Welcome, Guest
Username: Password: Remember me
  • Page:
  • 1
  • 2

TOPIC:

How to find out my active speaker's Xover details? 7 years 9 months ago #12754

  • phranky
  • phranky's Avatar Topic Author
  • Offline
  • Junior Member
  • Junior Member
  • Posts: 25
  • Thank you received: 0
I'm trying to linearize my active speakers with rePhase 0.9.7 to load the settings on a OpenDRC-AN. Now I want to set the correct crossover at the "Filters Linearization" tab. In my speaker's datasheet I found the crossover frequency of 2000 Hz. But nothing about the filter quality, if it has a LR 12, 24 or 36 dB/oct filter built in. Choosing these different filters affect the linearization a lot (see pic below). Google wasn't able to help me.
--> Any ideas how to find out my active speaker's crossover filter Q?

I have the Yamaha HS80M, actually one of the most popular studio reference speakers. So I hope to find any help :)
Thank you very much in advance for any hints!

Attachments:

Please Log in or Create an account to join the conversation.

How to find out my active speaker's Xover details? 7 years 9 months ago #12762

  • pos
  • pos's Avatar
  • Offline
  • Platinum Member
  • Platinum Member
  • Posts: 354
  • Thank you received: 200
Hello phranky

I cannot imagine it being less than 24dB/oct, and it probably is just that, but only a measurement would tell (no need for a real measurement microphone either: a simple uncalibrated dynamic mic would do, with REW or HOLM).

Why not try a 24dB/oct linearization? If it is a 36 or 48dB/oct filter then you would at least have compensated part of the phase shift.
That said, audibility of phase shifts at 2kHz is tenuous to say the least..

Are you planning on compensating the BR phase shift as well?
The following user(s) said Thank You: phranky

Please Log in or Create an account to join the conversation.

Last edit: by pos.

How to find out my active speaker's Xover details? 7 years 9 months ago #12768

  • phranky
  • phranky's Avatar Topic Author
  • Offline
  • Junior Member
  • Junior Member
  • Posts: 25
  • Thank you received: 0
Thanks, pos! I'll try it this way..
Do you think, I can equalize the phase alltogether with the Paragraphic Phase EQ and ignore the Filters Linearization tab, as this does just an approximate flattening?

What do you mean with "BR phase shift"?
I'm sorry, I'm pretty new to the phase issues. But I'm getting closer with your help. Thank you very much!

Please Log in or Create an account to join the conversation.

How to find out my active speaker's Xover details? 7 years 9 months ago #12774

  • pos
  • pos's Avatar
  • Offline
  • Platinum Member
  • Platinum Member
  • Posts: 354
  • Thank you received: 200
If you have measurement capabilities you should really still use the filter linearization tab and adjsut it to reduce phase shifts as much as possible (frequency will often be a tiny bit higher than the theoretical one, especially with horn systems), and then use the phase EQ for fine tuning only.
Having the proper polarity and time offset is key when looking at the measurement : the phase shift should look like the theoretical once as much as possible, otherwise something is wrong with those two parameters (or your measurement is broken).
Using the impulse auto centering feature in HOLM should oftentimes deal with the time offset, so you will only have to check polarity (and redo the auto centering afterward): the first impulse peak shall be positive, even if the second one is bigger.

BR is the bass reflex box tuning: if you know your bass reflex tuning frequency you can use the filter linearization tab to linearize its phase shift.
If you don't know the exact frequency then the simplest way to find it is to use a tone generator and find the frequency at which the cone as almost not moving but a lot of air goes through the port.
It is probably around 40Hz with this kind of monitor.

You should also check for any subsonic high pass filter (quite common with active monitors) that would add another phase shift under the BR.
Checking this is also easy: once you found your BR tuning frequency you can start decreasing the frequency of your tone generator, and if no subsonic filter is present you should see the excursion of the woofer rising quickly. If you don't that means that you have a subsonic filter engaged (commonly12 or 18dB/oct...), and it can be compensated for in the filter linearization tab.
The following user(s) said Thank You: phranky

Please Log in or Create an account to join the conversation.

Last edit: by pos.

How to find out my active speaker's Xover details? 7 years 9 months ago #12776

  • phranky
  • phranky's Avatar Topic Author
  • Offline
  • Junior Member
  • Junior Member
  • Posts: 25
  • Thank you received: 0
Yes, I made my measurements in REW with a Behringer ECM8000 (mic and soundcard calibrated). I don't have HOLM.
I don't have a separate bass speaker, just my two Yamaha HS80M (both in same distance to my head).

Having the proper polarity and time offset is key when looking at the measurement : the phase shift should look like the theoretical once as much as possible, otherwise something is wrong with those two parameters (or your measurement is broken).

--> How to achieve proper polarity and time offset?

Maybe a rePhase screenshot of my measurements helps:

You see the huge phase shifts above 3000Hz? I can't get rid of them with the Phase EQ. What might be wrong there and what can I do?
Attachments:

Please Log in or Create an account to join the conversation.

How to find out my active speaker's Xover details? 7 years 9 months ago #12777

  • pos
  • pos's Avatar
  • Offline
  • Platinum Member
  • Platinum Member
  • Posts: 354
  • Thank you received: 200
It is not normal to have such sharp discontinuities in your phase.
Please do post your measured impulse as a wav or txt file so I can look it up.

Please Log in or Create an account to join the conversation.

How to find out my active speaker's Xover details? 7 years 9 months ago #12778

  • pos
  • pos's Avatar
  • Offline
  • Platinum Member
  • Platinum Member
  • Posts: 354
  • Thank you received: 200

just my two Yamaha HS80M (both in same distance to my head).

Are you measuring both speakers at once? :ohmy:
The resulting comb filtering would explain those phase discontinuities up top for sure...

Please post the file of a raw impulse of one speaker playing at a 1 or 2m distance, with the speaker at its normal position.

Please Log in or Create an account to join the conversation.

Last edit: by pos.

How to find out my active speaker's Xover details? 7 years 9 months ago #12779

  • phranky
  • phranky's Avatar Topic Author
  • Offline
  • Junior Member
  • Junior Member
  • Posts: 25
  • Thank you received: 0
No no, not at once. This is only the left speaker.

Here they are:

File Attachment:

File Name: measurementFeb24.txt
File Size:1,175 KB

Thank you in advance, pos!
Attachments:

Please Log in or Create an account to join the conversation.

Last edit: by phranky.

How to find out my active speaker's Xover details? 7 years 9 months ago #12780

  • pos
  • pos's Avatar
  • Offline
  • Platinum Member
  • Platinum Member
  • Posts: 354
  • Thank you received: 200
I would have preferred an impulse instead of an amplitude and phase file, as it would have let me apply some smoothing and/or windowing, and look at the impulse.
I would also have made the polarity and time offset setting easier.
Here is the correction (no need for phase EQ, and low frequency phase measurement is most certainly not accurate enough to be corrected anyway)

Attachments:

Please Log in or Create an account to join the conversation.

Last edit: by pos.

How to find out my active speaker's Xover details? 7 years 9 months ago #12781

  • pos
  • pos's Avatar
  • Offline
  • Platinum Member
  • Platinum Member
  • Posts: 354
  • Thank you received: 200
By the way, the phase discontinuities with the -1440°/+1440° phase range appear to be a bug, I will have to fix that.
Thanks for the report ;)
In the meantime stay with the standard -180°/180° wrapped range... :whistle:

Please Log in or Create an account to join the conversation.

Last edit: by pos.

How to find out my active speaker's Xover details? 7 years 9 months ago #12782

  • phranky
  • phranky's Avatar Topic Author
  • Offline
  • Junior Member
  • Junior Member
  • Posts: 25
  • Thank you received: 0
Here's the impulse response txt:


I'm a bit confused now - how have you done the correction?
Attachments:

Please Log in or Create an account to join the conversation.

How to find out my active speaker's Xover details? 7 years 9 months ago #12783

  • pos
  • pos's Avatar
  • Offline
  • Platinum Member
  • Platinum Member
  • Posts: 354
  • Thank you received: 200

Here's the impulse response txt:

I was not able to load your impulse: it looks overly smoothed.

I'm a bit confused now - how have you done the correction?

I used your measurement, which was not ideal but still usable. It was also truncated to 15kHz up high.
Just for the sake of it I loaded it in HOLM and smoothed it a bit (I should have done this from the start ;)), and then applied the correction that I posted above:

Attachments:
The following user(s) said Thank You: phranky

Please Log in or Create an account to join the conversation.

How to find out my active speaker's Xover details? 7 years 9 months ago #12792

  • phranky
  • phranky's Avatar Topic Author
  • Offline
  • Junior Member
  • Junior Member
  • Posts: 25
  • Thank you received: 0
Now I wonder, if something with my measurement or settings is wrong or if rePhase just isn't able to linearize phase shifts of this extent?

Please Log in or Create an account to join the conversation.

Last edit: by phranky.

How to find out my active speaker's Xover details? 7 years 9 months ago #12799

  • phranky
  • phranky's Avatar Topic Author
  • Offline
  • Junior Member
  • Junior Member
  • Posts: 25
  • Thank you received: 0

and then applied the correction

Did you apply it in HOLM or in rePhase?

Please Log in or Create an account to join the conversation.

How to find out my active speaker's Xover details? 7 years 9 months ago #12802

  • pos
  • pos's Avatar
  • Offline
  • Platinum Member
  • Platinum Member
  • Posts: 354
  • Thank you received: 200

Now I wonder, if something with my measurement or settings is wrong or if rePhase just isn't able to linearize phase shifts of this extent?

It is right there, post #12780 of this thread .
Just enter those parameters and you are set.
Don't forget the polarity and time offset setting when importing the measurement: as already said setting these correctly is key when tackling a correction.

and then applied the correction

Did you apply it in HOLM or in rePhase?

HOLM can do a convolution between two responses.
One was your original measurement (in blue), the second one the generated correction from rephase (not shown), and you see the result of the convolution in red.
This is the response you should obtain when measuring the output of your openDRC (with correct polarity and time offset)
The following user(s) said Thank You: phranky

Please Log in or Create an account to join the conversation.

Last edit: by pos.
  • Page:
  • 1
  • 2
Moderators: devteam