>rbj
>it's a technical fact in a causal environment.  y[n] is a function of
y[n-1], y[n-2]... and x[n], x[n-1], x[n-2] ...

True as stated, but nothing there prevents anybody from defining y[n]
implicitly (i.e., as a function of the above *and* of y[n] itself). Indeed,
for some systems there is not going to be any equivalent algebraic
expression for y[n] purely in terms of previous y[n]'s and previous/current
x[n]'s.

So this seems to be more or less a semantic issue.

Personally, I am fine with the term "zero-delay feedback." An iterative
solver applied to an implicit equation seems like a pretty reasonable
approximation of the physical systems in question, wherein feedback paths
are connected without any explicit delay elements in them. I imagine the
iterations as corresponding to successive propagations of voltage/current
through the feedback path (of course these happen very quickly in the
analogue case, and the system never "holds still" at any point like in a
digital simulation, but you get the overall idea).

We could issue the same objection to analyzing feedback paths in analogue
circuits as "zero-delay," (since it always takes electricity some finite
time to propagate around the path) but it wouldn't make the zero-delay
approximation and associated fixed-point characterizations any less useful
or accurate. What's the point?

E


On Tue, Jun 24, 2014 at 8:37 AM, robert bristow-johnson <
[email protected]> wrote:

> On 6/24/14 6:00 AM, Urs Heckmann wrote:
>
>> You're right.
>>
>> I've been worked up ever since people post those silly and ignorant stabs
>> like this:
>>
>> On 09.04.2014, at 19:12, robert bristow-johnson<[email protected]>
>>  wrote:
>>
>>  if there is feedback, there must be at least one sample of delay,
>>> despite claims of zero-delay feedback i have read here on music-dsp and at
>>> other places.
>>>
>>
> it's a technical fact in a causal environment.  y[n] is a function of
> y[n-1], y[n-2]... and x[n], x[n-1], x[n-2] ...
>
> what's the next claim we'll be hearing about?  that we can get feedback of
> -1 or -2 samples "delay"?  why not?  it's just more extrapolation.
>
>
>  It occurrs that the topic triggers an immense resistance with those who
>> have stakes in DF-anything.
>>
>
> i never brought up DF-anything in this thread.  i *did* respond to it a
> little bit (mostly just to direct the discussion back to the topic).  i
> regret that i responded to it.  this thread has nothing to do with
> DF-anything.
>
>
>    It occurrs that no proof is enough to stop those people from stabbing.
>>
>
> "no proof" is what we're seeing.
>
>
>  Therefore I see no benefit in discussing the topic any further.
>>
>>
> it's one way to settle an unsupported claim.
>
>
> --
>
> r b-j                  [email protected]
>
> "Imagination is more important than knowledge."
>
>
>
> --
> dupswapdrop -- the music-dsp mailing list and website:
> subscription info, FAQ, source code archive, list archive, book reviews,
> dsp links
> http://music.columbia.edu/cmc/music-dsp
> http://music.columbia.edu/mailman/listinfo/music-dsp
>
--
dupswapdrop -- the music-dsp mailing list and website:
subscription info, FAQ, source code archive, list archive, book reviews, dsp 
links
http://music.columbia.edu/cmc/music-dsp
http://music.columbia.edu/mailman/listinfo/music-dsp

Reply via email to