1. i am assuming that the strap featured in barefoot's design is nixed as it would defeat the purpose of the sorbo pads, yes? am i missing out on some functionality that the strap added to the design?
There isn't much tension on the strap, and it is flexible, so it should not transfer much energy from the speaker to the frame: It will transfer some, yes, but since the frame is already very rigid and massive, and it will mostly be higher frequencies that get transferred, so I wouldn't worry about it too much. The other point here is that if the speaker is inside a box and there is Sorbothane between them to decouple, then adding the straps to the outside of the box to keep it in place does not affect the decoupling at all...
2. how far out from the speaker should the bezel extend? is there some ratio of baffle size as compared to speaker size we should be observing? in john's design the timber finish goes almost all the way to the top of the wall. however his example features a speaker placed much higher up the wall than ours. we will be placing 16" speakers at a height of 40" on a 109" wall.
Ahhhh! Now that's the 60 million dollar question, isn't it! ?
Technically, the bezel or soffit panel around the speaker is known as an "infinite baffle", and that gives you an idea of how big it should be in theory: infinite!
However, that really means "infinite with regard to the wavelengths in question".
Obviously, a truly infinite size baffle is impossible, and even one that is very large with respect to all audible wavelengths is also impossible, considering that the wavelength of 20 Hz is about 17 meters! It would be rather difficult to have a baffle 34 meters wide and 34 meters high with your speaker in the middle...
Many years ago, some speaker manufacturers would actually bury their speakers in the ground, pointing upwards at the sky with the front surface level with the ground, and take their test measurements like that. That's about as "infinite" as you can get, in the real world.
OK, so let's look at the theory here in more detail, to find out exactly what it is we are trying to accomplish, to then come up with a sensible size. What is the actual purpose of the soffit?
It works like this:
Sounds don't all behave the same. Higher frequencies tend to propagate more like beams of light, or rays, projecting straight out from the speaker in a sort of tight cone and gong forwards, without spreading out much to the sides. Low frequencies, on the other hand, act more like a balloon that is being inflated around the speaker, spreading out in all directions. Part of the reason that so much design work goes into tweeter horns, is to try to get the high frequencies to spread out wider and more evenly, and so much work goes into woofers to try to get the sound to NOT spread out so wide, so that the highs and lows match more reasonably (there are other reasons too, such as the major impedance mismatch between speaker cone and air, etc.)
In a perfect speaker, all of the sound, at all frequencies, would seem to come out the front of the speaker as a flat planar wave that just moves across the room. The only real-world speakers that come close to accomplishing this, are large electrostatic panels, and that's exactly how they work: the entire front panel (several feet high and wide) is a membrane that moves back and for all at once. And they sound fantastic! But they are expensive, complicated, and not able to produce very high levels of sound, so you don't see them too often these days. Unfortunately.
Another problem with such speakers is that they won't fit in your car to take them home from the store, and they won't even fit on the shelf in the store! They are huge.
So, since manufacturers cannot actually make a perfect speaker, they are stuck with just making "really good" speakers that try to sound much bigger than they really are. And that gets back to the problem of how sound propagates differently for different frequencies.
OK, so if highs go straight forwards but lows don't then obviously there is a power mismatch between the two. If you have, for example, 1 watt of high frequencies and 1 watt of low frequencies, well ALL of that 1 watt is going straight at your head for the highs, but only HALF of it is going towards you for the lows: the other half is spreading out behind the speaker, going AWAY from you. In technical terms, there is a difference of 6 dB in the sound power radiated forwards for highs, and all around for lows. The lows are said to be radiating into "full space" (a sphere all around the speaker), and the highs are radiating into "half space" (half a sphere, in front of the speaker).
The main reason this happens is because of the "baffle", the technical name for the front panel of the speaker: Just by being there, it forces all of the highs to go straight towards you, while the lows go all around. The
size
of the baffle determines where the changeover happens. The basic issue is that sound waves are only affected by objects that are substantially larger than their own wavelength, so waves that are smaller than the shortest dimension of the baffle tend to go forwards (since the baffle is bigger than the are, and prevents them from going backwards), but waves that are larger than the dimensions of the baffle don't even see it: they just "wrap around" behind it, as though it was not even there.
This whole effect is called the "baffle step response", because if you look at a graph of power vs. frequency, you'll see a "step" in the middle of the curve, at the point where the size of the waves is comparable to the size of the baffle. If you really ant to get technical about this, the mid point of that "step" occurs at the frequency f = 4560 / Wb, where "Wb" is the width of the baffle in inches. You probably don't need to know that, though. And if you are
To "fix" this power imbalance between lows and highs, manufacturers include a "baffle step correction" circuit inside the speaker, usually as part of the cross-over network, so that the speaker increases the power for low frequencies. In fact, it has to produce twice as much power for the frequencies that are lower than the baffle step, as compared to frequencies above the baffle step.
All of this problem is totally eliminated by putting the speaker inside a soffit, flush with the front surface. In effect, the soffit face is just a much larger baffle that moves the baffle step point to a much lower frequency, since it acts as an infinite baffle for waves that are smaller than the shortest dimension of the soffit. That's why professional speakers have a "bass roll-off" control on the back that reduces the power for low frequencies by 6 dB, so ti is the same as the power for high frequencies. In reality, what that control does is to turn off the "baffle step compensation" that is built into the speaker, since you don't need it any more if the speaker is very close to a wall, or soffit mounted.
OK, so getting back to your basic question: how big does the baffle have to be? Obviously, if you made a baffle just one inch bigger than the speaker, then you haven't accomplished much! And if you make it 100 feet bigger, then you have accomplished everything, and more! Basically, start by looking at the lowest "cut-off" frequency for your speaker. There's no point in making the baffle big enough to deal with frequencies that it doesn't even produce. So for example, if your speaker works down 80 Hz, then it only produces waves up to 14 feet long, so a soffit 14 feet wide would be fine. However, you don't actually need to cover a full wavelength: In fact, a quarter wave is fine, so you could get by with a soffit 3'6" wide, for a cutoff of 80 Hz. If your speaker goes down to 50 Hz then a soffit 5'8" wide would be good, theoretically.
Those are still pretty big for typical home studios, but that's what theory says. On the other hand, soffits normally blend into the side walls and the front wall, which makes them appear bigger than they really are, acoustically: that combination of "soffit plus walls" can even act somewhat like a very large horn, which is good as it helps with the impedance matching, to a certain extent.
So there's your answer: make the soffit as big as you can in all directions, within reason, and hopefully about a quarter wavelength of the lowest frequency that your speakers produce.
But! There's another issue here: the speaker should NOT go in the exact center of the soffit: if you do that, then it creates symmetrical "lobing" patterns, focusing some frequencies more strongly than others in certain directions. You should offset the speaker so that it is a different distance from each edge of the soffit. Normally that happens automatically in the vertical plane: the speaker is closer to the bottom edge of the panel than the top. But you also need to do the same left-to-right. Ideally, try to get the speaker to a location at about 2/5 of the width of the panel, but if you can't get it that far over, then any amount is better than no amount! It normally turns out that moving the speaker so far off center means that it won't fit inside the soffit any more: the back edge will be hitting the side wall, or the front wall, depending on which way you moved it. So just move it as far as you reasonably can off center.
That's just one of the hundreds of compromises that you have to make, when designing a studio!
If you REALLY want to get muddled up in the theory of all this, and confuse yourself even more, then here's a little program I found many years ago that helps predict the baffle step response effect for speaker designers, and also calculates the parameters of the baffle step correction circuit that would be needed for any give speaker shape/size.... :
Speaker-Baffle-Step-Calculator-The-Edge.rar
i look forward to showing you the room some day should travel ever bring you up this way from santiago.
Cool! I do get up to Canada every couple of years, so next time I'm there I'd love to drop in and take a look. Thanks for the invitation!
- Stuart -
