2CAudio Kaleidoscope | It's A Trip | Latest Update 1.1

[Gone soft]

I hoped there was a way to silence the input and just then hear the resulting tuned resonances. Not sure if that even makes sense. I thought it was already possible with the input/output and/or dry/wet settings. Yeah, i still do not understand Kaleidoscope really but it is a unicorn for
ambient textures and rhythmic stuff

Have you tried turning up the degree of noise? With 100% noise you remove all harmonies from the input but the input controls the volume of the noise fed to the resonators.

At 200% you get a constant sounding drone (as long as the pictures chosen doesn't silence parts) unrelated to the input at all. Values inbetween the extremes can be interesting too, but I'd suggest you start out with exploring the extremes of 0%, 100% and 200% noise til you're fairly sure that part of it kind of makes sense to you.

EDIT: rememeber to do all this with 0% dry signal so you only hear wath Kaleidoscope actually does to the sound.

[sdv]

The "theory" is simply to use the waveform as a pure data format. Don't think of it as an audio file. Think of it a collection of 2048 "relative ratios" between 0% and 100% which can then be fit to any frequency range:

That's clever!!

[Andrew Souter]

thanks. we try.

It is indeed quite cool/novel for creating all kinds of unusual/atonal sound FX and ambiences appropriate for scoring work etc.

[Sampleconstruct]

Three instances of Kaleidoscope sequencer patches in 11/8 time signature processing white/pink noise and a drum sequence and a fourth instance adding a pad sound (which uses the same tuning file).

https://soundcloud.com/sampleconstruct/ ... -patchpool

[Andrew Souter]

awesome! I like this funky side of things! Fresh!

[Sampleconstruct]

awesome! I like this funky side of things! Fresh!

Thanks, Andrew.
----
Although I mainly use KS as a “synthesizer“ feeding it with all sorts of external, non-tonal noise sources and percussive sounds to unfold it's magic, it also works very well on tonal sources/instruments. In this example KS processes an improvisation played on a sampled celesta (Iconica - section &players), B2/Breeze2 and UVI Relayer provide the space.
https://soundcloud.com/sampleconstruct/ ... -patchpool

[sqigls]

it's there waiting for you as soon as you have Cubase 8, 9, 10 or maybe even 7.5.

Cubase 6.5 came out in 2012 and was already 3-4 years old when KS 1.0 came out as we discussed a few pages back. KS uses very modern coding techniques and is multi-threaded etc. It is unknown if hosts this old are compatible. The current versions of all major hosts at the time of its initial release (January 2015) and all releases since then are checked and compatible. And if anything ever pops up with new host releases we are happy to check, but we can't check hosts from 7-8 years ago as we simply don't have them anymore, nor are they typically even compatible with the latest OS's and most likely if there are problems we probably could not even make it work if we wanted to as it is likely the host/os does not support the kind of modern techniques the KS code uses. Sorry.

i have a license of Cubase 9.5 now and KS still can't make it past the Blacklist.
this DAW version release is about 6 months old.
ACTUALLY... now ALL the 2C plugins are on the blacklist, except Precedence.
I can roll Breeze 2 back to 2.0.3 and Cubase can load it, but i'm kinda devastated about this.
i really don't want to update my OS just for B2 and an updated Breeze 2... and (possibly) Kaleidoscope?
It's not possible really, i need to keep Cubase 6.5 installed for a few projects for now...

B2 is the only thing keeping me from jumping whole heartedly into C 9.5 though.
I've started to write without any 2C plugins, which kinda sucks
but i'm losing my patience

[Andrew Souter]

Re-enable them manually in Cubase 9.5 and higher. They all work fine. (This issue has no functional change, it was just some thing Apple made Steinberg adopt, and plugs are forced to be updated to pass automatically also, but it is functionally little more than a small extra hoop to jump through, and does not actually affect the performance of the plugs in any way AFAIK. If you simply reenable them everything is fine.)

Breeze 2 and Precedence should already work without having to do this. The next updates to all plugs will have this fix/tweak/hoop-jump applied also so that you do not need to manually reenable them.

I am personally running all our plugs on Cubase 9.5 & 10.x, on OSX (and also Win10). No problems.

[sqigls]

Kaleidoscope crashes instantly on re-activation.
Adding it to whitelist will just cause Cubase to permanently crash upon opening, until the plugin is removed from the vst folder.

[Andrew Souter]

What OSX version? I have KS running in Cubase 9.5, on OSX.10 (Yosemite) as I write this.

[sqigls]

10.9.5

[Andrew Souter]

Woah!

I just tested the shipping version of KS 1.1 on an 18-core Intel Skylake-X 7980xe (Win10, Cubase 10). I can run 80 instances real-time with 512 active lines/voices, Dual Modulation (both images in use), Dual Spring resonator mode (ie. the maximum CPU usage setting possible for Spring resonators). This is equivalent to 40,960 synth voices each with 4 LFOs!

Dual String was a little less than half of that, roughly as expected.

https://ark.intel.com/content/www/us/en ... 0-ghz.html

The recent 9980xe refresh is even faster, as is the 28-core Xeon-W.

Exciting times for CPUs!

[dayjob]

The next updates to all plugs ....

what can we expect for Kaleidoscope in this next update? just fixes/tweaks or will it get further optimization for non-28 core CPUs and have some more modes beyond the strings etc. to me it's crying out for something that can get it further away from that plucked comb filter resonator type sound. for me personally it would be more useful if it did some other types of things. regardless, it still is a bit crushing to my 2.2 GHz Intel Core i7

[acousticglue]

Cubase wins again lol

[Andrew Souter]

what can we expect for Kaleidoscope in this next update?
....
and have some more modes beyond the strings etc. to me it's crying out for something that can get it further away from that plucked comb filter resonator type sound. for me personally it would be more useful if it did some other types of things.

I can't really discuss the features bc I don't know how long it will be until it is released bc it is not completely in my control. It is NOT a minor update. It is extremely major with many new features -- maybe even more features than were in 1.0 in total. Yes there is lots of new and different sonic potential.

This was more or less complete in 2017 before we switched to Breeze 2 and Precedence work, but it was not stable -- it crashed very frequently and the code is extremely complex so we had to pause on it for a while to do Breeze 2 and Precedence. I don't know yet the schedule for it. Right now we are just about done with another major Breeze 2 & Precedence update. Then we will see.

just fixes/tweaks or will it get further optimization for non-28 core CPUs
...
regardless, it still is a bit crushing to my 2.2 GHz Intel Core i7

I do NOT expect major performance optimizations compared to the 1.x version. We did already spend a year more or less optimizing it from 1.0 to 1.1, and this performance increase was quite major, and is a big part of the reason why Breeze 2 and Precedence are so fast now too. These tricks are all already mostly applied in KS. KS is already extremely efficient compared to the amount of work it does. It simply does a ton of work. Maybe there will be comparatively minor things that can still be done, but I am not sure, but I doubt anything as extreme as "twice as fast" etc is possible at this point using the same instruction set and CPU. Also, some of the new features definitely use more CPU actually.

Certainly an AVX-512 version is possible and that could have a major performance benefit, but it will only help new CPUs that have AVX-512 like the ones listed above.

Perhaps it would be possible to apply our "suspend on silence" logic to KS also --- that would be very useful as you could then create a host project with tons of KS instances and do an arrangement, like above with 80 instances or whatever, but it is unusual to have 80 instances all playing at the same time in real world projects -- so medium to sparse arrangements like typical real world music projects would end up a lot more efficient if everything is left real-time.