Help understanding mod matrix and shape mod

[godparticle]

I was watching a video on programing and was curious how things affect one another.

The shape mod of the oscillator was 20%
The modulation matrix was sourced to LFO1 and sent to shape mod
The amount in the modulation matrix was set at around 50

My question is, since the shape mod was initially set at 20%, how does the LFO frequency affect that value?  Does the LFO frequency affect how fast it goes between 0 and 20%?  And what does amount do in the mod matrix?

Sorry if these are really noobish questions. I'm learning the basics

[maxter]

This thread:

https://forum.sequential.com/index.php?topic=3203.0

[godparticle]

This thread:

https://forum.sequential.com/index.php?topic=3203.0

Thanks, but that's a bit over my head.

I guess my question boils down to am I understanding the basic elements.

Shape mod: controls how wide the pulse is
Frequency: how fast it triggers
Amount: how....high on the Y scale a pulse reaches(like if looking at an oscilloscope)? I don't really know what this does. It goes +/- 127

[creativespiral]

If you can, try and get an oscilloscope... there's some free ones available for most DAWs.  It will allow you to visually see how Shape Mod / Pulse Width Modulation works as you play with knobs or adjust modulation amounts.   

The best shape to experiment with first is the Square/Pulse wave... that is most commonly associated with shape mod / PWM.   On the Rev2, if you set the shape value to "50", that will give you classic symmetric square wave.   

With an LFO, set destination of Osc1 Shape and select the triangle for the LFO shape... that is a bipolar LFO shape, meaning it swings positive, then crosses zero point and swings negative, and back again.   This is the type of motion that works best for PWM (or a sine if available).   

Now, set the amount fairly high if you want to really hear the PWM (try something 60+ for amount... its scaled down when applied to shape)... the amount controls how far away -/+ from the base value the shape will swing (ie: starting point / middle point of 50)   Finally, set the Frequency of the LFO to taste... usually slower/lower frequencies for square/pulse modulation.   

[godparticle]

If you can, try and get an oscilloscope... there's some free ones available for most DAWs.  It will allow you to visually see how Shape Mod / Pulse Width Modulation works as you play with knobs or adjust modulation amounts.   

The best shape to experiment with first is the Square/Pulse wave... that is most commonly associated with shape mod / PWM.   On the Rev2, if you set the shape value to "50", that will give you classic symmetric square wave.   

With an LFO, set destination of Osc1 Shape and select the triangle for the LFO shape... that is a bipolar LFO shape, meaning it swings positive, then crosses zero point and swings negative, and back again.   This is the type of motion that works best for PWM (or a sine if available).   

Now, set the amount fairly high if you want to really hear the PWM (try something 60+ for amount... its scaled down when applied to shape)... the amount controls how far away -/+ from the base value the shape will swing (ie: starting point / middle point of 50)   Finally, set the Frequency of the LFO to taste... usually slower/lower frequencies for square/pulse modulation.

Ok, so maybe I was kind of right in thinking it was how far +/- it went. The only strange thing I see with that is if I set source/destination through the mod matrix, the amount can go +/- 127, but if I do it through the LFO section, it only goes 0-127. Thanks the the help

[godparticle]

Tried making a graphic to see if that makes more sense of my understanding

[Razmo]

You're right with your drawing... but just remember, that the values are ADDED to the destination parameter... this mens that if you set osc shape to 50%, and then set the amount to more than half the range, then the destination value will actually clip (stay at max/min value longer since it's clipped). the shape mod cannot go below zero or above maximum value when modulated by a mod source, this in the above case you would have to set the amount to a maximum of half of it's range for it not to clip.

Also remember, that the LFO destination, (that only have a positive amount parameter) has a much finer resolution than if you set up the same modulation in the mod matrix... the mod matrix amount range is actually 4 times bigger than the LFO's own destination amount... thus an amount of 1 in the mod matrix is the same as the value 4 in the LFO's mod amount... worth noticing, or you can get confused.

[godparticle]

You're right with your drawing... but just remember, that the values are ADDED to the destination parameter... this mens that if you set osc shape to 50%, and then set the amount to more than half the range, then the destination value will actually clip (stay at max/min value longer since it's clipped). the shape mod cannot go below zero or above maximum value when modulated by a mod source, this in the above case you would have to set the amount to a maximum of half of it's range for it not to clip.

Also remember, that the LFO destination, (that only have a positive amount parameter) has a much finer resolution than if you set up the same modulation in the mod matrix... the mod matrix amount range is actually 4 times bigger than the LFO's own destination amount... thus an amount of 1 in the mod matrix is the same as the value 4 in the LFO's mod amount... worth noticing, or you can get confused.

I kind of see what you're saying. Is there a way graphically I could see how amount is affecting the shape mod%?  For example, if I had shape mod at 40% and applied the LFO with an amount of 20, what would that make the shape mod%?

Also, in the LFO, is the amount going from 0 to 20 as the wave goes up and down(ex: triangle wave ramping up to 20 and ramping down to 0) in my example above?

[Razmo]

You're right with your drawing... but just remember, that the values are ADDED to the destination parameter... this mens that if you set osc shape to 50%, and then set the amount to more than half the range, then the destination value will actually clip (stay at max/min value longer since it's clipped). the shape mod cannot go below zero or above maximum value when modulated by a mod source, this in the above case you would have to set the amount to a maximum of half of it's range for it not to clip.

Also remember, that the LFO destination, (that only have a positive amount parameter) has a much finer resolution than if you set up the same modulation in the mod matrix... the mod matrix amount range is actually 4 times bigger than the LFO's own destination amount... thus an amount of 1 in the mod matrix is the same as the value 4 in the LFO's mod amount... worth noticing, or you can get confused.

I kind of see what you're saying. Is there a way graphically I could see how amount is affecting the shape mod%?  For example, if I had shape mod at 40% and applied the LFO with an amount of 20, what would that make the shape mod%?

Also, in the LFO, is the amount going from 0 to 20 as the wave goes up and down(ex: triangle wave ramping up to 20 and ramping down to 0) in my example above?

It depend on what waveshape you choose for the LFO... some are unipolar others are bipolar... the noise and triangle are bipolar... they go both negative and positive... the others are unipolar and only go positive.

generally it's just a matter of experimenting with the LFO amount... the shape parameter has values from 0 to 99, so if you set it to 50 which is the middle value, then the amount will swing around that value... so if you choose the amount to be say; 20, then it would be modulating the shapemod parameter from 40 to 60 with 50 being the center... if you instead chose the sawtooth waveshape which is unipolar, it would modulate the shape parameter from 50 to 70...

You will have to experiment with this... the best way to do this is by routing the LFO to an oscillator's pitch... that will make you able to hear the modulation quite easily if you slow down the LFO rate... clipping will be rather obvious this way.

One thing to note is that the LFO amount goes much higher than the 100 values of the shape parameter... therefore, in the case of modulating the shapemod parameter, the maximum LFO amount should be 100 I think since that would give you a full swing around the center (-50 to +50 or rather 0 to 100).

I know this sounds a bit complicated, but it's actually rather logical and simple as soon as you get how it works.

[godparticle]

You're right with your drawing... but just remember, that the values are ADDED to the destination parameter... this mens that if you set osc shape to 50%, and then set the amount to more than half the range, then the destination value will actually clip (stay at max/min value longer since it's clipped). the shape mod cannot go below zero or above maximum value when modulated by a mod source, this in the above case you would have to set the amount to a maximum of half of it's range for it not to clip.

Also remember, that the LFO destination, (that only have a positive amount parameter) has a much finer resolution than if you set up the same modulation in the mod matrix... the mod matrix amount range is actually 4 times bigger than the LFO's own destination amount... thus an amount of 1 in the mod matrix is the same as the value 4 in the LFO's mod amount... worth noticing, or you can get confused.

I kind of see what you're saying. Is there a way graphically I could see how amount is affecting the shape mod%?  For example, if I had shape mod at 40% and applied the LFO with an amount of 20, what would that make the shape mod%?

Also, in the LFO, is the amount going from 0 to 20 as the wave goes up and down(ex: triangle wave ramping up to 20 and ramping down to 0) in my example above?

It depend on what waveshape you choose for the LFO... some are unipolar others are bipolar... the noise and triangle are bipolar... they go both negative and positive... the others are unipolar and only go positive.

generally it's just a matter of experimenting with the LFO amount... the shape parameter has values from 0 to 99, so if you set it to 50 which is the middle value, then the amount will swing around that value... so if you choose the amount to be say; 20, then it would be modulating the shapemod parameter from 40 to 60 with 50 being the center... if you instead chose the sawtooth waveshape which is unipolar, it would modulate the shape parameter from 50 to 70...

You will have to experiment with this... the best way to do this is by routing the LFO to an oscillator's pitch... that will make you able to hear the modulation quite easily if you slow down the LFO rate... clipping will be rather obvious this way.

One thing to note is that the LFO amount goes much higher than the 100 values of the shape parameter... therefore, in the case of modulating the shapemod parameter, the maximum LFO amount should be 100 I think since that would give you a full swing around the center (-50 to +50 or rather 0 to 100).

I know this sounds a bit complicated, but it's actually rather logical and simple as soon as you get how it works.

That helped a lot. It's hard to find a good explanation of these things online. They either are super basic or jump to PHD level lol. I'll play around with it. One thing I wondered about is what clipping sounds like. Is it like a sort of popping sound?

[Razmo]

You're right with your drawing... but just remember, that the values are ADDED to the destination parameter... this mens that if you set osc shape to 50%, and then set the amount to more than half the range, then the destination value will actually clip (stay at max/min value longer since it's clipped). the shape mod cannot go below zero or above maximum value when modulated by a mod source, this in the above case you would have to set the amount to a maximum of half of it's range for it not to clip.

Also remember, that the LFO destination, (that only have a positive amount parameter) has a much finer resolution than if you set up the same modulation in the mod matrix... the mod matrix amount range is actually 4 times bigger than the LFO's own destination amount... thus an amount of 1 in the mod matrix is the same as the value 4 in the LFO's mod amount... worth noticing, or you can get confused.

I kind of see what you're saying. Is there a way graphically I could see how amount is affecting the shape mod%?  For example, if I had shape mod at 40% and applied the LFO with an amount of 20, what would that make the shape mod%?

Also, in the LFO, is the amount going from 0 to 20 as the wave goes up and down(ex: triangle wave ramping up to 20 and ramping down to 0) in my example above?

It depend on what waveshape you choose for the LFO... some are unipolar others are bipolar... the noise and triangle are bipolar... they go both negative and positive... the others are unipolar and only go positive.

generally it's just a matter of experimenting with the LFO amount... the shape parameter has values from 0 to 99, so if you set it to 50 which is the middle value, then the amount will swing around that value... so if you choose the amount to be say; 20, then it would be modulating the shapemod parameter from 40 to 60 with 50 being the center... if you instead chose the sawtooth waveshape which is unipolar, it would modulate the shape parameter from 50 to 70...

You will have to experiment with this... the best way to do this is by routing the LFO to an oscillator's pitch... that will make you able to hear the modulation quite easily if you slow down the LFO rate... clipping will be rather obvious this way.

One thing to note is that the LFO amount goes much higher than the 100 values of the shape parameter... therefore, in the case of modulating the shapemod parameter, the maximum LFO amount should be 100 I think since that would give you a full swing around the center (-50 to +50 or rather 0 to 100).

I know this sounds a bit complicated, but it's actually rather logical and simple as soon as you get how it works.

That helped a lot. It's hard to find a good explanation of these things online. They either are super basic or jump to PHD level lol. I'll play around with it. One thing I wondered about is what clipping sounds like. Is it like a sort of popping sound?

No, not really... it just means that if a modulation "clips" it has reached either it's maximum or minimum value of the parameter you are modulating... that means it will simply stay at min/max value until the modulation value comes below maximum, or above minimum value again... the effect of that is that it will most likely just sound like modulation is not happening for the time it's clipping... it's basically like if you if you walked down a hall with a long pole slowly raising it while you are walking... when it reaches the roof, it will not go any further, and it will stay along the roof until you start to lower the  pole again (this means that you are the modulating LFO here)... same goes for the floor, which in this case would be the negative swing... so to speak

This is also why I warned you about the fact that modulations are actually ADDED... this is important to understand, if you modulate the same parameter from more than one modulating source because then the maximum (and minimum) modulation value can potentially reach the combined value of the sources added to the destination parameter...

Also notice that the mod. matrix unlike the LFO amount can be set to a negative amount... this effectively invert the modulating source... also, it always work a 4 times that of the LFO's dedicated modulation amount (as I also wrote).. you need to take all these factors into account when you create modulations, and the more you add to a single parameter, the more complicated it'll get.

[godparticle]

No, not really... it just means that if a modulation "clips" it has reached either it's maximum or minimum value of the parameter you are modulating... that means it will simply stay at min/max value until the modulation value comes below maximum, or above minimum value again... the effect of that is that it will most likely just sound like modulation is not happening for the time it's clipping... it's basically like if you if you walked down a hall with a long pole slowly raising it while you are walking... when it reaches the roof, it will not go any further, and it will stay along the roof until you start to lower the  pole again (this means that you are the modulating LFO here)... same goes for the floor, which in this case would be the negative swing... so to speak

This is also why I warned you about the fact that modulations are actually ADDED... this is important to understand, if you modulate the same parameter from more than one modulating source because then the maximum (and minimum) modulation value can potentially reach the combined value of the sources added to the destination parameter...

Also notice that the mod. matrix unlike the LFO amount can be set to a negative amount... this effectively invert the modulating source... also, it always work a 4 times that of the LFO's dedicated modulation amount (as I also wrote).. you need to take all these factors into account when you create modulations, and the more you add to a single parameter, the more complicated it'll get.

That was a great explanation. All the little pieces I'm finding online are starting to come together. I guess the problem now is that conceptually, I'm starting to get it, but putting into practice and recognizing how it affects sound is something else entirely

[Razmo]

No, not really... it just means that if a modulation "clips" it has reached either it's maximum or minimum value of the parameter you are modulating... that means it will simply stay at min/max value until the modulation value comes below maximum, or above minimum value again... the effect of that is that it will most likely just sound like modulation is not happening for the time it's clipping... it's basically like if you if you walked down a hall with a long pole slowly raising it while you are walking... when it reaches the roof, it will not go any further, and it will stay along the roof until you start to lower the  pole again (this means that you are the modulating LFO here)... same goes for the floor, which in this case would be the negative swing... so to speak

This is also why I warned you about the fact that modulations are actually ADDED... this is important to understand, if you modulate the same parameter from more than one modulating source because then the maximum (and minimum) modulation value can potentially reach the combined value of the sources added to the destination parameter...

Also notice that the mod. matrix unlike the LFO amount can be set to a negative amount... this effectively invert the modulating source... also, it always work a 4 times that of the LFO's dedicated modulation amount (as I also wrote).. you need to take all these factors into account when you create modulations, and the more you add to a single parameter, the more complicated it'll get.

That was a great explanation. All the little pieces I'm finding online are starting to come together. I guess the problem now is that conceptually, I'm starting to get it, but putting into practice and recognizing how it affects sound is something else entirely

That will come with practice... simple as that no corners can be cut in that case.

[psionic11]

I guess my question boils down to am I understanding the basic elements.

Shape mod: controls how wide the pulse is
Frequency: how fast it triggers
Amount: how....high on the Y scale a pulse reaches(like if looking at an oscilloscope)? I don't really know what this does. It goes +/- 127

Not quite accurate the way you describe it, and therefore your understanding will be off a bit.  Try looking at it this way, from a physical point of view:

Shape mod:  changes the waveshape.  Twist the knob manually to hear the different shapes.
Frequency:  how fast you physically twist the knob back and forth around a set point.
Amount: how much you actually physically twist the knob back and forth.  Low amount, you barely twist the knob back and forth.  Medium amount, you make wider twists back and forth.  Max amount you twist the knob all the way counterclockwise then all the way clockwise, and back again, repeating. 

So for example, if you set the waveshape knob at noon, you get a certain sound.  If you twist the knob to 10 o'clock then to 2 o'clock, and keep going back and forth, this is a medium amount.  If you do this slowly, say back and forth slowly over 2 seconds, that is a slow LFO freq.  If you twist the knob back and forth quickly going from 10 to 2 o'clock 3 times a second, that's a quicker LFO rate.  Hope that clears it up for you.  It's really pretty intuitive, and should just "click" when you get it.

[godparticle]

That will come with practice... simple as that no corners can be cut in that case.

I'll definitely keep playing with it. I'm one of those weird people who has to know something inside/out, so I appreciate the patience

Not quite accurate the way you describe it, and therefore your understanding will be off a bit.  Try looking at it this way, from a physical point of view:

Shape mod:  changes the waveshape.  Twist the knob manually to hear the different shapes.
Frequency:  how fast you physically twist the knob back and forth around a set point.
Amount: how much you actually physically twist the knob back and forth.  Low amount, you barely twist the knob back and forth.  Medium amount, you make wider twists back and forth.  Max amount you twist the knob all the way counterclockwise then all the way clockwise, and back again, repeating. 

So for example, if you set the waveshape knob at noon, you get a certain sound.  If you twist the knob to 10 o'clock then to 2 o'clock, and keep going back and forth, this is a medium amount.  If you do this slowly, say back and forth slowly over 2 seconds, that is a slow LFO freq.  If you twist the knob back and forth quickly going from 10 to 2 o'clock 3 times a second, that's a quicker LFO rate.  Hope that clears it up for you.  It's really pretty intuitive, and should just "click" when you get it.

I think it's the amount...visualizing the amount that is throwing me off. I'm a visual person, so I have to see it plotted out. The other stuff I completely get. I found this online of how the LFO would affect the pulse wave. Even though it appears to be unipolar, how would the amount fit into that green line? It sounds more like the amount is the width of each green line segment(how long it takes to go from the bottom to top and back down)

[psionic11]

Ok, referring to your graph with the green line:

Imagine the green line as a straight horizontal line across the middle.  The shape of each pulse will be like the top graphic -- each pulse will be equal (meaning 50%, or a square wave).  The flat green line represents zero Amount.

Now imagine the green line almost straight, but just a bit like the bottom green line that is shaped like a rooftop (that is, imagine a very flat rooftop).  The shape of each pulse will still mostly be equal (mostly 50%), but the pulses at the peak of the (very flat) rooftop will be slightly less than 50% (narrower), while the pulses out at the edges of the rooftop will be slightly more than 50% (wider).

As the green line becomes more and more like a rooftop, which means more and more modulation Amount, the underlying original waveform deforms more and more, with the pulses under the peak of the roof becoming progressively narrower (and narrowest at the peak of the rooftop), while the pulses under the outer edges of the rooftop becoming wider.

Visual understanding is good.  But in this case don't let it override your ears and your fingers physically twisting the knob to understand modulation more intimately. 

Another way to say "twisting the knob back and forth" is to "oscillate" the knob back and forth.
Another way to say "how fast you twist the knob back and forth" is the "frequency" at which you twist the knob.

Hopefully you will get an "aha!" any moment now. =)

[psionic11]

By the way, it occurred to me that you might be trying to understand this only through the Mod Matrix.

Did you know that, in the LFO section with the 4 LEDs, you can press and hold an LED then twist the OSC1 or OSC2 Shape Mod knob?  This will quickly and directly assign that LFO to affect the Shape of that OSC.

Then you can easily experiment with the LFO Frequency and Amount knobs to intuitively see how they react and change the sound.  Make sure to turn off OSC2 so you can focus on listening to a single OSC.

Try putting OSC1 Shape Knob straight at noon. 

Also, start with very slow Frequency settings, with the Freq knob only at 9 to 10 o'clock at most.  Much faster than that and it begins to pulsate or buzz too quickly to hear the subtle changes.

[godparticle]

By the way, it occurred to me that you might be trying to understand this only through the Mod Matrix.

Did you know that, in the LFO section with the 4 LEDs, you can press and hold an LED then twist the OSC1 or OSC2 Shape Mod knob?  This will quickly and directly assign that LFO to affect the Shape of that OSC.

Then you can easily experiment with the LFO Frequency and Amount knobs to intuitively see how they react and change the sound.  Make sure to turn off OSC2 so you can focus on listening to a single OSC.

Try putting OSC1 Shape Knob straight at noon. 

Also, start with very slow Frequency settings, with the Freq knob only at 9 to 10 o'clock at most.  Much faster than that and it begins to pulsate or buzz too quickly to hear the subtle changes.

Yeah, I know how to do all of that stuff. That was what initially confused me because mod matrix has -127 to 127 amount while the LFO section has 0 to 127.

As far as your explanation of my graphic, it kind of sounds opposite of what I see. When I look at it, I read as having an initial square wave of 50%. Adding the LFO(triangle) amount makes the resulting square wave smaller on the edges, then fatter in the middle as the LFO peaks out, then back down

As far as sound, I just notice it sort of a warble that gets more pronounced to the point were it kind of sounds like a repeating on off square wave sound with amount turned way up

I hope y'all aren't getting sick of me yet lol

[Razmo]

Understanding sound design down to the deepest level is key to getting the most out of your synth, so do not see that as a disadvantage... when you know stuff all the way to the bottom, you're much better off than only understanding the "top layer" ... it's been my own driving force to be honest, going all the way back to the 80's and 90's where I coded hardware assembly language... i want to know it ALL!! 

A few newer synths actually have displays showing the shapes of a modulating source (and also the audio waveforms), something that is very handy... and something I'd like to see from Sequential in the future too... understanding a single LFO, how it swings etc. is not a problem, but when you start to modulate a destination with more than one source, it will be very handy to actually be able to see in real time, how that combined source is actually moving... but I guess we can only dream about that feature for now...

[psionic11]

Yeah, I know how to do all of that stuff. That was what initially confused me because mod matrix has -127 to 127 amount while the LFO section has 0 to 127.

As far as your explanation of my graphic, it kind of sounds opposite of what I see. When I look at it, I read as having an initial square wave of 50%. Adding the LFO(triangle) amount makes the resulting square wave smaller on the edges, then fatter in the middle as the LFO peaks out, then back down

You're right, I glanced too quickly at the graphic.  At the peak of the roof, the original square wave is fattest due to the LFO being at max modulation.

So to envision what negative modulation does, just flip the rooftop upside down into a V.  Now, at the most extreme modulation (the bottom of the V valley), the original square wave is narrowest.

The key thing to remember is that zero LFO modulation is represented by a flat horizontal line that divides the waveform graphic in half.  That is the zero crossing point, where LFO Amount = 0.

With 0 to 127 LFO amount, you get a rooftop like in the graphic.
With 0 to -127 LFO amount, you get an inverted rooftop (the V or valley).