Re: Questions about FM synthesis and amplitude

Radium wrote:
> robert bristow-johnson wrote:
>>  the last option is what emulation is.  if it generates
>> the same waveform (after bandlimiting to the human hearing range) and sounds
>> the same, who cares?
> 
> It will not generate the same waveform. That's the problem.
> 

In the context of audio, that being sound impulses heard and perceived 
by a human, then it makes no difference if the waveform is different, as 
long as it's perceived the same.

That's why you can use non-linear phase filters in audio, the ear is not 
(as) sensitive to phase distortion.  Are you therefore saying that 
putting some audio signal through a filter makes it a crap signal?

in article 1129266949.818412.230470@g47g2000cwa.googlegroups.com, Radium at
glucegen1@excite.com wrote on 10/14/2005 01:15:

> robert bristow-johnson wrote:
>> the last option is what emulation is.  if it generates
>> the same waveform (after bandlimiting to the human hearing range) and sounds
>> the same, who cares?
> 
> It will not generate the same waveform. That's the problem.

that is an engineering and psychoacoustic issue.  try to know what you're
talking about before making "authoritative" statements.

-- 

r b-j                  rbj@audioimagination.com

"Imagination is more important than knowledge."

Bevan Weiss wrote:
> In the context of audio, that being sound impulses heard and perceived
> by a human, then it makes no difference if the waveform is different, as
> long as it's perceived the same.

I've listen to actual synths and emulation of those synths. I notice a
considerable amount of difference. I don't perceive the emulation as
sounding the same the the actual synth.

> Are you therefore saying that
> putting some audio signal through a filter makes it a crap signal?

Not necessarily. I use high-pass filter all the time. I like the treble
but don't really care for the bass.

I dream of a filter that differentiate vocals from non-vocal audio and
that can differentiate the musical instruments from each other, and can
change the volume or eliminate certain instruments/vocals as desired by
the user.

Radium wrote:
> Bevan Weiss wrote:
>> In the context of audio, that being sound impulses heard and perceived
>> by a human, then it makes no difference if the waveform is different, as
>> long as it's perceived the same.
> 
> I've listen to actual synths and emulation of those synths. I notice a
> considerable amount of difference. I don't perceive the emulation as
> sounding the same the the actual synth.

That kind of comparison is rather invalid.  It is very obvious that you 
have a great persuasion towards actual synthesizers over emulation type 
audio production.  Therefore unless your comparison is actually 
performed via some blind test then it is highly biased.

I doubt there is much perceivable difference between the two, and 
believe that if a blind test were performed between the two, you would 
only be able to identify the correct production method on average half 
the time (that of course being the statistic of a guess between two 
possibilities with no a priori knowledge).

This sounds very much like the argument over valves and fets.  I believe 
that a fet circuit can be made to sound equivalent to a valve, and that 
fets are actually superior due to more accurate sound reproduction. 
It's all about perception however, and with many people just telling 
them an amplifier uses valves make them consider it to be better, even 
if they wouldn't otherwise.

robert bristow-johnson wrote:
> in article 1129266949.818412.230470@g47g2000cwa.googlegroups.com, Radium at
> glucegen1@excite.com wrote on 10/14/2005 01:15:
>  >>robert bristow-johnson wrote:
>>
>>>the last option is what emulation is.  if it generates
>>>the same waveform (after bandlimiting to the human hearing range) and sounds
>>>the same, who cares?
>>
>>It will not generate the same waveform. That's the problem.
> > 
> that is an engineering and psychoacoustic issue.  try to know what you're
> talking about before making "authoritative" statements.

  I think psychoacoustics should figure into these "debates" more often. 
The line between real science and bankrupt nattering about the merits or 
lacks of specific sound creation methods is a broad one. No casual 
listener cares about waveform symmetry and such; the only thing that 
really sucks is lousy music and that's a debate for the alt.fan lists, 
heh. Creative spark and heart first; sampling rate conversion afterwards.

--

  HellPope Huey
     Man cannot live by common sense, facts
      and reasoned discourse alone;
     he must also watch Cowboy BeBop.

     Listen carefully to first criticisms
      made of your work.
     Note just what it is about your work
      that the critics don't like --
       -- then cultivate it.
     That's the only part of your work
      that's individual and worth keeping.
                ~Jean Cocteau

    "Jean Cocteau was a heroin addict, you know."
    "Yes, but Cocteau was Cocteau
       before he was a heroin addict."
       ~ film dialogue, "Liquid Sky"

"robert bristow-johnson" <rbj@audioimagination.com> wrote in message
news:BF741CA6.B1BD%rbj@audioimagination.com...
> however the original subject is about analog *synthesis* and those old
> minimoogs sounded *great* and that sound is maybe only very recently
> beginning to be faithfully emulated in software.  there are tons of posts
to
> the music-dsp mailing list about generating bandlimited sawtooth waveforms
> (something Moog didn't need to worry about since his sampling frequency
was
> infinite)

In what universe is anything infinite?

>and emulating the 4-pole Moog voltage-controlled LPF with
> "regeneration".  it's very hard to do this well at 44.1 kHz.

So you don't use CD's either I guess.

MrT.

"Radium" <glucegen1@excite.com> wrote in message
news:1129235945.625147.326210@g14g2000cwa.googlegroups.com...
> Sample-playback synthesis is even worse than analog synthesis.
>
> In fact sample-playback synthesis is probably the worst. It is not even
> a type *synthesis*.

Exactly why it is *SO* much better for real instruments. However using
digital oscillators does not prevent all other types of synthesis, it just
makes the tuning stable for a change.

MrT.