Center Channel Speaker Design Additional Considerations

Discussion in 'Loudspeakers' started by admin, Apr 12, 2012.

  1. admin Audioholics Robot Staff Member

    admin
    Joined:
    Apr 16, 2004
    Messages:
    172
    Likes Received:
    1,612
    This article takes a broad look at center channel design and discusses the prior two articles we've already written on the topic and compromises associated with horizontally arranging drivers. The subject of proper center channel speaker design is not a simple one. There are many considerations: price, desired coverage area, aesthetics, and others. A manufacturer has a daunting task trying to balance many seemingly conflicting requirements, while for the consumer, education and information are the keys. Being aware of the strengths and weaknesses of the various design types gives the consumer the best opportunity to make the most satisfying choice.
    [​IMG]

    Discuss "Center Channel Speaker Design Additional Considerations " here. Read the article.
    Last edited by a moderator: Apr 12, 2012
    • Like Like x 6
  2. fuzz092888 Audioholic Warlord

    fuzz092888
    Joined:
    Feb 13, 2012
    Messages:
    7,840
    Likes Received:
    3,834
    Location:
    42.3464° N, 71.0975° W
    If only adding a third tower wasn't so darn inconvenient :)
  3. TLS Guy Audioholic Overlord

    TLS Guy
    Joined:
    Oct 30, 2007
    Messages:
    12,100
    Likes Received:
    4,875
    Location:
    Benedict MN
    You are preaching to the choir.

    In my research activities in the design of my system, the center speaker was by far the most difficult quandary.

    I think the best solution currently is a coaxial driver for the center. This is not perfect because of the increased dopler distortion of coaxials.

    What is really required to bring off a good center, is a wide band midrange driver that does not have to be crossed over in the speech discrimination band. Better still would be a good high power full range driver.

    As I have pointed out in two threads now, we need to be digital from microphone capsule to speaker, with no passive crossover.

    I was in Magnolia on Monday surveying what was on offer. It was awful and not one system really sounded decent and none could integrate all the speakers properly. All the centers were highly deficient and the worst of the bunch, although the subs ran them a close second.

    We need a radical change of approach and I think we are on the threshold of radical change. Big money is now working on it and engaged. The change is so radical that even the AES/EBU SPDIF standard is about to give way to AES 42.

    Woe to anyone who has wire in there walls with no conduit. It will be good for the sheet rock business.

    All interconnects will be optical, to make noise floor effectively zero, The standards are in place and way above Toslink.
  4. gene Audioholics Master Chief Administrator

    gene
    Joined:
    Jul 1, 2002
    Messages:
    7,526
    Likes Received:
    3,257
    Location:
    Florida
    I am sure many will disagree with me but I still find a well designed MTM can perform admirably horizontally oriented as a center channel provided you don't sit that far off axis. The advantage of an MTM over a W(T/M)W is you have more efficiency in the critical midrange area (500hz-2kHz) by having 2 larger woofers handling those frequencies as opposed to a single smaller midrange in a 3-way design.

    First reflections aren't as critical for a center channel as they are for a Left/Right speaker since the center speaker is usually so far away from the side walls anyways.

    It really boils down to trying various designs in your own listening environment and seeing how well each one works. I always maintain if you have a seat off to a sidewall, you're not getting good sound anyways and that seat should always be reserved for the mother-in-law :D
    gene,
    • Like Like x 3
  5. FirstReflection AV Rant Co-Host

    FirstReflection
    Joined:
    Aug 3, 2008
    Messages:
    1,260
    Likes Received:
    887
    Well, for a dedicated home theater, the solution has always been an acoustically transparent projection screen and three, identical, vertical speakers all with the tweeters at ear height.

    But yeah, for any flat screen setup or normal living room, a vertical center just isn't practical.

    While not ideal, there are some ways to improve upon the sound quality of a standard M-T-M horizontal speaker though. An extremely robust tweeter that can play lower than the average tweeter helps. And very steep cross-over filters between the midrange and tweeter drivers helps.

    Of course, both of those things cost more money, which usually goes against what a manufacturer with an M-T-M center is aiming for :p
    • Like Like x 4
  6. GranteedEV Audioholic Ninja

    GranteedEV
    Joined:
    Jul 14, 2009
    Messages:
    4,933
    Likes Received:
    2,352
    Location:
    Calgary, Canada
    Gene, you would lose some of that theoretical efficiency boost in a (passive) speaker because the crossover will compensate for baffle step losses (even if it's reduced due to center speaker boundary gain) via resistive attenuation.

    In a 3-way you can avoid said resistive attenuation in that midrange area by crossing the woofer right where the box starts to lose efficiency. You can also optimize the midrange driver parameters themselves to cover a shorter bandwidth with higher efficiency.

    As far as displacement goes, to hit 100db @ 100hz you need 76 cm^3 whereas to hit 100db @ 300hz you need less than an 8th of the displacement.

    Let's say you've got a 4" driver (75cm^2 of radiating surface area) - it need about 1.3mm of xmax to hit that target at its lower limit frequency around 300hz.

    Now let's say you've got two 6.5" drivers (133cm^2 of radiating surface area each). To hit the 100db target at 100hz (where such SPL demands are probably more common) those two 6.5" drivers need 2.8mm xmax.

    It seems to me that as far as displacement goes, the 3-way's mid has the advantage over the dual midbasses, because it's asked to move less overall at the same SPL at the bottom of its passband.

    So the 4" dedicated mid does not necessarily have the displacement disadvantage. With the right engineering it really should not have the efficiency disadvantage either, as it can get away with around 3-4X the FS and less box size limitations for a sealed midrange. You can also get away with using metal cones like aluminum without the same disadvantages, because a 4" driver's cone breakup would be a good bit higher than that of a 6.5" driver.

    Additionally, the 4" dedicated mid probably allows a higher crossover frequency to the tweeter for reasons mentioned in the article, which improves the tweeter's power handling.

    The dual 6.5" woofers do have an advantage as far as heatsinking area is concerned, but are also asked to cover almost 2 more octaves, where a lot of power is centered.

    Finally, it means you can use 8" sealed woofers in the lower mids and upper bass, which will blend in better to subs than vented 6.5" mids.

    *shrug* :D

    Also, let's say your center is about 9 feet from boundaries. That sound still arrives at your ears within 17 ms of the direct sound. It won't affect timbre/imaging but I would not suggest it does not affect our perception.
    Last edited: Apr 12, 2012
  7. agarwalro Audioholic Ninja

    agarwalro
    Joined:
    Jan 31, 2005
    Messages:
    3,839
    Likes Received:
    985
    Location:
    Pittsburgh, PA
    In this figure,

    [​IMG]

    imho, in the interest of simplifying the visual, you have completely incorrect depicted the physics, while using the term 'interference'. (Which you are obviously aware of.)

    Dr. Don Keele (AudioArtistry.com) has .mov files which show simulated lobbing as it changes with freq. Those will be considerable better in exemplifying the physics of interference.



    Here specific points of the simulation have been captured and converted into a self contained document.
    Last edited by a moderator: Oct 19, 2014
    • Like Like x 3
  8. LXIX Audiophyte

    LXIX
    Joined:
    Feb 6, 2006
    Messages:
    7
    Likes Received:
    2
    Coincident designs as center speakers?

    I am curious why this paper made no mention of coaxial/co-planar designs like the KEF Uni-Q array, Tannoy Dual-Concentric or the Thiel coaxial/coplanar design?

    In my opinion, these systems strike an elegant design compromise that addresses many of the concerns discussed in this paper.

    I have 2 different designs in my home, I use a Polk Audio 5.1 setup in my family room with a traditional MTM driver array and I hate it because the dispersion is so narrow, I almost need to lock my head in a vise to get proper high frequency reproduction.

    My basement theater uses KEF Uni-Q arrays and I have much better lateral dispersion. In addition, I have 2 rows of tiered seating and I notice little appreciable difference in treble response when I sit in the second row of seating. These seats are 9 inches higher than the front row.

    My only complaint with this setup is the attenuation from the front row to the back row. I suppose the only solution would be a line array but I am not ready for this just yet.

    Just my $.02
    LXIX,
  9. GranteedEV Audioholic Ninja

    GranteedEV
    Joined:
    Jul 14, 2009
    Messages:
    4,933
    Likes Received:
    2,352
    Location:
    Calgary, Canada
    There is one more solution.

    Get larger format coaxials IE 12" or 15" which hold pattern down to around 1khz. Aim it up ~35-50 deg upwards at the rear seats so that they are more on-axis than the front seats.
  10. gene Audioholics Master Chief Administrator

    gene
    Joined:
    Jul 1, 2002
    Messages:
    7,526
    Likes Received:
    3,257
    Location:
    Florida
    Excellent points as always. I think we all could agree, if we could match the 3 front speakers with vertical arrangement of drivers, that would be the best solution. But this involves using a Perforated screen and not mounting the screen against the wall, unless you use an in-wall center channel.
    gene,
  11. TLS Guy Audioholic Overlord

    TLS Guy
    Joined:
    Oct 30, 2007
    Messages:
    12,100
    Likes Received:
    4,875
    Location:
    Benedict MN
    The trouble with those type of units is that even the best of them have serious cone break up issues of the bass cones before crossover.

    I have been very pleased with my SEAS coaxial units. I use two cones in the bass. The voice clarity is excellent and sounds as if it coming right out of the screen. Dialog covers all the seating of this studio very well. What is even better they blend with the metal cones of the mains well. Sounds move seamlessly across the stage.

    Much to my surprise they keep up well with the mains in terms of output, since I Switched out the Quad 405-2 for a 909.

    Part of the reason I suspect is the high pressure and therefore cone control by the TL.
  12. Malcolm02 Audiophyte

    Malcolm02
    Joined:
    Apr 13, 2012
    Messages:
    4
    Likes Received:
    0
    A question that's bugging me

    This article and discussion are very interesting but there is one question that is really bugging me:

    If I'm listening to a stereo pair of speakers, a center image is produced by sending the same signal to each speaker and sitting in the middle. Now if I move a little bit to the side then one speaker is closer to me than the other. Is this not exactly the same phenomenon as having two drivers horizontally arranged in a center channel? With the stereo speakers the sound is coming from two tweeters horizontally arranged and two mids horizontally arranged, even though in each speaker they may be vertical. Why don't I get a rollercoaster frequency response when I move from side to side? Or do I, and I just don't notice it? If I don't notice it then why is it benign with the stereo pair and so terrible with a horizontal center speaker?

    This question is of more than theoretical interest to me as I'm considering adding a center channel to my main speakers and the only practical choice is an MTM design.
  13. TLS Guy Audioholic Overlord

    TLS Guy
    Joined:
    Oct 30, 2007
    Messages:
    12,100
    Likes Received:
    4,875
    Location:
    Benedict MN
    It has to do with the lobing pattern of the crossovers and speaker layout.

    A vertical MTM if done correctly has a nice horizontal dispersion patter, and limited vertical dispersion. A horizontal MTM has a very good vertical dispersion and poor horizontal dispersion, which is the opposite of what is required.
  14. scott911 Full Audioholic

    scott911
    Joined:
    Nov 28, 2007
    Messages:
    280
    Likes Received:
    30
    nice discussion...

    I actually use a phantom center to avoid the potential problems with a center's placement. :)

    I just sink the L-c-r money into the best L / R I can get. Although I often think of adding a center, I get the best imaging with my left/right that I've heard in a non dedicated theater.

    I like another poster's suggestion that the Thiel / KEf "point source" approc (can't remember the proper or marketing terms) can go a long way towards solving these issues.
  15. bwaslo Audiophyte

    bwaslo
    Joined:
    Dec 31, 2006
    Messages:
    9
    Likes Received:
    3
    Last edited: Apr 16, 2012
    • Like Like x 1
  16. DD66000 Senior Audioholic

    DD66000
    Joined:
    Jan 29, 2008
    Messages:
    607
    Likes Received:
    91
    Location:
    SA
    I can say using a coax center provides near perfect dialogue. Years ago I used three large 2-way LCRs. With the center having the same cab and 14" woofer, but with the tweeter coax. (LE14A for the L/R & LE14C for the center). Being the tweeter was mounted in the center of the cone, the 14" driver actually became a waveguide for the tweeter and provided a wide dispersion.

    The article was wrong on one point; AR was not the first to vertically align the drivers. JBL came out with the L212 in '77 that was a vertically aligned 3-way and that system also had the B212 sub. The pic in my posts are of a custom design of that same speaker with the arrays offset for mirror imaged pair.
    Last edited: Apr 16, 2012
    • Like Like x 1

Share This Page

  • rbhsound.com
  • BlueJeansCable.com
  • Emotiva.com
  • SVS Sound Subwoofers
  • Experience the Martin Logan Montis
  • CEDIA