Resonance filter in AL-1 : weird behavior

Liviou2004 · Post by **Liviou2004** » Mon Sep 03, 2018 6:14 am

Hello,

I've just noticed something with resonance setting wich seems to be an issue.
On my Kronos (Kronos 2 ver. 3.1.2), starting with an initialized Prog AL-1, nevermind the waveform, filter cutoff full opened (99), just increase the resonance setting and you will hear the volume decreases !!

I've not yet proceeded to the same test with other engines (Polysix, MS-20, etc...)

I've never noticed such a behavior in other synths (true analog as my MS-20, or my sledge va) and I do not consider this behavior as normal.

Did you noticed that issue ?

Gunnar · Post by **Gunnar** » Mon Sep 03, 2018 6:36 am

Hi Liviou2004,

I've noticed the same thing. The Moog Mother-32 has the exact same behavior, so the Kronos is not alone in behaving this way, and I suspect is intentional.

My Arturia synths, on the other hand, do not, but they use a very different filter circuit.

voip · Post by **voip** » Mon Sep 03, 2018 8:12 am

Opening the filter right up puts the resonance point at the top end of the audio spectrum, so not audible to anyone but a young person. Since the whole point of resonance is to emphasise a narrow frequency range, other parts of the audio spectrum need to be reduced to keep everything within the system headroom. Overload and distortion effects can be added later, whilst still remaining within overall system headroom.

BobTheDog · Post by **BobTheDog** » Mon Sep 03, 2018 8:23 am

I would say "most" filters drop volume as resonance increases unless they have volume compensation.

X-Trade · Post by **X-Trade** » Mon Sep 03, 2018 11:11 am

I know in the analog world, Moog filters are known to lose volume when resonance is turned up.

Many other filters don't. It depends how the resonance is implemented. The same can be applied to digital filters but it is less commonly discussed.

Some filters get really loud with high resonance and actively drop their output volume to avoid clipping.

Liviou2004 · Post by **Liviou2004** » Mon Sep 03, 2018 5:55 pm

Thanks for your responses.

I want to add some tests I just did with the other engines :

MS-20 Ex : No attenuation at all : LPF fully opened and full resonance, the volume stay the same.

Polysix : There is a tiny volume reduction with a full resonance.

AL-1, Mod-7, STR-1 : as these three engines use the same type of filter emulation, the result is the same : a huge volume reduction. The higher resonance the more volume reduction.

If our friend, Charles Ferraro, who knows how to do precise measures, red this thread, it would be nice of you if you could measure the quantity of volume reduction, in db. I am sure it's really important.

Anyhow, I don't find this cutoff behavior and I consider it as an issue. I suggest you to test it with your Kronos.

KK · Post by KK » Tue Sep 04, 2018 12:56 am

The Kronos offers a great emulation of both the MS-20 and Polysix. The original MS-20s used diodes in their Q sub-circuits, which retained some voltage and participated in the particular forementioned filter gain compensation at high Q. As for the original Polysixs they used SSM2044 VCFs, which offered particularly good amplitude/Q vs frequency curves.

But on many other vintage and newer synths, the filters did/do not behave that way with amplitude decreasing when Q goes up. Some preferred other brands and models for that reason and/or for their oscillators, other features, characteristics, timbre possibilities, etc. I once destroyed an amplification system with a certain analog synth messing with high resonance while playing low notes. In this way, it's actually good that many synths also decrease in amplitude at high Q.

The fact that the newer Korg filters behave differently than the MS-20 and Polysix is simply a matter of design, not a flaw/problem. Each synth has his particularities. The AL-1, STR-1 and MOD-7 all offer incredible possibilities and versatility.

Liviou2004 · Post by **Liviou2004** » Tue Sep 04, 2018 5:54 am

KK wrote:The Kronos offers a great emulation of both the MS-20 and Polysix. The original MS-20s used diodes in their Q sub-circuits, which retained some voltage and participated in the particular forementioned filter gain compensation at high Q. As for the original Polysixs they used SSM2044 VCFs, which offered particularly good amplitude/Q vs frequency curves.

But on many other vintage and newer synths, the filters did/do not behave that way with amplitude decreasing when Q goes up. Some preferred other brands and models for that reason and/or for their oscillators, other features, characteristics, timbre possibilities, etc. I once destroyed an amplification system with a certain analog synth messing with high resonance while playing low notes. In this way, it's actually good that many synths also decrease in amplitude at high Q.

The fact that the newer Korg filters behave differently than the MS-20 and Polysix is simply a matter of design, not a flaw/problem. Each synth has his particularities. The AL-1, STR-1 and MOD-7 all offer incredible possibilities and versatility.

Yes I agree.
Though, you can try : there is no volume compensation in MS20-EX Kronos engine.
And I absolutely agree with you : AL-1, MOD-7 ans STR1 offer huge possibilities.

However, I consider the volume filter compensation to be excessive. Did you really test it ?

voip · Post by **voip** » Tue Sep 04, 2018 8:09 am

Using a compressor effect should get round the low volume issue, unless the resonance is creating a high signal level outside the audible range.

Broadwave · Post by **Broadwave** » Tue Sep 04, 2018 10:43 am

It's quite common for the volume to drop at higher resonances with a 24dB VCF... The ARP Odyssey (MkII & III) both exhibit this behavior, but the MkI's 12dB VCF actually INCREASES the volume.

KK · Post by KK » Tue Sep 04, 2018 12:45 pm

Liviou2004 wrote:Did you really test it ?

Yep and as explained it's normal to me, as many synths behave that way.