Changing pitch...

A Former User

@Christoph-Hart do we need SNEX to make a varying delay line? or can we do it with fix_delay and some node wizardry?

A Former User

Sounds like dBlue glitch lol

Christoph Hart

Yeah, you will need a varying delay line - the fix_delay will not work because it fades between different fade times to avoid the Doppler effect, but you want precisely that effect for the pitch change.

Also you will need to use two parallel signal chains and fade between them in an intelligent manner in order to avoid the artifacts when going from the end of the variable delay line to the start. I've seen a matlab graph somewhere.

But this is actually a good idea for a tutorial project since it covers many things that you need for DSP development, so I might add it to the list of examples in the docs sometime.

A Former User

@Christoph-Hart That would be great! Is the FFT stuff for SNEX added yet? I think FFT and Pitch Shifting are among the most requested things that I've seen on the forum

Natan

@iamlamprey Mate, What is The Instructor For The Above Example??
The Screen is Cutted,

A Former User

@Natan You can click the video title to open it in a tab to fullscreen. I don't have SNEX workbench atm cause I formatted

Note: this is not even close to a pitch shifter, as shown in the video. It is kind of a fun glitchy thing to play around with but

Natan

@iamlamprey ah Thank You
It is Made On Snex? right?

But How i can Build Snex?
Where to Download?

A Former User

@Natan develop branch, then in HISE->Tools->SNEX workbench

Natan

@iamlamprey Thanks Man Will Check In A sec

ustk

@Christoph-Hart said in Changing pitch...:

Also you will need to use two parallel signal chains and fade between them in an intelligent manner in order to avoid the artifacts when going from the end of the variable delay line to the start. I've seen a matlab graph somewhere.

I've done this about a year ago but I never succeeded in Hise, always having artifacts despite the crossfades. I'm 99% sure my math was correct. But even when the crossfades reached -100 I still had artifacts from the delay nodes when responding to the ramp's edge, and no crossfade could cut them... I might try later in the new scriptnode ;)