Electrostatic speaker and sub help

Discussion in 'Subwoofers' started by Realrift, Dec 24, 2006.

  1. Realrift

    Realrift
    Standard Member

    Joined:
    Jul 17, 2006
    Messages:
    3
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    1
    Ratings:
    +0
    Hello all,

    I was wondering if I could get some tips on choosing and setting up a subwoofer to integrate with my Quad electrostatics. I have included an in room response curve I plotted after following a great guide I found here. I want a sub mainly for music and secondarily for HT (HT not a big concern to me).
    These speakers seem to roll off quite sharply below 45Hz. I have been looking at a Rel Strata or possibly a Velodyne DD-10 (maybe DD-12). I have just moved on from a SVS PB12+ which was excellent for HT but was not satisfying me musically.

    Thanks for your input. Happy festive season,
    Rift.
     

    Attached Files:

  2. recruit

    recruit
    Guest

    Products Owned:
    0
    Products Wanted:
    0
    Ratings:
    +0
    Moved to sub forum should get a better response here :thumbsup:
     
  3. Nimby

    Nimby
    Well-known Member

    Joined:
    Jan 27, 2004
    Messages:
    9,206
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    136
    Location:
    The Danish Bacon Factory
    Ratings:
    +671
    Hi Rift and welcome to the sub forum.

    Did you try to equalise your SVS to your room? Or were you limited to high level connections?
    One bad room mode can completely spoil the sound (and response curve) of a decent sub.
    Which crossover frequency setting were you using on the sub?
    How large is your room?
     
  4. kstrain

    kstrain
    Guest

    Products Owned:
    0
    Products Wanted:
    0
    Ratings:
    +0
    The following is based on experiments, over about a year, with Final electrostatic speakers and a Monolith. Perhaps some of this will help to sort out what is important for you to achieve what you want. Unfotunately I don't have a means to view .xls files at the moment, so I cannot see your response graph.

    Due to a move of house this was all done twice in quite different rooms, that helps a bit to separate what the equipment does and what the room does.

    Some things I found (some unsurprising, of course) :-

    the panels have to be located correctly for the upper-bass, lower-midrange, where their room interaction becomes very strong. I think that will usually be the case for dipole radiators, due to the strong back wave reflecting around the room (most surfaces reflecting that frequency range quite well). Getting this wrong easily made a complete mess of the frequency range between roughly 100 and 400 Hz. So if you have this sorted you need to find a sub that can work with the panels where they are. (Which you probably want anyway!)

    it was not possible to arrange a good crossover with the sub where the room modes were making a mess of the panels' response (near the deep notches) presumably this was because the small output from the sub was not masked (or was it just the rapid change in phase between the sources? anyway I could not make it work!) You are aiming for a lower crossover, but room modes will be a problem.

    sub position was quite important - in my new room (rectangular) the only places that I can get it to "work" is against the wall behind the panels (this effect was less pronounced in the old room which had a more complex shape, and was smaller). The solutions I always come up with have one very strong mode left with the listener near a null - and pretty terrible sound away from the sweet spot. The problem is that the sub has to be positioned to suit the room for deep bass, yet suit the (fixed) panel positions in the crossover region (see also below). Putting the sub in a corner left just one mode in this state, but the effect was very strong.

    I believe that a big sub is essentially always better, and that the box design and raw frequency response are pretty unimportant (compared to distortion at a given level), as the room will completely dominate the response anyway. I think it will be pretty much necessary to use a BFD or some other programmable filter (parametric EQ or an inverse response convolver) to correct the sub's response in this case, where a dipole and monopole are to be blended. Fortunately the basic response of the sub (2nd order for sealed box, 4th order for vented) is essentially "invertible" so it is possible to make compensating filters that work very well. Unfortunately the room response cannot be compensated in general, except in a very approximate way. Using the full bag of tricks, however, it was not too hard to blend the Monolith with the panels such that e.g. bass-percussion sounded convincing (and giving a good measured transient response).

    BTW: The Monolith was chosen as it gives good volume (l x w x d) per £, yielding reasonable efficiency with low distortion. Like any similar sub design it would likely be a terrible match to dipoles without EQ (or very good luck).

    I found it important to correct the time-delay to match the arrival time at the listening position. This makes a huge difference to the integration when the sub is more than foot or two farther from the listener than are the panels.

    For me, the whole point of doing this was to get much less distortion from the panels, so I had to push the crossover up quite high (as much as 120 Hz). The problem being that the sub could then start to be located (just). This meant using steeper than usual crossover filters (up to 7th order). A side advantage of this was that the crossover could be slotted into a benign frequency range (both rooms had a suitable region near 120 Hz).
    Crossing over lower down as I think you are aiming to do avoids the problem of being able to locate the sub, but may be just as challenging with respect to room modes.

    In summary, I think integrating a monopole sub with dipole panels is most likely to work with a fairly detailled approach to sub position and EQ. I doubt very much that there is a satisfactory "drop-in" solution, but someone else may know better ;)

    Background: my music is played (using Foobar 2000) on a PC with a nice external sound card and amplification. EQ is done via parametric filters (including a BFD for the sub, as it is very fast to use for the big room modes) and/or a "DRC" generated inverse impulse response. Measurements use an ECM8000 microphone, and listening is carried out with a wide range of music (not all of which I would listen to for pleasure).

    So I'm not sure that a change of sub is really the way to win here, but good luck in any case!

    Ken
     
  5. Realrift

    Realrift
    Standard Member

    Joined:
    Jul 17, 2006
    Messages:
    3
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    1
    Ratings:
    +0
    Hi Nimby,

    The PB12+ I had did not come with the PEQ. I was using the low level input and had the crossover at about 55-60hz. Unfortunately I had not plotted a response curve while I still had the PB12+.

    The room is about 7x7.5m with 8foot ceilings. There is a partitioning wall running down the middle of the room but going only half-way across. The partition starts at the back of the room and my audio setup is at the front of the room to the left.

    X-mas Cheers,
    Rift.
     
  6. Nimby

    Nimby
    Well-known Member

    Joined:
    Jan 27, 2004
    Messages:
    9,206
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    136
    Location:
    The Danish Bacon Factory
    Ratings:
    +671
    Rift,

    Your room shows a fairly smooth spread of resonant modes due to parallel boundaries. Only the 8 foot ceiling height might produce a noticeable ~80Hz mode. The modes due to length and width are too low to matter much. The partition wall is probably unimportant as it helps to break up the area of the opposed walls. The same can be said for large furniture in the vertical sense. Tables, sofas & settees and bulky armchairs all help the cause.

    The PB12+ is highly thought of at least in AV circles. It is probably too late to try something like a Behringer BFD or Velodyne SMS-1 on it since it sounds as if your have already parted company.

    The response curves of the PB12+ and Ken's Monolith are just too much alike to be worth trying to swap one for the other if you didn't get on with the PB12+.

    For music it might be worth looking at a sub with a broadly humped response peaking around 40Hz to match your Quads without the need for room equalisation. The more popular AV subs (which you hear so much about on here) tend to have flat responses which can be more difficult to match to difficult speakers.

    The obvious choice would be an IB subwoofer but not everybody wants or can manage a fixed installation. The idea of an IB is to insert some large speakers in a wall, floor or ceiling of the listening room. The rear output of the drivers is kept isolated from the front of the speakers.

    The raw response of an IB is very low in distortion and very gently humped. Usually somewhere around 40Hz. It takes up no space in the room itself.

    Have a look a the IB Cult's FAQs and galleries for the many possibilities. It may appeal. It may not. I know that a number of electrostatic speaker owners enjoy their IBs. ThomasW, The IB Cult forum owner certainly does. (Magnepans I think)

    http://ibsubwoofers.proboards51.com/index.cgi?
     
  7. kstrain

    kstrain
    Guest

    Products Owned:
    0
    Products Wanted:
    0
    Ratings:
    +0
    Thanks, that is more or less what I was trying to say, in too many words. I suspect anything in a similar sized box is going to match just as poorly.

    Does the peak come from a combination of the drivers' resonance and room leakage? I had not thought about tuning the IB response that way, hmm.... I'm still a long way from cutting holes in walls to try an IB, unfortunately.

    I wondered too about a dipole sub for dipole speakers, but don't see any particular merit (apart from it sounding as if the two should go together), as below the room modes it will just be a losing battle. Especially true when the main panels already get down fairly low.

    Ken
     
  8. Nimby

    Nimby
    Well-known Member

    Joined:
    Jan 27, 2004
    Messages:
    9,206
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    136
    Location:
    The Danish Bacon Factory
    Ratings:
    +671
    Dipole bass is a disaster waiting to happen.

    Extend the finite sized baffle to include the boundaries of a room and you have an Infinite Baffle. Without having a huge baffle in the room. :)

    Saw a hole in the floor instead? ;)
     

Share This Page

Loading...