KVR Audio

Dúnedain · Post by **Dúnedain** » Sun Dec 04, 2016 6:47 pm

How to make a more smooth saw in Spire, like the saw in the picture? That is: no abrupt transition from the highest value to the lowest value but, as the picture shows, a more smooth transition.

I know one of the LP filters in Spire can help me with this but I want to try it without filters only using a single oscillator.

I think I understand how a single oscillator in Spire works. You are in fact mixing two waveform but despite that and after playing with the ctrlA and ctrlB knobs I do not seem to get the smooth saw I want from the picture.

ere2learn · Post by **ere2learn** » Sun Dec 04, 2016 6:54 pm

Turn wt mix to the left or just use the init button at the top for a fresh patch

Dúnedain · Post by **Dúnedain** » Sun Dec 04, 2016 7:16 pm

Big apologies, it is there, despite that .... this is how it looks:

but this is what Spire suggests it is:

Which is definitely something different. The suggestion is a perfect harsh saw.

To make it complete a zoomed in view on the transition:

cron · Post by **cron** » Sun Dec 04, 2016 7:19 pm

The rippling at the transition is due to bandlimiting. A 'perfect' saw in the digital world would alias to buggery. The shape you see is what a saw with no frequency content above nyquist (half the sampling rate) looks like.

Dúnedain · Post by **Dúnedain** » Sun Dec 04, 2016 7:24 pm

cron wrote:The rippling at the transition is due to bandlimiting. A 'perfect' saw in the digital world would alias to buggery.

That part of the theory isn't 100% clear yet to me. Does it alias after conversion to analog?

The shape you see is what a saw with no frequency content above nyquist (half the sampling rate) looks like.

You are (probably) right. But then I wonder why the picture Spire shows in the oscillator doesn't match with this? What is the use of that picture after all?

AnX · Post by **AnX** » Sun Dec 04, 2016 7:45 pm

Its a general guide/indication of the shape. A square wave doesnt look anything like a square either....

exmatproton · Post by **exmatproton** » Sun Dec 04, 2016 8:09 pm

Dúnedain wrote:
cron wrote:The rippling at the transition is due to bandlimiting. A 'perfect' saw in the digital world would alias to buggery.
That part of the theory isn't 100% clear yet to me. Does it alias after conversion to analog?

The shape you see is what a saw with no frequency content above nyquist (half the sampling rate) looks like.
You are (probably) right. But then I wonder why the picture Spire shows in the oscillator doesn't match with this? What is the use of that picture after all?

nice explanation

https://www.youtube.com/watch?v=cIQ9IXSUzuM&t=1s

Dúnedain · Post by **Dúnedain** » Sun Dec 04, 2016 8:19 pm

exmatproton wrote:
Dúnedain wrote:
cron wrote:The rippling at the transition is due to bandlimiting. A 'perfect' saw in the digital world would alias to buggery.
That part of the theory isn't 100% clear yet to me. Does it alias after conversion to analog?

The shape you see is what a saw with no frequency content above nyquist (half the sampling rate) looks like.
You are (probably) right. But then I wonder why the picture Spire shows in the oscillator doesn't match with this? What is the use of that picture after all?
nice explanation

https://www.youtube.com/watch?v=cIQ9IXSUzuM&t=1s

I have seen it and have read the website as well, regarding this topic: what does it explain?

exmatproton · Post by **exmatproton** » Sun Dec 04, 2016 8:24 pm

Dúnedain wrote:
exmatproton wrote:
Dúnedain wrote:
cron wrote:The rippling at the transition is due to bandlimiting. A 'perfect' saw in the digital world would alias to buggery.
That part of the theory isn't 100% clear yet to me. Does it alias after conversion to analog?

The shape you see is what a saw with no frequency content above nyquist (half the sampling rate) looks like.
You are (probably) right. But then I wonder why the picture Spire shows in the oscillator doesn't match with this? What is the use of that picture after all?
nice explanation

https://www.youtube.com/watch?v=cIQ9IXSUzuM&t=1s
I have seen it and have read the website as well, regarding this topic: what does it explain?

It is explaining why the representation on screen is different from what is coming out of your speakers. He shows it with a square, i believe. Same principal.

Also, what bandlimiting is doing and what it is for.

*mind you, this is NOT an answer to your question in the first post. But to this one; "That part of the theory isn't 100% clear yet to me. Does it alias after conversion to analog?"

Dúnedain · Post by **Dúnedain** » Sun Dec 04, 2016 8:28 pm

Yes I know, but that is not the point being discussed. The question is why the saw in small oscillator the screen looks different to the saw bounced to the hard drive

exmatproton · Post by **exmatproton** » Sun Dec 04, 2016 8:30 pm

Dúnedain wrote:Yes I know, but that is not the point being discussed. The question is why the saw in small oscillator the screen looks different to the saw bounced to the hard drive

Yeah, i know. i've edited my post. It was merely an answer to;

"That part of the theory isn't 100% clear yet to me. Does it alias after conversion to analog?"

trying to be helpful

I don't know why Spire is displaying a different result in it's little display. Maybe because it is little ?

I dunno.

cron · Post by **cron** » Sun Dec 04, 2016 8:33 pm

Dúnedain wrote:Yes I know, but that is not the point being discussed. The question is why the saw in small oscillator the screen looks different to the saw bounced to the hard drive

Likely because users who don't understand (and indeed have no need to understand) what bandlimiting is will get annoyed that Spire (or any digital synth) doesn't produce a 'real' saw/square/etc wave. The analog equivalent is easier to understand intuitively, much cleaner visually, and it's what people are used to seeing.

Dúnedain · Post by **Dúnedain** » Sun Dec 04, 2016 8:34 pm

@ exmat

thank you
But even after Monty I still do not understand it: why does a perfect saw already in digital domain alias?

Dúnedain · Post by **Dúnedain** » Sun Dec 04, 2016 8:36 pm

cron wrote:
Dúnedain wrote:Yes I know, but that is not the point being discussed. The question is why the saw in small oscillator the screen looks different to the saw bounced to the hard drive
Likely because users who don't understand (and indeed have no need to understand) what bandlimiting is will get annoyed that Spire (or any digital synth) doesn't produce a 'real' saw/square/etc wave. The analog equivalent is easier to understand intuitively, and it's what people are used to seeing.

I am lost

cron · Post by **cron** » Sun Dec 04, 2016 8:41 pm

Dúnedain wrote:thank you
But even after Monty I still do not understand it: why does a perfect saw already in digital domain alias?

There is no such thing as a perfect saw in the digital domain. Such a thing would require an infinite sampling rate, as a perfect saw contains frequencies all the way up to infinity. Sure, you could draw samples in the shape of a perfect saw, but the D/A reconstruction of those samples doesn't simply 'join the dots' in linear fashion.

How to get a smooth saw in Spire