How to assess and interpret omnipolar speaker measurements for accuracy?

[MinusTheBear]

For the loudspeaker experts out there that reads these threads - Gene, Dr. Olive etc.

What do you look for in the family of measurements and why? Do different rules apply than the standard "flat frequency response" for omnipolar speakers?

 

[lsiberian]

Ideally you want measurements from the listening areas. But when evaluating third party measurements they would be similar assuming proper placement of omni polars.

 

[cschang]

I could swear I saw a write up on this very subject some time ago. Can't seem to find it now.

 

[AcuDefTechGuy]

I could swear I saw a write up on this very subject some time ago. Can't seem to find it now.

Does it say,"All the measurements on omni speakers are done incorrectly? "

 

[alphaiii]

All measurements I've seen of omnipolar speakers (by HT Mag, Sound & Vision, Soundstage) show a downward sloping response, and in each case, the review states that's what SHOULD be seen with the particular design.

The anechoic or quasi-anechoic measures should show lowered treble response - otherwise, a flat response in these types of measurements would equate to ear-bleeding bright in-room response due to all of the reflected high-end.

I agree with lsiberian though - measurements at the listening position seem to be the only way to know the the final tonal balance will look like. Of course, that will vary ALOT from room to room...

 

[lsiberian]

All measurements I've seen of omnipolar speakers (by HT Mag, Sound & Vision, Soundstage) show a downward sloping response, and in each case, the review states that's what SHOULD be seen with the particular design.

The anechoic or quasi-anechoic measures should show lowered treble response - otherwise, a flat response in these types of measurements would equate to ear-bleeding bright in-room response due to all of the reflected high-end.

I agree with lsiberian though - measurements at the listening position seem to be the only way to know the the final tonal balance will look like. Of course, that will vary ALOT from room to room...

A good speaker designer will have his room in mind. You are correct that the slope should be downward in some designs, but omnipolar leaves a lot of room for a lot of different designs.

 

[zieglj01]

Interesting, on the Omni 250
Mirage Omni 250 Surround Speaker System Measurements | Home Theater

Soundstage, Omni 60
http://www.soundstagemagazine.com/measurements/mirage_omni60/

Audioholics, Omnisat V2
http://www.audioholics.com/reviews/speakers/floorstanding/mirage-omnisat-v2-fs/page-2

 

[alphaiii]

A good speaker designer will have his room in mind. You are correct that the slope should be downward in some designs, but omnipolar leaves a lot of room for a lot of different designs.

Well, remember that "Omnipolar" is a trademarked term of Mirage, and I was specifically thinking of Mirage when referring to the downward sloping response.

Mirage distinguishes its "Omnipolar" designs from true omnidirectional designs. Here's an interesting article discussing the design philosophy with Mirage's former chief designer, Andrew Welker:
SoundStage! Getting Technical Back-Issue Article (2/2004)

As for omnidirectional designs - that may be a different story, and I don't know specifics about target measured responses of those designs.

 

[3db]

All measurements I've seen of omnipolar speakers (by HT Mag, Sound & Vision, Soundstage) show a downward sloping response, and in each case, the review states that's what SHOULD be seen with the particular design.

The anechoic or quasi-anechoic measures should show lowered treble response - otherwise, a flat response in these types of measurements would equate to ear-bleeding bright in-room response due to all of the reflected high-end.

I agree with lsiberian though - measurements at the listening position seem to be the only way to know the the final tonal balance will look like. Of course, that will vary ALOT from room to room...

This is a valid point. So if the designer purposely slopes off the treble to avoid excessive brightness, how can that be viewed as a poor frequency response? The criteria has changed from that of conventional loudspeakers and therefore it's frequency response.

Since omnis are even more room dependent than conventional speakers, I think anechoic measurements should be taken by the reviewer (logistics nightmaree I know) to get any real sense of its performance. Measured anywhere else is adding too much room interaction to make the specs meaningful.

 

[AcuDefTechGuy]

Wow, this Mirage Omni looks awesome in the NRC anechoic chamber:

SoundStage! Measurements - Mirage Omnisat v2 FS Loudspeakers (6/2005)

How does that differ from the Audioholic non anechoic chamber measurement?

http://www.audioholics.com/reviews/speakers/floorstanding/mirage-omnisat-v2-fs/page-2

 

[alphaiii]

Wow, this Mirage Omni looks awesome in the NRC anechoic chamber:

SoundStage! Measurements - Mirage Omnisat v2 FS Loudspeakers (6/2005)

How does that differ from the Audioholic non anechoic chamber measurement?

Omnisat v2 FS Measurements and Analysis — Reviews and News from Audioholics

Note that the Audioholics' measurements were done @ 2m and not at the listening position, presumably to minimize room interaction, so that's probably why they look similar.

Still, the Audioholics measurements show portions of the treble elevated compared to the Soundstage measurements, so I'm guessing there is room influence in those measurements that is not seen with Soundstage's anechoic measurements.

From Audioholics' explanation of the measurements:
"Focusing now on that 6kHz - 20kHz portion of the spectrum, the generally downward slope of the amplitude response plot is quite noticeable. With a front-firing loudspeaker this HF downward slope would indicate, subjectively speaking, a very recessed top end. However, when you take into account that the v2s (when operating in this portion of the spectrum) by design radiate energy all over the room, what appears to be a recessed top end will in fact sound like a clean, respectably smooth HF response characteristic. And it does so because its not only the on-axis response that counts but also the total amount of energy dumped into the room, on and off-axis by the system. It's that sum, perceived by the listener, that counts.

EQ this portion of the v2s direct sound amplitude response flat (as I did for the fun of it) and you'll end up listening to what'll sound like a pair of loud, annoyingly efficient tweeters. So, for a system with the radiation characteristics such as that possessed by the v2s, that downward sloping amplitude response plot is exactly what you want to see in a measurement." - mark, Audioholics

 

[AcuDefTechGuy]

What I mean is both the anechoic and in room FR look like crap.

I'm not saying the speakers will sound like crap.

But the measurements just look bad no matter where they measure it.

To some people, this fact doesn't bother them. But it does bother some.

 

[3db]

I have a question... How do you measure off-axes when the radiation pattern is supposedly 360 degrees?

 

[3db]

What I mean is both the anechoic and in room FR look like crap.

I'm not saying the speakers will sound like crap.

But the measurements just look bad no matter where they measure it.

To some people, this fact doesn't bother them. But it does bother some.

Where the measurements modified to deal with omni? If measured correctly, did the designer purposely roll off the high end to avoid excessive brightness in a typical room? Would this be a good frequency response for an omni?

The point I'm trying to make is use the proper tool for the application.

 

[cschang]

The associated review with the v2's measurements is a good read.

Welcome to www.SoundStageAV.com

 

[lsiberian]

Well, remember that "Omnipolar" is a trademarked term of Mirage, and I was specifically thinking of Mirage when referring to the downward sloping response.

Omnipolar is the common terminology used within DIY circles. I've not heard omnidirectional used. Of course I prefer bipolar(sounds crazy)

 

[AcuDefTechGuy]

Omnipolar is the common terminology used within DIY circles. I've not heard omnidirectional used. Of course I prefer bipolar(sounds crazy)

Hey, as long as they sound great, who cares how they measure.

I'm sure that's what Floyd Toole, Sean Olive, Siegfried Linkwitz, Jim Salk, and Dennis Murphy have in mind.

R---------I----------G--------------H------------T

 

[3db]

Note that the Audioholics' measurements were done @ 2m and not at the listening position, presumably to minimize room interaction, so that's probably why they look similar.

Still, the Audioholics measurements show portions of the treble elevated compared to the Soundstage measurements, so I'm guessing there is room influence in those measurements that is not seen with Soundstage's anechoic measurements.

From Audioholics' explanation of the measurements:
"Focusing now on that 6kHz - 20kHz portion of the spectrum, the generally downward slope of the amplitude response plot is quite noticeable. With a front-firing loudspeaker this HF downward slope would indicate, subjectively speaking, a very recessed top end. However, when you take into account that the v2s (when operating in this portion of the spectrum) by design radiate energy all over the room, what appears to be a recessed top end will in fact sound like a clean, respectably smooth HF response characteristic. And it does so because its not only the on-axis response that counts but also the total amount of energy dumped into the room, on and off-axis by the system. It's that sum, perceived by the listener, that counts.

EQ this portion of the v2s direct sound amplitude response flat (as I did for the fun of it) and you'll end up listening to what'll sound like a pair of loud, annoyingly efficient tweeters. So, for a system with the radiation characteristics such as that possessed by the v2s, that downward sloping amplitude response plot is exactly what you want to see in a measurement." - mark, Audioholics

This makes perfect sense. Although direct measurements look very bad initially, the sum of the mid to high frequncy energy will have the affect of having a conventional speaker with a flat frequency response. This makes it quite plain to me that the technic of measuring conventional loudspeakers fails for omnies and any published frequency response curves as a result of the conventional method are an INACCURATE indicator of how the speaker will actually perform in the mid to high frequency section. This is a perfect example where a speaker outperforms its objective measurements.

 

[alphaiii]

What I mean is both the anechoic and in room FR look like crap.

I'm not saying the speakers will sound like crap.

But the measurements just look bad no matter where they measure it.

To some people, this fact doesn't bother them. But it does bother some.

I would agree they look terrible when judging them by the same criteria as monopole speaker measurements...

But I think we need to remember that measurements of this type of design cannot be interpreted using the same criteria...

 

[cschang]

In a specific room, it would be interesting to see the FRs of an omnipolar speaker vs monopole at the listening position.