ForumsSoftware ← Kaivo Beta

Hey it's the Kaivo beta thread! 1.0b1 is going out to folks on the beta list right now. The beta distribution is closed now, but feel free to discuss anything you like about it here.

If you made any audio you would like to share, please throw it on the Soundcloud group and post a link here. I wish we could embed Soundcloud content here, but we can't. You can embed text links using Markdown---the syntax for that is described here.

Link to Soundcloud group

Just installed Kaivo and tried to play around with it in Logic Pro X. Got the following in Console for grains FM:

1/19/14 3:12:43.148 PM Logic Pro X[39976]: Audio engine overloaded,

(Repeating until plug-in turned off or another preset selected.)

I don't get those errors for the other presets, but the only presets that make any sound are Koto, Water Koto and Dulcimer — all roughly the same low volume, highly distorted.

Did you try using only one voice? You can also turn the body to "none", but then you don't get a body resonance of course.

The next beta will be much faster, so you can just try again when that comes out if you like.

I made three examples in Live and added them to the Soundcloud group, to start us off.

Koto and Water Koto sound very different when dropped to 1 voice. Dulcimer no different and all others still silent. No errors logged this time.

System info:
2.3 Ghz i7
16 GB DDR3

If "default" doesn't work then I'm guessing you don't have any samples loaded.

I just installed and ran the AU in Studio One (latest beta build). Firstly, the presets folder has to go in the USER library/audio/presets folder, not the root drive library folder as in the email. Also, the right arrow preset browser in Kaivo always goes back to the "arf" preset rather than moving to the next preset. Those are the first 2 things I noticed. It definitely hits the CPU pretty hard as mentioned. Looking forward to digging deeper into this beast!

Just got the beta, thanks for having me, Randy :)
The Koto, Water Koto and Dulcimer, seem to be the only ones that work, indeed. Only things I've found so far, are mentioned in the email. I'm on a new W8 x64 laptop(Asus-Core i5 4 GIG RAM). I have not run Kaivo on W7 X64 (about 3 years old)(Core i7 6 GIG RAM)I will to night, thanks :)

Firstly, the presets folder has to go in the USER library/audio/presets folder, not the root drive library folder as in the email.

I'm looking at the email I sent and it says to put them in the user Library.

Anyone else have only Koto, Water Koto and Dulcimer working? Try "default", does that work? Do you get sounds like the example clips?

Do other presets work if you set the res mode to metal string?

Heya Randy,

Got all the presets working, noticed the heavy CPU spikes but am running all 4 voices.

Stupid questions:
Does the release CV Input on Envelope 2 work?

Noticed that when using the preset arrows around the preset box, it always goes back to ARF, using left or right.

Figuring this thing out and making patches right now. Is there room for suggestions? Mainly in the gate section, having a CV in for leak would be incredible.

I should probably mention what kind of feedback I'm looking for right now!

I don't care about CPU usage reports because of course it's a horrible hog now and that will get better soon.

I'm interested in feedback on the design though frankly things are pretty well nailed down at this point. If I head some idea that was super earth-shakingly awesome I could include it. But realistically, probably not until a 1.1 release.

I'm very interested in feedback on the sound! I realize that will get more possible as CPU use gets better.

If you make patches now, they should work without too much modification in the future. I will tune the levels and so forth, and may change how some of the models respond, but the existing complement of dials is not going to change.

I expect most people will wait until the next beta to get serious about making patches.

Apologies, I didn't have samples in the correct place. All sounds are working when lowered to one voice. Will look into making a patch or two.

No apologies needed, we are all trying to figure this out together

Great sounds can be had with this gem.

Randy you are a genius! I could get all patches to work at 4 voices (besides the Dulcimer ofc) and some at even 6-7 with no problems on my system. (Intel Xeon CPU @ 3.10 Ghz)

Here is the first experiment I've recorded. Using just a Sine wave and playing with the various modulation options as well as the "resonators"

Here's the patch (be sure to put the left arrow at beginning and right at the end dunno why it doesn't post with them)

Kaivo pluginVersion="65536" presetName="oneone[1]" scaleName="12-equal" key_voices="2" key_mod="2" key_bend="0" key_glide="0" seq_trig="1" seq_rate="2.3229997158050537" seq_offset="10" seq_pw="62" seq_value_delay="4.5" seq_pulse_delay="4" seq_value0="0.23999999463558197" seq_value1="0.68999999761581421" seq_value2="0.85999995470046997" seq_value4="0.40999999642372131" seq_value5="0.82999998331069946" seq_value6="0.11999999731779099" seq_value9="0.50999999046325684" seq_value10="0.93000000715255737" seq_value12="0.94999998807907104" seq_value14="0.56000000238418579" seq_value15="0.87000000476837158" seq_pulse0="1" seq_pulse8="1" lfo_offset="0.64999997615814209" lfo_offset_x="0.039999961853027344" lfo_offset_y="-0.12000000476837158" lfo_rate="0.45000007748603821" lfo_level="0.80000001192092896" lfo_offset_p="0.39999997615814209" lfo_rate_p="1.1999998092651367" lfo_level_p="0.93999993801116943" noise_peaks="5" noise_width="1.7999999523162842" noise_falloff="0.53999996185302734" noise_offset="0.099999964237213135" noise_rate="185.68002319335938" noise_level="0.72999995946884155" noise_auto="1" noise_filter="1" noise_offset_p="0.99999994039535522" env1_decay="0.12999998033046722" env1_sustain="1" env1_release="0.049999993294477463" env1_level="0.34999999403953552" env2_repeat="11.329689025878906" env2_delay_p="0.33999991416931152" env2_repeat_p="-1.9600000381469727" gran_position_x="0.68000000715255737" gran_position_y="0.2800000011920929" gran_rate="55" gran_density="1" gran_pitch_env="2.9802322387695313e-008" gran_pan="-0.20000004768371582" gran_follow="1" gran_smooth="1" gran_wrap="1" gran_position_x_p="0.27999997138977051" gran_position_y_p="0.99999994039535522" gran_rate_p="0.99999988079071045" gran_pitch_p="0.99999994039535522" gran_pitch_env_p="0.99999994039535522" gate_decay="0.81000000238418579" gate_leak="0.029999999329447746" gate_level_p="0.35999998450279236" res_position="0.78999996185302734" res_brightness="0.59999996423721313" res_pitch="220.00001525878906" res_nonlin="0.62999999523162842" res_sustain="0.79999995231628418" res_pan="-0.26999998092651367" res_out_pos="0.059999998658895493" res_lambda="0.93999999761581421" res_length="277" res_sp="8" res_w0="0" res_w1="0" res_m="7.0399999618530273" res_bridge_height="1.6000000238418579" res_bridge_len="0.053000003099441528" res_bridge_amount="0.019999999552965164" res_mode_num="1" res_position_p="0.55999994277954102" res_brightness_p="0.25" res_pitch_p="0.99999988079071045" res_nonlin_p="0.49999997019767761" res_sustain_p="0.21999996900558472" res_pan_p="0.12000000476837158" res_mix_wet="0.61000001430511475" res_mix_dry="0.90999996662139893" body_position_x="0.79999995231628418" body_position_y="0.74000000953674316" body_nonlin="0.099999994039535522" body_tone="0.45999997854232788" body_pitch="0.65999996662139893" body_sustain="0.77999997138977051" body_pitch_mode="2" body_mode_num="3" body_width="8" body_height="4" body_position_x_p="0.49000000953674316" body_position_y_p="0.29999995231628418" body_nonlin_p="0.039999961853027344" body_tone_p="1" body_sustain_p="0.24000000953674316" body_mix_wet="0.15999999642372131" body_mix_dry="0.50999999046325684" output_tilt="4.4703483581542969e-008" patcher_input_1="00000000000000000000010001000000000000000000" patcher_input_2="00000100000010001000000000000000000000000000" patcher_input_10="00000010000000000000000000000000000000000000" patcher_input_13="00000000000000000000000000010000000010100000" patcher_input_14="00000000000000000000000000000000000010010000" patcher_input_15="00000010000000000000000000000000000000000000" patcher_input_16="00000000000000000000000000000010000000000000" patcher_input_17="00000000000000000000000000000000000000000010" editor_x="3" editor_y="21" editor_width="1152" editor_height="624" editor_num="1" editor_anim="1" lfo_function="circle" res_mode="metal string" body_mode="metal plate" sample_name="Oscillators/sine.wav"/

Just installed it and playing it now with my morning coffee. I am getting sounds I have not heard yet from this computer box, bravo!

Bug Report (Using MacOSX 10.7.5, Live 9.1 Suite, AU Version)

In the BODY Section:
when I up the pitch to + 1.00 with the "small wood box" the output is only crackling white noise. All other impulses retain their original sound.

In the Sequencer:
When set to INT, starting and stopping playback of Live 9.1 stops the sequencer. Has to be set to HOST and back to play again.

My first patch


(lowering the number of voices in an adaption of the "harmonic sweep patch" caused Kaivo to stop outputting sounds, even though the sequencer was still running. I had to open another patch and then get reload this to make it play again.)

NOT A BUG - I had a gate from the "KEY" section still patched in.

second patch, playing around with the body, running two voice

Wow Randy, this synth is absolutely amazing. I can't get over how awesome it is to control physical models with samples like water and human voices. I spent about (no joke) 12 hours fiddling with it into the early morning hours and I still haven't scratched the surface. Just posted 3 new demos of Kaivo in the soundcloud group

This is so much fun! I uploaded two more patches to the Soundcloud group.

Sound suggestions:
I am interested in more bodies, if possible instrument bodies with nice resonant zones such as violins. Also, some more resonator models would be beautiful.

The limiter is useful, but maybe it could act a bit more dirty and go into a wilder overdrive? Of course I could simply insert a plugin behind it, but the way an instrument overdrives by itself has a huge appeal to me.

That's all for now!

Some nice demos here guys, congrats !

One suggestion for the limiter, it would be nice to not just have a built in limiter but also a compressor. For both of them though something visualizing the gain reduction taking place would be helpful, so you know how much it's going over the zero so to speak. A plugin called Pianoteq has a built in limiter and is a good example of what im talking about in terms of visualizing the gain reduction.

  • yes on more bodies
  • and more models like 'pipes'
  • some kind of brown noise or s&h random stepper oscillator/generator would be helpful on top of the controller sources already here, so you can have something like a random decay/sustain time for every time the note is plucked or struck
  • some kind of feedback loop mode where you can route the end of the chain back into itself at the beginning, for physical modeling this might be tricky since it's so unpredictable and 'spikey' already but Kaivo is probably one of the smoothest, if not the smoothest physical modeling synths i've used with this much tweakability.

the only thing i haven't been able to create effectively that i like doing in physical modeling is flute and horn type sounds, but maybe i'm missing a technique or trick on how to achieve this.

Running through a few patches I conjured up.

Running through a few patches I conjured up.


Thanks fluorescentgrey!

  • The limiter is going to be replaced with a "real" limiter by my friend Eric and tuned to cope with what Kaivo is putting out.
  • more bodies and models are coming. And if I don't get pipes (for example) into 1.0 I will roll them out later.
  • You can turn off 'auto' on the noise module. In this mode it slews to a new random value every time it gets a trig pulse.
  • feedback is planned !
  • would love to add a visual level display but this is post 1.0

Thanks everyone for putting up with this bare-bones forum software we wrote. I plan to put in some work on it this year. I have many usability problems noted for later fixing, believe me.

I've got it installed ok but it's bogging my system too much to do anything, so need to wait until next beta. Looks awesome though!
Win7 64bit, Intel i7-3930k, 32gb ram, Focusrite Scarlett 18i20, FL Studio 11

Here is a patch where I tried to go less for gnarly sounds, but rather more clean and shining tones.

New York Cello Player

@Randy: It does sound very good, much better than any of the PM synth I tried. I can't any more wishes despite what has been said already. Maybe if I play a bit more with it something will come up. For now, I am rather happy.

Wow, what a beast! Some sounds:

Drums + gamelan:

Modular groaning and giant didgeridoo:

Mutant woodwinds, reeds, and brass:

Physical modelling is generally not something that grabs me, but this is definitely an exception. Also wasn't expecting it to be able to get as warm and organic (dare I say "analog"...?) as it can.

Very brave to leave conventional lfos, filter, etc. out of it, and I commend the decision.

Fabulous work as usual, Randy! Big thanks for all of this.

Randy, just wanted to throw this out there. Have you played with the Logic Audio plugin Sculpture? I ask because of all the physical modeling synths I've used (and I've used them all.. hardware included: vl1, vl70, technics wsa1, korg wavedrum) Kaivo has most similarities with Sculpture. I'm only comparing the timbre characteristics and tweakability of them both. They share a similar lineage. Since Sculpture is exclusive to Logic I'm sure it's not going to offer more competition, but I'd suggest checking it out to see what overlap there is and what things can be implemented into Kaivo to differentiate it and make it ultimately a much more usable instrument. So far Sculpture is a lot less controllable and is a harder beast to tame. My 'test' for doing this was setting up a random CC array to change each knob to a random position, and with Sculpture you get usable results about 1 out of every 5 'random' bursts, with Kaivo I'm getting good results about 1 out of every 2. Sculpture divides things up differently, to be able to use audio clips to control the models it has to be loaded as an effect, where yours does not and that is already a huge advantage. Quite easily though I'm sure you could also turn yours into some kind of insert effect.

Sculptures strengths are percussion, plucked strings and rolling/jittering sounds. Right now Kaivo has the same strengths and I'd argue it's stronger as it already stands for these types of sounds, but what they both lack is the ability to do convincing wind brass and bowed string sounds, and i think if you could really dial these types of sounds in before the final release you will have a totally unstoppable beast with no comparison in the software realm.
edit: and I know you're already working out the CPU hogging nature of Kaivo, but Sculpture on my same system can be run in about 5-6 instances, sometimes even up to 10 without maxing out my cpu.