[Soundminded]

I'm interested I had assumed that reflections would generally make the system non-minimum phase, but this sounds worth looking into 

I think only the direct field applies to this minimum phase theory. In the reverberant field it's an entirely different story. But even in the direct field it doesn't work unless two criteria are met. In the crossover region the time of arrival of sound from both drivers operating at the same frequency must coincide at the listener. This means time delay compensated for differences in group delays between drivers. The other is that the drivers must be coaxial and you must be on axis. When these criteria are met there will be a direct correlation between amplitude and phase response as a function of frequency. Irregularities in phase respose are audible as frequency response variations where sustained tones arrive in and out of phase to reinforce and cancel each other. The exact effect is a function of the geometrical relationship among the drivers and between the drivers and you, the difference in group delay between drivers, and the frequency. There can be anything from a 3 db reinforcement to complete suckout at any frequency. However, being that drivers are not exactly point sources, your ear is not a point receiver, and there are reflections. much if not all of this is almost certainly inaudible.

[medium jim]

I think only the direct field applies to this minimum phase theory. In the reverberant field it's an entirely different story. But even in the direct field it doesn't work unless two criteria are met. In the crossover region the time of arrival of sound from both drivers operating at the same frequency must coincide at the listener. This means time delay compensated for differences in group delays between drivers. The other is that the drivers must be coaxial and you must be on axis. When these criteria are met there will be a direct correlation between amplitude and phase response as a function of frequency. Irregularities in phase respose are audible as frequency response variations where sustained tones arrive in and out of phase to reinforce and cancel each other. The exact effect is a function of the geometrical relationship among the drivers and between the drivers and you, the difference in group delay between drivers, and the frequency. There can be anything from a 3 db reinforcement to complete suckout at any frequency. However, being that drivers are not exactly point sources, your ear is not a point receiver, and there are reflections. much if not all of this is almost certainly inaudible.

Soundminded:

Thanks for the clarification of the Minimum Phase Theory and applied application, makes sense to me!

Jim

[medium jim]

Not trying to start WWIII, 124hz is very close to the note B.  It would be somewhat obvious if there was a suckout if either the B was missing in a passage or very faint.  Look, I'm not negating the value of measuring, and it had nothing to do with "audiophile ego" on my part, just stating a fact.

Jim

[Tyson]

One thing to keep in mind is that measuring is not a simple thing.  For example, you may have a suckout and a peak in the FR at one listening point, but if you move your head 6 inches forward or backward, the peak and suckout will often change.  So, if you "fix" the peak and suckout at the listening spot, it actually makes the total room response worse.  So you have to be very, very careful when doing EQ, and use your ears in addition to the graphs.  If you can to multiple measurements from a variety of different spots clustered around your listening position, then you get a much better idea of what is actually going on.

[*Scotty*]

Tyson, are you talking about response anomalies in the midrange or the highs where wave-lengths are short compared to frequencies below 200Hz.
Scotty

[werd]

I'm interested I had assumed that reflections would generally make the system non-minimum phase, but this sounds worth looking into 

Yes non-min phase issues are always reflection issues. The best way to deal with this is by late-nite listening, Low volume, and we are talking quite low shuts down any reflections. Room issues are almost non existence unless its a really tiny room. If you want to hear what you system sounds like with zero phase issues or suck out of freq this is how you do it. Listening position is paramount but the pay off is huge. The sound stage will expand after a few minutes and then pace and resolution is maximum.  The only adjustable reflection and is floor bounce and is solved using wool rugs.

[Ethan Winer]

you may have a suckout and a peak in the FR at one listening point, but if you move your head 6 inches forward or backward, the peak and suckout will often change.

Exactly. In the graph below, the peak at 42 Hz varies by 3 dB over a distance of only four inches, and there's still a 1 dB difference even as low as 27 Hz. Nulls also tend to occupy a relatively narrow physical space, which is why the nulls on either side of the 92 Hz marker have very different depths. Indeed, the null at 71 Hz in one location becomes a peak at the other.

--Ethan

[JohnR]

I found that in my room the response was fairly consistent over the designated listening area up to 70 Hz, above which it diverged wildly. I averaged the readings (7 of them) and one of the readings was pretty close to the average above 70.

[JohnR]

With regard to minimum phase, here I think is the section of the REW manual andy_c referred to:

http://www.hometheatershack.com/roomeq/wizardhelpv5/help_en-GB/html/minimumphase.html

Something to try ...

So this brings up the question of how to determine the frequency ranges for which the system behavior is minimum-phase.  The REW author, John Mulcahy, has found an innovative way to do this, which is described in the REW manual.

[andy_c]

With regard to minimum phase, here I think is the section of the REW manual andy_c referred to:

http://www.hometheatershack.com/roomeq/wizardhelpv5/help_en-GB/html/minimumphase.html

Something to try ...

Yes, that's it.  After re-reading that writeup, I realized that REW author John M. explains it a lot better than I could.

It's worth mentioning the possible confusion that could come up in discussion of minimum phase.  Under Bode's definition, a system is either minimum-phase or it is not, there is no in-between.  At some point, and I don't know who originated this, the notion of "minimum phase over a frequency band" came up.  The first I noticed it mentioned was in Floyd Toole's book.  I didn't even know what he meant by that when I first read it.  Then I looked up Bode's book (DJVU file), and in chapter 14, he has a formula that can be used to compute the phase shift of a minimum-phase system, given only its magnitude response.  So I just assume that "minimum-phase over a frequency band" means that the measured phase over that band matches very well with the value computed from the magnitude response using Bode's formula.  Then later, I read the material on this subject from the REW manual that JohnR linked above, and was blown away by what the REW author had done.  It's certainly worth investigating, even if it's only for the purpose of determining through some objective approach where not to attempt EQ.

I'm assuming we're only talking subwoofer EQ here, as EQ at frequencies above that is problematic for other reasons already mentioned in this thread.

[Donald]

Here are my Legacy Whisper XD's, measured  individual speakers using OmniMic 1/6 octave at listening position approx. 12 ft. away. 

Have a pair of Dunlavy 4A's beside the XD's. Measured them also at listening position. Again using OmniMic 1/6 octave. Individual measurement's. Will reverse speaker positions in the future and remeasure.





HTR is fully carpeted, with two over stuffed couches, treatments at first reflection points and home made corner traps. (rolls of large insulation stacked approx. 5 ft. high)

XD's come with Xilica processor and have the benefit of being equalized flat to seating position by installer.     

[*Scotty*]

Here are couple of measurements I took this afternoon. I am still feeling my way around with REW software
1/3 oct. measurements obscure a lot of information.

1/48 oct. measurements are very interesting.

The high frequency measurement limit of 16.6kHz is also a little puzzling. The the CrossSpectrum sourced UMM-6 mic is calibrated to 25kHz and with 16/44 sampling I would have thought I could have had measurements out to 20kHz with no problem.
The bass response reflects the contributions of two subs operating out of phase in two different locations in the room. I am going to try to relocate one of them to try to fill in the suckout centered on about 100Hz. and see what happens.
Scotty

[JohnR]

Hi Scotty, looking pretty good  Setting the graph scales (perpendicular arrow button, next to the "gear" button) may be all you need to get up to 22/24k.

You may want to try a measurement sweep instead of RTA.

[jimdgoulding]

That's pretty nice, Donald.  Yer doin something right.  Those are some fine speakers you have and in what I'm guessing is a large room what with 12' away seating, mind blowing.  Welcome to the forums, too.

[kevinh]

Even multiple high quality subs distributed throughout the room to even out room nodes? I use that setup for HT but haven't yet figured out a way to use my subs (3) with both my 2 channel as well as the HT. I think I could get better results, but maybe not.

You night want to check out the E Geddes methodology for setting up multiple subs to work with a stereo setup.

[poseidonsvoice]

Scotty,

I agree with doing a full measurement to get a general idea. Play the speakers at least at 75 dB or above to swamp out room noise.  Also change your vertical scaling from 20 dB to 5 dB  Notice that Ethan Winers graph posted above has 3dB vertical scaling.

After that, limit the sweep to just 300-400 Hz to concentrate on the bass alone. 1st with no smoothing at all. Then add smoothing a little at a time and see how it changes, i.e how resolution is lost. Do all your tweaking with the unsmoothed graph, in particular with frequencies below 300-400 Hz.

Do all this before moving subs around.

Best,

Anand.

[AJinFLA]

Agree with Anand, your vertical scale is too large, try reducing the upper/lower limits to around an 50-70db window (from the current 100+) to better see what is happening. Won't look as pretty but will be much more informative.
Are you measuring from LP/seat? Is your room "treated"? What speakers are being measured?
The midbass dip is possibly a measurement artifact, but the HF attenuation seems a bit much.

cheers,

AJ

[Ethan Winer]

Also change your vertical scaling from 20 dB to 5 dB  Notice that Ethan Winers graph posted above has 3dB vertical scaling.

Exactly. This article shows some guidelines for displaying the various graph types:

Room Measuring Primer

Also, 1/3 or 1/6 octave smoothing is appropriate and useful above 400 or 500 Hz or so, but for the bass range it's best to disable smoothing altogether.

--Ethan

[*Scotty*]

I have managed to take a few more measurements.





I am going to play with relocating one of the subs to try and fill in the hole around 125Hz. and see what happens.
Scotty