Best Modular DAW?

[pdxindy]

it is under the hood yes, and this will be opened from version two..



http://www.bitwig.com/en/bitwig-studio/forthcoming.html

What's not clear at the moment is whether that will make Bitwig a modular DAW or just a regular DAW with a modular device building environment that can interact with the DAW (as with M4L). In a fully modular DAW (like MuLab and Usine) the whole signal path is modular, including the routing of audio and midi in and out of tracks, sequencers, arps, audio recorders and other forms of processing component.

Since all the Bitwig devices are made in the modular environment and they include sequencers and routing devices, is there much difference between the two?

Unlike Live, Bitwig does not allow feedback loops. Besides that limitation seems like you would be able to do most anything in Bitwig v2

[camsr]

How about Usine Hollyhock.
(Sorry if it was mentioned earlier)
http://www.sensomusic.org/
they are working on a vst version of it.

For me it will be an instant buy when this happens.

the bottom image shows feedback?
http://www.sensomusic.org/wiki3/doku.ph ... de:plugins
Ok not sure what you mean by feedback.

As long as parameters adjust to vertical mouse movements, and possibly true digital feedback, me too

[Trancit]

Can please somebody tell me something about benefit of having feedback loops???

TBH I cannot imagine any scenario where I would wish that....

[pdxindy]

Can please somebody tell me something about benefit of having feedback loops???

TBH I cannot imagine any scenario where I would wish that....

I create feedback loops all the time in Bazille... for example, feeding the output of the filter back into itself.

[pljones]

Is the feedback latency-free? I'm just wondering how many cycles the host can manage before it has to fail to get the audio to the output... If it's taking it as the next frame of input (i.e. one buffer cycle of latency added), then I can see it working but not latency-free unless I'm missing something. (Honest interest in the problem, btw.)

[pdxindy]

Is the feedback latency-free? I'm just wondering how many cycles the host can manage before it has to fail to get the audio to the output... If it's taking it as the next frame of input (i.e. one buffer cycle of latency added), then I can see it working but not latency-free unless I'm missing something. (Honest interest in the problem, btw.)

From what I hear, that is the problem with DAW level feedback loops... you can lose plugin delay compensation

I never did feedback loops in Live and don't miss it in Bitwig. IMO, the DAW is not sampling at a high enough rate to make it sound good.

[whyterabbyt]

How about Usine Hollyhock.
(Sorry if it was mentioned earlier)
http://www.sensomusic.org/
they are working on a vst version of it.

After 3 years or so, Ive given up waiting for that.

[whyterabbyt]

Can please somebody tell me something about benefit of having feedback loops???

TBH I cannot imagine any scenario where I would wish that....

Do you switch them off in all your delay plugins then?

[camsr]

Is the feedback latency-free? I'm just wondering how many cycles the host can manage before it has to fail to get the audio to the output... If it's taking it as the next frame of input (i.e. one buffer cycle of latency added), then I can see it working but not latency-free unless I'm missing something. (Honest interest in the problem, btw.)

If you have any of the software in question, simply set up a feedback loop and input a one sample spike (dirac spike). The output should look like a DC signal, because it should feed back the spike every sample.
And make sure the sample of the spike is not sample rate converted also

[camsr]

Can please somebody tell me something about benefit of having feedback loops???

TBH I cannot imagine any scenario where I would wish that....

Do you switch them off in all your delay plugins then?

Ohhhh that's edgy

[jens]

Can please somebody tell me something about benefit of having feedback loops???

Delay?

[pljones]

(I'm going a bit OT, sorry )

If you have any of the software in question, simply set up a feedback loop and input a one sample spike (dirac spike). The output should look like a DC signal, because it should feed back the spike every sample.
And make sure the sample of the spike is not sample rate converted also

I don't, hence the interest.

I just feel analogue feedback must be difficult to emulate - with analogue, the signal is "just there", it's not being sampled at all, no buffering, where there's connection, there's electrons .

Digital feedback to me must mean delay. The worst case, of course, is if you loop externally - you get two buffers of delay (out then in). But whatever, you just don't know the result of "(function of inputs)" until the inputs are available. If you've the output of "(function of inputs)" as an input... hm... isn't the minimum delay the duration of "function"? If you're able to do "function" for all of the buffer within the time of a buffer (in "real time"), then you're what's generally known as zero latency (that's "no added sugar", not "sugar free" ), of course... But with feedback you've got at least that minimum delay before one of your inputs is available each time... haven't you? And that means your final input isn't there until the start of the next buffer...

At which point my universe starts to collapse in on itself...

[camsr]

(I'm going a bit OT, sorry )

If you have any of the software in question, simply set up a feedback loop and input a one sample spike (dirac spike). The output should look like a DC signal, because it should feed back the spike every sample.
And make sure the sample of the spike is not sample rate converted also

I don't, hence the interest.

I just feel analogue feedback must be difficult to emulate - with analogue, the signal is "just there", it's not being sampled at all, no buffering, where there's connection, there's electrons .

Digital feedback to me must mean delay. The worst case, of course, is if you loop externally - you get two buffers of delay (out then in). But whatever, you just don't know the result of "(function of inputs)" until the inputs are available. If you've the output of "(function of inputs)" as an input... hm... isn't the minimum delay the duration of "function"? If you're able to do "function" for all of the buffer within the time of a buffer (in "real time"), then you're what's generally known as zero latency (that's "no added sugar", not "sugar free" ), of course... But with feedback you've got at least that minimum delay before one of your inputs is available each time... haven't you? And that means your final input isn't there until the start of the next buffer...

At which point my universe starts to collapse in on itself...

Maybe the buffers are only 1 sample big? Inefficient but functional.

[pljones]

Maybe the buffers are only 1 sample big? Inefficient but functional.

I guess that depends on the function, of course -- some work on multiple samples. I'd expect VSTs in general to expect to have access to a full buffer (as IIRC, that's what the protocol gives them). So feedback would need to be at buffer level?

[Trancit]

Oh come on ... the delay answers are lame... and is available in so many plugins (even with modular feedback path like MUX have) ...no need for another implementation

And feedback back into the filter... I guess, this will sound quite drastic and nasty...
No more examples... and this is worth to judge every modular system to good or bad...?