TZF Flanger question

[Digital Larry]

What it lacks is the effect DrAlx mentioned in post10 where it goes through zero then quickly turns round and goes back through it the other way before spending more time the other side. I think that's the missing ingredient.

Ah-ha, that is choosing where to position the fixed tap, cool idea.  The samples I put up have the tap in the middle.  Having it at the end creates more drama.  You approach the zero point slower.  I'll have to figure out how to do that (make it so that I can tweak the width but one edge is always on or hanging over the zero point).

[anotherjim]

Larry -  What's the maximum delay of the sweep in those flanger clips? Currently mine is only 4ms, so I suppose it explains why I don't hear that "down the pipe" tone?

I've just realized that it'll be hard to A/B between delayed dry and zero delay dry with the same buffer length - the former only sweeps part of the buffer each way while the latter can use the whole length.

[Digital Larry]

Hi Jim,

I was goofing around with times between 5 and 10 msec I think.  By the way this was done using the simulation mode of the SpinCAD Designer software.  See my sig line below for a link.  Although it specifically creates DSP code for the Spin FV-1, you could use its simulation mode to audition concepts that could be taken into the analog domain or other DSP or brute force digital implementations.  As SpinCAD Designer is a Java program, it works on Windows, Mac, and Linux.

Cheers.
DL

[anotherjim]

Thanks, I should probably take a look at SPINcad.

For now, I'm cutting my teeth using a basic MCU.

I tried using a fixed delay (half way behind write pointer) for dry and can't hear any difference between this and zero delay dry - except, for the same buffer size, the sweep is halved for the former! Tried offset from centre too, not too keen on the quick turnaround around effect for that. Maybe on full range audio it would be good, but on guitar - not so much.

Funny thing, I don't get cancellation with negative regen at zero delay. My regen sample is the previous cycle output mix subtracted from the newest sample from ADC. Should I do it differently? That said, I don't really want total cancellation.

As I've said earlier, I only get total cancellation at zero delay when the sweep delay sample is subtracted from the dry sample for the output mix, but I would expect that.
I've made a sweep option that reduces the sweep range and has a small fixed offset behind the write pointer so I can stop delay going all the way to zero and don't  hear total cancellation.
All my mix ratios are 50:50 to keep things simple (don't have hardware mul/div!)
And I'm using triangle LFO (actually a timer/counter in up/down mode) - also to keep things simple.

[Mark Hammer]

Make sure you're using the right sort of signal to be processed.  Whether it passes through zero or stays firmly planted on one side of it, flangers always sound more dramatic when the signal covers a wide bandwidth.  White noise, or at least a ten-fingered string-synth chord, is often the best sort of test signal.

[anotherjim]

OK, but I'm only aiming at guitar fx - even though the full glory of flanging will evade it. A lowish sample rate of 31.25Khz and a desire to keep filtering simple means it's not going to be a studio processor. I'm aiming for the simplicity of an Electric Mistress - and simpler still.