Maker Pro
Maker Pro

Amplitude-Modulation Synthesis

R

Radium

Jan 1, 1970
0
Hi:

I know of many types of synthesis for music. I am curious about two
types -- FM and AM.

Does FM synthesis have any advantages over AM synthesis? If so, what
are those advantages?

NOTE: FM and AM synthesis are *not* to be confused with FM and AM
radio.


Thanks,

Radium
 
T

Tim Williams

Jan 1, 1970
0
As in synth notes, etc?

Well duh, if you knew any musical terminology (which begs why you're asking
here, once again, three in a row mmm?), you'd know that vibrato is a
variation in pitch (FM) and tremelo is a variation in loudness (AM).
Indeed, according to Google:
"A periodic change in amplitude (unlike vibrato, which is a periodic change
in frequency)."
www.modularsynth.co.uk/glossary.htm
Over the long term, DC coupled FM control is generally referred to as a
control voltage (CV) with pitch proportional to voltage.
Likewise, DC coupled AM control sets loudness and envelope shapes (cresc.,
decresc., etc.).

So, yeah, really, when ARE you going to actually *use* Google yourself? Too
scared of leaving your diapers?

Tim
 
R

Radium

Jan 1, 1970
0
Tim said:
As in synth notes, etc?

Well duh, if you knew any musical terminology (which begs why you're asking
here, once again, three in a row mmm?), you'd know that vibrato is a
variation in pitch (FM) and tremelo is a variation in loudness (AM).
Indeed, according to Google:
"A periodic change in amplitude (unlike vibrato, which is a periodic change
in frequency)."
www.modularsynth.co.uk/glossary.htm
Over the long term, DC coupled FM control is generally referred to as a
control voltage (CV) with pitch proportional to voltage.
Likewise, DC coupled AM control sets loudness and envelope shapes (cresc.,
decresc., etc.).
So, yeah, really, when ARE you going to actually *use* Google yourself?

I did my best trying to search about the audio quality of AM synthesis
vs. FM synthesis. Google didn't give crap.

Sure I found definitions for AM and FM synthesis [and how those
synthesis types work] -- which I already knew long before.

However, I didn't find any website comparing the audio quality of of AM
synthesis vs. FM synthesis.
 
J

jasen

Jan 1, 1970
0
Hi:

I know of many types of synthesis for music. I am curious about two
types -- FM and AM.

Does FM synthesis have any advantages over AM synthesis? If so, what
are those advantages?

NOTE: FM and AM synthesis are *not* to be confused with FM and AM
radio.

AM lets you chage the amplitude (loudness) of the tone
FM lets you change the frequency (pitch).

you usually need both.

Bye.
Jasen
 
B

Bob Masta

Jan 1, 1970
0
Hi:

I know of many types of synthesis for music. I am curious about two
types -- FM and AM.

Does FM synthesis have any advantages over AM synthesis? If so, what
are those advantages?

I am not aware of any popular AM synthesis algorithms.
But FM synthesis is what is used in the much-maligned
Yamaha OPL2 and OPL3 synth chips in early sound cards.
Note that FM synthesis is *not* like vibrato; vibrato is typically
a very low frequency modulation (maybe 7 Hz or so) of the
"carrier" (the tone being played). Vibrato doesn't change the
underlying timbre of the tone, just makes it quiver a bit.

FM synthesis uses a modulator and carrier that are typically
related by a ratio of small integers like 2:3 or such. The idea
is that the carrier is so heavily modulated that the waveshape
itself is changed. This effect was originally called "Chowning
Modulation" after the guy who wrote it up. He found that it
was very effective for creating brass sounds.

When digital synths started getting popular, this was the
method of choice. I think it pretty much ruled until sampling
technology came along. The beauty of FM was that really
complex timbres could be produced with only a few parameters.
The downside was that *nothing* about the process was
intuitive; every new sound was a trial-and-error creation.

One surprising reason that FM was popular was that it was
easy to do. A pure sine wave was created by walking around
a circular sine table with a step size proportional to the desired
output frequency. To apply FM, you just had to modulate that
step size... which was *also* done by walking around the same
table to get the new step. So everything was done by simple
lookups. Compare that to AM, which typically required that
the output of the initial table lookup be multiplied by the
modulator waveform. Multiplication was really slow on early
hardware, so it was avoided at all costs. I even recall a method
to get AM for tremolo by looking up the same carrier twice with
variable phases that could be added to create the effect, just
to avoid a multiply.

Earlier analog synths used subtractive synthesis, which was
a lot more intuitive in that you started with a harmonic-rich
sawtooth or whatever and filtered out what you didn't want.
Additive synthesis was used on a few high-end systems,
where you built up the tone by adding each harmonic separately;
Intuitive, but really cumbersome.

Incidentally, the FM carrier and modulator didn't have to be
sine waves. I recall half-sines and a few other strange waveshapes
that were apparently easy to make with the hardware of the day.

Anyway, I'm sure you could use AM in the same way as FM with
small integer ratios to create new waveforms. But I suspect that
it won't be as useful as FM, probably because it would tend to make
a conventional harmonic series, whereas FM allows a much richer
spectrum. Just a guess.

Best regards,


Bob Masta
dqatechATdaqartaDOTcom

D A Q A R T A
Data AcQuisition And Real-Time Analysis
www.daqarta.com
Home of DaqGen, the FREEWARE signal generator
 
M

Michael Black

Jan 1, 1970
0
Tim Williams" ([email protected]) said:
As in synth notes, etc?

Well duh, if you knew any musical terminology (which begs why you're asking
here, once again, three in a row mmm?), you'd know that vibrato is a
variation in pitch (FM) and tremelo is a variation in loudness (AM).
Indeed, according to Google:
"A periodic change in amplitude (unlike vibrato, which is a periodic change
in frequency)."
www.modularsynth.co.uk/glossary.htm
Over the long term, DC coupled FM control is generally referred to as a
control voltage (CV) with pitch proportional to voltage.
Likewise, DC coupled AM control sets loudness and envelope shapes (cresc.,
decresc., etc.).

So, yeah, really, when ARE you going to actually *use* Google yourself? Too
scared of leaving your diapers?

Tim
There's a garbling here in the terminology.

In the beginning, there were analog synthesis, which might be considered
subtractive synthesis, since you'd start with a waveform that often had lots
of harmonics, and then you'd filter it according to the sound you wanted.

"FM synthesis" was a completely different method, using digital techniques.
It was the cat's meow when it came along. Freqency modulate one oscillator
with another audio frequency oscillator, and it provides a different sound.
This isn't the vibrato of the analog synthesizers where a very low frequency
was applied to a vco to get some vibrato. I've never paid much attention
to the process, but it's well documented.

I gather a lot of sound cards may still use the method, unless they've all
been superceded by wavetable synthesis (which when it came along, was
considered better than FM Synthesis.

I've never seen the term "AM Synthesis".

Given that the original poster seems to be making up terms, there's little
reason to bother to compare. And if he did a little reading first, then
he'd be in a far better place to ask questions that could be answered (and
are worth answering).

Michael
 
B

Bob Masta

Jan 1, 1970
0
I am not aware of any popular AM synthesis algorithms.
But FM synthesis is what is used in the much-maligned
Yamaha OPL2 and OPL3 synth chips in early sound cards.
Note that FM synthesis is *not* like vibrato; vibrato is typically
a very low frequency modulation (maybe 7 Hz or so) of the
"carrier" (the tone being played). Vibrato doesn't change the
underlying timbre of the tone, just makes it quiver a bit.

FM synthesis uses a modulator and carrier that are typically
related by a ratio of small integers like 2:3 or such. The idea
is that the carrier is so heavily modulated that the waveshape
itself is changed. This effect was originally called "Chowning
Modulation" after the guy who wrote it up. He found that it
was very effective for creating brass sounds.

When digital synths started getting popular, this was the
method of choice. I think it pretty much ruled until sampling
technology came along. The beauty of FM was that really
complex timbres could be produced with only a few parameters.
The downside was that *nothing* about the process was
intuitive; every new sound was a trial-and-error creation.

One surprising reason that FM was popular was that it was
easy to do. A pure sine wave was created by walking around
a circular sine table with a step size proportional to the desired
output frequency. To apply FM, you just had to modulate that
step size... which was *also* done by walking around the same
table to get the new step. So everything was done by simple
lookups. Compare that to AM, which typically required that
the output of the initial table lookup be multiplied by the
modulator waveform. Multiplication was really slow on early
hardware, so it was avoided at all costs. I even recall a method
to get AM for tremolo by looking up the same carrier twice with
variable phases that could be added to create the effect, just
to avoid a multiply.

Earlier analog synths used subtractive synthesis, which was
a lot more intuitive in that you started with a harmonic-rich
sawtooth or whatever and filtered out what you didn't want.
Additive synthesis was used on a few high-end systems,
where you built up the tone by adding each harmonic separately;
Intuitive, but really cumbersome.

Incidentally, the FM carrier and modulator didn't have to be
sine waves. I recall half-sines and a few other strange waveshapes
that were apparently easy to make with the hardware of the day.

Anyway, I'm sure you could use AM in the same way as FM with
small integer ratios to create new waveforms. But I suspect that
it won't be as useful as FM, probably because it would tend to make
a conventional harmonic series, whereas FM allows a much richer
spectrum. Just a guess.


By the way, you can use my freeware DaqGen signal generator
to easily experiment with AM or FM synthesis and see the resultant
waveform or spectrum while you listen to the results.

Best regards,


Bob Masta
dqatechATdaqartaDOTcom

D A Q A R T A
Data AcQuisition And Real-Time Analysis
www.daqarta.com
Home of DaqGen, the FREEWARE signal generator
 
Top