Kenji's K5000 Message Board Digest - Overview K5000 Resources - Overview
The Eat at Joe's Kawai K5000 Message Board Digest
AM Synthesis on the K5000


AM Synthesis
 Thursday, 04-Dec-97 03:56:19

      Message: writes:

      Here's what the manual says about AM Synthesis:

      "R2 AM
      Selects sources for Amplutude Modulation. One source can be set to modulate
      an adjacent source, i.e., 1>2."


      So what is AM modulation? Why is it different from FM modulation? And why
      did Kawai use AM modulation instead of FM modulation?

      My thoughts on the subject - I have a casio CZ101 which can do two source
      FM modulation, and this AM modulation sounds alot like the FM on the
      CZ101. It's good for getting weird, crusty tones, but without knowing the
      math, it's impossible to have the slightest clue as to what I'm doing.

      Does anyone understand this feature?



Re: AM Synthesis
Thursday, 04-Dec-97 10:20:25 writes:

     I think its closer to ring modulation than to FM, but that's about all I know. 


Re: AM Synthesis
Thursday, 04-Dec-97 16:40:41 writes:

     I looked at it wrong - my CZ101 does ring modulation, which seems almost identical
     to amplitude modulation, although I haven't quite figured out the
     difference yet.

     Lookee here:

     You can do realtime AM tweaking of an oscilliscope picture here (you can drag
     things in the windows and move them around). It's great! Substitute "liveAM"
     with "liveFM" in the URL above and you can do it for FM synthesis too.

     This is the formula, where m is the amount of modulation and x and y are the
     two sources of the modulation (sine waves).

     (1 + m sin(x))sin(y)

     With simple sine waves, AM will produce two sidebands (extra tones that are
     audible), one at frequency x+y and the other at frequency x-y. You also still
     hear one of the original tones - think it's the second one on the K5000S, but
     I can't remember exactly. With saw waves, it will probably do some odd things,
     because with saw waves, we are modulating more than two frequencies, because
     each overtone is a new sine wav (does anyone know what starts happening

     FM synthesis uses a different formula and produces a much larger number of
     sidebands, which explains why FM synthesis is more useful for creating "normal"
     synth sounds.

     So what does all of this mean? It means that we can emulate a "tuning the AM dial"
     sound, and we can also get some very bizarre sounds. I think I will
     attempt the later of the two this weekend. I think this could be very powerful if
     we could learn to grapple with it - I think it essentially turns two oscillators into
     three notes and possibly some very strange overtones...



Re: AM Synthesis
Friday, 05-Dec-97 05:33:40 writes:

     As far as I remember, in the K5K, the first source is the "modulator", the second
     is the "carrier", as it is called in telecommunication engineering. The carrier is
     the component that always remains in the signal. Just split your formula: 

     1 * sin(y) + m * sin(x) * sin(y) ,

     and you see: carrier + sidebands.

     What is ring modulation? The same without the carrier; a simple multiplication
     thus. (You don't need m then.)

     And modulating more frequencies? The sidebands term becomes the sum of all
     possible pairs of the x components and the y components multiplied. Each of these
     products generate the sum and difference frequencies that you mentioned, Kenji,
     and soon you have a nice collection of freqencies.

     If both inputs are subsets of the harmonic series of some tone, the ring
     modulation (or AM) result is guaranteed to lie in the same harmonic series,
     otherwise, you get those non-harmonic bell sounds.

     And FM? Something like:

     sin(x + m * sin(y)) .

     Don't try to solve an FM formula by hand! The spectra you get are VERY complicated!
     But the above-mentioned harmonic series argument is still valid.

     Jens Groh 


What-I-Found-Out #3: Ring Modulation
Saturday, 14-Mar-98 06:41:16 writes:

     The K5000 can do ring modulation! (Did anyone know?) 
     Recall our amplitude modulation discussion. Assume, "AM" (in the "Common" menu of
     your K5000) is set to "1->2". That is, SOURCE2 will be the
     carrier and SOURCE1 will be the modulator. The AM formula told us: 



     Well, the Kawai people seem to have followed a different concept. They use the
     level values of both sources and get a more flexible AM formula: 


     Make an experiment: 
     Set the number of sources to "2" and AM to "1->2". Take loud, but simple tonal
     signals for both source 1 and source 2 and set the pitches to some odd
     interval. You get some metallic sound. Now go to the "Control" menu and play
     with the volume values: With the SOURCE1 volume you can adjust the
     modulation amount, as could be expected. But with the SOURCE2 volume you can
     adjust the "carrier", too. If you turn it completely down to zero, the
     pure "sideband" signals remain, and that is nothing else but ring modulation: 


     With AM, you could not get rid of the carrier frequencies, but now you can. 

     Jens Groh 


Re: What-I-Found-Out #3: Ring Modulation
Monday, 16-Mar-98 10:17:37 writes:

     So you can fade from AM to RM by fading out the carrier, source 2? 


Re: Re: What-I-Found-Out #3: Ring Modulation
Monday, 16-Mar-98 10:34:46 writes:


     Jens Groh