hi,
i have to show how work am and fm radio modulation.
can PD do this (in abstraction) ?
cheers
-
Am and fm radio modulation
-
Hi, the patch help browser >> Pure Data >> 3.audio.examples >> H09.ssbmodulation shows single sideband amplitude modulation, an advanced form of AM. FM is slightly more complicated, you need to calculate instantaneous frequencies from the phase differences between successive samples. Use [hilbert~] to complexify the signal and [cyclone/atan2~] to find phase and see attached abstraction [instantphase~].
Katja
-
thankyou for your help
i will look in the patch tonight
-
Hi there,
I'd like to make a FM synthesis patch using [hilbert~] (or [olli~] derived from what was proposed by Olli Niemitalou and relayed by Katjav here (http://puredata.hurleur.com/sujet-6428-hilbert, but I can't find the address of the abstraction atm).
I found this : http://www.ni.com/white-paper/3361/en , but don't figure out how to implement it in pd. I tried cross-multiplication than addition of complexified signals, but it doesnt seem to work (kinda ringmod scheme but multiplicating real and imaginary together, thus swapping two cables).
Should I use [instantphase~] ? How ? Is it possible to leave both sources freewheel or should I constantly act on the carrier phase ?
The context is fm guitar effect using an lfo.
Has someone an idea ?
Thanx,
Nau
-
Why are you trying to use [hilbert~]? I believe that's only needed for demodulation. FM can be done on real signals and doesn't require a conversion to a complex signal.
This might help you with what you're looking to do:
-
Yop,
I just want to have fun and learn new things, using parcellar information given by guitar pedals constructors as hints to wander through topics.
For now the "quadrature" stage of the "Strymon mobius", and its fm mode is under inspection
: I've been playing around with am modulation (ringmod and ssb with [hilbert~] is pretty straightforward), and now trying to play with fm.
They say they use "quadrature oscillators", so I'm trying to see what is possible to do with this in mind, even if it's maybe just a "commercial" statement...
Maelstorm : thank your for the link, I'll check it out.
Nau
-
Hi there,
I tought you might be interested in this:
http://www.acoustics.hut.fi/publications/papers/dafx09-cm/Also, M. Kleimola's page is kind of mindblowing:
http://www.acoustics.hut.fi/~jkleimol/
lots of good stuff and some code for pd externals in "website" links. -
Thank you all. Many paths ... didn't find mine yet.
I get back readingNau
-
Yop,
here is a small workbench providing simple multiplication and quadrature oscillators. (open quadam_workbench.pd)
Simple multiplication leads to a 'classic' ring modulation, ie double sideband. Quadrature oscillators allows us to select single sideband or double sideband. Single sideband is described by katjav at http://www.katjaas.nl/quadmixing/quadmixing.html . It deals more with pitch modulation than dsb does.
Both schemes can be antialiased, upsampling at twice the velocity (seems sufficient as the max frequency to be created should be 40khz), with 10th order Bessel filters. Antialiasing in not that efficient in stict unity gain bandwidth point of view, but much cleaner to the ear than 10th order Chebyshev filters. I used the latter in the antialiasing section of a disto previously and was kinda happy with them, but when the signal needs to be cleaner they just fail. They gave me ugly artifacts at transients. (there's a dummy filter transfer function approximation in the patch for the ones curious). Anyway I barely can find antialised audio neatly superior to 'normal' audio. Seems like a kind of luxury.
Suggestions welcome !
I will now try to follow one the route showed by Kyro, with http://www.khm.de/~cp/ (I play violin)
Thanx !
Nau
-
thanks to all
