ScriptNode Band Splitters - Control Frequencies?

[griffinboy]

@griffinboy

https://1drv.ms/u/c/6c19b197e5968b36/ESv7Pm8TL7FDqO6I521nz4UBp-jlUb6w0fVZpwXutzk_aA?e=zho87t

C++ Multiband split.

These filters *may" have some latency (linear phase). But I haven't checked yet to see if that's true.
(Should be easy to check just run two sine generators at once: one with Multiband splitters on it, one without. Then check for phase cancellation between the two sines)

[DanH]

@griffinboy Thanks. Do you have any without latency?!

[griffinboy]

@DanH

I'm not sure if it's possible without latency if you want the bands to sum perfectly. Maybe if using convolution / FIR?

You can try measure the latency of the filters I've not actually tested it, I assume there to be additional latency but I may have been mistaken.

[DanH]

@griffinboy cool I'll give them a whirl

[Lindon]

@DanH said in ScriptNode Band Splitters - Control Frequencies?:

I'm just trying out the pre-made Band Spitters in Scriptnode and for some reason when I compile the network I can't change the crossover frequencies - the controls do nothing. Before compiling it works fine.

Any ideas?

what do you mean by "Band Splitter" ? you mean the multiband filters LWR?

[griffinboy]

@Lindon

Linkritz filters I assumed

[Lindon]

@griffinboy said in ScriptNode Band Splitters - Control Frequencies?:

@Lindon

Linkritz filters I assumed

then these LWR filters work fine...and have done for several years(I think)_

[griffinboy]

@Lindon

When compiled? I've had problems where changing the cutoff doesn't work

[DanH]

@Lindon @griffinboy exactly this

When compiled? I've had problems where changing the cutoff doesn't work

[resonant]

@griffinboy said in ScriptNode Band Splitters - Control Frequencies?:

@griffinboy

https://1drv.ms/u/c/6c19b197e5968b36/ESv7Pm8TL7FDqO6I521nz4UBp-jlUb6w0fVZpwXutzk_aA?e=zho87t

C++ Multiband split.

Thank you for sharing this. It sounds smoother than stock Linkwitz Riley node.

In Plugin Doctor, I noticed a few decibel bumps at the crossover frequencies of the bands. Is this intentional?

[griffinboy]

@resonant

Thanks for doing an analysis!

The reason I haven't shared this dsp on a proper post, is because it's not quite done.
I'll take a look and find the cause, it's supposed to be flat! Maybe there is a latency after all that I need to compensate for, either that or the math isn't quite right.

Does oversampling it help?

[resonant]

@griffinboy said in ScriptNode Band Splitters - Control Frequencies?:

Does oversampling it help?

Just compiled with 4x Oversampling, same behavior. I hope it is solved, because it has a smooth crossover which is great.

[griffinboy]

@resonant

I'll have time to look at it tomorrow