[sdiy] scanned pots "jumping"

[cheater cheater]

Joel,
> 2 when updating actual changes to sound generation parameters, if the
> parameter changes by a large value(jumps) then increment change value
> every(reasonable number of microseconds) until the new value is reached - in
> other words portamento for parameters.

The definition of the word 'updates' here is a question of research
being done by people for the last several decades. Depending on how
the 'updates' are calculated the behavior of the filter can be
completely different. It could end up creating zipper noise in the
signal and make the value oscillate, and keep it jumping back and for
the all the time. Or, it could end up being the smoothest thing ever.
If I understand that you mean just adding a constant value every n
milliseconds, then this would end up pretty pretty bad. If you use a
low pass filter it will be pretty good and good enough. If you use an
average, it will be pretty good too and quite responsive as well, and
very simple to calculate using a simple bit shift as was said earlier.

On the other hand you could get for example do something like this:
take your current value stored in the memory (call it y[n-1]). Compare
it to the current value read from the potentiometer (call it x[n]). If
x[n] = y[n-1], do nothing. If x[n] < y[n-1], then change the current
value in memory, called y[n], to y[n-1] - c. The number c is constant.
If x[n] > y[n-1] then y[n] = y[n-1] + c. Calculate y[n] every 50
milliseconds. Does this sound good? It describes what you said, right?

Now consider
y[n-1] = 100,
x[n] = 102+(-1)^n    the -1 is to describe jitter
c = 5

Your filter starts oscillating, and creates zipper noise.

Hope this helps understand something about the topic Joel. If you are
further interested, I suggest reading the pages of Julius Orion Smith,
who's a professor of DSP on a university, and is a sort of genius on
the topic.


-------------------------------------------------------------------------

Adam,

> Care to cite some specific references to a few of these "hundreds" of
> books
> on the subject?

here you go:
http://www.amazon.com/s/ref=nb_sb_noss?url=search-alias%3Daps&field-keywords=digital+signal+processing&x=0&y=0
each and every of those books contains an explanation of the one pole
rc filter, of a smoothing filter, or a one-pole digital filter
algorithm of some kind. Every single engineer to come out of school in
the last quarter decade knows what a low pass filter is, and I am not
only talking about electronic engineering and peripheral disciplines.

Here is some more for your viewing pleasure:
http://www.lmgtfy.com/?q=low+pass+filter+code

And just so that you know you'll definitely get code:
http://www.google.com/codesearch?q=smoothing
http://www.google.com/codesearch?q=lowpass+filter

So as you see, if you just try to search you find hundreds of
examples. And those are in no way difficult to find, just need to
think a little.

I never said that everyone here is grossly ignorant on any subject.
However, if you feel attacked, then you probably have a reason to feel
so, but that's your own subconscious problem.

Next time before you're calling someone out on knowledge, make sure
you're not taking them on an absolutely basic topic such as low pass
filtering. Are you going to call me out on being able to tie my shoes?
This sort of behavior makes you look quite stupid.

John,
it's simple to join a mocker when you have no real arguments of your
own. If you have any specific things to say to me, you can always use
my email address. I felt compelled to give Adam information about some
books and such, but if we're supposed to talk there's no reason to
annoy the list members.

D.

On Wed, Feb 3, 2010 at 20:33, John Mahoney <jmahoney at gate.net> wrote:
>
>> Care to cite some specific references to a few of these "hundreds" of
>> books
>> on the subject? Since everyone else on this list seems to be so grossly
>> ignorant about this topic, I'm sure we could all benefit by being given a
>> couple of specific examples of the sources of your extensive knowledge.
>> Thanks in advance.
>>
>> --Adam
>
> I was curious as to the dictionary which defines "uncomplicated" as "having
> already been invented," as used below:
>
>> ... (and even core fusion is not that complicated anymore,
>> since it has been invented already).
>
> John
>
> _______________________________________________
> Synth-diy mailing list
> Synth-diy at dropmix.xs4all.nl
> http://dropmix.xs4all.nl/mailman/listinfo/synth-diy
>