I found a way to do feedback routings

[u-u-u]

I know, a bit much but here are some good (Kraftwerk-inspired? no, actually just playing around ) siren sounds. (playing around with preset in previous post and using a Resonator in the feedback path)

[Taika-Kim]

What I was doing yesterday:

The camera for some reason lost focus, probably when I was walking past it after turning it on, and never refocused back So sorry for the blurry video. If you have questions, I'll be happy to ask. The sound is Phase-4 played through my Resonant Octaves preset.

https://youtu.be/36yGrKtbPTY

[Tj Shredder]

cool...

[Hez]

This is the best thread on this forum for a long time! Cheers for the Polynom in the feedback path tip, it's wild how rapidly you can change the timbre & behaviour with it. It's a really good way to sit right on the fence between chaos and nothing happening - change one variable slightly and the whole thing blows up in a good way.

Can we all petition the devs for a slew limiter modulator next release? I've had a little bit of success hacking one together with a Steps mod but a fully featured one would be really handy. I've had the best results with the low pass filter right down on the audio rate modulator to get some more 'macro' level features out of the results (instead of just varying the timbre of some fairly brutal noise), but it would be really great to be able to do that at any point in a chain of modulators.

EDIT: My god, I put a flanger in the Phase-4 FX (pre-feedback loop) and start modulating feedback and time, this is the closest anything I've ever made has come to Qebrus (RIP)...

EDIT2: Zero effort put into this, just started bashing notes (can you tell I'm not a keyboard player...) and fiddling with a few controls. All Phase-4, Bitwig Flanger (inside the phase-4 FX slot which is pre-feedback loop), xfer Ott outside the feedback path (just to bring some of the detail up) and a bit of limiting. It's definitely in noise territory rather than music but I'm amazed how many interesting timbres resulted from such a basic patch. Watch your ears, it gets pretty grim. https://www.dropbox.com/s/8yi45w9t54ciw ... 3.mp3?dl=0

[Taika-Kim]

The sample and hold module can also be used to smooth the modulation signals. But a slew with separate settings for up and down slopes would be good.

Care to post that patch with the flanger? That sounds interesting.

[Hez]

https://www.dropbox.com/s/pasl8nhiw5ztw ... wclip?dl=0

Patch includes xfer ott which I just disabled, I've replaced it with the rebuild of Ott which a user uploaded here a while ago (somethingtohide I guess, based on the patch author?) which sounds roughly the same.

I'm running quite a high buffer size at the moment (2048) so it may sound completely different at normal buffer sizes. I also screwed with the depth of some of the modulation while I was recording; I've tried to set it back to a roughly cool range. There's some automation recorded along with the clip which covers a few of the parameters I found interesting. Another essential one would be gain of the audio rate modulator generating the feedback path obviously, but I accidentally deleted that lane and lost the undo.

The original patch was based on u-u-u's feedback -> button -> osc 1 phase idea, but I ended up getting rid of the button & stopping the oscillator with ratio 0 like somebody suggested earlier in the thread which helps to get some of the really janky stop start behaviour going.

One cool thing about it is you can fade in another oscillator which isn't stopped to get actual notes if you want them (I used osc Y). Depending on the various interactions, you can get a bit of consonance & melody in there but as soon as you drop the gain of the 'musical' osc a tiny bit it heads back into chaotic territory. With a lot of refinement I think you could maybe get some interesting neuro basses or psy leads out of it.

I will watch your vid in a second Kim, looks really interesting!

I tried to use S+H for slewing in the past but I couldn't quite figure out how to get useful slew behaviour/timings out of it - I assume the smooth time is based on the clock rate in the modulator which defines the sampling rate, but it didn't really behave like I expected.

[justin3am]

But a slew with separate settings for up and down slopes would be good.

Before Bitwig 2.0 was released, I had access to builds with the modular system enabled. I put together a modulator device that was like a cross between the Random and Sample and Hold modulators (largely based on the concepts of the Buchla SOU). It had independent rise and fall times for the slew. It did a lot of neat tricks.
Boy that was a fun device (but it was like the old v1 modulators, where you had to have the target devices within the modulator's device chain). I'll have to dig that out to see if it still works.

[u-u-u]

To bring a new aspect to the thread: With the feedback routing, a sample can now scratch itself...

Sampler Self Scratchin'.JPG

Edit: and here a demo clip + preset

Sampler Self Scratch.zip

[dioxine]

Interesting, why Bitwig guys almost cancelled that feedback feature? this is a great founding i use mostly in Ableton to create wonderfull echoing atmos. Ehh, very pity.
Another thing i really miss here is – groove pool. If you are not great player, in Ableton, you can play stuff live badly, for example, piano, but with expression, and then save it as a groove pattern. Then, using mouse, create identical, but ideal melody and put that saved groove on it. It becomes alive

Except this, Bitwig just far ahead with it's modulations and now with The Grid. Ok, M4L is a great stuff, but Max is too hard for understanding, to create something like simple synth.

[x.iso]

Interesting, why Bitwig guys almost cancelled that feedback feature? this is a great founding i use mostly in Ableton to create wonderfull echoing atmos. Ehh, very pity.
Another thing i really miss here is – groove pool. If you are not great player, in Ableton, you can play stuff live badly, for example, piano, but with expression, and then save it as a groove pattern. Then, using mouse, create identical, but ideal melody and put that saved groove on it. It becomes alive

Except this, Bitwig just far ahead with it's modulations and now with The Grid. Ok, M4L is a great stuff, but Max is too hard for understanding, to create something like simple synth.

It's possible to recombine two midi sources, where you have one playing melody and other the rhythm pattern. Not the same thing exactly, but can be used in real-time.

Easiest way to go is using Channel Filter device, turning channel 1 off (or others if they're used) and adding Note Sidechain modulator with 0% curves and 100% sustain to turn channel on whenever notes come from rhythmic pattern. In case of adding notes to percussion, you add note receiver device first, to receive notes from melodic source, then apply what I've described above.

[brownerthanu]

So the latest official Bitwig video shows that you can route through the Long Delay module to create feedback loops in the Grid with 0.2 ms delay time.

On one hand that’s great! On the other hand Bitwig really sucks at community building. Why didn’t someone just say this earlier when people were bursting their heads complaining about it? After all this is the official Bitwig support forum. Anyhow, I’m glad to see it.

[Tj Shredder]

It was mentioned pretty early, maybe even in the introducing session of Dave at super booth. Anybody should have seen all of Daves videos...
Even I published an example which was necessary for doing algorithm 4 and 6 of my GriDX7 project...
Steinberg sucks at community building, not Bitwig...
Btw. the minimum delay is not 0.2 ms, its exactly one block size of your dsp settings. This is the same for all software which allows feedback. The absolute minimum would be 1 sample delay... (as block size it would bring your CPU to halt...)

[brownerthanu]

It was mentioned pretty early, maybe even in the introducing session of Dave at super booth. Anybody should have seen all of Daves videos...
Even I published an example which was necessary for doing algorithm 4 and 6 of my GriDX7 project...
Steinberg sucks at community building, not Bitwig...
Btw. the minimum delay is not 0.2 ms, its exactly one block size of your dsp settings. This is the same for all software which allows feedback. The absolute minimum would be 1 sample delay... (as block size it would bring your CPU to halt...)

Got it. Thanks for the correction!

[Tj Shredder]

Important with dsp audio feedback at that minimum delay, the result will sound different for different block sizes and sampling rates and certainly different than pure analog feed back like you can have in Eurorack modulars...
But in most cases it doesn‘t matter too much, as results are in the area of unpredictability anyway...

[x.iso]

It was mentioned since first betas that there are 2 ways of feedback routing - through Long delay and modulator out. both apply block size, but Long Delay seem to have fixed minimum delay as block size even without feedback connection used.