Devialet Phantom Reactor 900 Measurements (now with Spinorama)

[QMuse]

It looks like @napilopez was right, LR4 at 525Hz seems to correct the phase response:

And GD indeed looks good:

Step response is also "by the book":

 

[QMuse]

Phase without crossover correction:

 

[ofrappier]

Hello,

i've rephase too...

But how could be GD fine, if phase is not linear at 0° between 20 and 20000Hz

Explain me, i've to learn more

And can you explain me the step response graph that seems good ?

 

[QMuse]

But how could be GD fine, if phase is not linear at 0° between 20 and 20000Hz

GD is 1st derivation of phase vs frequency so it won't change much when you do phase correction of passive XO. What will change after correction is excess phase (difference between actual phase response and minimum phase).

And can you explain me the step response graph that seems good ?

That one is indeed tricky as it looks like the dirvers already are fully time aligned.

 

[QMuse]

Here is actual phase response, minimum and excess phase. Quite interesting..

 

[ofrappier]

Hi,

Thanks... i've just discovered this option in REW thanks to you.

I use it with optical, and i could add a FIR DSP (MINI-DSP-DI) (optical in-optical out) before Reactor , need i correct the excess phase ?
(I've rephase, i can produce linearize FIR)

or it's already good and useless to correct?

Regards,

O.

 

[napilopez]

I personally don't think you'd get much from correcting for that bit of excess phase, but you can always try =]

 

[QMuse]

Hi,

Thanks... i've just discovered this option in REW thanks to you.

I use it with optical, and i could add a FIR DSP (MINI-DSP-DI) (optical in-optical out) before Reactor , need i correct the excess phase ?
(I've rephase, i can produce linearize FIR)

or it's already good and useless to correct?

Regards,

O.

Frankly, I'd probably leave it as it is.

 

[ofrappier]

Hi,

Without FIR correction at 73Hz / 281 HZ:

With Rephase FIR SETTINGS :

With those linearize rephase linear filter FIR between DSP optical -> to Reactor, i've this :

Better ?

Regards,

O.

 

[QMuse]

Hi,

Without FIR correction at 73Hz / 281 HZ:

View attachment 62028

With Rephase FIR SETTINGS :

View attachment 62030

With those linearize rephase linear filter FIR between DSP optical -> to Reactor, i've this :

View attachment 62036
Better ?

Regards,

O.

Yep, much better!

GD also probably got a little better at the very bottom of LF.

 

[QMuse]

I've just corrected, the delay alignement : Crossover at 73Hz and 281 Hz, LR4....

How did you get those 2 numbers? Is it because of the phase wraps on the first graph you posted?

 

[QMuse]

Hi,

Without FIR correction at 73Hz / 281 HZ:

View attachment 62028

With Rephase FIR SETTINGS :

View attachment 62030

With those linearize rephase linear filter FIR between DSP optical -> to Reactor, i've this :

View attachment 62036
Better ?

Regards,

O.

When I import your REW file in rePhase and apply your filters I'm not getting your graph in REW:

 

[QMuse]

And when i export that filter into REW and apply it to your measurement I'm getting the same graph as in rePhase (and that is how it should be):

So, I really don't understand how you got that result with your filter because when i apply your filter to the REW measurement you posted I'm not getting anythign similar to your new measurement. (see my previous post)

 

[ofrappier]

Hi,

I search. i will make new measurements today with a 525 linearize LR4 FIR filter....

But when i import my REW measurement, with same WRAP settings than REW -180 /180 :

i've LR4 like Rotation -180 / +180 at 73 Hz and 281 Hz

and if i linearize this imported measurement at 73/281, the phase become ok, much better than with linearize at 525. Hz, not same results in real life with my reactor 600.

I will make other measurements...

 

[QMuse]

Hi,

I search. i will make new measurements today with a 525 linearize LR4 FIR filter....

thanks

Thank you! It would also be nice if you can make another measurement of uncorrected response so we can check it with first one.

Btw, I sincerely doubt they are 2 crossover points, what we are doing here is phase correction and that comes with the price. As you can see using 2 crossover compensation filters already introduced some pre-ringing artifacts in step response. Now, this one is still probably inaudible but adding more of them would cause audible artifacts.

 

[ofrappier]

Hi,

I doubt there is two crossover point, too

with another placement i've 44Hz /271 Hz

The left/right woofer are very sensitive, i have to change the placement

 

[ofrappier]

Hum now i found LR 60 dB at 170 Hz not LR 24 dB crossover at othher values....
i will order new support..... for new measurements, outdoor...

Thanks for the share, and help

 

[QMuse]

Hum now i found LR 60 dB at 170 Hz not LR 24 dB crossover at 170Hz....
i will order new support..... for new measurements

Did you try measuring with LR4 at 525Hz?

Btw, how far is your mic from the speaker? I suggest you try with app 1m-1.2m.

 

[ofrappier]

Hi,

the woofer in Phantom classic have a bandwidth : 14 to 300 Hz

see :

 

[napilopez]

Definitely not two crossover points. Did you try taking measurements of the individual drivers? I could have sworn I measured the HF unit but can't seem to find it. Here's a nearfield of the woofer though: