The difference between ramp and saw is really important

[Architeuthis]

I just noticed in KarmaFX version 1.18 in the additive module, the saw is low on the left side and high on the right side. If something goes from low to high, isn't that a ramp? So that is mislabeled.

Isn't a ramp low to high?
Isn't a saw high to low?

Searching google images 'ramp waveform' and 'sawtooth' produce the same images.

However, there's one image that labels both as sawtooths and then makes the distinction between positive and negative ramp or "ramp up, ramp down". That seems sensible.



I hope everyone understands that all your synths sawtooths need to be all positively ramping or all negatively ramping or those sawtooths will interfere into metallic square waves.

What is the proper terminology? Whatever it is, it needs to be universal.

[Architeuthis]

Ok, interestingly enough if you build a sawtooth out of harmonics using sin and the phase of those sines are 0, you get what I think is the true sawtooth which is ramp down. I prefer ramp down because I like to think in terms of transients and usually modulating something from high to low is more often done (I bet).

[declassified]

It's also imperative that all your synths square waves need to be all top-to-bottom or all bottom-to-top or those squares will interfere into dismal silence.

[Architeuthis]

for those of you who are just starting out, the above is a joke.

[foosnark]

Ramp up and ramp down are the same thing with inverted phase. I don't think there's any particular standard for calling them "ramp" vs "saw", and they sound identical.

The main difference is going to be with LFOs, or when used to modulate the phase of another oscillator. Also a downward ramp gives cleaner triggers if you're using it in modular patching.

A saw core analog VCO or DCO is going to generally produce an upward ramp. A capacitor charges (at a rate that varies with the input voltage). When the voltage hits a certain threshold, a transistor closes a circuit that discharges the capacitor, which creates the sudden drop. Take a look at Stargirl's article on the Roland Juno DCO for a pretty good explanation.

For triangle core analog VCOs, the method used by the tri-to-saw shaper is going to determine which way the ramp goes but upward is also pretty typical. Minibrute for instance generates an upward saw. Tri-to-saw usually involves converting the tri to a square, and using the high/low state of the square to switch between the triangle or an inverse of the triangle.

But it's very easy to flip the phase and get a downward saw. Sometimes it's the filter that does that. So you'll see both upward and downward saws from analog synths.