rePhase, a loudspeaker phase linearization, EQ and FIR filtering tool - Page 11 - diyAudio
Go Back   Home > Forums > Loudspeakers > Multi-Way

Multi-Way Conventional loudspeakers with crossovers

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 28th December 2012, 04:49 PM   #101
pos is offline pos  Europe
diyAudio Member
 
pos's Avatar
 
Join Date: Feb 2008
Location: Paris
You are absolutly right John, but I think these natural LP filters above the nyquist frequency should be ignored if one wants to visualize and understand the phase behavior of its loudspeakers, and be able to correct them.

Here is the same LR24 @ 1khz simulation with an additional 2nd order LP at 30khz.
When choosing the offset in HOLM it is easy to make this final phase shift (caused by the 30khz LP) completely disappear (as with any LP, which is often a trap when doing this with a woofer measurement...), and end up with the same curve as the above HOLM screenshot.
And I think this should be done, as there is not advantage in correcting this phase shift anyway (and that would suppose its phase behavior is known, and that the mic and measurement gear are perfectly corrected), and in my opinion it just makes things difficult to look at and understand.
Attached Images
File Type: png LR24 1khz + LR12 30khz.PNG (11.6 KB, 304 views)

Last edited by pos; 28th December 2012 at 05:16 PM.
  Reply With Quote
Old 28th December 2012, 04:52 PM   #102
diyAudio Member
 
Join Date: Feb 2009
Location: UK
@pos

Many thanks. I am using a 100uF NP series cap for tweeter protection which may, or may not be adequate...

So what is the general situation with a driver's phase when driven without an electrical filter? Is it usually zero-ish at the low end, and lagging at the high end, for example?

I have been assuming that I can, at a minimum, match the drivers' phase in the crossover regions, but that I might also maintain the time domain coherency between drivers for not only steady state waveforms, but also transients. My idea was that I could correct to a flat 0 degree phase shift for each driver, then tweak the relative delay between them to align the drivers in the time domain (at some defined listening position). Phase shifts of >360 degrees even having optimised the delay would appear to thwart that. Is this a question of minimum vs. excess phase? Why would a nearfield tweeter measurement not give a pure minimum phase result?
  Reply With Quote
Old 28th December 2012, 05:03 PM   #103
pos is offline pos  Europe
diyAudio Member
 
pos's Avatar
 
Join Date: Feb 2008
Location: Paris
Quote:
Originally Posted by CopperTop View Post
So what is the general situation with a driver's phase when driven without an electrical filter? Is it usually zero-ish at the low end, and lagging at the high end, for example?
Ignoring the natural LP of your tweeter, as John pointed out, you should aim for zero-ish at the high end, and the phase should slowly rise as you go lower in frequency (in part due to the cap that will produce a 1st order HP filter somewhere low in frequency, as well as the natural roll off of your tweeter).

As for the other questions, what kind of crossover are you using?
Passive? Active analog? Active DSP IIR? PC based with one convolution per channel?...

To get your filtering correct (ie phase coherent between drivers at and around the crossover points) the easiest thing (and best thing IMHO) to do is to aim for acoustical LR symmetrical slopes (for both amplitude and phase, which is not that easy to do in IIR) and align the delays between your drivers based exactly on the geometrical distance (in depth) of their diaphragms (or geometrically align them, which is often better when it can be done without tradeoff)

That way you will get a text book behavior, that will also be easy to correct using rephase.
  Reply With Quote
Old 28th December 2012, 05:23 PM   #104
pos is offline pos  Europe
diyAudio Member
 
pos's Avatar
 
Join Date: Feb 2008
Location: Paris
John,

I did some additional testing with that LP @ 30khz thing, and your remark was absolutely relevant: it changes the phase behavior up high and cannot be totally ignored by playing with offsets in HOLM.
In fact at best you end up with a phase behavior somewhere in between the two HOLM curves I posted (post #99, auto vs manual offset).

This makes 0 at nyquist a reasonable target to look for in practice, instead of fuzzing with that (purely theoretical, and false as you pointed out) asymptotic target.

Last edited by pos; 28th December 2012 at 05:27 PM.
  Reply With Quote
Old 28th December 2012, 05:32 PM   #105
diyAudio Member
 
john k...'s Avatar
 
Join Date: Aug 2004
Location: US
I don't know pos. With the UE what I do is measure all the drivers with the mic at the listening or design position. Then I removed excess delay from the mic to some point on the baffle surface. This leaves the minimum phase and some excess phase embedded in the SPL data. Then, when I set up the UE it will automatically correct each driver/filter combination so that it is linear phase relative to that arbitrary position on the baffle. The result is that all the drivers + crossover filters are made linear phase relative to positions which are equal distant from the design point. I don't have to compensate for AC offsets as the phase linearization does that.

(It may not be clear, but when the same excess delay is removed for each driver during measurement, even though the ACs may be at different distances, the phase linearization of the embedded excess phased corrects for the driver offsets. I usually use the distance from mic to tweeter flange to set the amount of excess phase to remove. I don't worry about what the phase is at the nyquist.)
__________________
John k.... Music and Design NaO Dipole Loudspeakers.
  Reply With Quote
Old 28th December 2012, 05:45 PM   #106
pos is offline pos  Europe
diyAudio Member
 
pos's Avatar
 
Join Date: Feb 2008
Location: Paris
Yes I see what you mean: any measurement can be corrected, even if it embeds too much phase rotations (delay).
You will loose some precious taps for just delay correction though.
  Reply With Quote
Old 28th December 2012, 05:52 PM   #107
diyAudio Member
 
john k...'s Avatar
 
Join Date: Aug 2004
Location: US
Quote:
Originally Posted by pos View Post
You will loose some precious taps for just delay correction though.
Not really because if you reduce all the drivers to minimum phase and then linearize relative to the driver's ACs you will still have to introduce delays to compensate for the AC offsets. If you keep the small embedded excess phase I referred to in the measurement this just happens automatically. It's, as they say, 6 of one or 1/2 dozen of the other.
__________________
John k.... Music and Design NaO Dipole Loudspeakers.
  Reply With Quote
Old 28th December 2012, 07:30 PM   #108
diyAudio Member
 
Join Date: Feb 2009
Location: UK
So you chaps don't recommend nearfield measurements as the basis of your phase correction, but gated measurements at the listening position?

Not entirely following the part about phase linearisation if your measurement has several phase rotations in it. Is such an impulse response invertible, or is it that you may be losing the absolute time domain coherency of transients?

My system is homebrew PC-based DSP software with linear phase crossover filters which I am modifying to achieve individual driver correction, rather than an overall correction. As I understood it, I can pre-convolve the signal with the driver's inverse impulse response (maybe with some frequency domain smoothing etc.) to correct it, but only if the impulse response is minimum phase.
  Reply With Quote
Old 28th December 2012, 09:37 PM   #109
diyAudio Member
 
Join Date: Nov 2011
Location: Cooktown, Oz
Quote:
Originally Posted by john k... View Post
For example, if the tweeter rolls off 2nd order at 30 K Hz the phase would asymptot to -90 degrees.
Actually the asymptotes are 0 & 180.

The attached pic shows this behaviour for a LR 2nd order filter though this and the other LR pseudo stuff was well known to speaker designers well before they came on the scene.

Phase is in radians. Pi radians is 180 degrees.
__________________

I also attach an excerpt from an early paper on the subject Is Linear Phase Worthwhile? which may help people to distinguish between the 3 types of 'phase' response.

There are many caveats with Digital EQ some of which are discussed in Simple Arbitrary IIRs including the "phase@Nyquist" issue.

My own crude efforts date back to the early 90's (or even late 70's if you include non-digital but sampled data systems )
__________________

Coppertop, your example in #89 shows the IR delay has been overcompensated. IM very HO, any delay correction should make the measured phase very close to Minimum Phase.

This is removing the exp(-jwT) term in PhaseResp.gif

For a single unit with simple passive xovers, the resultant Minimum Phase response should be always falling with frequency unless the amplitude response is seriously wonky.

If it is rising as in your example, you have overcompensated.
Attached Images
File Type: gif 2ndOrder.GIF (24.5 KB, 290 views)
File Type: gif PhaseResp.GIF (52.2 KB, 278 views)
  Reply With Quote
Old 28th December 2012, 10:21 PM   #110
diyAudio Member
 
john k...'s Avatar
 
Join Date: Aug 2004
Location: US
Yes, my bad. A typical tweeter will roll off 2nd order below Fs and 2nd order, or steeper at some frequency above audible. Phase will go from +180 at DC to -180 (or more) at infinity.
__________________
John k.... Music and Design NaO Dipole Loudspeakers.
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
FIR linear phase plugin for MiniDSP? diyjb01 miniDSP 13 7th February 2014 02:24 AM
FIR filter design tool for Loudspeaker magnitude equalization ttmusic Software Tools 3 24th May 2013 09:30 PM
FIR Filtering experiences Olombo PC Based 8 10th February 2013 04:45 PM
AVX based FIR VST, crossover / EQ / DRC and delay KOON3876 PC Based 97 26th November 2012 08:18 AM
Phase EQ using FIR filters Grasso Multi-Way 2 2nd July 2003 11:37 PM


New To Site? Need Help?

All times are GMT. The time now is 04:08 AM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2