Pros & Cons of Various Center Channel Designs

[TheWarrior]

I wouldn’t say it’s a waste of a midwoofer, maybe just a compromised implementation of it. 60% of the entire soundtrack is anchored to the center, therefore greater demand is placed on the center than any other speaker in your setup. A second woofer gives an extra 3dB of headroom, ensuring it can better handle those demands. My setup consists of two rp 150ms for the left and right, and a rp 250c for the center, because of the extra woofer, my left and right speakers are calibrated at 3dB louder than the center to match. Since the horn loaded tweeter can already play stupidly loud in comparison to the direct radiating woofers and has likely been turned way down with an L-pad, it’s the woofers that have greater demand placed on them. Even with a low xover point, the range from 100hz-1khz is the most demanding for two reasons, for 1, a majority of content falls in this range, and two, it’s generally the point where woofers see their lowest impedance, requiring as much as two to three times the power as the upper midrange and treble. Adding a second woofer ensures the center can cope with those greater demands without distortion.

It wouldn’t necessarily have to be an mtm, a tmm with woofers arranged vertically or horizontally would actually work better. If you’ve got a set of tmm towers, a 2.5/3way mmtmm or a similar is needed to keep up with those towers in a home theater setting. The center channel should be the biggest, baddest speaker in your setup in my opinion, and should have more headroom than even the left and right speakers.

Unfortunately, all speaker designs involve a compromise. The only way to completely avoid off axis issues is to use a coaxial alignment, but then you have to deal with greater inter modulation distortion, most home theaters use LCD displays vs perforated projection screens, so the center has to fit below the tv. This either requires a compromise in size, for example, a small bookshelf paired with large towers, or a compromise in driver orientation. It also depends on the behavior of the mtm speaker off axis, and how far off axis your seating goes. For example, I don’t experience any nulls or cancellations up to 45 degrees off axis with the center I’m using now. All of my main seats are within 25 degrees of center axis, with the worst seat being about 40 degrees off axis, so it would make no difference if I were to use a tm design. An mtm design that only managed to remain uniform out to 30 degrees in a larger room with seats further off axis would be an issue however.


Sent from my iPhone using Tapatalk

I failed to specify 2-way MTM in my response, as I did in the original comment. (thought the conversation would stay on topic, I guess) A 2-way MTM is a waste of a midrange. 2.5 - 3 way crossovers are not the same because each driver is playing a different band of frequencies.

The distance between two mid range drivers in an MTM is equal to the wavelength of some mid range frequencies, and will cause cancellations in addition to narrower dispersion. Using multiple woofers to handle bass is fine because yes the power demands are higher, but the wavelengths are so large the cancellations are seldom a problem for woofers.

Coaxial speakers by nature create resonances because of the way they have to be stacked and attached to each other. This shouldn't need explanation when you consider the lack of high end audio designs incorporating coaxial's. KEF's LS50 has published NRC measurements that show the same resonances @ 800 hz and 2 kHz in all of the curves. A peak in on axis and all the off axis measurements centered on the same frequency is a resonance. (I'm clarifying because offense was taken when I made that comment previously. Sorry, but that's how anechoic measurements work!)

 

[shadyJ]

I failed to specify 2-way MTM in my response, as I did in the original comment. (thought the conversation would stay on topic, I guess) A 2-way MTM is a waste of a midrange. 2.5 - 3 way crossovers are not the same because each driver is playing a different band of frequencies.

The distance between two mid range drivers in an MTM is equal to the wavelength of some mid range frequencies, and will cause cancellations in addition to narrower dispersion. Using multiple woofers to handle bass is fine because yes the power demands are higher, but the wavelengths are so large the cancellations are seldom a problem for woofers.

Coaxial speakers by nature create resonances because of the way they have to be stacked and attached to each other. This shouldn't need explanation when you consider the lack of high end audio designs incorporating coaxial's. KEF's LS50 has published NRC measurements that show the same resonances @ 800 hz and 2 kHz in all of the curves. A peak in on axis and all the off axis measurements centered on the same frequency is a resonance. (I'm clarifying because offense was taken when I made that comment previously. Sorry, but that's how anechoic measurements work!)

I wouldn't be so quick to attribute those resonances to the coaxial aspect of the driver. Look at the other KEF speakers that use the Uni-Q driver in soundstage's measurements. Nice smoothing response on nearly all of them.

 

[TheWarrior]

I wouldn't be so quick to attribute those resonances to the coaxial aspect of the driver. Look at the other KEF speakers that use the Uni-Q driver in soundstage's measurements. Nice smoothing response on nearly all of them.

http://www.soundstagenetwork.com/index.php?option=com_content&view=article&id=16&Itemid=18

Really? R100 R500 LS50 Reference One are not smooth from just under 1 kHz to above 2 kHz. Those would be the wavelengths related to the varying size midrange/tweeter coax.

In casual listening KEF sounds wonderful. But there is no denying when you get a chance to A/B them, there are sounds missing.

 

[shadyJ]

I am just not seeing what you are seeing. They aren't perfectly flat but few speakers are. I see a relatively good response in all cases, although not perfect. And what imperfections that exist, I don't see how you are making a connection to the coaxial driver.

 

[gene]

I failed to specify 2-way MTM in my response, as I did in the original comment. (thought the conversation would stay on topic, I guess) A 2-way MTM is a waste of a midrange. 2.5 - 3 way crossovers are not the same because each driver is playing a different band of frequencies.

The distance between two mid range drivers in an MTM is equal to the wavelength of some mid range frequencies, and will cause cancellations in addition to narrower dispersion. Using multiple woofers to handle bass is fine because yes the power demands are higher, but the wavelengths are so large the cancellations are seldom a problem for woofers.

Coaxial speakers by nature create resonances because of the way they have to be stacked and attached to each other. This shouldn't need explanation when you consider the lack of high end audio designs incorporating coaxial's. KEF's LS50 has published NRC measurements that show the same resonances @ 800 hz and 2 kHz in all of the curves. A peak in on axis and all the off axis measurements centered on the same frequency is a resonance. (I'm clarifying because offense was taken when I made that comment previously. Sorry, but that's how anechoic measurements work!)

Some of the best speakers I've ever heard have been MTM's. Dual mids offer more dynamic range and efficiency than a single mid. Narrow vertical dispersion is a good thing for controlling unwanted reflections. If executed right, they are a special breed of speaker. Don't just go by theory, try to demo a few good designs from guys like RBH Sound and Triad. You may be pleasantly surprised.

 

[Dennis Murphy]

I am just not seeing what you are seeing. They aren't perfectly flat but few speakers are. I see a relatively good response in all cases, although not perfect. And what imperfections that exist, I don't see how you are making a connection to the coaxial driver.

Right. I've worked with a number of coaxial drivers, and the tell-tale symptom of a poorly executed design is a sharp dip in the upper highs caused by cancellations in the "tunnel" formed by the woofer cone. KEF appears to have solved that problem. They just haven't been able to design a perfectly flat speaker. I'm not sure anyone's done that.

 

[TheWarrior]

Some of the best speakers I've ever heard have been MTM's. Dual mids offer more dynamic range and efficiency than a single mid. Narrow vertical dispersion is a good thing for controlling unwanted reflections. If executed right, they are a special breed of speaker. Don't just go by theory, try to demo a few good designs from guys like RBH Sound and Triad. You may be pleasantly surprised.

The contempt of MTM's comes from 'Sound Reproduction' where polar plots were used to show the narrow dispersion of an MTM 2-way center channel vs. a 3 way. When using an MTM 2-way horizontally, I will maintain that it is a bad choice due to the cancellations of midrange frequencies.

Vertical dispersion, however, can be good or bad depending on application. Totally agree there!

My RBH WM30's are surprisingly capable speakers. But even they are barely able to satisfy listeners on one couch.

Right. I've worked with a number of coaxial drivers, and the tell-tale symptom of a poorly executed design is a sharp dip in the upper highs caused by cancellations in the "tunnel" formed by the woofer cone. KEF appears to have solved that problem. They just haven't been able to design a perfectly flat speaker. I'm not sure anyone's done that.

My issue with KEF comes from A/B testing of the LS50 that had wonderful sound for casual listening. But there were obvious sounds missing, that were in the frequency range I am talking about.

 

[gene]

The contempt of MTM's comes from 'Sound Reproduction' where polar plots were used to show the narrow dispersion of an MTM 2-way center channel vs. a 3 way. When using an MTM 2-way horizontally, I will maintain that it is a bad choice due to the cancellations of midrange frequencies.

Vertical dispersion, however, can be good or bad depending on application. Totally agree there!

My RBH WM30's are surprisingly capable speakers. But even they are barely able to satisfy listeners on one couch.



My issue with KEF comes from A/B testing of the LS50 that had wonderful sound for casual listening. But there were obvious sounds missing, that were in the frequency range I am talking about.

I'm mostly talking about vertical MTM'S BUT used horizontally with a listener more than 25-30 deg off-axis a W(T/M)W is a much better choice.

 

[TheWarrior]

I'm mostly talking about vertical MTM'S BUT used horizontally with a listener more than 25-30 deg off-axis a W(T/M)W is a much better choice.

Unrelated, but glad to see you still have power and internet!

 

[ATLAudio]

I'm mostly talking about vertical MTM'S BUT used horizontally with a listener more than 25-30 deg off-axis a W(T/M)W is a much better choice.

And as a wise man once said, "...it is important to note that most center channels are rarely listened to at even 30 degrees off-axis. In fact, 20 degrees off-axis response is quite often the worst case scenario."

And also, "...it is important to counter the argument against horizontally mounted MTM designs and note that a well executed MTM can still offer a level of performance that equals or exceeds a “dedicated” (WTMW) center channel speaker design..."

It would seem that your goals for center-channel off-axis listening are crucial to what design would work best for you.

 

[TheWarrior]

And as a wise man once said, "...it is important to note that most center channels are rarely listened to at even 30 degrees off-axis. In fact, 20 degrees off-axis response is quite often the worst case scenario."

And also, "...it is important to counter the argument against horizontally mounted MTM designs and note that a well executed MTM can still offer a level of performance that equals or exceeds a “dedicated” (WTMW) center channel speaker design..."

It would seem that your goals for center-channel off-axis listening are crucial to what design would work best for you.

Except that the data from Harman's tests indicates listener preference for wide dispersion, despite the listener being seated on axis.

 

[Dennis Murphy]

Except that the data from Harman's tests indicates listener preference for wide dispersion, despite the listener being seated on axis.

I think this would be more a cause for concern if an MTM were use horizontally as main L and R for music. As a center channel that mainly will be localizing and giving greater clarity to dialog, I'm not convinced that broad dispersion is a necessity. I wouldn't offer an MTM center if I thought it was seriously compromised in normal use (with no listeners far off axis).

 

[gene]

Except that the data from Harman's tests indicates listener preference for wide dispersion, despite the listener being seated on axis.

That's for left/right stereo pairs. Lateral reflections for a center speaker are not as critical.

See: http://www.audioholics.com/loudspeaker-design/center-channel-speaker

 

[mtrycrafts]

Gene, you are posting in the middle of a hurricane? LOL

 

[rojo]

Gene, you are posting in the middle of a hurricane? LOL

Berta: Sweetie, help me check our stock on flashlight batteries, bottled water, gas for the generator, energy bars, medicine, and all this other stuff.

Gene: Just a sec, honey. Somebody's wrong on the Internet. *tap* *tap* *tap* Aaaaand... Post! Refresh... Refresh...

 

[yepimonfire]

Andrew Jones has written that coaxial speakers do not have substantially more IMD than conventional designs.

Do you have a link? If that’s the case I may look into them, I’ve always avoided them because I find IM distortion to be intolerable.


Sent from my iPhone using Tapatalk

 

[yepimonfire]

@gene

Controlling dispersion can be better achieved by a horn or waveguide in my opinion. If you look at the response of the old 90x60 Klipsch horns, you’ll see out past 30 degrees vertically the highs above the xover drop off rapidly. In an mtm design, vertical cancellations are not always even.

There is a reason pros use horns to control directivity and reduce reflections from walls and ceilings.


Sent from my iPhone using Tapatalk

 

[killdozzer]

Coaxial speakers by nature create resonances because of the way they have to be stacked and attached to each other. This shouldn't need explanation when you consider the lack of high end audio designs incorporating coaxial's. KEF's LS50 has published NRC measurements that show the same resonances @ 800 hz and 2 kHz in all of the curves. A peak in on axis and all the off axis measurements centered on the same frequency is a resonance. (I'm clarifying because offense was taken when I made that comment previously. Sorry, but that's how anechoic measurements work!)

Why didn't you tell me this when I was buying them?!?

This is somewhat unnerving.

So what would be the speaker without this resonance that retains other, desirable acoustic attributes of LS 50?

 

[ATLAudio]

Berta: Sweetie, help me check our stock on flashlight batteries, bottled water, gas for the generator, energy bars, medicine, and all this other stuff.

Gene: Just a sec, honey. Somebody's wrong on the Internet. *tap* *tap* *tap* Aaaaand... Post! Refresh... Refresh...

...and you just won the internet

 

[ATLAudio]

That's for left/right stereo pairs. Lateral reflections for a center speaker are not as critical.

See: http://www.audioholics.com/loudspeaker-design/center-channel-speaker

Moreover, it's important to note just how wide a 30 degrees off axis span is by the time it reaches your listening area. I have wall to wall listening at 30 degrees off axis at the distance I view from.

@gene is this why we see more MTM selection? It seems that you gain max SPL, and few would truly benefit from a different configuration.