Tilt filter?

[Tusker]

I vaguely recall someone (Rob?) talking about the usefulness of a shallow slope lowpass filter (1/db/octave?) in creating a warm sound. Was one ever created by the NM community? How was it done? Anybody remember?

Thanks in advance,

Jerry

[cebec]

Hi. Here you go. I think this is what you're looking for:

http://mail.electro-music.com/pipermail/nord-modular/2005-July/003023.html

[Tusker]

Hey thanks Cebec. That's exactly what I was looking for.

Jerry

[davep]

Hi all,

Just to clarify, the NM1 filter described in the link is not a "tilt" filter, it is a narrow notch that filters out a narrow band of high frequencies centered at around 12kHz. Technically, it's called an eliptical filter, which is a notch that doesn't come back up all the way on one side of the notch - the "high" side of the notch is a few dB lower than the "low" side of the notch.

The "tilt" filter often used on the G2 is not a notch, it is a gradual 'tilt' of the entire audio spectrum, slightly boosted at the low end & gradually attenuated at the high end. And I don't think it can really be done on the NM1 because it requires allpass filters (on the G2, the BPF output of the multimode filter becomes an allpass filter when the module is set to the 6dB/Oct mode).

But the G2 doesn't seem to have the exact same characteristic 12kHz hash that the NM1 has, so the notch filter works best on the NM1 and the allpass tilt filter works best on the G2.

[Tusker]

The NM file was uncanny. I had been doing 6db lowpass filters at about 8 Khz, but was always aware of the trade-off between warmth and clarity. But this dials into the problem area very specifically. Thank you so much.

Of course, I'm going to ask the obvious question. How do we know that a LPF/HPF combination dialed in at 6.64 Khz, is going to do something for the hash at 12 Khz? Was this a trial and error thing, or did you arrive at these settings analytically? Gratefully,

Jerry

[Kassen]

I'm quite proud to have been beta testing Rob's analogue version of the tilt filter over the past months. This beta testing was quite easy; I just had to use it in day to day life and do such involving projects as taking it on the road, playing through it and setting it loose in secret on unsuspecting DJ's then making mental notes of the results.

I concluded it was good, need relays to make it transparent in the case of it's power failing and that was my job... :¬)

Oops, back on topic; I think the G2 version depends on a deterministic execution order which means the NM can forget about it. The NM simply isn't too strong in homebrew filters compared to the G2. At controll rates you can brutally clock the whole loop and force execution order that way but that doesn't work on high frequency audio. Other ways of working with psychoacoustics are still very much open in the NM. One area where I think the NM realy shines in psycho acoustics is dynamic spatialisation using the interpolating delay.

[cebec]

I wanna beta test Rob's tilt filter, too!

On a related note, I tweaked a version of a patch used to remove aliasing artifacts on certain G2 patches at audible frequencies a bit more so that it creates a deep notch that manages to include the area around 24 kHz -- completely eliminates the huge spikes @ 24 kHz one might see on plots of the G2s outputs (or NM Classic, for that matter) when running patches that are more prone to aliasing or when using control signals as audio.

[davep]

[altemark]

I don´t seem to be able to download the pch from the mailinglist archive - can anyone help out?

[blue hell]

For the tilt filter patch and a discussion about it look here.