Cytomic "The Drop" Resonant Filter

[andy-cytomic]

I noticed some kind of faint 'ringing' when the HP filter is enabled ... Like there was some kind of oscillator going on, it seems to be also affected by the env. This is even with resonance values way below self oscillation (actually it happens even when reso is at zero). I will make some audio clip soon to explain the problem.

I am experiencing this as well, and it's not so faint under certain circumstances. Here's Diva's default preset with the cutoff on its filter rolled off a bit, into The Drop's default preset with HP filter activated, HP freq automated from 0 - 20 kHz over 8 bars. No compression, just normalized to 0dbfs.

https://www.dropbox.com/s/iqps8af70jpz3 ... inging.mp3

Thanks for spotting this! It looks like a bug when I detect silence and bypassing processing, where I don't clear out a buffer of previous audio. It should be easy enough to spot and to fix. Can you please try sending a low level hiss to The Drop all the time and see if this prevents the behaviour on your machine? This is not for a workaround, this is to detect the cause of the issue.

PS: this shows that the high pass filter is doing it's job, and completely cutting low frequencies to silence quickly (ie not ringing)!

[btw]

I noticed some kind of faint 'ringing' when the HP filter is enabled ... Like there was some kind of oscillator going on, it seems to be also affected by the env. This is even with resonance values way below self oscillation (actually it happens even when reso is at zero). I will make some audio clip soon to explain the problem.

I am experiencing this as well, and it's not so faint under certain circumstances. Here's Diva's default preset with the cutoff on its filter rolled off a bit, into The Drop's default preset with HP filter activated, HP freq automated from 0 - 20 kHz over 8 bars. No compression, just normalized to 0dbfs.

https://www.dropbox.com/s/iqps8af70jpz3 ... inging.mp3

Thanks for spotting this! It looks like a bug when I detect silence and bypassing processing, where I don't clear out a buffer of previous audio. It should be easy enough to spot and to fix. Can you please try sending a low level hiss to The Drop all the time and see if this prevents the behaviour on your machine? This is not for a workaround, this is to detect the cause of the issue.

PS: this shows that the high pass filter is doing it's job, and completely cutting low frequencies to silence quickly (ie not ringing)!

Yes, I think you've got it. I'm sending audio from a synth (now Bazille ) as well as from Logic's test oscillator to the same bus, then using the sum as side chain input to The Drop on a third track. Even a -96dB signal from the test oscillator (white noise, sine wave, anything) is sufficient to eliminate the problem. Disable the test oscillator, and there's the problem again.

It's triggered most strongly by impulses or sounds with a fast envelope, which makes intuitive sense to me given what you said, but I'd be curious to hear your explanation. The ringing is typically at two resonances, with fundamentals jumping from approximately Bb3 to F4, plus what sounds like square-wave-like harmonics. For most HP filter models, the speed of that resonance jump increases when cutoff frequency is set above 5 kHz and becomes vanishingly fast above 6 kHz, sounding only at F4. It starts to break up for low cutoff frequencies under 150 Hz or so. In between, the resonant behavior is pretty consistent among filter models except for PRD HD, where it breaks up a bit at the jump, and MS1 HD, where it often jumps up and down between Bb3 and F4 a few times. Interesting!

[btw]

…and if I turn up the HP filter resonance, I can get the ringing to play me a nice little half diminished arpeggio, G3 Bb3 Db4 F4 Bug, or FEATURE???

[andy-cytomic]

…and if I turn up the HP filter resonance, I can get the ringing to play me a nice little half diminished arpeggio, G3 Bb3 Db4 F4 Bug, or FEATURE???

Self oscillating filters are definitely a feature. Press the Safe button above to Res knobs to block this from happening.

[andy-cytomic]

Just had a small session with the drop, it's really amazing. I was surprised with the vast possibilities shown in the presets, especially the SFX section. Are there any plans for some kind of presets sharing? Maybe through a "company forum" here on KVR?

I'm using a drupal from the cytomic web page, so I'll enable the "forums" module so and add a preset sharing area

[ElevatedBeing]

Just had a small session with the drop, it's really amazing. I was surprised with the vast possibilities shown in the presets, especially the SFX section. Are there any plans for some kind of presets sharing? Maybe through a "company forum" here on KVR?

I'm using a drupal from the cytomic web page, so I'll enable the "forums" module so and add a preset sharing area

Sounds great

[sqigls]

AWESOME thank you Andy!

I wish the filter names were a bit more descriptive = prefer it if they said sallen, ladder etc

totally

Yeah mousewheel over dropdown menus is cool. See u-he (I know you're friends with them, Andy ).

Mouse wheel sounds like a great solution! I'll add it

please add support for Macbook pad also

[andy-cytomic]

AWESOME thank you Andy!
...
please add support for Macbook pad also

Thanks! Yep, whichever way you do a vertical scroll will be translated to the correct message to change filter circuit or any other label based control like the LFO2 steps / pattern.

[btw]

…and if I turn up the HP filter resonance, I can get the ringing to play me a nice little half diminished arpeggio, G3 Bb3 Db4 F4 Bug, or FEATURE???

Self oscillating filters are definitely a feature. Press the Safe button above to Res knobs to block this from happening.

Oh, agreed of course about self-oscillating filters, but I was having a little fun further exploring/describing the HP filter bug. This is with resonance set below the point of self oscillation and is not expected behavior, but it seems you have a good idea of the cause in failing to clear a buffer.

[andy-cytomic]

Oh, agreed of course about self-oscillating filters, but I was having a little fun further exploring/describing the HP filter bug. This is with resonance set below the point of self oscillation and is not expected behavior, but it seems you have a good idea of the cause in failing to clear a buffer.

Gotcha. Yeah, definitely bug, but please feel free to make it into a feature while it lasts

[aMUSEd]

Here's a preliminary template for controlling The Drop using the NI Kore 2 hardware - basically you can control all knobs and buttons with this apart from the 6 macro knobs which at the moment it isn't possible to automate:

https://dl.dropboxusercontent.com/u/564 ... mplate.ksd

[andy-cytomic]

Here's a preliminary template for controlling The Drop using the NI Kore 2 hardware - basically you can control all knobs and buttons with this apart from the 6 macro knobs which at the moment it isn't possible to automate:

https://dl.dropboxusercontent.com/u/564 ... mplate.ksd

Thanks for contributing this! I'll let you know what NI say about support for mapping macro knobs that change target controls.

[ekwipt]

Just chiming to say loving the drop, ran my TR-808 through some of the drum presets and umm, wow that's all I did all night!

Any hints in your next project? I'm hoping for a Oscilator/LFO based on different synths, maybe some Eurorack, Serge and Buchla modules would be a good way to start.

Cytomic looks to be a software modular powerhouse!

[andy-cytomic]

Yes, I think you've got it. I'm sending audio from a synth (now Bazille ) as well as from Logic's test oscillator to the same bus, then using the sum as side chain input to The Drop on a third track. Even a -96dB signal from the test oscillator (white noise, sine wave, anything) is sufficient to eliminate the problem. Disable the test oscillator, and there's the problem again.

It's triggered most strongly by impulses or sounds with a fast envelope, which makes intuitive sense to me given what you said, but I'd be curious to hear your explanation. The ringing is typically at two resonances, with fundamentals jumping from approximately Bb3 to F4, plus what sounds like square-wave-like harmonics. For most HP filter models, the speed of that resonance jump increases when cutoff frequency is set above 5 kHz and becomes vanishingly fast above 6 kHz, sounding only at F4. It starts to break up for low cutoff frequencies under 150 Hz or so. In between, the resonant behavior is pretty consistent among filter models except for PRD HD, where it breaks up a bit at the jump, and MS1 HD, where it often jumps up and down between Bb3 and F4 a few times. Interesting!

The buzzing sound isn't to do with ringing / resonance, but its sound will change with different filter settings. It is just a looping buffer that hasn't been cleared, so it has mostly silence but a little bit of audio from when the sound decayed just before it went silent. The pitch of the sound is to do with the length of the buffer being looped, which depends on a few things, but has a maximum length of around 1.45 milliseconds, which is around 689 hz.

Thanks for spotting this bug and confirming it was what I guessed it was, I'll have a fix for it in the next release

[Mace404]

The authorization information is stored in the roaming part of a user profile and this gives problems with roaming profiles.
I use the same account on different computers in a domain so I can easily sync my settings across multiple systems.
Had to exclude this folder from roaming to prevent activating every time I logged onto another system.

Since it's using machine based variables (I read hard-disk ID somewhere earlier) I think the local part of a user's profile is more suitable for storing this.

I'm sorted for now but something to consider optimizing.