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

[TexLoco5]

After hearing the results of KS ... I'm certainly intrigued ... but the talk seems to indicate that a good deal of computing power is needed ... and I'm definitely planning to upgrade. I've seen the specs in the manual ... which are rather general ... but for those of you that are using KS without any hiccups ... what are your specific processor specs?

KenB

[Andrew Souter]

Dry:

https://soundcloud.com/2caudio/bright-chords-dry


one instance KS. Nothing else:

https://soundcloud.com/2caudio/bright-c ... eous-lazer


factory preset "aqueous lazer groove" Ref pitched changed to D#, lowered Feedback some.


Interesting, actually this demonstrates a few different things:

1) rhythmic against textural pollyrhythym which created the interesting variations in the textual/rhythmic composite thing.

2) using atonal tonal with lowish feedback applied to tonal input...

If feedback is low, we get filtering effects as explained earlier.

So, actually, this kind of thing is interesting to so with waveform tuning also... (sorry to confuse matters... )

[rustman]

After hearing the results of KS ... I'm certainly intrigued ... but the talk seems to indicate that a good deal of computing power is needed ... and I'm definitely planning to upgrade. I've seen the specs in the manual ... which are rather general ... but for those of you that are using KS without any hiccups ... what are your specific processor specs?

KenB

I'm running an i7 2600K clocked to 4.6 Ghz. 16 Gb RAM on Win 7-64. Cubase 8.05 64 bit. I set buffer size to 2048 on my audio interface, which matches the default setting in KS. Cubase required me to set ASIO guard from medium to high to use this setting.

[KBSoundSmith]

After hearing the results of KS ... I'm certainly intrigued ... but the talk seems to indicate that a good deal of computing power is needed ... and I'm definitely planning to upgrade. I've seen the specs in the manual ... which are rather general ... but for those of you that are using KS without any hiccups ... what are your specific processor specs?

KenB

As I mentioned in my post a few pages back, I'm running a 2.53 GHz duo core, pre-Intel iX-series (it's a P8600 if I recall correctly). On my computer, many presets stutter badly, just not enough juice (plus I'm on 32-bit Windows Vista, which isn't officially supported). So in the piece of music I made with KS, I had to keep bouncing down audio in small chunks.

But while I can't run it in real-time, I'm not sure it's the biggest issue. I think for a lot of applications, bouncing down audio and assembling the portions you like best is a good workflow for KS, since the resulting audio tends to be a touch on the unpredictable side. If you have a really fast computer and have the power to run multiple instances, I can totally see it working as a send effect and having it be an interesting background to the foreground material being sent. But using it to generate brand new foreground audio and then cutting/pasting can get you into very new sonic territory and allows you to be really extreme with the results you obtain.

I might still buy it, despite my under-powered and not-officially-supported computer. The results, I think, are worth the workflow hassle.

[Andrew Souter]

more from the same source:

https://soundcloud.com/2caudio/bright-c ... ugs-in-the

https://soundcloud.com/2caudio/bright-c ... eat-box-14

https://soundcloud.com/2caudio/bright-c ... ter-walker

https://soundcloud.com/2caudio/bright-c ... lemetry-02

https://soundcloud.com/2caudio/bright-c ... ing-breeze

https://soundcloud.com/2caudio/bright-c ... he-yard-b2

Added Breeze or B2 to some... (although apparently I can't spell the same of my own product... )

just some quick simple examples for musical electronica kinda stuff...

I guess the ambient examples are pretty well covered from my full pieces, which are entirely KS and our verbs using White source.) Or you want more of that?

Full audio demos tomorrow.

[Andrew Souter]

How do you get 50% off the 64 bit? It isn't showing up in my shop. I bought the 32bit.

Use this coupon code:

ArchitectureWaveformsCG


(this will actually give 50% off ANY format if you have any other format...)

[Andrew Souter]

After hearing the results of KS ... I'm certainly intrigued ... but the talk seems to indicate that a good deal of computing power is needed ...

from the manual:

System requirements

Minimum Recommended Hardware: 4-core CPU with SSE2 support running at least at 2.0GHz and 4GB of RAM.

Kaleidoscope is a massively parallel algorithm with dynamic CPU usage. CPU usage scales linearly depending on the number of enabled lines (voices) used in a particular preset. The minimum number of active lines is 1 and the maximum is 512. Using 512 lines requires roughly 512 times as much CPU power as using 1 line! Additional factors such as the current Oversampling Ratio, Resonator Mode, Modulation Mode, and others also affect CPU usage.

Kaleidoscope is multi-threaded and can distribute its CPU load across a variable number of threads and cores depending on the current setting in the Info Page Preferences. It can be set to use all available cores on your system including virtual logical cores from hyper-threading. Kaleidoscope offers extreme control over threading behavior to allow you to find the best match for your current hardware, OS, host, and project needs.

Kaleidoscope has been optimized for two things: extreme Real-Time performance relative to the amount of computational complexity it represents, and extremely fast Offline render times even when using the most extreme settings. This is achieved both by multi-threading as well as offering a variable Buffer Size preference that allows users to choose between maximum efficiency and lowest latency.

Using the maximum number of lines together with Oversampling at low Buffer Size settings can require a large amount of CPU resources. Oversampling offers only a very minor benefit to the Spring Resonator Modes, and no benefit at all to the FIR Resonator mode; it offers the most benefit to the String Resonator Modes. We therefore recommend simply running at 1X Oversampling for Real-Time use, and reserving Oversampling for Offline bounces.

Kaleidoscope is a very different type of tool compared to 2CAudio reverb products. It is generally assumed that reverb products will be used live on track inserts or auxiliary busses and potentially many instances will be kept continuously running in a host session for many hours in the working day. Tools such as reverb effect mix decisions and therefore are typically kept live in sessions. Kaleidoscope is designed more as a sound-design and special FX tool. It should be thought of as content creator and extreme FX processor. While it is possible to run multiple instances live in projects, particularly when using large Buffer Size settings, the recommended work flow it to bounce or freeze KS instances once a desirable result has been achieved. Running many instances of Kaleidoscope live with maximum settings will quickly consume all available resources even on the fastest computers available in 2014-2015. It is important to get acquainted with the process of rendering audio FX in your host application via bouncing, freezing, or rendering-in-place to get the most out of Kaleidoscope. This is particularly true when using a low Buffer Size preference. Using the maximum Buffer Size setting VERY significantly increases efficiency, and dramatically lowers CPU usage at the expense of added latency.

Sound designers and power users may optionally run Kaleidoscope in a stand-alone stereo editor application such as Sony Sound Forge or Steinberg Wavelab in parallel to their multi-track DAW host or as a separate part of the production process. In such hosts, it is recommended to always work with the largest Buffer Size setting. In such scenarios the processing requirements of Kaleidoscope are surprisingly reasonable despite its incredible computational complexity.

Kaleidoscope is designed to offer new possibilities to the world’s best sound-designers, composers, and producers who are working on major Hollywood blockbusters and similar projects. A fast computer is assumed in such environments.

In summary, Kaleidoscope can be very demanding on CPU resources when using the maximum settings and small Buffer Sizes. When using these settings, the best user experience will be found by using the newest multi-core processors with high core counts, large caches, and high clock frequencies. Intel Xeon-class workstations are an ideal-case example.

[Andrew Souter]

But while I can't run it in real-time, I'm not sure it's the biggest issue. I think for a lot of applications, bouncing down audio and assembling the portions you like best is a good workflow for KS, since the resulting audio tends to be a touch on the unpredictable side. If you have a really fast computer and have the power to run multiple instances, I can totally see it working as a send effect and having it be an interesting background to the foreground material being sent. But using it to generate brand new foreground audio and then cutting/pasting can get you into very new sonic territory and allows you to be really extreme with the results you obtain.

This is this perfect advice for underpowered systems. This is exactly what the best sound-designers have been doing for a long time: generate content first as a separate process, and then assemble into your DAW as audio blocks/stems/parts etc. This a great way to work... even with faster machines.

If you were to run KS in a editor like Sound Forge to make content, and this is the only thing happening, it is actually pretty blazingly fast even on laptops. We put a lot of effort into making exports/renders very fast for this exact reason. This is one of the expected ways to work with it.

We really don't expect people to run many instances real-time. I did it myself for some of the ambient pieces, but that was in part just to see how far I could push things before something broke.


Also we expect some additional optimizations of the Sting resonator as mentioned earlier in the thread.

Ok, that's it for me for the night.

[
As I mentioned in my post a few pages back, I'm running a 2.53 GHz duo core, pre-Intel iX-series (it's a P8600 if I recall correctly).

I might still buy it, despite my under-powered and not-officially-supported computer. The results, I think, are worth the workflow hassle.

thanks.

[Sampleconstruct]

Exploring the FIR-Mode some more, here I use it on a reversed and pitch-modulated actor's voice from one of my musical theatre works, video starts dry, if you're somewhat paranoid, don't watch it:



All vids are here:
http://www.youtube.com/playlist?list=PL ... 2PxoUbW05k

[Sampleconstruct]

And here is something more dreamy, a simple sequence played with an acoustic guitar flageolet synth (MachFive 3) processed with a KS patch reminding of downsampling and tremolo:

[Beatworld]

And here is something more dreamy, a simple sequence played with an acoustic guitar flageolet synth (MachFive 3) processed with a KS patch reminding of downsampling and tremolo:

Nice, thanks.
That makes me think I would like to hear some Andreas Vollenweider fed into KS

[Sampleconstruct]

And here is something more dreamy, a simple sequence played with an acoustic guitar flageolet synth (MachFive 3) processed with a KS patch reminding of downsampling and tremolo:

Nice, thanks.
That makes me think I would like to hear some Andreas Vollenweider fed into KS

That would probably cause trillions of New Age-gasms around the world.

[tonkatodd]

Loving the demo. How long is it specially priced?

[jackmazzotti]

This maybe 2.0 level stuff but I would love to have the transpose menu have a midi input option that would allow for midi controlled transposition in real time allowing you to "play" the image that Andrew described doing with the mouse. KS I believe has opened a door into a new renaissance era of computer generated music. Endless thanks.

[jackmazzotti]

dont forget if you live 9 you can convert audio content in to midi tracks which opens a whole other realm of using KS for composition purposes