Anyone got a ramp generator that cubically or spline-ally interpolates between control points?

(Yeah, "you can roll your own," but I'm in a hurry and would prefer to save time if possible -- it would almost be faster to run PluginCollider via vstplugin~ and generate control signals by LFDNoise3 but that's maybe a bit much.)

hjh

Doesn't [tabread4~] apply cubic interpolation? If you store a ramp to a table, you could use that with [line~] to read through it, couldn't you?

@alexandros hmm... so, rotate control points through a table, that could work. Let me try later.

hjh

@ddw_music +1 to @alexandros' suggestion, but also--do you remember having this exchange? https://forum.pdpatchrepo.info/topic/14762/fidelity-of-tabread4/2
tabread4c~ doesn't have those non-differentiable points if that matters to you.

Tried to use an array as a signal-rate shift register -- it ends up in the right place but when the phasor~ resets, the shift isn't properly synced and the output glitches.

"The series of [pd] objects is meant to force scheduling order" but clearly it doesn't:

Can't avoid the cyclone reference I think b/c Vanilla has no equivalent of poke~. (Actually I wonder if poke~ doesn't update the array on time.)

lfdnoise3~.pd

hjh

@jameslo said:

@ddw_music +1 to @alexandros' suggestion, but also--do you remember having this exchange? https://forum.pdpatchrepo.info/topic/14762/fidelity-of-tabread4/2
tabread4c~ doesn't have those non-differentiable points if that matters to you.

Good catch, but the shift register has to work first.

hjh

"moon-lib" has a tabread inbetween-er. the bazier formulas are available in pd vanilla somewhere. P1 P2 foci fomulas. one of them is in "context"-lib

@ddw_music I didn't stare holes at your patch, but it looks to me like you are trying to generate ultra-low frequency noise. If I've guessed right, here's my solution: ULFNoise~.pd ULFNoise~-help.pd

Instead of a shift register, I just maintain a 64 sample array of noise and refill it when I'm nearing the end, taking care to copy the samples relevant to the last read point to the beginning of the array so it can pick up where it left off.

@jameslo said:

Instead of a shift register, I just maintain a 64 sample array of noise and refill it when I'm nearing the end, taking care to copy the samples relevant to the last read point to the beginning of the array so it can pick up where it left off.

Yes, low frequency noise, just not linearly interpolated.

I was gradually arriving at the idea of a longer array.

But I wonder why the shift register doesn't work -- pretty sure the same logic in SuperCollider, with BufWr, would be fine as long as the unit generator order is right. If poke~ is broke, maybe it should be fixed.

hjh

@ddw_music said:

But I wonder why the shift register doesn't work -- pretty sure the same logic in SuperCollider, with BufWr, would be fine as long as the unit generator order is right. If poke~ is broke, maybe it should be fixed.

Ah, I see it... It isn't poke~ that's broke. When poke~ writes a new value in the middle of a block, any downstream tabread~ of any variety will read the new value for the whole block. (And SuperCollider would behave the same here -- my guess was wrong.) So the only way to make a shift register work is [block~ 1], which I won't do here. At any larger block size, the control point reads (not writes) are quantized to block boundaries and out of sync with the phasor.

So David's approach is right.

hjh

@ddw_music Just for fun, while I'm waiting on marching orders from a choreographer:
ULFNoise~ shift register version.pd

Hi, my array reading and other objects in ELSE use spline interpolation by default, like in SuperCollider.

I had it on my list to also clone their cubic low frequency noise objects. If people want it, let me know.

@porres said:

I had it on my list to also clone their cubic low frequency noise objects. If people want it, let me know.

Yeah, that's exactly what I wanted. I don't actually mind it being an abstraction; in the current situation (tight deadline) I'd have preferred not to have to spend time trying and failing to build it myself.

@jameslo I'd not have thought of the rzero-rev~ trick. Nice. (But I'd not use tabread4~'s formula, since tabread4c~ is better.)

hjh

BTW I ended up still using the array for initialization -- with the samphold~ - rzero_rev~ approach, d won't receive a nonzero value until the third cycle, maybe not OK for the very slow control signals I need in this project.

But this works!

lfdnoise3~.pd

hjh

@jameslo said:

@ddw_music Just for fun, while I'm waiting on marching orders from a choreographer:
ULFNoise~ shift register version.pd

My ELSE version. I don't get the interpolation formula, which one is it?

It was easy to adapt [rampnoise~] to use spline interpolation.

@porres said:

I don't get the interpolation formula, which one is it?

That's the "bad" one from tabread4~ where, in my abstraction, I used the "good" one from tabread4c~ -- see https://forum.pdpatchrepo.info/topic/14762/fidelity-of-tabread4/7

hjh

@ddw_music Just to throw in another way of doing the same thing …

Shift register with precalculated basis functions (or rather: basis functions read from a precalculated kernel): interpolated-noise.pd

I used this for an interpolated version of the Gendyn stochastic synthesis algorithm, which can be used as a simple noise generator as well … gendyn-interp.pd

@ddw_music said:

That's the "bad" one from tabread4~

So that is 'lagrange' interpolation, and it seems different than the one I use, but ok. For reference, here it is

double interp_lagrange(double frac, double a, double b, double c, double d){
double cminusb = c-b;
return((t_float)(b + frac * (cminusb - (1. - frac)ONE_SIXTH *
((d - a - 3.0cminusb) * frac + d + 2.0*a - 3.0 * b ))));
}

Now, what is bad about it?

where, in my abstraction, I used the "good" one from tabread4c~

Where is the reference to this formula and what is "good" about it?

I didn;t know about "LFDNoise3", is it a new one? Wow, I can't believe they have so many
I see this one is dynamic. Hard to to know exactly by looking, the documentation kinda sucks in clarity, but it seems the other one "hangs" on low frequency in a way that it can't speed up follow a frequency increase from a [phasor~] or something? Wow, that's bad  
So, like I said, I easily adapted [rampnoise~] to use spline, so it's a 4 point interpolation. Now, I wanted to see what kind of interpolation SC uses and it's also not clear or well documented. It just says "quadratically interpolated" or "cubic interpolated", and for someone not that much into math, this is not enough, can anyone help? What kind of interpolation is spline? I don't wanna look into their source code

@manuels said:

Shift register with precalculated basis functions (or rather: basis functions read from a precalculated kernel): interpolated-noise.pd

Hey, this is cool, I've never seen this. But hey, I don't know what you mean by "Shift register with precalculated basis functions (or rather: basis functions read from a precalculated kernel)" I don't know what "Shift register" is supposed to mean here, or "precalculated basis functions", or "basis functions read from a precalculated kernel", or what is "kernel" here I just wanted to make music, no one told me there'd be math

What I see is that you have 6 point interpolation, even for linear interpolation, and this is really something new and confusing, a whole new approach to me, as I'd just use two points here, and you have at least 3 for that. Anyway, I also see "quadratic" and "cubic", which was my question from before...

I guess this is an interesting approach that works well for didactic reasons. I think that if I really get what is going on here I'll better understand these kind of interpolations, but it doesn't seem like a good way to code something in 'C', huh? Or is it?