Mixed Front and Center speaker - Not a good idea!

[ctrl]

As @amirm and @hardisj measure more and more center speakers and speakers that are potentially suitable for this, people want to mix them with their existing front speakers. So it's time to take a closer look at the facts.

I am sure that many of you have already read somewhere that front and center speakers can't be combined in any way - everyone should really take that to heart.
Sometimes even the manufacturers don't really know what they are doing.

For those who are impatient and don't want to read wordy posts, the crucial thing is that the phase frequency response of the speakers must match each other - if you want to know more about this, you have to read on. Of course everyone is free to say f... the facts!

To show what can go wrong when using different speakers for front and center, here are a few examples.

A) Front and center same speaker (2-Way speakers)

If we are very well-behaved, we use the same speaker as front and center speaker. When acoustically summing the speakers, what you would expect happens, the sound pressure simply increases.
Since the speakers are identical, we only see one phase frequency response and and in blue the summed FR.




B) Front and center same speaker but different crossover frequency (2-Way speakers)

This can happen if the manufacturer significantly lowers the crossover frequency for better horizontal dispersion of the center speaker.
If the difference in crossover frequency (between front and center speaker) is less than half an octave (only intended as a rule of thumb), it's not a big problem.
But if the difference is an octave or more, it gets nasty due to the different phase frequency responses in the crossover frequency range.

a) crossover frequency less then a half octave apart - [email protected] versus LR4@2kHz (LR4 = Linkwitz-Riley crossover 4th order - the most used crossover filter)
The frequency responses of the two speakers are almost identical and are at 90dB. The phase frequency responses (green and red curves) are shifted by 800Hz to each other. In blue the summed total frequency response is shown.

The different phase frequency responses lead to interference and in this case to a one dB dip.

b) crossover frequency one octave apart - LR4@2kHz versus LR4@1kHz.

In this case, the shifted phase frequency responses lead to a five dB dip. This would ruin the home cinema experience, probably even your whole life and it will probably start raining frogs.



C) Front and center same crossover frequency but different filter order (2-Way speakers)

Assuming your front speakers have a crossover frequency around, say, 2kHz, then you might think that if the center speakers also have a crossover frequency of 2kHz, everything is fine.
This can fit, but only if the filter order is also identical - if both speaker use e.g. a Linkwitz-Riley fourth order crossover filter.

Assuming two manufacturers specify the crossover frequency of a 2-way front/center speaker as 2kHz, but use a different filter order, one second order and one fourth order, the summed sound pressure might look like this:

Again, the frequency responses of the two speakers are nearly identical at 90dB.
But for the loudspeaker with the second-order crossover frequency to succeed in summing the filter slopes, the polarity must be inverted for one driver - in the example it is the tweeter.
Thus, at high frequencies, the phase frequency responses of the two loudspeakers are 180° apart, resulting in complete sound cancellation.


Different filter types with same filter order is for the common used filter types (Linkwitz-Riley, Butterworth) less a problem.



D) Front and center speakers other combinations

If you now combine 3-way front speakers with 2- or 3-way center speakers, it quickly becomes clear that a lot can go wrong. Even manufacturers often offer only 2-way center speaker to 3-way front speaker.

An example of this.
The 3-way front speaker has crossover frequencies at 800Hz and 3400Hz, the 2-way center speaker at 2600Hz. If all crossover filters use LR4, there are no problems with the polarity of the drivers, only the phase frequency responses are shifted against each other because of the different crossover frequencies.

As always, the frequency responses of the two speakers are almost identical (there are minor differences due to the mutual influence of the filter slopes). But due to the different phase frequency responses, the example combination of 3-way front with 2-way center speaker results in a cancellation around 1kHz, which will be audible. In this case, it would probably be better to spend the money on sex, drugs and rock'n'roll instead of a center speaker (but that's just my opinion)
How much the interference will be audible certainly depends on the individual case. The theoretical impact is explained on the basis of the examples.



E) Update
As others point out in posts below, these potential theoretical influences are not subjectively perceived by many.

A simple test to exclude a possible influence (by the center speaker) on the timbre would be to first run pink noise through the front speaker and then switch on the center speaker with pink noise too.
In the best case, only the volume impression should change, in no case should the timbre change.



F) How can you tell in advance whether a combination of front and center speaker will work?
To be continued...

 

[Blumlein 88]

I think there is a conceptual mistake here. Your idea of summing identical speakers only works if they are at the same location, or equi-distant from the listener and there are no room reflections or resonances that matter. The first is no use for multi-channel audio. The latter is only possible in an anechoic chamber, outdoors, or a heavily damped and very large space which still only gets close. Further all of this only works for one point in space for the listening position. All conditions that pretty much won't exist in real world use.

Also the level of ill effects described for mismatches are also only that severe in the same generally never seen conditions described above.

I've had a 5 channel system with 5 identical speakers, and there are benefits to it. However my current setup has identical center and surround with different front left and right. All things considered it is better. Mainly because the front left and right are better speakers. So as always in such a complicated world there are trade offs.

In essence however your summing is a wrong concept for what you are trying to show here. When your premises are wrong.................

 

[richard12511]

As @amirm and @hardisj measure more and more center speakers and speakers that are potentially suitable for this, people want to mix them with their existing front speakers. So it's time to take a closer look at the facts.

I am sure that many of you have already read somewhere that front and center speakers can't be combined in any way - everyone should really take that to heart.
Sometimes even the manufacturers don't really know what they are doing.

For those who are impatient and don't want to read wordy posts, the crucial thing is that the phase frequency response of the speakers must match each other - if you want to know more about this, you have to read on. Of course everyone is free to say f... the facts!

To show what can go wrong when using different speakers for front and center, here are a few examples.

A) Front and center same speaker (2-Way speakers)

If we are very well-behaved, we use the same speaker as front and center speaker. When acoustically summing the speakers, what you would expect happens, the sound pressure simply increases.
Since the speakers are identical, we only see one phase frequency response and and in blue the summed FR.
View attachment 119362



B) Front and center same speaker but different crossover frequency (2-Way speakers)

This can happen if the manufacturer significantly lowers the crossover frequency for better horizontal dispersion of the center speaker.
If the difference in crossover frequency (between front and center speaker) is less than half an octave (only intended as a rule of thumb), it's not a big problem.
But if the difference is an octave or more, it gets nasty due to the different phase frequency responses in the crossover frequency range.

a) crossover frequency less then a half octave apart - [email protected] versus LR4@2kHz (LR4 = Linkwitz-Riley crossover 4th order - the most used crossover filter)
The frequency responses of the two speakers are almost identical and are at 90dB. The phase frequency responses (green and red curves) are shifted by 800Hz to each other. In blue the summed total frequency response is shown.
View attachment 119363
The different phase frequency responses lead to interference and in this case to a one dB dip.

b) crossover frequency one octave apart - LR4@2kHz versus LR4@1kHz.
View attachment 119365
In this case, the shifted phase frequency responses lead to a five dB dip. This would ruin the home cinema experience, probably even your whole life and it will probably start raining frogs.



C) Front and center same crossover frequency but different filter order (2-Way speakers)

Assuming your front speakers have a crossover frequency around, say, 2kHz, then you might think that if the center speakers also have a crossover frequency of 2kHz, everything is fine.
This can fit, but only if the filter order is also identical - if both speaker use e.g. a Linkwitz-Riley fourth order crossover filter.

Assuming two manufacturers specify the crossover frequency of a 2-way front/center speaker as 2kHz, but use a different filter order, one second order and one fourth order, the summed sound pressure might look like this:
View attachment 119368
Again, the frequency responses of the two speakers are nearly identical at 90dB.
But for the loudspeaker with the second-order crossover frequency to succeed in summing the filter slopes, the polarity must be inverted for one driver - in the example it is the tweeter.
Thus, at high frequencies, the phase frequency responses of the two loudspeakers are 180° apart, resulting in complete sound cancellation.


Different filter types with same filter order is for the common used filter types (Linkwitz-Riley, Butterworth) less a problem.



D) Front and center speakers other combinations

If you now combine 3-way front speakers with 2- or 3-way center speakers, it quickly becomes clear that a lot can go wrong. Even manufacturers often offer only 2-way center speaker to 3-way front speaker.

An example of this.
The 3-way front speaker has crossover frequencies at 800Hz and 3400Hz, the 2-way center speaker at 2600Hz. If all crossover filters use LR4, there are no problems with the polarity of the drivers, only the phase frequency responses are shifted against each other because of the different crossover frequencies.
View attachment 119455
As always, the frequency responses of the two speakers are almost identical (there are minor differences due to the mutual influence of the filter slopes). But due to the different phase frequency responses, the example combination of 3-way front with 2-way center speaker results in a cancellation around 1kHz, which will be audible. In this case, it would probably be better to spend the money on sex, drugs and rock'n'roll instead of a center speaker (but that's just my opinion).



How can you tell in advance whether a combination of front and center speaker will work?
To be continued...

What if it's the same speaker, but at different distances? I have identical LCR, but the left and right speakers are out further into the room and are about 30cm closer to the listening position than the center. I find that dialog glues to the TV so much better when the center speaker is closer to it.

 

[Sancus]

There was a study done in '98 which basically showed that if your center dispersion pattern doesn't match the L/R it's very bad, people hate it -- significantly worse than just about any matching combination.

(question 3 is: "How natural is the projected presentation?")

 

[HooStat]

But if 90% of the information is coming from the center channel with most TV shows and many movies, does it really matter?

 

[Vasr]

As @amirm and @hardisj measure more and more center speakers and speakers that are potentially suitable for this, people want to mix them with their existing front speakers. So it's time to take a closer look at the facts.

I am sure that many of you have already read somewhere that front and center speakers can't be combined in any way - everyone should really take that to heart.
Sometimes even the manufacturers don't really know what they are doing.

For those who are impatient and don't want to read wordy posts, the crucial thing is that the phase frequency response of the speakers must match each other - if you want to know more about this, you have to read on. Of course everyone is free to say f... the facts!

To show what can go wrong when using different speakers for front and center, here are a few examples.

A) Front and center same speaker (2-Way speakers)

If we are very well-behaved, we use the same speaker as front and center speaker. When acoustically summing the speakers, what you would expect happens, the sound pressure simply increases.
Since the speakers are identical, we only see one phase frequency response and and in blue the summed FR.
View attachment 119362



B) Front and center same speaker but different crossover frequency (2-Way speakers)

This can happen if the manufacturer significantly lowers the crossover frequency for better horizontal dispersion of the center speaker.
If the difference in crossover frequency (between front and center speaker) is less than half an octave (only intended as a rule of thumb), it's not a big problem.
But if the difference is an octave or more, it gets nasty due to the different phase frequency responses in the crossover frequency range.

a) crossover frequency less then a half octave apart - [email protected] versus LR4@2kHz (LR4 = Linkwitz-Riley crossover 4th order - the most used crossover filter)
The frequency responses of the two speakers are almost identical and are at 90dB. The phase frequency responses (green and red curves) are shifted by 800Hz to each other. In blue the summed total frequency response is shown.
View attachment 119363
The different phase frequency responses lead to interference and in this case to a one dB dip.

b) crossover frequency one octave apart - LR4@2kHz versus LR4@1kHz.
View attachment 119365
In this case, the shifted phase frequency responses lead to a five dB dip. This would ruin the home cinema experience, probably even your whole life and it will probably start raining frogs.



C) Front and center same crossover frequency but different filter order (2-Way speakers)

Assuming your front speakers have a crossover frequency around, say, 2kHz, then you might think that if the center speakers also have a crossover frequency of 2kHz, everything is fine.
This can fit, but only if the filter order is also identical - if both speaker use e.g. a Linkwitz-Riley fourth order crossover filter.

Assuming two manufacturers specify the crossover frequency of a 2-way front/center speaker as 2kHz, but use a different filter order, one second order and one fourth order, the summed sound pressure might look like this:
View attachment 119368
Again, the frequency responses of the two speakers are nearly identical at 90dB.
But for the loudspeaker with the second-order crossover frequency to succeed in summing the filter slopes, the polarity must be inverted for one driver - in the example it is the tweeter.
Thus, at high frequencies, the phase frequency responses of the two loudspeakers are 180° apart, resulting in complete sound cancellation.


Different filter types with same filter order is for the common used filter types (Linkwitz-Riley, Butterworth) less a problem.



D) Front and center speakers other combinations

If you now combine 3-way front speakers with 2- or 3-way center speakers, it quickly becomes clear that a lot can go wrong. Even manufacturers often offer only 2-way center speaker to 3-way front speaker.

An example of this.
The 3-way front speaker has crossover frequencies at 800Hz and 3400Hz, the 2-way center speaker at 2600Hz. If all crossover filters use LR4, there are no problems with the polarity of the drivers, only the phase frequency responses are shifted against each other because of the different crossover frequencies.
View attachment 119455
As always, the frequency responses of the two speakers are almost identical (there are minor differences due to the mutual influence of the filter slopes). But due to the different phase frequency responses, the example combination of 3-way front with 2-way center speaker results in a cancellation around 1kHz, which will be audible. In this case, it would probably be better to spend the money on sex, drugs and rock'n'roll instead of a center speaker (but that's just my opinion).



How can you tell in advance whether a combination of front and center speaker will work?
To be continued...

Are you talking about using three speakers with the same signal going in all three channels or three separate channels with their own sounds some of which overlap to provide stage location (not all sounds in center speaker come from the center only) spanning the entire width (and in surround set ups 360 degrees)?

For your assertions to make sense, I assume you are talking about the latter where there is some overlap between center and either left or right to place images left or right of center and how they may interfere with each other. This is no different from the L and R interfering with each other for common content by being out of phase. Which is why in room correction you align phases (which is also influenced by setup geometry relative to MLP). You do the same with center and mains as well. That takes care of the first order phase differences due to distances from MLP.

Now, the second order issues are from the different phase shifting characteristics of speakers relative to each other and perhaps even frequency dependent or crossover dependent as you point out. I am not sure this is more than just academic interest with very little practical impact.

I have used many different mix and match of speakers for multi-channel over the years and the thing that seems to make the most difference in seamless integration is similar tonal balance rather than phase characteristics. Phase differences affects volume but in most multi-channel, there is seldom equal content going into two different speakers to have noticeable (audible) artifacts. Most artifacts, if any, only manifest in transitions - when the virtual sound source appears to be moving and so transitioning between speakers.

What you don't want to happen is a plane sounding like a drone appearing from the left and sound like a jet as it transitions to center or vice versa.

So that you are not misunderstood can you set the context better of what kind of content this is an issue in? Pure multi-channel, Stereo played in all speakers as "All Stereo" mode, a test tone from an analyzer played through all speakers, etc.

 

[ctrl]

I think there is a conceptual mistake here. Your idea of summing identical speakers only works if they are at the same location, or equi-distant from the listener and there are no room reflections or resonances that matter.

Which is why in room correction you align phases (which is also influenced by setup geometry relative to MLP). You do the same with center and mains as well. That takes care of the first order phase differences due to distances from MLP.

Exactly, just as an isosceles stereo triangle is important for optimal stereo listening, the center speaker should also be placed "correctly" or adjusted by delay.

Room reflections are, of course, always present and influence the envelopment, image and clarity; the considerations made in the first post are, as with stereo listening, primarily about the direct sound.

What if it's the same speaker, but at different distances?

Then you get comb filter effects, similar to offset stereo speakers. How much this is noticeable or distracting is difficult to predict, perhaps it is not distracting at all and is only indirectly more "strenuous" for the brain.
But as with stereo listening, why deliberately make a "mistake" when it is easily avoidable.

But if 90% of the information is coming from the center channel with most TV shows and many movies, does it really matter?

If the signals between center and front are uncorrelated, then it does not matter.

But if the center speaker is involved in left-right effects (as others have said), then the phase frequency response between front and center does matter - how much is likely to depend on the individual case.
If, for example, the woofers at the center are connected with inverse polarity to the front speaker, then this will be noticeable by cancellations in the low frequencies.
Not as dramatic as with stereo speakers connected with different polarity, but under certain circumstances it should be noticeable.

Now, the second order issues are from the different phase shifting characteristics of speakers relative to each other and perhaps even frequency dependent or crossover dependent as you point out. I am not sure this is more than just academic interest with very little practical impact.

Judging by your reactions and experiences, you might be right.

I think we have to distinguish between the different cases. If only the crossover frequencies between front and center are somewhat different and there are dips with high Q, then this should not be a problem because the hearing simply ignores them - I hope that my ironic exaggerations in the first post were not taken literally - at these points I should probably edit my post.
It might be different if the phase frequency response of the center is in phase opposition to the front speakers over large frequency ranges.

However my current setup has identical center and surround with different front left and right. All things considered it is better. Mainly because the front left and right are better speakers. So as always in such a complicated world there are trade offs.

It would be interesting to see if the loudspeakers match each other from their phase frequency responses, or if it is even perceived as positive if they do not match at all.


If someone has either exact manufacturer data for his front and center speakers, such as crossover frequencies with (filter type and) filter order, or if there are measurements of the step response for the speakers (with crossover frequency), then one could simulate how front and center add up.
I would be very interested in what comes out.

 

[Blumlein 88]

My current HT setup is a pair of 3 ways right and left and a two way for the center. While I think matched all the way should have a few advantages I don't know how important they'll turn out to be. For stereo separation for instance a difference of 15 db gets you maximally left or right. It will be a small amount for center to left or center to right. So it is possible the various phase effects you are worried about aren't very noticeable by us as it takes little to maximally shift where we hear things from.

The way to experiment is setup an all identical front three speakers, and introduce the dips and things artificially into the signal and see if those are as audible as you think they might be.

 

[ctrl]

The way to experiment is setup an all identical front three speakers, and introduce the dips and things artificially into the signal and see if those are as audible as you think they might be.

It would be even easier to use pink noise.

A simple test to exclude a possible influence by the center speaker on the timbre would be to first run pink noise through the front speaker and then switch on the center speaker with pink noise too.
In the best case, only the volume impression should change, in no case should the timbre change.

 

[Blumlein 88]

The FR of your ears is slightly different straight ahead vs 30-45 degrees to each side. So you would in fact expect a bit of difference even with pink noise.

 

[Rick Sykora]

My experience suggests this problem is more complex than the simply using a matching speaker. I have done this more than a few times with less than optimal results. Often you hear that the center anchors the sound. If that is the preferred answer, then maybe am wrong as these centers did anchor my attention to them way more than I like.

I prefer when the center blends well with the LR pair and sounds are transparently emanating from the source on the screen. Otherwise it seems distractingly annoying to me. I have had centers from Dayton, Jamo, Mirage and used "matching" speakers by HTD, Zaph and Signet. Despite trying different locations around my flat screen, the only one that met my needs (in a more traditional system) was the cheap Dayton. It is shaped in an arc and has the desirable mid/tweeter combo.

Today, I use a Mirage OMD-5 with my CBT24s and that works well. The CBTs did not have a matching center, so had to be creative. It sits on a shelf just above the tv. My family is happy with the sound too. I use OMD-5s as surrounds too. So there is a match there as well. So suggest it may be important to define what the desirable outcome is.

From my experience matching speakers does not necessarily result in matching sound!

 

[ctrl]

Today, I use a Mirage OMD-5 with my CBT24s and that works well. The CBTs did not have a matching center, so had to be creative. It sits on a shelf just above the tv. My family is happy with the sound too. I use OMD-5s as surrounds too. So there is a match there as well. So suggest it may be important to define what the desirable outcome is.

From my experience matching speakers does not necessarily result in matching sound!

Rick, thanks for the information.

With this information one can already start something.
I read somewhere that the crossover frequency for the Mirage OMD-5 speaker is 2.3kHz. I now just assume that this speaker uses a fourth order crossover filter (LR4 is certainly the most widely used crossover filter).

With the CBT24 used as front speaker it is especially easy, because this is a full range speaker.

If the center speaker is optimally adjusted to the sweet spot via delay (or placement), the following sound addition results (maybe I should have lowered the sound pressure level of the center speaker a bit):

As has already been said, outside the sweet spot, as with reflections, comb filter effects occur which our brain can actually handle very well - if they are sufficiently "randomly distributed".
Exactly in the sweet spot the combination has hardly any effect, the dip rather contributes to a particularly "warm" timbre.

Then the crossover tuning does not play a role after all! I am not convinced of this, despite the many dissenting voices.

If Rick would invert the phase or reverse the polarity of his center speaker, then the following addition of front and center speaker would result:

Exactly in the sweet spot frequency ranges are amplified, which rather provide for a harsh timbre. Again, comb filter effects occur in the high frequencies, outside the sweet spot. But in the lower midrange and in the bass range, comb filter effects (around the sweat spot) hardly have any effect due to the large wavelength. There it comes to (partial) cancellations.

If these considerations made by me do not play a role, because due to the reflections, possible effects are completely masked, then it should not matter with which polarity a center speaker is connected.

 

[Frank Dernie]

As @amirm and @hardisj measure more and more center speakers and speakers that are potentially suitable for this, people want to mix them with their existing front speakers. So it's time to take a closer look at the facts.

I am sure that many of you have already read somewhere that front and center speakers can't be combined in any way - everyone should really take that to heart.
Sometimes even the manufacturers don't really know what they are doing.

For those who are impatient and don't want to read wordy posts, the crucial thing is that the phase frequency response of the speakers must match each other - if you want to know more about this, you have to read on. Of course everyone is free to say f... the facts!

To show what can go wrong when using different speakers for front and center, here are a few examples.

A) Front and center same speaker (2-Way speakers)

If we are very well-behaved, we use the same speaker as front and center speaker. When acoustically summing the speakers, what you would expect happens, the sound pressure simply increases.
Since the speakers are identical, we only see one phase frequency response and and in blue the summed FR.
View attachment 119362



B) Front and center same speaker but different crossover frequency (2-Way speakers)

This can happen if the manufacturer significantly lowers the crossover frequency for better horizontal dispersion of the center speaker.
If the difference in crossover frequency (between front and center speaker) is less than half an octave (only intended as a rule of thumb), it's not a big problem.
But if the difference is an octave or more, it gets nasty due to the different phase frequency responses in the crossover frequency range.

a) crossover frequency less then a half octave apart - [email protected] versus LR4@2kHz (LR4 = Linkwitz-Riley crossover 4th order - the most used crossover filter)
The frequency responses of the two speakers are almost identical and are at 90dB. The phase frequency responses (green and red curves) are shifted by 800Hz to each other. In blue the summed total frequency response is shown.
View attachment 119363
The different phase frequency responses lead to interference and in this case to a one dB dip.

b) crossover frequency one octave apart - LR4@2kHz versus LR4@1kHz.
View attachment 119365
In this case, the shifted phase frequency responses lead to a five dB dip. This would ruin the home cinema experience, probably even your whole life and it will probably start raining frogs.



C) Front and center same crossover frequency but different filter order (2-Way speakers)

Assuming your front speakers have a crossover frequency around, say, 2kHz, then you might think that if the center speakers also have a crossover frequency of 2kHz, everything is fine.
This can fit, but only if the filter order is also identical - if both speaker use e.g. a Linkwitz-Riley fourth order crossover filter.

Assuming two manufacturers specify the crossover frequency of a 2-way front/center speaker as 2kHz, but use a different filter order, one second order and one fourth order, the summed sound pressure might look like this:
View attachment 119368
Again, the frequency responses of the two speakers are nearly identical at 90dB.
But for the loudspeaker with the second-order crossover frequency to succeed in summing the filter slopes, the polarity must be inverted for one driver - in the example it is the tweeter.
Thus, at high frequencies, the phase frequency responses of the two loudspeakers are 180° apart, resulting in complete sound cancellation.


Different filter types with same filter order is for the common used filter types (Linkwitz-Riley, Butterworth) less a problem.



D) Front and center speakers other combinations

If you now combine 3-way front speakers with 2- or 3-way center speakers, it quickly becomes clear that a lot can go wrong. Even manufacturers often offer only 2-way center speaker to 3-way front speaker.

An example of this.
The 3-way front speaker has crossover frequencies at 800Hz and 3400Hz, the 2-way center speaker at 2600Hz. If all crossover filters use LR4, there are no problems with the polarity of the drivers, only the phase frequency responses are shifted against each other because of the different crossover frequencies.
View attachment 119455
As always, the frequency responses of the two speakers are almost identical (there are minor differences due to the mutual influence of the filter slopes). But due to the different phase frequency responses, the example combination of 3-way front with 2-way center speaker results in a cancellation around 1kHz, which will be audible. In this case, it would probably be better to spend the money on sex, drugs and rock'n'roll instead of a center speaker (but that's just my opinion)
How much the interference will be audible certainly depends on the individual case. The theoretical impact is explained on the basis of the examples.



E) Update
As others point out in posts below, these potential theoretical influences are not subjectively perceived by many.

A simple test to exclude a possible influence (by the center speaker) on the timbre would be to first run pink noise through the front speaker and then switch on the center speaker with pink noise too.
In the best case, only the volume impression should change, in no case should the timbre change.



F) How can you tell in advance whether a combination of front and center speaker will work?
To be continued...

If any of this is true a multi-channel system can just f*ck off an leave me alone

 

[Rick Sykora]

Rick, thanks for the information.

With this information one can already start something.
I read somewhere that the crossover frequency for the Mirage OMD-5 speaker is 2.3kHz. I now just assume that this speaker uses a fourth order crossover filter (LR4 is certainly the most widely used crossover filter).

With the CBT24 used as front speaker it is especially easy, because this is a full range speaker.

If the center speaker is optimally adjusted to the sweet spot via delay (or placement), the following sound addition results (maybe I should have lowered the sound pressure level of the center speaker a bit):
View attachment 120949
As has already been said, outside the sweet spot, as with reflections, comb filter effects occur which our brain can actually handle very well - if they are sufficiently "randomly distributed".
Exactly in the sweet spot the combination has hardly any effect, the dip rather contributes to a particularly "warm" timbre.

Then the crossover tuning does not play a role after all! I am not convinced of this, despite the many dissenting voices.

If Rick would invert the phase or reverse the polarity of his center speaker, then the following addition of front and center speaker would result:
View attachment 120950
Exactly in the sweet spot frequency ranges are amplified, which rather provide for a harsh timbre. Again, comb filter effects occur in the high frequencies, outside the sweet spot. But in the lower midrange and in the bass range, comb filter effects (around the sweat spot) hardly have any effect due to the large wavelength. There it comes to (partial) cancellations.

If these considerations made by me do not play a role, because due to the reflections, possible effects are completely masked, then it should not matter with which polarity a center speaker is connected.

since it is in the same room as I do Directiva measurements, when I take a break, will see if your theory holds.

 

[Rick Sykora]

If any of this is true a multi-channel system can just f*ck off an leave me alone

There is so much processing going on in multi-channel, as long as the subs integrate well, I judge it more subjectively than a setup where am trying to listen critically.

 

[Rick Sykora]

Rick, thanks for the information.

With this information one can already start something.
I read somewhere that the crossover frequency for the Mirage OMD-5 speaker is 2.3kHz. I now just assume that this speaker uses a fourth order crossover filter (LR4 is certainly the most widely used crossover filter).

With the CBT24 used as front speaker it is especially easy, because this is a full range speaker.

If the center speaker is optimally adjusted to the sweet spot via delay (or placement), the following sound addition results (maybe I should have lowered the sound pressure level of the center speaker a bit):
View attachment 120949
As has already been said, outside the sweet spot, as with reflections, comb filter effects occur which our brain can actually handle very well - if they are sufficiently "randomly distributed".
Exactly in the sweet spot the combination has hardly any effect, the dip rather contributes to a particularly "warm" timbre.

Then the crossover tuning does not play a role after all! I am not convinced of this, despite the many dissenting voices.

If Rick would invert the phase or reverse the polarity of his center speaker, then the following addition of front and center speaker would result:
View attachment 120950
Exactly in the sweet spot frequency ranges are amplified, which rather provide for a harsh timbre. Again, comb filter effects occur in the high frequencies, outside the sweet spot. But in the lower midrange and in the bass range, comb filter effects (around the sweat spot) hardly have any effect due to the large wavelength. There it comes to (partial) cancellations.

If these considerations made by me do not play a role, because due to the reflections, possible effects are completely masked, then it should not matter with which polarity a center speaker is connected.

So, it is lunch and so briefly tore down speaker measurement and did room measurement setup. My initial results showed the response was pretty flat regardless of whether the center was contributing or not.

Here is how I did the measurement. Microphone is at listening position pointed towards ceiling, mic cal is set for 90 degrees and SPL set to 75 db. First sweep was with center disconnected and second sweep was taken with center connected. The output from pc sound card is set for R+L and is connected via stereo to the AUX on my AVR. DSP mode on AVR is set to All Channel Stereo. This is the only setting where I can get the center to have output.

This older Onkyo AVR has Audyssey and has been calibrated. To have output from the center without the DSP would have to connect some separate amps. Would take some doing as I do not have 3 matching ones.

 

[ctrl]

Here is how I did the measurement. Microphone is at listening position pointed towards ceiling, mic cal is set for 90 degrees and SPL set to 75 db. First sweep was with center disconnected and second sweep was taken with center connected.

Have you also made a measurement with the center speaker connected with reverse polarity?

 

[Rick Sykora]

Have you also made a measurement with the center speaker connected with reverse polarity?

Yes, minimal impact with polarity reversal too. Without the CBTs, the output of the center alone is 70 dB btw.

 

[Rick Sykora]

Have you also made a measurement with the center speaker connected with reverse polarity?

Remembered I built an NC250 mono amp. So once I figured out a 3-way splitter, ran sweeps with straight amplification. No control over relative level, but comparable results whether center is same phase or not. I even brought the center down off the shelf and on a stand closer to ear level.

Only diff from AVR is more high freq rolloff. This is to be expected with no eq on CBT24s.