Center channel theory

[Feanor]

Reading this thread prompts me to wonder...is this type of center design only for aesthetics?



In other words, is this a compromised design that would this actually be outperformed by an exact match of the front speakers, which look like this?

A damned good question. My understanding is that multi-speaker arrays limit amplitude a long the axis of the array: this being true, two woofers arranged horizontally would limit side to side dispersion -- not a good idea.

So again, if true, the sideways arrangement would be purely cosmetic and actually counter-functional.

[Sir Terrence the Terrible]

Reading this thread prompts me to wonder...is this type of center design only for aesthetics?

This design is problematic, but less so than most center speakers. What makes this speaker light years ahead of the typical center is the symmetrical vertical placement of the tweeter and midrange. The wide placement of the woofers virtually assures lobing off axis, and that is the problem with horizontally mounted center speakers.

This design is mainly to satisfy the WAF. It seems that the THX mandated similar speaker layout created problems aesthetically when placed vertically over or under a television set. So in response, manufacturers start making lower profile center speakers which could fit atop a television without poking up in the air like a 10 story building. The trade off was a non-uniform off axis response(which causes dialog intelligibility issues for viewers seated off axis). To make this issue less troublesome, some manufacturers added a midrange drivers under the tweeter so the woofers could be crossed over before lobing happens. Lobing still occurs, just not over as wide a frequency as the single tweeter design.

Two woofers are in center channels to make them more power compatible with tower speakers, or speakers that deploy more than one bass driver in their cabinets.

In other words, is this a compromised design that would this actually be outperformed by an exact match of the front speakers, which look like this?

I would say it is a better match frequency and dispersion wise, but would probably suffer power wise to dual mid/woofer designs.

There is no problem with having two bass/mid drivers in a center speaker, as long as the drivers are lined up in a vertical fashion, not a horizontal setup. THX had this one right.

[Mr Peabody]

According to this http://www.audioholics.com/education...peaker-designs the OP would have the optimum set up using 3 matching speakers, for left/right/center. He pretty much found out what Sir T was saying.

[JoeE SP9]

I've tried various center channel speakers. Even though my house is my man cave, adding another large ESL is too much for even me. After I tried a succession of "monkey coffins" I went back to a phantom center. The timbre difference between my stats and anything else was to much. When I win the lottery I may look at ML ESL centers. Until then it's phantom for speech.

[Feanor]

According to this http://www.audioholics.com/education...peaker-designs the OP would have the optimum set up using 3 matching speakers, for left/right/center. He pretty much found out what Sir T was saying.

Interesting, Mr P.

So I think this tends to vindicate my decission to use a single mid-woofer in my DIY center. There would be some interference between the mid-woofer and tweeter, but this would be limited to the crossover overlap, and thus would be much less than between two mid-woofers carrying the same frequencies. No??

Of course an over/under orientation of the Tweeter <> Mid would best but I've got the usual space limitations.

[Mr Peabody]

Feanor, did you use the same drivers as in your mains? If not, even though you may have a vertical design your set up is still not correct because you aren't timbre matched. I feel some off axis issues are far better than mismatched speakers. The OP was going to use 3 matching speakers as the guy in the article determined 3 matching bookshelves was better than 2 matching bookshelves and a matching horisontal speaker.

I'm not sure if it will work but if so I want to try watching a movie with my center vertical to see if any difference can be noticed. It may only be those off to the side though that can tell any difference.

[Feanor]

Feanor, did you use the same drivers as in your mains? If not, even though you may have a vertical design your set up is still not correct because you aren't timbre matched. I feel some off axis issues are far better than mismatched speakers. The OP was going to use 3 matching speakers as the guy in the article determined 3 matching bookshelves was better than 2 matching bookshelves and a matching horisontal speaker.

I'm not sure if it will work but if so I want to try watching a movie with my center vertical to see if any difference can be noticed. It may only be those off to the side though that can tell any difference.

No, my mains are Paradigm MiniMonitor V.3's. Doubtless they are not a perfect match with my DIY, but the results are quite enjoyable just the same.

However no doubt the best advice to our OP is to use three identical speakers for L-C-R. Furthermore it is best that they be all on the same horizontal plane.

[Sir Terrence the Terrible]

One point has to be recognized within the context of this discussion. Not all horizontal center speakers are created equal. Some designers have gotten quite clever in creating 3 way horizontal center speakers that do not have the lobing and driver interference issues the typical two way two woofer and tweeter designs have. Some designs have a vertical arrayed tweeter and midrange drivers, while using a tapered array for the horizontal woofers.

A taper array would roll off one of the woofers at 120-130hz, just below the major interference pattern between the two drivers, thereby reducing its effect to just a notch, rather than a broadband interference. This way you have the power response of a two driver system where you need it(in the lower frequencies), and as the frequency goes up, the power demand of the single driver will keep it aligned to those of a larger tower speaker.
This is also possible with a two way(or two and a half way) system with a spaced woofer as well. The object is to filter out the frequencies that lie at the point where the drivers do not couple together like they do in the bass frequencies. With the typical 6 1/2" driver, that would be around 150hz. The best example of this design I have seen it in Aerial Acoustics center speakers(the CC3 if I am not mistaken), and one of the best of the old school designs the Academy Center from Klipsch, which allowed that speaker to be placed both vertically and horizontally.

I am very surprised not to see this design used more often

The effect of a mismatched speaker is not just limited to off axis. The differing frequency response of a mismatched front array will make any sound change, or stick out differently in a mix. The most typical thing I hear from mismatched center speaker is a noticeable collapse of the image as it moves from center to the sides. It is especially audible during equal power response pans across the front sound stage, which causes a narrowing of the image as it passes through the center. While the lively acoustics of the typical room can hide this effect to a certain extent, in an acoustically treated room this effect will stick out like a sore thumb.

I can understand the position the audiophile guys have in having to incorporate a center channel into their more optimized two channel rigs. When you have great left/right high quality speakers not purchased as part of a high end home theater system, it is VERY difficult to find a center channel for that sort of set up. Compromises have to be made not by their owning doing, so the thing is to make the compromises as audibly palatable as you can. It's a tough choice to make, but that is why I have chosen to invest only in mid to high quality surround systems to avoid having to make that compromise. I guess starting my studio mixing career in film sound also played a big role in this decision as well.

[eisforelectronic]

I love this thread! The design of "traditional" center speakers has always bothered me quite a bit. I always wondered how much of a compromise it was because so many of the center speakers I heard didn't seem to perform like they were supposed to. My plan has always actually been to use the bookshelf version of whatever floor standers I had as a center. The problem is getting someone to sell me a single speaker.

[Mr Peabody]

Usually the higher quality or more boutique type brands sell individual rather than pair but it still may be a special order as most shops probably order them in pairs. Unless you get them off Amazon or some place.

You could open another can of worms and talk about dual center speakers then one wouldn't go to waste if you bought a pair. Or, just do 6.1 with a rear center.

[eisforelectronic]

Well, I'm picking up a pair of used Audio Physic Tempo 6's and since a friend of mine is an Audio Physic dealer, I'm hoping to see if he can order a single "Step" for me. Otherwise, I'll either go 6.1 or convince someone else to use one as a center as well.

[RoadRunner6]

I have a pair of Athena S2s as my mains (for surround sys), they are very good speakers, especially when accompanied by a Sub. I use the matching C1 center. But the S2 is a better speaker than the C1 as far as sound quality and power handling. Wouldn't a 3rd S2 be better as a center channel? It would definitely blend better (not that the C1 doesn't blend well) and it handle base/power better.

Yes, a third S2 would be better, but only if you place it also in the vertical orientation just like the left and right S2's. Having all three front speakers (LCR) be the exact same speaker is the prefect blended match across the front for seamless front stage sound and perfect pans. You would want the center as close as possible in height to the left and right or the center tilted up or down to insure that all three tweeters are aiming at your ears.

Do not get a third S2 and place it on its side in the horizontal orientation. It is not designed to play on its side. Otherwise, you should keep the C1 since it has a matching tweeter to the S2. Be sure it is aiming at your ears. Also that it is not back from the leading edge of the surface that it is placed on or set back in an enclosed part of an entertainment cabinet. The center speaker must be set forward flush or slightly ahead of the front edge of the surface it is on. Re-check the speaker setup tone that cycles thru all 5 speakers and make sure that the center speaker is matched perfectly volume wise with the L&R speakers (use an analog Radio Shack sound level meter if you have access to one.....it is much more accurate than your ears). Some will turn up the center channel just a tad during movies for enhanced dialog. You need a powered sub with these speakers for both music and movies.

Do not rely solely on an auto setup system in your receiver. They make lots of errors. Do a "manual" check on all speakers. Be sure all of your 5 speakers are set to"small" in this setup.

[kexodusc]

Caution against using a 3rd L/R speaker as a center channel speaker. L/R speakers are designed to operate in free space, and have baffle-step compensation assumptions built into the crossover network.

Using that same speaker in a center channel capacity on top of, or near a large TV display or front wall will have a significant affect on midrange down to mid-bass. EQ can compensate.

A friend of mine uses 3 Revel Performa M22's - when he cycles through some of the sound modes on his Anthem processor to shift dialogue from phantom to center channel, there's an immediately noticeable differences in sound - listen to vocals especially, it can be quite obvious. Sucks for him - his wife won't let him move it two feet further out in the room, which would solve the problem. Still, it's not unbearable, but panning left to right across the front changes the apparent "timbre" of the speaker as much as if it was a different model altogether.

Center channels are (hopefully) designed to work better given that speaker placement constraint. If you have the space to bring the speaker out sufficiently however, I'd prefer a 3rd, identical speaker.

[Feanor]

Caution against using a 3rd L/R speaker as a center channel speaker. L/R speakers are designed to operate in free space, and have baffle-step compensation assumptions built into the crossover network.

....

I understand and when I built my DIY center I used less BSC than I would have for free-standing.

I have occasionally mentioned the option of using self-powered (active) monitors as HT speakers. Depending on the chosen actives, they can have these attributes which are arguably an advantage:

• "Accousitc space" control to accomodate free standing/wall/corner placement
• Controlled dispersion to limit first reflections
• Build-in amplifier(s) to ensure appropriate power to match the speaker's potential.

Thus a person on a medium budget (more than me) could put to together a nice system comprised of:

• An Emotiva UMC-1 prepro
• 5 or 7 M-Audio CX5's or CX8's
• A subwoofer

... see the CS5's controls HERE.

[Mingus]

I also have a mismatch center and fronts and don't seem to have a problem. In my basement HT system I have KEF C95 fronts and Polk CSi40 Center.

[02audionoob]

Anyone ever heard a setup like the Martin Logan Vignettes?

http://www.martinlogan.com/products/vignette

It's discontinued, so perhaps it didn't sell.

[Feanor]

Anyone ever heard a setup like the Martin Logan Vignettes?

http://www.martinlogan.com/products/vignette

It's discontinued, so perhaps it didn't sell.

What's wrong with that picture?

Well, though electrostatic, the Vignettes have same line-source type dispersion issues as dynamic speakers. (Also, they aren't arranged in same horizontal plane which is arguably almost as big a problem is timbre mismatch.)

[Luvin Da Blues]

What's wrong with that picture?

Well, though electrostatic, the Vignettes have same line-source type dispersion issues as dynamic speakers. (Also, they aren't arranged in same horizontal plane which is arguably almost as big a problem is timbre mismatch.)

...and not to mention the esthetics are butt frugin' ugly....(IMHO)

[Sir Terrence the Terrible]

Caution against using a 3rd L/R speaker as a center channel speaker. L/R speakers are designed to operate in free space, and have baffle-step compensation assumptions built into the crossover network.

Using that same speaker in a center channel capacity on top of, or near a large TV display or front wall will have a significant affect on midrange down to mid-bass. EQ can compensate.

Kex, this is a problem whether we are speaking of L/R speakers, or dedicated center speakers. I know of no speaker designer(well one anyway) that voices his speaker sitting on a television. They are usually voiced in free space as well, since the designer does not know where his speaker is going to sit. One of the major reasons why dialog intelligibility issues crop up comes from sitting even a horizontal center speaker directly on the television. Hence my point about the center being in a different "environment" than the L/R speakers.

A friend of mine uses 3 Revel Performa M22's - when he cycles through some of the sound modes on his Anthem processor to shift dialogue from phantom to center channel, there's an immediately noticeable differences in sound - listen to vocals especially, it can be quite obvious. Sucks for him - his wife won't let him move it two feet further out in the room, which would solve the problem. Still, it's not unbearable, but panning left to right across the front changes the apparent "timbre" of the speaker as much as if it was a different model altogether.

Wouldn't the answer to baffle step diffraction be to place the speaker closer to the wall to compensate for the loss in bass? Does near wall placement give the speaker a 3db(or near) boost to compensate for the loss of bass bending around the cabinet? My logic tells me that total compensation for the step loss is impossible to account for since the designer does not know how close(or far) the speakers will be to the walls. It seems to me that bass that is loss due to cabinet diffraction would not be totally lost in the room - it would just be added to and cancelled out from internal room reflections. While I understand the concept, I am baffled as to how it can be totally accounted for with so many variables in speaker set up in most rooms.

Also, it seems to me there should be a shift in sound when going from phantom to hard center speakers, even matched ones. Then reason being is that phantom imaging has frequency deviations that are naturally occurring as an effect of distance between our ears, the width apart the speakers sit, and the wavelength of the mono signal being reproduced. Phantom imaging usually has frequency suckouts between 1-4khz as a result of the same signal(mono) hitting both ears simultaneously. It is a combination of the distance between our ears(and the speakers width apart) correlating with the wavelength of the signal itself. With a hard center, that effect disappears because now there is a hard center there to interrupt the HRT effects of the same signal coming from two different locations. Based on this, there should be no timbre change, but a change in the spaciousness of what we hear. Because of the cancellation, the voice should sound like it has more space around it than with a hard center.

Center channels are (hopefully) designed to work better given that speaker placement constraint. If you have the space to bring the speaker out sufficiently however, I'd prefer a 3rd, identical speaker.

In an effort to keep my speakers in identical environments, I prefer the all to have identical placement from hard boundaries.

[audio amateur]

lol, I agree with you LDB

[kexodusc]

Kex, this is a problem whether we are speaking of L/R speakers, or dedicated center speakers. I know of no speaker designer(well one anyway) that voices his speaker sitting on a television. They are usually voiced in free space as well, since the designer does not know where his speaker is going to sit. One of the major reasons why dialog intelligibility issues crop up comes from sitting even a horizontal center speaker directly on the television. Hence my point about the center being in a different "environment" than the L/R speakers.

I think 10 years ago this was almost always the case - we use to see a lot of L/R/C speakers out there, not as much these days. Actually, there are very many companies that make the extended baffle assumption - Ascend Acoustics even came up with some buzzword for it in their marketing literature. I think you'd be surprised how many companies do this now. It's probalby not something they advertise in product brochures often because the average buyer wouldn't get it anyway.

You're right, not all though take it into consideration.

Wouldn't the answer to baffle step diffraction be to place the speaker closer to the wall to compensate for the loss in bass? Does near wall placement give the speaker a 3db(or near) boost to compensate for the loss of bass bending around the cabinet? My logic tells me that total compensation for the step loss is impossible to account for since the designer does not know how close(or far) the speakers will be to the walls. It seems to me that bass that is loss due to cabinet diffraction would not be totally lost in the room - it would just be added to and cancelled out from internal room reflections. While I understand the concept, I am baffled as to how it can be totally accounted for with so many variables in speaker set up in most rooms.

The loss isn't perfectly uniform or sloped, and isn't a hard +3 db jump or anything. It's a gradual curve, that depends on the drivers themselves. The compensation network applies a transfer function that, well, eliminates that unwanted curve (which is different for every speaker/baffle combo) and flattens response in free air (for L/Rs). Actually, you don't even need a dedicated filter, sometimes you can get better results just by adjusting the woofer's crossover point and slope and achieve the same effect (this cuts down on xo parts). Even moving it against the wall wouldn't fully address it as well as the crossover can, and probably introduces some negatives since, as you know, that's not necessarily the best place for speakers to be in a lot of rooms. I think on-wall speakers often just rely on the wall boundary though...

Also, it seems to me there should be a shift in sound when going from phantom to hard center speakers, even matched ones. Then reason being is that phantom imaging has frequency deviations that are naturally occurring as an effect of distance between our ears, the width apart the speakers sit, and the wavelength of the mono signal being reproduced. Phantom imaging usually has frequency suckouts between 1-4khz as a result of the same signal(mono) hitting both ears simultaneously. It is a combination of the distance between our ears(and the speakers width apart) correlating with the wavelength of the signal itself. With a hard center, that effect disappears because now there is a hard center there to interrupt the HRT effects of the same signal coming from two different locations. Based on this, there should be no timbre change, but a change in the spaciousness of what we hear. Because of the cancellation, the voice should sound like it has more space around it than with a hard center.

I don't know much about that, so I take your word for it, but the difference from the unecessary BSC in the sound I heard extended well into the midbass as well - I'm sure it's partially due to the effects you mentioned, but it also has to do with the proximity to the wall - the speakers sounded much, much similar when we moved the center out further into the room away from the wall. I guess, if phantom mode produces a suckout, and putting an L/R to close to the wall produces a boost, this is exactly what one might expect!

In an effort to keep my speakers in identical environments, I prefer the all to have identical placement from hard boundaries.

I think this is best, but unfortunately I think the vast majority of HT's can't accommodate this - commercial speaker designers probably start with a broad assumption about what the situation of the majority of their buyers is rather than how people should optimize their living rooms for HT.

[Feanor]

....
The loss isn't perfectly uniform or sloped, and isn't a hard +3 db jump or anything. It's a gradual curve, that depends on the drivers themselves. The compensation network applies a transfer function that, well, eliminates that unwanted curve (which is different for every speaker/baffle combo) and flattens response in free air (for L/Rs). Actually, you don't even need a dedicated filter, sometimes you can get better results just by adjusting the woofer's crossover point and slope and achieve the same effect (this cuts down on xo parts). Even moving it against the wall wouldn't fully address it as well as the crossover can, and probably introduces some negatives since, as you know, that's not necessarily the best place for speakers to be in a lot of rooms. I think on-wall speakers often just rely on the wall boundary though...
...

This is consistent with what I've heard.

Basically, amplitude begins to drop off at the frequency where the half-wave length becomes greater than the width of the baffle, (I guess it can be influenced by the shape of baffle too). Thus if you have a 10" baffle the effect will start to kick in around 700 Hz -- this assumes the speaker is some distance from the wall behind.

I think the customary BSC network is comprised of an inductor in parallel with resistor whose value is equal to the impedance of the speaker. But based on the simple modeling I've done using X-Over Pro, you can achieve the effect by a combination of crossover point and fiddling (increasing) the value of the larger inductor used in the low-pass filter ahead of the mid-woofer -- this saves you the cost of an additional inductor.

[Sir Terrence the Terrible]

I don't know much about that, so I take your word for it, but the difference from the unecessary BSC in the sound I heard extended well into the midbass as well - I'm sure it's partially due to the effects you mentioned, but it also has to do with the proximity to the wall - the speakers sounded much, much similar when we moved the center out further into the room away from the wall. I guess, if phantom mode produces a suckout, and putting an L/R to close to the wall produces a boost, this is exactly what one might expect!

Not quite on this one. The suck out is between 1-4kHz where the wall plays no role in the output. You are going to have a bass boost with the suck out intact if you place the speakers close to the wall.

The only way to avoid this kind of suck out is to use a hard center speaker to replace the phantom center image.

[kexodusc]

Not quite on this one. The suck out is between 1-4kHz where the wall plays no role in the output. You are going to have a bass boost with the suck out intact if you place the speakers close to the wall.

The only way to avoid this kind of suck out is to use a hard center speaker to replace the phantom center image.

Oh ok, like I said, I didn't know anything about a suckout in phantom mode - i don't use it in my system because I've always prefered CC's. Either way, the difference between the 3rd L/R at center and just phantom mode that I tuned into wasn't in the 1-4kHz range so much but more 200Hz -1.5 kHz range or thereabouts...if phantom mode had a suckout, it wasn't as bad as the sound from the CC up against the display. Then again, we were focusing on vocals, mostly dialogue, and were probably more attuned to the boost than the omission.

What is going on in the phantom mode that's causing that suckout that doesn't occur in regular stereo music listening? Can you explain it to dummies like me?

[Mr Peabody]

Kex, I didn't understand your post. You said your friend has three Revel Performas which are matching speakers, then toward the end you said you would rather have 3 matching speakers. Are you saying he laid one on it's side, or he has the right speakers just not a good set up?