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

[Sound Author]

You can only switch the tuning manually for now, it can't be automated or assigned to Midi which would be a great feature for an update.

Thanks - so if I want to record, say, an acoustic guitar finger picking around A, E & C, I would record the guitar track, then turn on Kaleidoscope and manually set root notes appropriately - how do I record that though?

My experience with resonators is that if you have a resonant filter tuned to a specific note, and then you play the same note (feeding your instrument's output into the resonator) you will likely experience an unpleasant "howling" effect that usually clips over 0db. You're probably better off filtering the musical data from your recording, keeping only the noise of the pluck, and then feeding it into KS.

Resonators work well with noise, which is why there's a built-in noise oscillator. With musical information, you can get some very unpredictable results. However, I believe you can tinker with granular delays in FIR mode, which is a whole different ballgame. But I'm pretty sure FIR mode won't process your input the way the resonant filters will. You might have better luck with FIR if you want to process an acoustic instrument.

Otherwise, the resonators will probably just scream at you all day.

[Sound Author]

Now that I think of it, would be pretty cool to trigger the slider for the white noise oscillator with a sidechain signal coming from a guitar output. Hehe. MWAHAHAHAAAA!!

[Banquet]

Thank you both for your help and advice. Kaleidoscope is very dreamy and addictive. I just need to get my head around how to build a complete composition with it. I can see that you could record each chord manually and render it, but that seems cumbersome. It's almost like you're not supposed to change the note or else they'd have built in a way of doing it via midi.

[lnikj]

You can only switch the tuning manually for now, it can't be automated or assigned to Midi which would be a great feature for an update.

Thanks - so if I want to record, say, an acoustic guitar finger picking around A, E & C, I would record the guitar track, then turn on Kaleidoscope and manually set root notes appropriately - how do I record that though?

My experience with resonators is that if you have a resonant filter tuned to a specific note, and then you play the same note (feeding your instrument's output into the resonator) you will likely experience an unpleasant "howling" effect that usually clips over 0db. You're probably better off filtering the musical data from your recording, keeping only the noise of the pluck, and then feeding it into KS.

Resonators work well with noise, which is why there's a built-in noise oscillator. With musical information, you can get some very unpredictable results. However, I believe you can tinker with granular delays in FIR mode, which is a whole different ballgame. But I'm pretty sure FIR mode won't process your input the way the resonant filters will. You might have better luck with FIR if you want to process an acoustic instrument.

Otherwise, the resonators will probably just scream at you all day.

Very informative. Thanks!

[Sound Author]

Cool, it works! I just sidechained the peak signal from a drum loop to the white noise slider. Could probably do the same with the wet and dry sliders as well. Cool beans!

[Andrew Souter]

You can only switch the tuning manually for now, it can't be automated or assigned to Midi which would be a great feature for an update.

Thanks - so if I want to record, say, an acoustic guitar finger picking around A, E & C, I would record the guitar track, then turn on Kaleidoscope and manually set root notes appropriately - how do I record that though?

My experience with resonators is that if you have a resonant filter tuned to a specific note, and then you play the same note (feeding your instrument's output into the resonator) you will likely experience an unpleasant "howling" effect that usually clips over 0db. You're probably better off filtering the musical data from your recording, keeping only the noise of the pluck, and then feeding it into KS.

Resonators work well with noise, which is why there's a built-in noise oscillator. With musical information, you can get some very unpredictable results. However, I believe you can tinker with granular delays in FIR mode, which is a whole different ballgame. But I'm pretty sure FIR mode won't process your input the way the resonant filters will. You might have better luck with FIR if you want to process an acoustic instrument.

Otherwise, the resonators will probably just scream at you all day.

The trick, as mentioned a few times earlier in the thread and also in the manual, when applying KS to highly tonal material is simply to use lower feedback if the KS tuning aligns with the notes in the source material....

When using low feedback KS acts like a very complex filter, and there is not much ringing or strong resonance.

when using String Mode, Feedback = 0.0 essentially coverages into FIR mode. There is NO feedback, and therefore no resonance. However when using String mode each delay "tap" still has it's own Damping filter. Technically if there is no feedback, there is no damping per-se bc damping is action that should occur over time... So If using String mode with Feedback = 0.0 ow have a bank of 512 (1024 actually as it is a stereo process) delayed filters.... you can use Band Pass Filters to get some very cool things this way... Also String mode with zero feedback can achieve non-integer sample delays for each of the delays.

FIR mode simply gives literally pure integer sample delays and nothing else. But when you sum a ton of them and their gain and panning can change over time in complex manners, this still gives VERY interesting results...

Also for highly tonal input material, you can use atonal tunings that do NOT align with the source frequency content. You can use low to medium feedback values, say up to 50% or so, and this will give results similar to the modal response of some weird instrument body and can color your musical input in very interesting and time-varying ways...

But yes, if you feed a resonator that tuned to 440z with high feedback, a sine way that is at 440z, you will get a LARGE increase in gain. This can be cool. But if it is not desirable the above are the suggestions to avoid it.

Also WHITE is your friend in this cases. White is not simply adding white noise. White up to 100% "erases" pitch data in the input signal while maintaining rhythmic data. If the input is silence, while 100% will NOT do anything for example. This is important to realize and is a very powerful too. It produces enveloped white noise that is enveloped by your input signal!

White 200% is "always on" blanket of white noise for input with total disregard of the input signal. This is used when you want to use KS for purely generative synthesis applications...

[Andrew Souter]

Now that I think of it, would be pretty cool to trigger the slider for the white noise oscillator with a sidechain signal coming from a guitar output. Hehe. MWAHAHAHAAAA!!

You don't need to do this. White 0 to 100% already effectively does this...

White > 100% starts to mix in the "blanket" of "always on" white noise.

Up to 100% it will be enveloped, and will follow your input signal exactly like a side-chain comp scheme, well actually not exactly as side chain has it's own slow behavior with attack/release/sustain/threshold/etc and in our case all of this is unnecessary and white is exactly perfectly in time with no attack etc.....

[Andrew Souter]

Thank you both for your help and advice. Kaleidoscope is very dreamy and addictive.

Thanks! Don't forget to eat!

I recommend protein shakes. This way you can have one by your side for hours and never have to leave the computer.

Personal favorite:

Met-Rx Protein Plus
Almond Milk (or soy)
One Banana
Frozen Blue Berries, Rasberries, Black Berries (from Trader Joes is in the USA)
Walnuts (or almonds, or other nuts)
Ice.
Blend. Drink slowly over 2-3 hours.

This is a very potent brain (and body) food. You can do great work with such fuel!

I just need to get my head around how to build a complete composition with it. I can see that you could record each chord manually and render it, but that seems cumbersome. It's almost like you're not supposed to change the note or else they'd have built in a way of doing it via midi.

Changing the frequency of (up to) 1024 resonators requires a "shitload" (yes that is a technical term ) of calculations to be performed. This is why we don't make it easy to change such as allowing MIDI input. If you are constantly changing the freq of all resonators, this would hugely increase CPU usage over what it is already. This is a short term goal however. We are working on it.

For musical use in composition etc. There are two basic approached:

1) Follow existing music by using chordal tunings, and yes, render each chord that you need and then cross-fade them in your DAW.

2) Use Melodic/Scale tunings and allow KS to generate new an unexpected melodies and chord progressions, and then compose and perform your human elements around this.

Both approaches are great. I use both personally. See earlier in the thread for more details on this.

[Phorous]

Having some fun trying to make a lush sounding evolving pad / texture with B2.

https://soundcloud.com/voice303/celestial-kaleidoscope

[Andrew Souter]

Speaking of CPU usage:

As mentioned earlier in the thread. There was an issue with CPU usage in high Freq String Resonators.

We have just fixed that. String resonators are now MUCH more efficient. They are now almost as efficient as Spring resonators.

Low freq string resonators are about 33% more efficient, and high freq resonators are something like 1300% more efficient... which is utterly ridiculous, almost so much that we should consider the current string resonator CPU usage to be a bug. Also strings can now go up to Nyqist instead of 1/2 Nyquist, which also ultimately saves CPU usage as it no longer requires Oversampling to cover the whole audio range.

I suspect a lot of the CPU usage and "crackling" reports used String resonant modes.

This is now resolved, and we will upload a new demo and full version later tonight or tomorrow.

(And we are continuing with some other optimizations too... and might be able to obtain another significant boost... not quite as crazy as this one though...)

[Sampleconstruct]

Low freq string resonators are about 33% more efficient, and high freq resonators are something like 1300% more efficient... which is utterly ridiculous

WOW

[arachnaut]

This is now resolved, and we will upload a new demo and full version later tonight or tomorrow.

Looking forward to this.

[Aiynzahev]

I'll try to give it another go in that case, nice work!

[Banquet]

Thank you both for your help and advice. Kaleidoscope is very dreamy and addictive.

Thanks! Don't forget to eat!

I recommend protein shakes. This way you can have one by your side for hours and never have to leave the computer.

Personal favorite:

Met-Rx Protein Plus
Almond Milk (or soy)
One Banana
Frozen Blue Berries, Rasberries, Black Berries (from Trader Joes is in the USA)
Walnuts (or almonds, or other nuts)
Ice.
Blend. Drink slowly over 2-3 hours.

This is a very potent brain (and body) food. You can do great work with such fuel!

I have beer at the moment, which is making Kaleidoscope even more wondrous and mind-boggling, but I think your suggestion is much more healthy!

Thanks for the tips, I'll check earlier in the thread to find out more. I also have to get my head around Kaleidoscope being an effect track which I can't record, even though it's making a sound! What is the best way to record it with just itself generating white noise? Do I create an audio track and feed it with just one note to make sure it records?

[arachnaut]

Personal favorite:
Met-Rx Protein Plus

I get this at Trader Joe's as well.
Another good one is dotfit leanMR.
I prefer plain vanilla.