GoodSoundClub - Romy the Cat's Audio Site

In the Forum: Audio Discussions
In the Thread: It’s mad, mad, mad... electricity.

Post Subject: Just trying to understandPosted by Lx_ on: 4/13/2011

Romy the Cat wrote:
You absolutely incorrectly, in my view, look at the data. The files of cause do not have the same start mark. Why shell they? I did not subtract one from another. You did in phase the files and did the subtraction but what the subtraction shows? You see the delta dose not indicate anything that we as listeners understand. [...] I did not run subtraction but instead I run global picks extraction into 3D. This way the difference in timing marks is irrelevant.

I thought that you had used the audio file comparer to get the signal difference and used that as input for the 3D view. That is why I reacted and spoke about subtracting signals. I was wrong, thank you for explaining what you did to produce the 3D graphics (though I do not see how you can get a 3D view of global peak extraction; also why not run the 3D analysis directly?).

Romy the Cat wrote:
What I did was zooming the file to the point where the resolution is high, the difference between the file is visible but the resolution still present the note or a phase in it’s completeness. Then you can see how the difference relates to Sound.

I still do no see how the first picture you posted can be used to show any meaningful difference between the two signals. Here is another, slightly more zoomed in view of the same signals:

(first 3.5s of signal instead of 7s; same amplitude around +/-1.5; one frame for the good electricity signal, another for the bad one)
Obviously, there is a slight DC offset (the time offset is also visible). But would you say that this picture indicates a dramatic difference between the good and bad electricity signals?
Note that the DC offset exists only at the beginning, it quickly disappears pretty much entirely.

As to the 3D views, they do not seem to have enough resolution to be meaningful either. For example, this seems to have much better resolution, but is in 2D (this is for the entire good electricity signal):

Note that it would be pretty easy to turn the 2D view into a 3D one.
Does anyone know of a software capable of producing high-res, 3D temporal spectrum analysis?

Rerurn to Romy the Cat's Site