Hi,
I'm trying to figure out the logic behind the attached patch. [vd~] receives continuous data from a [phasor~]. so the delay value is continuously changing??? why?
Also, I wonder why the sawtooth wave is inverted (by the [*~ -1] and [+~ 1])
Thanks a lot!

http://www.pdpatchrepo.info/hurleur/3-7-2-1-granular-live.pd

It looks to me like he's using the delay to play each grain backwards. That's why the phasor is inverted.

Hi Psyko,
Thanks for replying. I'm not sure about the [vd~] object. I thought you just need to give it a value of the delayed time, but in the patch continuous changing values are sent to [vd~]. Not sure what it means.
thanks

[delread~] is for fixed delay times. [vd] is interpolated to permit delay time modulation, for pitch shift, chorus effects, etc.

Hi acreil,
Thanks. I guess I'm missing something here. If you change the delay time (or the starting point of the buffer) with a [phasor~] how can you get to the point where the buffer's values are sent out? you just jump from one starting point to another, no?
And when you write interpolation, do you mean that the input values are interpolated?
Sorry for being a bit slow here. Thanks a lot

Picture a delay as a circular buffer. [delwrite~] is the write pointer, with the size of the buffer given by its argument. [delread~] is the read pointer, with the distance from the write pointer given by the delay time. You can move the read pointer relative to the write pointer to change the delay time, but you'll get clicks because it's not continuous. [vd~] permits continuous modulation, but since it's resampling the signal in the delay buffer, it needs to be interpolated. Anyway, it's not resetting anything or clearing the delay buffer, or anything like that when you change the location of the pointer.

Hi acreil,
Thanks, it's clear now!
Going back to the patch I posted earlier...If I understood correctly, it goes like this:
The write pointer is moving along the buffer, in a continuous way - when it reaches the end (after 5 sec), it jumps back to the beginning. Meanwhile, the read pointer is following the write pointer, starting from the length of the window (in ms), getting closer and closer to it (the delay time is becoming shorter by the inverted [phasor~] till it reaches 0 by the end of a period...). Is that right???
THANKS!

Yeah that seems right. I find it pretty interesting that you think of the write pointer as something that travels along with and in front of the read pointer. . . while I have a much easier time thinking of the write pointer as something that's fixed while read pointer moves in front of it.

Different thinking, same idea I suppose.

OK, Thanks wufu...the credit goes to acreil. it makes sense since the buffer's size is fixed, so in order to record the write point moves within the buffer's boundaries.

"you think of the write pointer as something that travels along with and in front of the read pointer. . . "

like a tape head!

Exactly. I really just have a bunch of tape flying around in my head when I think of this stuff.

At least now I know a bit more about how a software buffer actually works.