Alignment tool questions

user44455555

Registered
Thread Starter
Joined
Jan 16, 2021
Messages
138
I try in this version the alignment tool and i did not know wy this happen. I use on older versions also sometimes alignment tool but notice this not. See video there happen also

that the dip at 500 change in strength depend on gain alot . I also not understand wy i see 3 curves and 2 moves. I select that 2 are visible
 

John Mulcahy

REW Author
Joined
Apr 3, 2017
Messages
7,536
the dip at 500 change in strength depend on gain alot
Dips occur where measurements have a large phase difference and similar levels. If the phase difference is 180 degrees and the measurements have exactly the same level the result of the sum is zero, so infinitely deep on a log scale (just limited by the noise floor). For more detail google "phasor addition".

I also not understand wy i see 3 curves and 2 moves.
The 3 curves are the two measurements being aligned and their sum. Changing the gain of one of the measurements changes the level of that measurement and the sum.
 

user44455555

Registered
Thread Starter
Joined
Jan 16, 2021
Messages
138
Dips occur where measurements have a large phase difference and similar levels. If the phase difference is 180 degrees and the measurements have exactly the same level the result of the sum is zero, so infinitely deep on a log scale (just limited by the noise floor). For more detail google "phasor addition".


The 3 curves are the two measurements being aligned and their sum. Changing the gain of one of the measurements changes the level of that measurement and the sum.

in the spl view i have select unwrap phase. seem the all spl did always do wrap phase and because the wrap of the phase in 1 measure and a little later on 2. measure this large drop happen. can all spl switch to unwrap phase ?

wrap phase.jpg
 

user44455555

Registered
Thread Starter
Joined
Jan 16, 2021
Messages
138
that the large dip happen i think is because REW calc 1 measure for example at phase 179 degree second is unwrapped 185 degree . so unwrappet very few phase diffrence but wrappet it is -175. so lots phase diffrence that which can explain the large dip . the trace arithmetric a+b tool do the same large dip . i attach the measure so you can see yourself . maybe need some phase filtering or room problems ? . is there trace arithmetric a+b possible without phase influence ?
 

Attachments

  • 500 hz dip.mdat
    1.9 MB · Views: 4

user44455555

Registered
Thread Starter
Joined
Jan 16, 2021
Messages
138
i notice clarity display look also not ok see screenshot. happen also with older measures other speakers in same way

clarity.jpg
 

user44455555

Registered
Thread Starter
Joined
Jan 16, 2021
Messages
138
No, 185 degrees and -175 degrees are the same angle, the phase difference is the same. Phase is periodic.

I mean 185 degree is 1 cycle + 5 degree in unwrap mode and in wrap mode it get to 0 cycle 5 degree for 1 measure and the other measure have 0 cycle 179 degreee which cause huge phase diffrence between those to measure in wrap mode. and in unwrap mode it is small . diffrence 0 cycle 179 degree to 1 cycle 5 degree
 

John Mulcahy

REW Author
Joined
Apr 3, 2017
Messages
7,536
You are missing the point. Full cycles of phase rotation have no effect on the result. Take the earlier hint and read about phasor addition.
 

user44455555

Registered
Thread Starter
Joined
Jan 16, 2021
Messages
138
You are missing the point. Full cycles of phase rotation have no effect on the result. Take the earlier hint and read about phasor addition.

does REW for frequency repsonse calculation also use phase values ?. I see at 500 hz gate with 9 ms and more do lots of phase change at 500 hz very fast in compare to gate with 8 ms . my impulse have very few reflections because of treated room. and there is nothing larger see between 8 and 9 ms. seem phase and GD values of measures of this measure algo many software use is very sensitive to reflections and in general not trustworthy . it is not REW only audionet carma show simular values . with arta i did not know how it can show unwrapped phases. so i think to a b measures should also possible that it do not use phase values to do a+b

impulse.jpg

gate 8 ms.jpg
gate 9 ms.jpg


rt 60 t20.jpg
 
Top Bottom