Looper / Recorder FX Plugin

[Christoph Hart]

@oskarsh no you would need to use one recorder and then probably a SNEX node which does the custom playback / slicing. I'm currently working on a demo project in the hise_tutorial repo that you can use as starting point.

[d.healey]

Slight tangent but how are things looking for getting some kind of time stretching? Would be very useful with a looper.

[oskarsh]

@Christoph-Hart Thanks will check out the demo once its ready!

[Christoph Hart]

@oskarsh Alright, I've pushed a project to the tutorial repo. Make sure to use the latest develop branch build, I had to fix a few SNEX things (the new backend still needs some polishing) and the recorder node is also better now.

https://github.com/christophhart/hise_tutorial/tree/master/RecorderDemo

[AHTV]

@Christoph-Hart said in Looper / Recorder FX Plugin:

@oskarsh Alright, I've pushed a project to the tutorial repo. Make sure to use the latest develop branch build, I had to fix a few SNEX things (the new backend still needs some polishing) and the recorder node is also better now.

https://github.com/christophhart/hise_tutorial/tree/master/RecorderDemo

Hi Chris, this demo make my HISE crash (I tried like 4-5 times), but other demo in the github is loading fine, can you check? Thanks

[Christoph Hart]

this demo make my HISE crash

Have you compiled the latest develop branch?

Slight tangent but how are things looking for getting some kind of time stretching? Would be very useful with a looper.

Actually I've found a super-awesome time stretching library which is MIT licensed:

https://github.com/Signalsmith-Audio/signalsmith-stretch

I've played around with it a bit and to my surprise the sound quality is on par with the Complex Pro algorithm from Ableton (which uses Elastique). The performance isn't quite there but the library isn't optimized very well (I think the author preferred simplicity and "educationalibility" over performance).

There are a few low hanging fruits (eg. replacing the custom FFT implementation with the fastest FFT implementations on each OS (IPP / vDSP / FFTW) and some other stuff, but I think this is the best fit so far.

[d.healey]

@Christoph-Hart Oh yeah that sounds good, I've just been watching the ADC presentation, guy seems to know what he's talking about :)

Do you see this being added as an effect or would it be a parameter of the sampler/group/samples? Or both?

[Christoph Hart]

@d-healey I'm currently testing it with a simple buffer player node but I can see it being used at multiple places.

[Adam_G]

hey @Christoph-Hart this is cool! how do we compile it? im getting these errors

[d.healey]

@Adam_G

https://youtu.be/24vH-oxX1m4

[Adam_G]

@d-healey thank you. when tying to complete this im getting an error during the compile dsp networks step "error C2386: 'snex_node_t'
: a symbol with this name already exists in the current scope". any idea what it means?

[d.healey]

@Adam_G I'm not familiar with that error (or snex) so can't help you with this one.

@Matt_SF Any idea?

[Matt_SF]

@Adam_G I'm not a c++ wizard so this error doesn't mean anything to me out of the blue. If you could share a snippet or a screenshot (even in private if you want) I might be able to spot what's wrong

[Christoph Hart]

Yeah, that's on me - the code generator chokes when trying to compile a SNEX node that has a comment attached. Either wait for the fix (incoming), or remove the comment from the SNEX node in the network.

[Adam_G]

@Christoph-Hart thank you i shall wait for your update @Matt_SF thanks ill definitely take you up on that next time. its just the recorder demo atm though

[Matt_SF]

@Christoph-Hart @Adam_G haha I don't know if I could have found this one