Minimum delaytime is not 0.02ms for feedbacdk loops , but 6ms

Official support for: bitwig.com
Post Reply New Topic
RELATED
PRODUCTS

Post

I noticed that the minium delay is NOT 0.02 ms as the long delay module says but actually 6.13ms.
Which is way too mutch for feedback loops
Image
In the screenshot you can see a phasor module going into a sine lookup table , feeding back into itself
Increasing the feedback should give a saw .

To be absolutely sure I did the same thing in reaktor , and used a delay set to 6ms in a feedback loop (phasor/ramp going into sine lookup feeding back into itself )
Results were identical .
This is bad news
Eyeball exchanging
Soul calibrating ..frequencies

Post

you're doing a feedback connection and that is always delayed by 1 block, so minimum delay would depend on your audio settings, lower the latency - lower the minimum available delay. It cannot be 0 anyway, even in Phase-1 module or Phase-4 synth feedback is delayed by 1 sample.

If Reaktor also implementing 4x oversampling, then it's a matter of time when that delay will be eliminated. If not, then that's why it can.

Post

The minimum delay time is not 0.02ms
Doesn't matter if you do a feedback loop or not
See here , noise into envelope , panned left and right
Right channel goes into delay module set to minum , the recorded output for the right channel is delayed by 6ms .
Reason why I posted the feedback loop before was because I could instantly hear it was more then 0.02ms when doing a feedback loop .
This example proves it

Image

Image

You can easily test this yourself
Eyeball exchanging
Soul calibrating ..frequencies

Post

Did you try with the Mod Delay? Mod Delay doesn't do feedback at all... so the delay value might be correct all the way down.

EDIT: I just checked with Mod Delay and Mod Delay goes all the way down to the lowest value... because it cannot do feedback loops

Post

Yes it seems the the incorrect readout is only with the long delay module, which is the only delay that is usable for feedback loops >
Still really a bummer :roll:
Eyeball exchanging
Soul calibrating ..frequencies

Post

It depends on your settings. The standard setting is 256 samples which is at 44.1 kHz pretty exactly 6.0090702947858 ms. If you want it shorter, you can set it as low as 32 samples which will make your CPU (and you) cry and still won’t go down to 0.02, but at least down to 0.7256... I simply assume, that the block size within the Grid is 4 times the one outside...
But squeezing the wave form is so much easier with the Grid oscillators, I would not insist doing it with a feedback PM anyway...
Your patch does what Phase-1 is doing, but with 1 sample delay btw...

Post

As pointed out above, the minimum feedback time cannot be shorter than the current buffer size for your audio interface. If you want a shorter effective feedback time, you have to make your buffer as small as your system can handle.

Post Reply

Return to “Bitwig”