Sound Quality Subwoofers (SQ)

  • Participant
    stryke on #636

    Hi everyone, I’ve been meaning to stop by the thread for a few days but have been quite busy with things. When people talk about SQ what should really be talked about is accurate reproduction. The goal for any speaker system is to accurately reproduce what is on a recording without altering it in any way. In terms of that, there are really two things to consider. The first is flat response and the other is low distortion.

    Lets take these in reverse though from the woofer design side first. I hear all the time that it’s all about putting the woofer in the right box. If it’s put in the right box any woofer can sound good. While it is important to put the woofer in the right box, there are things that won’t change no matter what box you have. You will always have non-linearities in the woofer that make it sound different at higher levels. As you put in power you get thermal distortion issues and the large distortion issues due to flux modulation. As the driver moves to high excursions you get variances in inductance and you get decreasing motor strength. If the woofer has not addressed these issues in the design stages, no enclosure is going to correct them. Attention should be paid to pulling heat from the coil quickly, lowering and linearizing inductance, providing a symmetrical BL curve, and lowering flux modulation. While the woofer needs to be in the right enclosure, a woofer that addresses all these issues and is put in the right enclosure will always be able to more accurately reproduce the signal input to it than a poorly designed woofer in the “right” enclosure.

    That said, we can see that distortion comes from a few places. One, from power being applied to the driver. This creates the thermal issues and the current going through the coil modulates the permanent magnetic field. Therefore the less power we need to put in to get any given SPL level the lower the distortion and the more accurate it will be. We can see that “efficiency” is key to this, although not necessarily the 1W/1M spl at lower frequencies. We also want to keep the driver operating at lower excursions to be in a more linear range to reduce the inductance variances and BL variances. Quite simply, the more output you can get with less power applied and while moving the driver less, the more accurate it will be. Up higher the driver efficiency is the biggest part of this efficiency. Mass, motor strength, and cone area are the determining factors. Below efficiency is determined by the enclosure per Hoffman’s Iron Law.

    This is why, contrary to popular belief, a well designed vented/PR enclosure will almost always be lower distortion than a given sealed enclosure at the same SPL levels. The vented box will be up to 9-10dB more efficient at the tuning frequency. This means as little as 1/10 the power needs to be applied for the same SPL level. Thermal issues are cut to a fraction and so is flux modulation. This also means the driver doesn’t need to move nearly as far for any given SPL around the tuning frequency. In fact it barely moves at all at tuning. As a result you will most always get the lowest distortion output at a given frequency from the vented box. (until you get below tuning)

    Of course you may now be thinking “ported boxes don’t sound good though, they’re boomy.” That leads me to the second part of accurate reproduction and that is flat response. When wanting to get flat response, it is important to realize that it is the whole system that has to be looked at and not just the drivers alone. The woofer is part of the system but without taking into account the vehicle transfer function and the enclosure, the woofers response doesn’t mean much. This happens in a home theater room, recording studio, night club, etc. As a result there are two things you can do if you want flat response.

    Again keeping in mind our goal is for the lowest distortion, we look at using EQ to correct the response of the mentioned ported woofer. EQ is your friend. We really don’t care what the response was before because we’re going to flatten it anyway. We were only concerned with getting a given SPL with the lowest distortion. Putting that vented box in the vehicle may give far too much low end output. That is ok though. In fact it is great! If you have 20dB extra headroom at 30hz for example, that means you can get away with literally 100x less power at that point that at say 75Hz. Using EQ to flatten the response, the low end is brought down to the appropriate level. Any bumps in the response curve are corrected for. What is left is now a system where the woofer doesn’t need much power, doesn’t need to move much, and as a result has extremely low distortion output with a ton of headroom for transient peaks. In most cases where EQ is available I will design a system this way. You do also want to remember to tune low enough to get the rise in group delay at tuning down to an inaudible frequency. I will often tune with PR’s in the area of 20hz. The low B on the bass guitar is around 32hz. Low A on the piano is 29hz. Most electronic music doesn’t reach much lower. As a result nearly everything is covered cleanly well above the tuning frequency.

    Now the other method, which is most often used is to try to get the enclosure rolloff to approximate the inverse of the cabin gain. Often times this is done with a sealed box which will begin to roll off much before the vented box would. This will help to tame down the low end but can’t help to deal with the bumps and still won’t really compensate for the cabin gain. A sealed box will roll off at 12dB per octave. At some points it won’t be enough. At others it will be too much. The following was the transfer function in BumpinBuick’s 4 runner. From 67-53hz the gain increases at about 30dB per octave. Below there it drops to about 12dB.

    The sealed will give the closest approximation to flat for a box just put into the vehicle. This approach is easier than EQ’ing the system and less costly. However, the properly EQ’d vented/PR system will overall have the ability to play the cleanest and have the most headroom for dynamic transients down low.

    Also keep in mind again that the lower the input power for a given output level, the lower the distortion. A woofer that needs 4000W to get the same levels as another woofer that only needs 1000W is not a good thing. While it can “handle” a lot of power, the distortion goes through the roof! Flux modulation is huge, thermal issues are huge, and the amp is working much harder than it needs to. Quite simply though, most woofers are not designed for low distortion and clean output. They’re designed to take abuse. I could easily stack 4 spiders on an AV woofer to stiffen up the suspension so it would take 4x the power before hitting it’s physical limits. However this then means that you also need 4x the power to get the same SPL levels that can be achieved with the single spider. IMO this is not a benefit. Does the softer suspension make them easier to damage? Yes. Are they easy to damage if used properly? No. A highpass(subsonic) filter at or right below the tuning frequency of the driver keeps excursion under control just fine.

    Regarding SPL levels of the AV woofers, we did some playing around at BumpinBuick’s place last year. We played at high levels for about an hour with no damage to the woofer at all.

    Here is a video of a 30sec bassrace which did 140.7dB

    Playing test tones I was able to get 146.1dB from the single AV15

    We were also able to get 129dB at around 80hz where there is very little cabin gain at all. This was among other testing with music, test tones from 37hz up to 80hz+. The woofer then shipped out to a customer the next day. The AV woofers are designed for accurate reproduction and are not designed as an SPL woofer. However, they can play quite loud when called upon to do so.


    BruceH on #6890

    Can I assume that the Infinite Baffle setup will be similar to a ported enclosure for the above description?

    Is there a recommended EQ for balancing the subwoofer? Is it necessary to have a 32 band parametric EQ or can one get away with a small EQ? Perhaps that is a loaded question as it may depend on how bad the interior acoustic environment alters the sound output.
    Would any manufacturer’s EQ work or must it be a high-end unit? Or rather perhaps I should ask for recommendations on a typical EQ that would do the job.

    While it is winter here and still with temperatures of -20F, there is not too much action on the car audio front in terms of installs. I am waiting until spring for the IB15.

    simon5 on #6891

    The better the EQ, the better the results will be, usually. On the other hand, a very good EQ can be affordable, no need to spend big dollars, just need to choose well.

    Also remember than when you are EQing, you are often EQing at the exact spot where your input microphone is. So it’s a world of compromises…

    boy.kroy on #6892

    What about Infinite Baffle? Isnt it the best choice if you have a trunk(and not a hatch)? *You mentioned a 4Runner

Viewing 4 posts - 1 through 4 (of 4 total)

You must be logged in to reply to this topic.