Electrical phase: any good correcting it?

[daftcombo]

Hi,

Soundstagenetwork.com provides electrical phase measurement of speakers.

Here's an example (Focal Aria 906)

source: https://www.soundstagenetwork.com/i...cal-aria-906-loudspeakers&catid=77&Itemid=153

You can get that kind of curve at home with REW. Perhaps it is something different though, like the acoustic phase?

Anyway, a tool like RePhase would allow to correct it (create an inverse curve so that both curves "add" to get a flat line), save it in a convolution file and use it in Foobar or JRiver.

Do you think any audible improvement would come out of that?
In the past I've already tried that with other speakers and didn't hear any difference (except when I fucked the phase by 3600° rotating it on purpose).

 

[NTomokawa]

I found this here paper that more or less goes beyond my understanding.

I figure phase correction (or the goal of achieving linear phase) has to have something to it, else Technics wouldn't have come up with these funky speakers.

Waiting for more authoritative responses.

 

[Berwhale]

I'd be interested to understand why phase varies with frequency before asking if (and why) it should or could be corrected.

My basic understanding of phase is based on college study of things like QAM.

 

[daftcombo]

Some monitors like the Neumann KH80 have very straight crossover slopes (48dB / oct) to keep phase flat.

 

[Hipper]

I found this, post 3:

https://forums.stevehoffman.tv/threads/impedance-electrical-phase-for-speakers.448812/

 

[daftcombo]

I found this, post 3:

https://forums.stevehoffman.tv/threads/impedance-electrical-phase-for-speakers.448812/

That is between objective and subjective.

 

[briskly]

Anyway, a tool like RePhase would allow to correct it (create an inverse curve so that both curves "add" to get a flat line), save it in a convolution file and use it in Foobar or JRiver.

RePhase cannot act on the electrical phase of the loudspeaker. It is for adjustment of the acoustical signal phase, which is wholly distinct. You would try to correct something that is not related to the scope of what it does.
The phase of electrical components describes the angle that the voltage leads the current flow. This stems from the stored magnetic energy and capacitance of components, intentional or otherwise. The driver velocity is linked into the circuit by the motor, so the impedance includes the driver's mechanical resonances.

 

[daftcombo]

Someone else posted this quotation:

"The persistence of a bump in the spatially averaged curves [of the spinorama] is confirmation. There is no need to look at a waterfall diagram, because if there is a resonant peak, there will be ringing; no peak, no ringing. As was reported in Toole and Olive (1988) and discussed here in Section 4.6, it is the spectral bump that is the most reliable indicator of audibility. Loudspeaker transducers behave as minimum-phase devices, and therefore resonances in them can be attenuated by (minimum-phase) parametric equalization in the electronic signal path. Done correctly, the amplitude and phase response of the loudspeaker is corrected: the resonant peak is reduced, as is the time-domain misbehavior."

link:https://www.audiosciencereview.com/...m-00-powered-monitor-review.10859/post-304182

Can it be related to this thread?

 

[briskly]

An electrical signal can be expressed as a voltage or a current. The phase of the system as it responds to signals is not well related to the phase of the impedance, which is the ratio of voltage to current. The system's transfer function, what it does to signals, is what can be decomposed into the minimum, linear delay, and all-pass components.