Page 2 of 8 FirstFirst 123456 ... LastLast
Results 11 to 20 of 74

Thread: Comb Filtering

  1. #11
    Senior Member
    Join Date
    Sep 2010
    Posts
    430
    You can get 2 drivers on a flat plane the same distance from the ear. That is basic geometry. All you are doing is making a triangle with the two drivers and the ear.

    Three drivers can not do this in a straight line. The line would have to be curved to do that.

  2. #12
    Founding Member
    Join Date
    Aug 2010
    Posts
    496
    Which is precisely the experiment run by one of the older folks who was on the forums years back, Christopher Evert. Experimented by suspending a string where he wanted the drivers in phase, attaching a pencil to it, then drawing arcs on baffles he placed where he was going to mount drivers.

    He tried it in both MMTT style, and TT MM style where the T's and M's were at different lengths. Claimed both worked very well, as in required minimal or zero correction compared to mounting locations resulting in higher PLDs.

  3. #13
    Senior Member
    Join Date
    Sep 2010
    Posts
    112
    the answers to the simple quiz are :

    1. yes, it is possible for two point sources to be the same distance from a (third) listening point. It is, however, not possible for two point sources to escape a straight line drawn between them Also ... and this is VERY important ... you can absolutely have two drivers that are different distances from a (third) listening point. But in any case, the most important "dimension" for a two-driver array is NOT the distance between the two drivers, but rather the difference in distance from each driver to your nose (ear).

    2. no, it is not possible for three point sources in a line to be the same distance from a (fourth) listening point.

    Understanding this most simple geometry will give you the first clue as to why the spacing between the midrange drivers in a D'Appolito MTM arrangement is NOT a primary concern ... if, IF the distance from each midrange to your ear is the SAME.

    So .. let's do another quiz :

    I've got TWO small full-range drivers that are 6 inches apart from each other. Where's the first null in the frequency response due to "comb filtering"?

    (careful, it's a trick question ... but one that would confuse many of those that love to blindly regurgitate "ctc rules" for arrays)

  4. #14
    Senior Member
    Join Date
    Sep 2010
    Posts
    430
    Roughly 4500 hz if they are playing full range.

    I think.

  5. #15
    Senior Member
    Join Date
    Sep 2010
    Posts
    112
    Quote Originally Posted by winslow View Post
    Roughly 4500 hz if they are playing full range.

    I think.
    really ....

    if they are 6 inches APART FROM EACH OTHER ... what is the difference in distance to your ear?

    can you tell me ... ?

  6. #16
    Senior Member
    Join Date
    Sep 2010
    Posts
    430
    nill none nadda zilch

    Same with the comb filtering
    Last edited by winslow; 02-20-2011 at 02:03 AM.

  7. #17
    Senior Member
    Join Date
    Sep 2010
    Posts
    112
    Quote Originally Posted by winslow View Post
    nill none nadda zilch

    Same with the comb filtering
    nope ...

    if the drivers are 6 inches APART FROM EACH OTHER ... then the difference in distance form each driver to your ear might be 2 inches. Might be 0.073 inches. Might be 4.97 inches.

    Draw this on a piece of paper. Two point sources, 6 inches apart. Can you find a point that's 4 inches from one driver, and 3.9 inches from the other? YES. Can you find a point that's 4 inches from one driver, and 4 inches from the the other driver? YES.

    The distance BETWEEN our two point sources tells you NOTHING about the difference in distance to your ear. This is not acoustics. This is simple geometry. The ONLY thing you know, about two drivers 6 inches APART FROM EACH OTHER, is that the MAXIMUM difference in distance to your ears is 6 inches. But it might be 5 inches. Might be 0.0194 inches.

    And it's the difference in distance that determines what the frequency response "comb filter" will look like.

  8. #18
    Founding Member
    Join Date
    Aug 2010
    Posts
    496
    But I always thought of comb filtering as something like the double slit experiment shown in every physics high school classroom. The one that invariably shows a dispersion pattern as a series of dots, or if 2D, a series of concentric circles.

    So using that visual model, the very center of the pattern would be where the PLD is 0 units. And it doesnt matter where, what, or how the "slits" are configured (full range, band-limited, large diameter, small diameter, large driver spacing, right next to each other, etc), as long as those PLD's are 0, then the response will always be the same, yes? No interference pattern visible (in this case audible) at any frequency from the perspective of the center dot position, which will always be of amplitude 2x.

    But also using that model, we could have a TERRIBLE interference pattern, either at one frequency or many, but we'd just never hear it. right?

    Where am I getting confused?

  9. #19
    Senior Member
    Join Date
    Sep 2010
    Posts
    112
    "comb filering" will necessarily be a function of two things : frequency, and position.

    Let's say you can have two sources displaced by some distance D (this is how far apart from each other they are). And you also have a THIRD point somewhere in space, this is the listening point. Let's say the distance from the listener to the first point source is D1, and the distance from the listener to the second point source is D2. Finally, we'll define the most interesting difference in distance as :

    DD = D1 - D2

    Ultimately, it is DD ... and not D ... that determines the "comb filtering" frequency response.

    So yes ... you can have two "fixed" sources displaced from each other by D. What's the comb filter frequency response? You don't yet know ... because each point in space around these two fixed sources will have different DD's ... and therefore, different listening positions around these two fixed sources will have very different comb filtering frequency responses.

    PLEASE take all the time required to understand this. It's not abstract acoustic theory, it's simple geometry

  10. #20
    Senior Member
    Join Date
    Sep 2010
    Posts
    430
    So we are wanting the difference between the hypotensuse and the opposite side to find out where the comb will be?

    The adjacent side will be 3 if assume symmetry...which would make it easier for your example I guess.
    Last edited by winslow; 02-20-2011 at 02:23 PM. Reason: additional question

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •