MFM questions ;)

[fgimian]

Hey Urs,

I've never gotten my hands dirty with MFM but thought I'd try (and help test) it recently. It's pretty amazing mate, nice work!

I was wondering, is there a way to assign 2 LFOs to the Ratio of a single delay line? Often to simulate a nice modulated tape delay, it's nice to have the usual sine wave smoothly providing vibrato while a random hold LFO is breaking up the sound a little bit.

Another silly question, is it possible to use both effects on a single delay line feedback loop? It would be great to have a combo of SoftClip (for distortion) and Decimate for some added noise.

And finally, can I have 2 filters in the same feedback path too?

Here's what I'm trying to build, basically a really cool sounding tape delay:

Input -> sine modulation -> random modulation -> highpass filter -> lowpass filter -> soft clip -> a little noisy (decimate) -> Feedback to start of chain -> Output

Thanks a lot for your help

[Urs]

If you use two delay lines in series, e.g 1+3 and 2+4 you can have 2 filters and 2 fx for a single delay line.

You can furthermore modulate 1+2 with a sine and 3+4 randomly, without even touching the ModMatrix

All you need to do is this:

- set input of delays 1+2 to input signal, delays 3+4 to no input
- set output of delays 1+2 to zero
- choose "user matrix" as feedback matrix

set the feedback matrix up like this:

Code: Select all

. . + .
. . . +
+ . . .
. + . .

Where . is zero and + is +100%

There you go, now delays 1+3 act as one, as do 2+4.

Now you just need to set each delay to half of the time you want

Cheers,

Urs

[fgimian]

If you use two delay lines in series, e.g 1+3 and 2+4 you can have 2 filters and 2 fx for a single delay line.

You can furthermore modulate 1+2 with a sine and 3+4 randomly, without even touching the ModMatrix

All you need to do is this:

- set input of delays 1+2 to input signal, delays 3+4 to no input
- set output of delays 1+2 to zero
- choose "user matrix" as feedback matrix

set the feedback matrix up like this:

Code: Select all

. . + .
. . . +
+ . . .
. + . .

Where . is zero and + is +100%

There you go, now delays 1+3 act as one, as do 2+4.

Now you just need to set each delay to half of the time you want

Cheers,

Urs

You know you're a genius right?

Thanks mate, works a treat! I'll send you a preset or two soon to show you the result of this.

Cheers

[fgimian]

Hey Urs, just one more question if I may, so sorry to bother you.

Is there any option to use a higher quality interpolation algorithm when modulating the delay ratio via LFO? I notice that when modulating this with extreme LFOs (lots of depth), the sound is degraded significantly.

Comparing to Kjaerhus Classic Chorus, it's very noticeable that there is perhaps something (intentionally) different (almost lo-fi) about the interpolation algorithm used in MFM2.

Here's an audio sample demo of this, the source sound is the same, first processed by Classic Chorus and then by MFM2 beta demo. Notice the aliasing?

Audio Sample: http://www.sendspace.com/file/vouw9z
MFM2 Preset: http://www.sendspace.com/file/ln81mb

Please let me know

Thanks again!

[Urs]

Hmmm, the only Lo-Fi you get should be happening when Delay length goes below 16 samples. Otherwise the interpolation is the same as in Zebra's Comb filters or Uhbik-A's reverb.

Please post an mp3 in the beta forum then.

[fgimian]

Hmmm, the only Lo-Fi you get should be happening when Delay length goes below 16 samples. Otherwise the interpolation is the same as in Zebra's Comb filters or Uhbik-A's reverb.

Please post an mp3 in the beta forum then.

Oopsies, just edited my post above with audio & a preset, happy to post on the beta forum too if you like.

[Urs]

Ah, ok, yes, MFM2 isn't oversampled yet. You can probably match the Chorus sound by adding a filter on the input (as the Chorus certainly does, if it's remotely modelled after analogue stuff).

This has nothing to do with the delay interpolation, the delay interpolation is pristine and as good as any

However, fast LFO modulations of delay length result in pitch changes, and pitching up an ocatve results in aliasing... naturally...

Hmmm... with today's machines we can probably add an oversample option for the next revision...

Urs

[fgimian]

Ah, ok, yes, MFM2 isn't oversampled yet. You can probably match the Chorus sound by adding a filter on the input (as the Chorus certainly does, if it's remotely modelled after analogue stuff).

This has nothing to do with the delay interpolation, the delay interpolation is pristine and as good as any

However, fast LFO modulations of delay length result in pitch changes, and pitching up an ocatve results in aliasing... naturally...

Hmmm... with today's machines we can probably add an oversample option for the next revision...

Urs

Thanks a lot for clarifying mate, will try the filter for now

Obviously you have a lot of 64-bit fun to get through before thinking about MFM 2.2 I imagine.

[fgimian]

Just thought I'd let you know that using a lowpass filter on the input didn't really help I'm afraid. Here's a more extreme example of delay modulation:

http://www.sendspace.com/file/7u8q5r

1) Classic Chorus
2) Cubase Chorus
3) MFM2.1 beta

If you need me to try anything, please let me know mate

[Urs]

Just thought I'd let you know that using a lowpass filter on the input didn't really help I'm afraid. Here's a more extreme example of delay modulation:

http://www.sendspace.com/file/7u8q5r

1) Classic Chorus
2) Cubase Chorus
3) MFM2.1 beta

If you need me to try anything, please let me know mate

Well, that doesn't sound like it. Can you send the patch? Seems to be rather a bug than aliasing tbh.

Cubase Chorus clicks. Is that a known bug?

[fgimian]

Just thought I'd let you know that using a lowpass filter on the input didn't really help I'm afraid. Here's a more extreme example of delay modulation:

http://www.sendspace.com/file/7u8q5r

1) Classic Chorus
2) Cubase Chorus
3) MFM2.1 beta

If you need me to try anything, please let me know mate

Well, that doesn't sound like it. Can you send the patch? Seems to be rather a bug than aliasing tbh.

Cubase Chorus clicks. Is that a known bug?

Hey mate, thanks for all your help.

Here's the preset: http://www.sendspace.com/file/47u91p

Regarding Cubase' Chorus, I don't think Steinberg will give a hoot about it, though to be fair, I've never officially reported it. Just assumed it was some LQ algorithms. I HATE the Cubase Chorus for the very reason you mentioned and never use it.

More than happy to test anything you throw at me

[Urs]

Just checked... Well, I guess those fast & deep delay modulations are simply out of the scope of MFM2 atm... unless you do use it for glitch...

[fgimian]

Just checked... Well, I guess those fast & deep delay modulations are simply out of the scope of MFM2 atm... unless you do use it for glitch...

Thanks mate, to be honest I noticed these issues with a low depth LFO when I was trying to build a chorus unit. I think with the current limitation MFM2 works really great for delays but not really for choruses.

Hope you might consider this in future.

Thanks again for all your help