Spectrum derivative (or spectrum acceleration) - page 12

 
Mathemat:

I still don't understand what you mean by phase, trollolo. You talk about it, you talk about it, but you don't give any definition.

What is a phase?


Later on, I'll try to describe what I mean by it on the basis of this picture. (I'll read more about digital spectrum analysis, maybe I'll express myself more accurately)

 
trollolo:


I'll try to describe what I mean by this picture later. (I'll read more about digital spectrum analysis, maybe I'll express myself more accurately)

The phase is the offset of the beginning of the harmonic from the beginning of the period, measured in radians (the whole period equals 2 * pi). Write it down somewhere on a piece of paper or on your forehead, so that next time you ask a trivial question here you don't have to make a picture gallery.
 
Trololo:

As I was pondering, the idea occurred to my paralon brain that I should compare 2 identical series using the ACF from the rest stop.

I wanted the dependence of the averaging period increase rate on the phase change rate...

What do you think, Alexei?

Are you sure it's the phase shift and not your phase measurements relative to the beginning of the period? After all, with each new bar, if you take this very bar as a new reference, the phase will also shift. The beginning of the period also shifts by one bar, i.e. by some number of radians.

Are you sure that everything outside the car window has "moved" and not your train has moved?

 
Reshetov:

Are you sure that the phase is shifting and not your phase measurement relative to the beginning of the period? After all, with each new bar, if you take that same bar as a new reference, the phase will also shift. The beginning of the period also shifts by one bar, i.e. by some number of radians.

Are you sure that everything outside the car window has "moved" and not your train has moved?

If so, then the phase change would be the same with each new bar, i.e. the fixed harmonic phase graph would be strictly sawtooth... it is easy to check that this does not quite correspond to reality.
 
alsu:
If this were the case, then the phase change would be the same with each new bar, i.e. the fixed harmonic phase plot would be strictly sawtooth... it is easy to check that this does not quite correspond to reality.
Even if the BP function is strictly periodic, but the period is not chosen correctly, the phases would "float". But here BP is non-periodic and the period is taken out of the box and the beginning of the period is taken from zero bar, so it is not surprising that with so many misunderstandings this result is obtained.
 
Reshetov:
The phase is the offset of the beginning of the harmonic relative to the beginning of the period, measured in radians (the whole period equals 2 * pi). Write it down on a piece of paper or on your forehead somewhere so that next time you ask a trivial question, you don't make a picture gallery here.
If we're talking about the phase of a sine wave, we need to be more specific. It can be either the phase shift of a sinusoid relative to another (for example, between current and voltage) or the phase of a point on a sinusoid (the current value of a sinusoid) relative to the transition point of that sinusoid through zero from bottom to top.
 
Trololo: In signal processing, the autocorrelation function (ACF) is defined by the integral wiki verb.

That's fine. What's next?
 
Trololo: In signal processing, the autocorrelation function (ACF) is defined by the integral wiki verboten.

How does this integral wiki verbosity relate to the amount of money earned on the euro-dollar exchange rate? )))
 
Mathemat:
That's fine. What's next?
The next step is to offer the person you are talking to a cake, as a reward for learning to use the wiki and quoting clever formulas from it.
 
All right, keep fiddling with your phases, hertzels and trend breaks. I won't understand it anyway, it's too complicated for me.