (I do wonder what Crossfading 0° 180° would truly sound like, since it passes through zero unlike 'normal' trem. Probably very choppy and not that useful, musically).

I'm still having trouble visualising what you're proposing, using text alone. But have you looked into how stombox phasers work, like the MXR phase-45 and phase-90? They do it by first creating vibrato, then mix some of the original signal into the output. The core of those circuits is just vibrato accomplished in a different way (ignore the mixing part). Looking at the Phase-45 vs. Phase-90 emphasises the point that you can keep adding more phase shift stages to increase the amount of pitch shift in your vibrato, theoretically without limit, but you don't reach a 'new type of effect'.

Once again I'm recommending the vibrato edition of stompboxology: https://app.box.com/s/fc83561b9625e7f4958f

OK, I took a look
The theory of operation is what gave me this idea.


If we have a 440hz sine wave and a 440hz wave that is 90° out of phase we basically have a sine wave and a cosine wave, is that right?

[Merlin]

Now picture a train blowing its whistle as it approaches, just as it is about to pass you, it stops blowing its whistle, but, another train is coming towards you from the same direction. It starts blowing its whistle, but stops blowing its whistle as it gets to you, and here comes another train, blowing its whistle.

So instead of vibrato that wobbles above and below the original pitch, we have vibrato that wobbles from above down to the original pitch, then above again, never going below?

[jeff]

Now picture a train blowing its whistle as it approaches, just as it is about to pass you, it stops blowing its whistle, but, another train is coming towards you from the same direction. It starts blowing its whistle, but stops blowing its whistle as it gets to you, and here comes another train, blowing its whistle.

So instead of vibrato that wobbles above and below the original pitch, we have vibrato that wobbles from above down to the original pitch, then above again, never going below?

Not exactly. There's two parts to this, and it's hard to explain one without the other because one is dependent on the other so, please stick with me. There's two different concepts I need to explain.


First is the doppler effect, how it works.
Imagine a train miles and miles away. And just to remove the volume change from the equation, let's say its whistle is extremely loud far away but decreases in volume as it gets closer. The loudness it tied to how far away from you it is, so you perceive it to be the same loudness however far away it is from you.
Just to simplify- you perceive no change in volume.
Got it?

[jeff]

So, the train is at a dead stop and blows its 440hz whistle. Takes a bit of time for the sound to get to you but when it does, you hear 440hz. Now, the train starts to accelerate chugga chugga chugga. It gets up to 20 mph you hear that pitch rise as the train accelerates. 40mph pitch still rising. 60mph, pitch keeps rising. The train gets to its top speed 100mph you hear the whistle at, I don't know 500hz(???guess). Anyway, higher than its actual 440hz. Whatever that pitch is you hear when the train is doing 100mph is the pitch you'll hear as long as that train maintains that speed. No matter how many miles away that train is, as long as it's maintaining its speed, as long as it's coming at you at 100mph, with no acceleration/deceleration, you'll hear the whistle at 500hz, not 440hz. You'll hear THAT pitch up until the train reaches you.


In other words, once the train is at a constant speed, the pitch doesn't go up up up up until only dogs can hear it.
 That's important to the next part.



So,
Constant velocity=constant deviation in pitch.
Acceleration towards you = rise in pitch, (but maintaining that constant speed the pitch is raised, but NOT rising.)
Opposite true after it passes you. You'll hear the same note, the same pitch, until the train decelerates.


That mmmmmMMMMMMMMMNNNOOOoooooohhh is just as its passing.
As it, approaches you, passes you, keeps going.

That's the first half

[jeff]

No for the phase stuff and how that relates:
Say we have a 440hz sine wave, we also have a 440hz cosine wave. Humans can't hear phase, not by itself. A sine wave and a cosine wave sound identical, the only diffrence is they are 90° out of phase. Only real difference between the to is one wave is 'delayed' by one quarter of its cycle. If the sine wave cycles 0V to +1V to 0V to -1V to 0V its cosine wave cycles +1V to 0V to -1V to 0V to +1V. Identical looking wave form, identical sound, same 440hz, same wiggley wave but when one wave is at its peak, the other is at 0V and vise versa.


The thing about mixing a sine wave and a cosine wave is you will always get a wave with the same frequency, and the same amplitude... but at a different phase. So if you take a 440hz sine wave(0°) and a 440hz cosine wave(90°) you'll get a 440hz sine wave X°out of phase. UNLIKE a sine wave and an inverted sine wave(180°). Mixing a 440hz sine wave with a 440hz wave 180° out of phase will cancel out.
50/50 mix of 0° 180° = silence
20/80 mix of 0° 180° = 3/4 of the 180°
80/20 mix of 0° 180° = 3/4 of the 0°


But mix a 440hz sine wave with a cosine wave,
50/50 you get the same volume 440hz wave at a different phase°
80/20 you get the same volume 440hz wave at a different phase°
20/80 you get the same volume 440hz wave at a different phase°


No matter how you mix the to 90°out of phase you get the same pitch, same volume. The combination of the to only changes the phase, and on its on we can't hear phase. 440hz 0°, 440hz 90°, 440hz ?° sounds exactly the same.


So, let's take the sine wave(0°) and put it on one end of a slider, a back n forth pot, whatever they call it. Put the cosine wave at the other.

Put it so 0(0°end) is facing you and 10(90° end) is away from you. So you push it and pull it towards you. This mixes the two signals. Any point on the slider is 440hz. No, really! Slider at 0, 440hz. At 3.5, 440hz. At 9.576, 440hz. Anywhere along that slider is 440hz...at some shift in phase.


So to get back to our train, we can think of that slider as our train blowing it's 440hz whistle. Anywhere along the track IS 440hz. Anywhere along that track as long as the train/slider is not moving we hear 440hz.


Start the train in the middle/half way point of the track. That's our vibe circuit with the LFO off. Just sitting there, not moving, blowing its 440hz whistle. We got a 440hz signal on the grid of one half of the mixer tube and another equal strength 440hz 90° out of phase signal on the other. Just like sitting in the mid point of our slider, just like the train halfway up the track blowing it's 440hz whistle.


Gotta go, to be continued....

[jeff]

Ok, where was I?
So there's a slider in front, that's just like standing on the track.
As long as the train/slider in stationary you hear 440hz.
440hz when the sliders closest to you.
440hz when its farthest
440hz when its dead center.
 just a differert phase of 440hz
Closest to you, you hear 440hz 0°
Furthest from you, you hear 440hz 90°
You can't hear phase on its own, just pitch. So no matter where the crossfaders slider stops you hear 440hz.


Motion! It's the train IN MOTION that cause the whistle to sound higher or lower. It's the Indiana Jones'lifting the idol/putting down the bag of sand' that changes the pitch.
It's crossfading the sine in/cosine out and vise versa that causes the pitch to change.


Say out slider/ train is half way down the track, MOVE the slider(you're doing the job of the LFO). This is exactly like the train starting to move. As it accelerates towards you, just like the trains whistle, you hear the pitch rise. Push it away the pitch lowers. At any point the train/slider stops moving you hear the 440hz as true 440hz in some degree of phase modulation.


So you ARE doing the job of the LFO. You control the blend of 0°/90° signals.
Push it towards you,
Stop,
Push it away,
stop,
Push it towards you....etc.
C
....Vibrato!


Remember the slider IS the train blowing its 440hz whistle. It's half way up the track(center on the slider) move the slider from center to 1/4 as the slider accelerates the pitch rises above 440hz. You stop, the pitch returns to 440hz. How can this be? If dead center is 440hz and you just moved the slider towards you, how can it still be 440hz. Because just like the trains whistle is ALWAYS blowing at 440hz, every single point on the slider produces a blend of sine/cosine that results in 440hz....just phase shifted. Just as accelerating the train towards you causes you to hear a higher pitch accelerating the slider does to.


Wiggle from 5-4-6-5
Now 5-3-7-5
Now 5-0-5-10
This is you turning up the intensity of the vibe.
The more it accelerates towards you the more the pitch deviates from 440hz, but it's all the same idea. As it approaches 0° the pitch rises, as it stops the pitch returns to 440hz(phase unknown) as it accelerates past where it started the pitch drops, as it stops the pitch returns to 440hz(different phase but still 440hz) as it returns to 5 pitch rises. You stop at 5...440hz.


So if the slider fades in the sine/fades out the cosine pitch rises with acceleration.
As the slider fades in the cosine/fades out the sine pitch lowers with acceleration
At any point you stop, you hear 440hz.


Remember the first part I talked about? Acceleration = rising pitch ..but.. Constant velocity = a constant pitch lower/higher than 440hz. How do we get up to a constant velocity in one direction. How do we get the train to ALWAYS be moving AWAY from us at a CONSTANT rate.....?


.....A Worm Hole!!!!...
More later, don't want to get to far ahead of myself.


(Not to get 'side tracked'...but imagen our train with you on one end of the track and me on the other end of the track. If the train accelerates towards you/away from me, you'll hear the pitch rise, I'll hear the pitch lower, the engineer will hear 440hz.
Think about that for a sec)


So it's the acceleration AND direction that causes the pitch to change.

[jeff]

Ok, All That Explanation Just to Explain This:
We're on the tracks the train accelerates away we hear pitch drop.
We run out of track so the train stops, we hear 440hz
The train reverses we hear the pitch rise
We run out of track the train stops, we hear 440hz.
The train accelerates away.... yadda, yadda, yadda,
Vibrato.


Here's the Idea I Was Trying to Explain, What I Wanted to Try.


We open a portal in space and time!
We create a worm hole!
We put the entrance at the far end of the track and the exit right in front of us.


The train starts to accelerate, we hear the pitch fall
When the train gets to the end of the track, it enters the portal.
It pops out right in front of us still accelerating away from us.
The train is ALWAYS moving away from us, goes through the 'worm hole', gets teleported back in front of us, travels away from us, gets teleported....Etc. Etc.
As long as the train accelerates we hear the 440hz getting lower n lower(acceleration = dropping pitch)


As it goes faster n faster, always AWAY from use that 440hz sounds lower and lower.


As soon as that note drops to the pitch we want, we tell the engineer, "Stop accelerating and maintain that exact speed!"(constant velocity = constant pitch)


We've turned our vibrato to a pitch shifter!
All we need to do is create the portal.
Hmmm.
With our LFO creating a sign wave we
Send the train down the track
Hit +V
Stop the train and reverse it
Hit 0V
Hit -V
Stop the train and reverse it.


If we want to send the train to the portal, it pops out, send it back through...etc. etc. We don't need a sine wave.
We need a wave that starts at some -V, constantly increases in voltage, hits its peak +V and starts the cycle all over again. A continuous ramp up. Over n over
Like a Sawtooth Wave!
That was the original idea.


There might be a click/noise/problem because the ends aren't the same phase. Maybe a ticking every trip through the portal, but I thought of another way that might work if the Sawtooth Wave doesn't(involving 0 90 180 270 waves)
If you see any errors in my reasoning, I'd appreciate if you can point them out, so I can learn from my mistakes.


 let me know what you think so far.


Thanks for listening
Jeff
 
 

[HotBluePlates]

But wait,
Isn't the whole point of those filters to get two signals?
Two signals whose phase is 90° apart from each other at all frequencies?

No.  The point is changing phase, which gives pitch-shift.

Re-read about Doppler: all vibrato circuits simply "do Doppler" but with electrical circuits instead of sound-waves.

  - When phase advances, pitch goes up.
  - When phase retards, pitch goes down.

The Wurlitzer/Vox circuit has outputs at 0º and 180º because they come out of a split-load inverter.  When one gets louder, the total-phase moves close to that-output's phase.  When the other output gets louder, the total-phase moves closer to this new output's phase.  They're not "90º apart" and staying there, but "moving from 90º to 140º" and then "moving from 140º to 90º and on to 40º" and back again.


And in Vox amps, they simply disconnect one of the 2 signals, which leaves tremolo.  That's because there is no 2nd-signal to compare against to cause the pitch-shift to materialize.

...
Here's ... What I Wanted to Try.
...
As soon as that note drops to the pitch we want, we tell the engineer, "Stop accelerating and maintain that exact speed!"(constant velocity = constant pitch)


We've turned our vibrato to a pitch shifter!
...

AFAIK, pitch-shifters don't work that way.

That said, I didn't bother to study any of them because most seemed to be related to sample/hold circuits and/or digital.  And because if I wanted to accomplish that task, I'd rather buy a pedal than make a more-expensive/less-good home-brew version.

[jeff]

  - When phase advances, pitch goes up.
  - When phase retards, pitch goes down.

But... You need to mix two waves that are 90° not two waves that are 180°


Put a 440hz through the PI to get a signal from the plate and cath.
You have 440hz(0°) and an inverted 440hz(180°)
Bypass those filters so you're mixing the two 180° signals
You get trem right?
Any %mixing of a 0° and 180° just cancels out.
If the LFO was off you'd hear nothing because the output is 50% wave, 50%inverted wave, right?


If we needed two signals 180° out of phase... We got that before the filter, right?
The filter does not take two signals 180° out of phase and convert them to two signals 180° out of phase. There's no need for a filter, the signals going in ARE 180° out of phase.


The point of the filter is to convert those two 0° 180° signals to two signals 90° out of phase.
Mix 0° and 180° you get cancellation.
Mix 0° and 90° you get phase modulation(pitch shift)
 
That's the point of the filters...we HAVE two 180°oop at the PI
We NEED two 90°oop to mix to get vibrato

[Merlin]

The train starts to accelerate, we hear the pitch fall
When the train gets to the end of the track, it enters the portal.
It pops out right in front of us still accelerating away from us.
The train is ALWAYS moving away from us, goes through the 'worm hole', gets teleported back in front of us, travels away from us, gets teleported....Etc. Etc.
As long as the train accelerates we hear the 440hz getting lower n lower(acceleration = dropping pitch)


As soon as that note drops to the pitch we want, we tell the engineer, "Stop accelerating and maintain that exact speed!"(constant velocity = constant pitch)

That checks out as long as the train is accelerating. But the moment it pops out of the worm hole and maintains constant speed, you are back to a starting point of 440Hz for each subsequent pass. I think...

[jeff]

Think of it this way,
You have a crossfader
As you turn the sine wave down and the cosine wave up the pitch drops.
As you turn the cosine wave down and the sine up the pitch rises.


That's vibrato, right?
A crossfader between a sine wave and a cosine wave.
Push towards cosine, pitch drops.
Get to the end pitch returns to true pitch
Push toward sine, pitch rises.
Get to the end pitch returns to pitch.


The direction of the motion is what causes the pitch change.
Any point on that crossfader IS 440hz, just in a different phase.


Imagine you're at 5 and do a little wiggle from 4 to 6 to 4 to 6.
That would sound exactly like starting at 8 and wiggling from 7 to 9 7 to 9
That would sound exactly like starting at 3 and wiggling from 2 to 4 to 2 to 4


Direction = raise or drop


Another example
You move the crossfader 0 to 10 to 0 to 10 very slowly
A nice smooth wide vibrato
But you have shakey hands
So you make tiny wabbles as you go from 0 to 10 to 0 to 10


What would that sound like?
 
 

[jeff]

The train starts to accelerate, we hear the pitch fall
When the train gets to the end of the track, it enters the portal.
It pops out right in front of us still accelerating away from us.
The train is ALWAYS moving away from us, goes through the 'worm hole', gets teleported back in front of us, travels away from us, gets teleported....Etc. Etc.
As long as the train accelerates we hear the 440hz getting lower n lower(acceleration = dropping pitch)


As soon as that note drops to the pitch we want, we tell the engineer, "Stop accelerating and maintain that exact speed!"(constant velocity = constant pitch)

That checks out as long as the train is accelerating. But the moment it pops out of the worm hole and maintains constant speed, you are back to a starting point of 440Hz for each subsequent pass. I think...

Ok, so you crossfade from 0° to 90°
Left to right
As soon as you get to the end the slider is teleported to the left.
You crossfade from left to right.


If right, stop, left, stop, right, stop, left, stop is up, down, up, down in pitch, what's right, right, right, teleport, right,right, right?

It can't be up n down
It can't be 440hz
What's would it be?

Remember, every single point on the crossfader is 440hz
0. 440hz 0°
1. 440hz 9°
2. 440hz 18°
3. 440hz 27°
4. 440hz 36°
5. 440hz 45°
6. 440hz 54°
7. 440hz 63°
8. 440hz 72°
9. 440hz 81°
10. 440hz 90°


Any one point IS 440hz.
Sliding is changing the phase
Changing phase changes pitch
Cross fading right drops pitch
Stopping plays pitch(440hz)no matter where on the slider you are
Cross fading left raises pitch.

[Merlin]

If right, stop, left, stop, right, stop, left, stop is up, down, up, down in pitch,what's right, right, right, teleport, right,right, right?

440Hz up up up, 440Hz up up up

[jeff]

If right, stop, left, stop, right, stop, left, stop is up, down, up, down in pitch,what's right, right, right, teleport, right,right, right?

440Hz up up up, 440Hz up up up

Ok, so the second part of the plan.
This is why I wanted the Quad filter! (Thank you again for that BTW)
Have 4 parallel tracks/sliders in front of you.
0° 90°
90° 180°
180° 270°
270° 0°


Just like before, worm hole at each end. The portal at the end of one track connects to the start of the neighboring track.
Start at 0° go to 90°, hit the portal, teleport to the 90° 180° track, go to 180°, hit the portal, teleport to the 180° 270° track, go to 270°, hit the portal, teleport to the 270° 0° track. Repeat.


It's a full cycle.
Sine wave crossfaded to its cosine at a constant velocity = doppler effect drop in pitch.
The sine wave is crossfaded with its cosine. Once it's fully faded out:
The cosine wave is crossfaded with ITS cosign, the cosine's cosine.
The cosine's cosine is crossfaded with the cosign's cosign's cosine
And the cosine's cosine's cosine crossfades with the cosign's cosign's cosign's cosine. (which is the original sine, repeat)
You're always fading out your wave for its cosign. Do that at a constant rate(LFO) you'll always hear the train's whistle the way it sounds as it's going away.

[jeff]

Merlin,
What was the program you used to simulate the phase splitter? I'd love to play around with it.
I tried doin the math to make a filter with parts I have on hand, this what I came up with.

Filter 1:
R1 33K
R2 198K(round to 180, or 200K, or 220K?)
R3 66K(round to 68K)

C1 .00324(.003uf)
C2 540pf(500pf or 560pf)
C3 .00108(.001uf)
Filter 2:
R's 1,2,3 the same as above
C1 .0147uf(.015uf)
C2 .00244uf(.0022uf)
C3 .00488(.0047uf)


Could you plug those values in for me to see if they'd would work in place of Vox/Wurlitzer's response?


I'd greatly appreciate it.
Thank you
Jeff

[Merlin]

I use Tina TI: https://www.ti.com/tool/TINA-TI

The free version doesn't come with tube models, but you can download a circuit file from my site and just copy-paste the tubes into Tina circuits (bottom of this page). http://valvewizard.co.uk/links.html

[jeff]

Thanks,
I connect to the internet thru my Xbox, so I'll have to get access to a computer to download and use that.

One more question, This sounds counter intuitive, but I need to ask, just to be sure I understand how pitch shifting vibrato works:
Is it true that, throughout the LFOs cycle, the pitch IS 440hz ONLY when the LFO is at its most positive AND when it is most negative?
When the LFO changes direction?

AND

When the LFO passes through 0V, THAT's when the pitch is at its highest OR at its lowest.

I know that sounds counterintuitive at first... but am I understanding that part correctly?

EXAMPLE:
When the LFO is off, You have a 50/50 mix of 440hz at 0° and 90° and you hear 440hz.
As the mix goes to 60/40, the doppler effect raises that pitch.
As the mix hits 75/25 and is peak positive on the LFO, you hear the pitch return to 440hz.
As the mix passes through 60/40, 50/50, 40/60, you hear the pitch lower
As the mix hits 25/75 you hear the pitch return to 440hz
As the mix passes through 60/40, 50/50, 40/60 on its way to 25/75 the pitch rises.

So, as counter-intuitive as it sounds, when the mix is 50/50, the note will ONLY sound 440hz if the LFO is off(0 intensity), but once you turn it on, 50/50 mix(LFO @0v in its cycle) will sound:

Higher than 440hz during the first half cycle of the LFO(as the wave is going + to -)
AND
Lower than 440hz during the second half of the LFOs cycle(as the wave is going - to +)

EDIT: Fixed mixed font-size -- HBP

[Merlin]

One more question, This sounds counter intuitive, but I need to ask, just to be sure I understand how pitch shifting vibrato works:
Is it true that, throughout the LFOs cycle, the pitch IS 440hz ONLY when the LFO is at its most positive AND when it is most negative?
When the LFO changes direction?

This makes my head hurt. OK, thinking about it, the output frequency will be 440Hz when the mixer is pushed fully over to the sine, or cosine, signals.
The frequency will be higher or lower as the mixer is passing through the centre where both sine and cosine are mixed equally (higher or lower depending on which direction it is coming from)