Talk:Frequency modulation synthesis

From Wikipedia, the free encyclopedia
Jump to: navigation, search


Hi!

Anybody has a reference to the original Yamaha FM-patent? This would be interresting because the DX7 is clearly an implementation of phase modulation and *not* frequency modulation (see discussion under phase modulation.)

mvh // Jens M Andreasen

(well, never mind. I found it and have added it to the article. DS)

--- Given the existence of the slightly older and more sophisticated DX-1, is is true to say that the DX-7 was the Yamaha flagship? Mark Wilson --- Mark! Dx7 was the most succesful, but the flagship at the time was the Dx5. Dx1 was a prototype of which only a few samples was ever build. Before the Dx series there was an even older 8 operator monster with slightly different algorithms. Stevie Wonder had one of those I remember. /jens



I made a couple of changes regarding the dates ... I hope that I did it correctly. I discovered FM synthesis (with my ears) during the acadmic year 1967-68 (I am not sure when exactly) The following is what I recently wrote about the discovery. "However, I have always been quick to point out that FM synthesis was not a theoretical discovery, but rather a discovery of "the ear." While performing some not very rigorous experiments with vibrato in 1967-68, I suddenly realized that with extreme vibrato, I was not hearing changes in pitch but rather changes in timbre. Within a short time I had produced a variety of different timbres … all lively with an internal dynamism that was unique … and with only two oscillators (operators)!! They were “magical moments” … having “cracked the code” with my ears. I then after learned about the theory of FM radio transmission and how the theory could explain, in a somewhat unusual application, what I had discovered." John Chowning JohnChowning 17:28, 26 November 2005 (UTC)


Hi John!

Good to see you here. Your user page isn't working though :)

/jens

Missing text[edit]

The intro seems to end abruptly - can anyone fix this?

Cheers —The preceding unsigned comment was added by 82.31.178.2 (talkcontribs).

Reverted to an earlier version. Thanks for pointing it out. Femto 15:42, 20 April 2007 (UTC)

Phase Modulation?[edit]

Hi there,

I was wondering why Phase Modulation is not realy covered by this article? It seems PM has some interesting features compared to FM. PM is also what most synth people understand as FM (or yamaha style FM). —Preceding unsigned comment added by 83.87.224.111 (talk) 14:54, 11 January 2008 (UTC)

Indeed, there is not enough information about FM versus PM in this article. As I understand it, from experimenting with the Nord Modular and Reaktor, the difference is this: With FM, the frequency of the carrier is modulated by the modulator; when the modulator signal is positive/negative, the carrier is sped up/slowed down; when the modulator signal goes through zero, the carrier oscillates at its original frequency. The difficulty here is that the perceived (average) frequency of the carrier changes if the modulator signal is not precisely symmetrical (centered around zero); in a graph of the modulator signal, the area inside the positive part of the curve should equal the area inside the negative part of the curve. This is why it is difficult to get stable FM with analog oscillators. With PM, the phase of the carrier is shifted by the modulator; when the modulator signal rises/falls, the carrier is sped up/slowed down; when the modulator signal goes through a maximum or minimum peak or stays constant (at any value, not necessarily zero), the carrier oscillates at its original frequency. The advantage here is that the modulator signal need not be symmetrical for the carrier to have a stable frequency. Now if someone with a PHD in DSP could explain this in the article, with the proper scientific terms and a good reference, I think a lot of synthesis newbies would find this artile much more useful. 84.198.246.199 (talk) 02:56, 16 April 2010 (UTC)

Many frequencies?[edit]

"Clearly a modulating waveform containing many frequencies (e.g. "partials") will create many FM sidebands." Is it clear? Only if we are assuming some sort of linear relationship, which may or may not be the case without delving into the mathematics... Segrub —Preceding unsigned comment added by Segrub (talkcontribs) 13:55, 20 February 2008 (UTC)

Mess[edit]

This article is a mess. Looks like the only way to synthesize FM signal is to use something that is related to _sound_ generation patent. Wrong. — Preceding unsigned comment added by 178.36.189.88 (talk) 20:45, 21 October 2012 (UTC)