Physical Modeling with MuLab

Official support for: mutools.com
Post Reply New Topic
RELATED
PRODUCTS

Post

Hi Folks

I read recently that combinations of bandpass filters, delay lines, and feedback can be used to physically model waves from a sound source bouncing around inside a space (like a guitar body or drum). I admittedly should have read more carefully and started this patch a bit earlier in the evening. But, perhaps someone could explain (1) how to experiment with physical modeling (such as Madrona Labs' Kaivo synth, which also uses granular synthesis) in MuLab, and (2) why the attached patch is making a crackling noise even though the source is a sine oscillator.

Thank you!
-d. vyd
You do not have the required permissions to view the files attached to this post.

Post

You are using a sine wave and several delays with bandpass filters.
At the moment you have one sine wave delayed once by several delays and the one bandpass that fits the sine wave freq repeats the sine.
As the delays finish abruptly you get cut off sine waves, which click.

Try noise or a saw wave to you get better results.
I would also try and increase the bandpass resonance to 1..2.

I guess what you really want is Karplus Strong synthesis.
A highly recommended example:
https://www.youtube.com/watch?v=fZ1rviedGvc
(if you want to dig deeper into modular patching, I recommend the rest of Nikol's videos)

I tried this setup once. See the attached.
You do not have the required permissions to view the files attached to this post.

Post

Thank you Andreas. These modular synthesis videos look fantastic!

Karplus Strong is indeed one of the algorithms I was thinking of. What are the impulse and impulse envelope modules in your patch? Are they renamed core modules? If not, I can't figure out how to open them to see how they work.

-d. vyd

Post

Yes, they are partly renamed standard modules. Taking the patch a bit apart:
Karplus Strong.jpg
Relating to the video, the red module is the noise source, the green module is the envelope generator to shape the noise (triggering the VCA in the noise module) and the blue module is the actual delay.
The low pass filter in the video is part of the delay and called damping. This is a 6dB lowpass, but the patch is from before the feedback delay module was available. Today you can use the feedback delay instead of the mono delay.
The rest of the patch is keytracking, which is done as described in the video by modulating the delay time. The first mod mapper is to square the note signal (note to freq is exponential, although this setup does not track well), the next is setting a propper level and the last green connection is inverting the modulation to get hight pitch from higher notes.
I often put "DO NOT TOUCH" modules in experimental patches. These are simple high pass filters set at a low frequency to kill any DC singnals that could kill speakers.
You do not have the required permissions to view the files attached to this post.

Post

Are amplifiers and speakers not protected by themselves? I think they should. I'm afraid adding such highpass filters in MUX may kill the bottom of sounds cause the filter slope is not steep enough to kill everything between say 0 and 20 Hz but keep everything intact from 20 Hz up. Right?

Post

Thank you Andreas. I've placed a multimode filter inside the modular feedback delay (MFD), but I'm not certain how to keytrack the delay time. Please see the attached image. The MFD has two inputs, one for audio and one for events. I have a note to modulation module--can I turn that into MIDI CC somehow and then map that controller to delay time? Also, how does one "tune" the delay time to that key presses and pitches match?

-d. vyd
You do not have the required permissions to view the files attached to this post.

Post

@Jo:
Speakers are not protected against DC. If you drive a woofer with DC it will move 'out', stay there, heat up and might die of overheating.
Before I check all my gear (sound card-which is DC coupled, mixer, amp, etc) i rather put a HP at the lowest freq. A 6dB at 1Hz would do.
And how would an amp block DC other than using a HP.

@d.vyd:
MFD does not have a modulation input, so you need a little extra with the parameter event generator (under event generators). You should check that the delay times stay short (<50msec)
Parameter Event.jpg
You do not have the required permissions to view the files attached to this post.

Post Reply

Return to “MUTOOLS”