John Sayers' Recording Studio Design Forum Archive

Hi!

After moths of being just the reader of this forum I decided to ask some questions. I'm trying to figure out the best way for flush mounting my monitors and prepare as best as I can before building process. I've found the formula below posted few years ago by Eric Best.

Eric Best wrote:
f3 = 4650/x where x is the baffle width in inches.

So for the 7" baffle we used you have 4650/7=651.43hz.

I understand that 7" and 18" width baffles with single driver flushed in that baffles are just examples and showing how frequency roll off can vary. If we change values to let's say 50" width (similiar in size to real baffle), the roll off frequency will start from 93Hz, which means (if I understand correctly) that all below 93Hz would be boosted +6db when flush mounted. Or maybe I've messed all this?!

1. If monitor have low shelf control starting from 300/400Hz wouldn't it make a huge frequency drop between 93 and 300/400hz?

2. What about the height of the baffle, or the surface size in m2? Isn't it required?

3. What is the meaning of 4560 number?

4. If my understanding is correct, should I build crossover starting the shelfing from 93Hz down instead of monitors build in low shelf?

I hope I've put it clear as my english is very poor. It would be great to hear from you guys, and Eric.

which means (if I understand correctly) that all below 93Hz would be boosted +6db when flush mounted. Or maybe I've messed all this?!

It won't all be rolled off: it's not a sudden drop in response, but a gradual change, over a couple of octaves. Also, the equation gives you the center point of the roll-off: the point where it is -3dB. So there's a couple of octaves below that point, and a couple above it.

What about the height of the baffle, or the surface size in m2? Isn't it required?

No. because the baffle step is all about the smallest dimension of the baffle, and with most speakers, that is the width. If you have a speaker that is wider than it is high, then you would use the height for your calculation, not the width. Whichever dimension is smaller.

What is the meaning of 4560 number?

It's a constant. It is 4560 if you are measuring in inches, 380 if you are measuring in feet, and 115 if you are measuring in meters.

If my understanding is correct, should I build crossover starting the shelfing from 93Hz down instead of monitors build in low shelf?

If the equation gives you 93Hz, that's the center point of the curve: the point where the roll-off is already 3 dB. The roll-off would start 2 octaves higher, at about 372 Hz, and it would end 2 octaves lower, at about 23 Hz. so your shelving filter would be designed for those parameters: start the roll-off at 372 Hz, -3dB at 93 Hz, and -6 dB at 25 Hz. Note (the 2 octave things is just a very rough estimate: there are more accurate ways of calculating the end points.)


- Stuart -

The more I read, the more questions I have, and I was hoping you could clarify some things for me.

1. Why does John's soffit design require the monitor to be placed inside a box, and Thomas's does not? Are there drawbacks or advantages to either?

2. I see no provisions for ventilation using Thomas's design. Does filling the cavity with loose fluffy pink stuff allow enough ventilation as is? Or should an airway be provided?

3. Should the bezel be as large as possible both horizontally and vertically? Ie. Floor to ceiling/wall to wall using Thomas's design? Or could a rigid backed broadband absorber be placed above and below the bezel? If so, how tall vertically would the bezel need to be before adding a little absorption?

4. I've read that the monitor should be centered in the bezel and I have also read that it should not. Which is correct, or is that dependant on the flush mounting design implemented? If an offset is preferred, how great or little of an offset is ideal?

5. For example, if I created a bezel that was 48" wide and 24" tall, would the center frequency be approx 93 or approx 227?

1. Why does John's soffit design require the monitor to be placed inside a box, and Thomas's does not? Are there drawbacks or advantages to either?

John's design holds the speaker very rigidly in place, since the speaker is attached to the baffle via the box. John's design is massive and rigid all around, keeping things very firmly in place. Thomas's design still keeps the front baffle very rigidly mounted, but doesn't need to be as tight on the speaker, since they are independent with his concept.

2. I see no provisions for ventilation using Thomas's design. Does filling the cavity with loose fluffy pink stuff allow enough ventilation as is? Or should an airway be provided?

You still need ventilation. I think Thomas does not show that for clarity, but you do still need it. You can't enclose a speaker that is generating hundreds of watts of heat in a small area with no ventilation and expect it to keep on working.

3. Should the bezel be as large as possible both horizontally and vertically? Ie. Floor to ceiling/wall to wall using Thomas's design? Or could a rigid backed broadband absorber be placed above and below the bezel? If so, how tall vertically would the bezel need to be before adding a little absorption?

That's a little tougher to answer!

The basic concept of flush mounting id to create an "infinite baffle" for the speaker. One that appears infinitely large acoustically. In other words, much lager than the longest wavelength that the speaker can produce. Obviously, in the real world that is impossible, since we are talking about waves that are dozens of feet long. So next best is just "as big as possible". But even that isn't strictly necessary, as you can see from the baffle-step equation. The real purpose of the baffle is just to prevent the wave from "wrapping around" behind the speaker (technically, it prevents it from radiating into full space and forces it to radiate into half-space), so as long as your baffle does that, you are fine. So if your baffle extends out to the side wall and front wall, you are fine horizontally, and if it extend up to the ceiling and down to the floor, you are fine vertically.

But even that isn't necessary: There's another issue involved here: the further away you go along the face of the baffle, the smaller the effect is, and the lower the problematic frequency is, so it gets to the point where level is low enough and the frequency is low enough that it doesn't really matter too much any more.

That's why John can do exactly what you suggest: set up a hard-backed absorber below the speaker to deal with reflections from the back of the desk. You could, in fact, do that on all four sides, starting a a decent distance from the speaker center. The further you get away from the speaker cone, the deeper the absorption can be.

The above applies to all soffit designs, not just John's or Barefoot's

With my soffit designs, I follow John's concepts for the part below the speaker, then I take the baffle up almost to the ceiling, but I do generally leave a gap of many inches, and I use that area for bass trapping. There's a "lid" on top of the area where the speaker is, of course, and on top of that lid I put a lot of absorption, then a fabric front on it. For high-end rooms, I borrow form both John and Barefoot to create a rigid box around the speaker but with suitably designed resilient pads all around, so that there is no contact at all between the speaker and the enclosing box: the speaker "floats" inside the box. I tune the mounting system to a low enough frequency such that resonance will not be an issue.

4. I've read that the monitor should be centered in the bezel and I have also read that it should not. Which is correct, or is that dependant on the flush mounting design implemented? If an offset is preferred, how great or little of an offset is ideal?

If you center it in the baffle, then there will be symmetrical focused lobing patterns forming for some frequencies. If you offset it from the center, that effect is reduced. A good rule of thumb is to center it about 2/5 of the width of the baffle, but that's also not written in stone. I used to think that centering it was ideal, but in reality it isn't, so you might still find some of my older posts from many years back, that say it should be centered. Ignore that!

5. For example, if I created a bezel that was 48" wide and 24" tall, would the center frequency be approx 93 or approx 227?

I think you mean baffle, not bezel, and a baffle that was much wider than it is high would be unusual anyway! But the baffle step response is always related to the SMALLEST dimension of the baffle, so it will be the higher frequency.

- Stuart -

Thank you, Stuart. I appreciate your generosity with your time and knowledge. It helps a great deal.

- Stuart -

Informative thread for me.

[SPAM SIGNATURE REMOVED BY MODERATOR]

viktorokiro wrote:Informative thread for me.

Yup! Very true... you informed us that you are a signature-spaming low-life twerp, so that information got you blocked, banned, and deleted... informatively!

- Stuart -

I have low end studio monitors that are back ported. Will flush mounting work with speakers that are not front ported.

Dirkg wrote:I have low end studio monitors that are back ported. Will flush mounting work with speakers that are not front ported.

Absolutely yes! The majority of rear-ported speakers can, in fact, be flush mounted, as long as the flush mount system is designed correctly. Here's an example: http://www.johnlsayers.com/phpBB2/viewt ... =2&t=20471 The speakers in that studio are Eve SC-407s, which have rather large ports on the rear. As you can see, they were flush mounted very successfully, and
the performance is outstanding.

So very likely your speakers can also be flush mounted. What speakers do you have? (make and model). Unless they are very unusual in some sense you should be OK.

- Stuart -

cabenegogna wrote:Based on the very well explained Genelec flush mounting techniques, and the techniques mentioned on this forum, i was wandering if i can make the outer wall from concrete blocks, and fill the inner space with absorbant material, and also place an opening (like a speaker port) near the bottom of this concrete wall, so that the air can enter and by convection cool a bit the speakers. In this case, do i get any significant advantage on using concrete blocks instead of wood or mda? should i calculate this port based somehow on the speaker size? thanks in advance guys!!!

Cheers from Argentina!!

Any recommendations on the sound absorbing material to deaden the cavity? Would 50mm (2") lets say 46kg/m3 (no idea in imperial sorry) be enough or something more broadband like perf wood is better for low freq abs?

Is there any rule of thumb on the minimum dimensions of the speaker cove? The smaller (with sound absorbing material) the better?


Thanks

As long as the monitors never come in contact with the front baffle, any method of decouplig the monitors from the framing of the baffle should suffice, correct? The goal is to limit vibration transfer between the two? Would suitably matched monitors and sorbothane hemispheres be sufficent since they stop 96% of energy transmission? I know another well respected studio designer uses them to decouple the monitors in his flush mount designs, but your thoughts would be appreciated.

For instance, in the following scenario, placing the monitor on the shelf, but decoupling with sorbothane vs buildig a rigid box to hold the monitors in place

JasonFoi wrote:Would suitably matched monitors and Sorbothane hemispheres be sufficent since they stop 96% of energy transmission?

Hi Jason,

I'm not sure where you got that 96% figure, but it's certainly not true for low frequencies. This graph shows transmission data for a 1.77 inch thick Sorbothane foot. The thickest hemispheres I can find are only 1.25 inches.




Thomas

Thanks for the reply, Thomas! I do not have any frequency specific data. I reviewed the manufactures claims, and they state, "absorbs 94.7% of vibration", not the 96 i previously stated. Thats at 1" hemisphere with a durometer of 50 (they have hemispheres up to 2" on their site). 6 of those would have a rating of 24lbs, the same as my monitors, eve audio sc208's. Considering the cutoff frequency of the baffle step at 447hz, and the fact that i will be using a subwoofer as well, would that be effective isolation? Or should i mount the monitors in a rigid box?

Or is it that the problem is the baffle transfers energy to the monitor instead of the problem being the monitor transfering energy to the baffle? If thats the case then decoupling the monitor as in your method is the superior choice? Correct?

Thanks for the reply, Thomas! I do not have any frequency specific data. I reviewed the manufactures claims, and they state, "absorbs 94.7% of vibration", not the 96 i previously stated. Thats at 1" hemisphere with a durometer of 50 (they have hemispheres up to 2" on their site). 6 of those would have a rating of 24lbs, the same as my monitors, eve audio sc208's

As Thomas pointed out (Hi Thomas! Welcome back!!! ), it's not as easy as that. And manufacturer's claims should be taken with a large grain of salt!

That said, I do use Sorbothane pads (not the hemispheres!) in my speaker mount designs, but it's a real pain to get it right: As the graph posted by Thomas show, you need to get a lot of weight on the pads in order to get the frequency down low enough... but you also have to ensure that you don't over-compress (or more accurately "over deflect") the rubber, as that will trash your isolation too.

Also, do take into account that in acoustics, we don't often talk about isolation in terms of "percent", as that's misleading. It's more common to talk about it in decibels. I always get skeptical when I see isolation claims specified in percentages, as it is easy to mislead like that. Someone who claims their product "stops 75% of sound" is not saying much at all! It sounds like a LOT, but it rally isn't. If your car fuel tank is 75% full, you have plenty of fuel, but if your speaker isolator gives 75% transmission loss, that's pretty poor. This conversion scale might help illustrate:
It's the reciprocal of 94% that you want here, so look at the 6 percent mark on that graph: it's about 24 dB. So if you could tune your mounting system to isolate across the full audio spectrum according to that 94% figure, you'd be getting 24 dB reduction. And that's assuming that you really do get 94%! If it turns out to only be 85%, then the reduction is more like 15 dB.

But what does that even mean? The isolation is different for every frequency on the spectrum, so when they say "94%", do they mean at 100 Hz? Or at the resonant frequency? Or maybe one octave higher? Or 4 octaves higher? They don't say, so you can only guess...

Saying that a product "isolates 94%" is basically meaningless, unless that is qualified with more data.

For example, lets' assume that the 94% figure is true for a tone of 1 kHz. But isolation rolls off ever more steeply at lower frequencies, for reasons that are a bit complex to go iot here (but you can see that on the graph Thomas posted), so even though the product might really provide 94% isolation at 1 kHz, it very surely will not be 94% at 100 Hz! Maybe 80%, with luck... And at 50 Hz, it will be far worse... The isolation gets worse, the lower you go.

You also need to factor in your cross-over (assuming you are using a sub). You might think that because you have a sub, your mains are not sending out any low frequencies, and all of the lows will be coming only from the sub. But that's not correct. For example, if you have your cross-over set to 80 Hz, that does not mean that your main speakers stop dead cold at 79 Hz! They are still emitting abundant sound down to much lower frequencies. Your SC-208's are spec'd to get down to 36 Hz (-3dB), so they are very much capable of putting out a lot of energy at low frequencies, even considering the cross-over. Let's say your cross-over is second-order: it's rolling off 12 dB per octave, so at 40 Hz (one octave below the cross-over point) the level is only 12 dB down: still quite a lot of energy there. Even if you use a fourth-order cross-over, rolling off at 24 dB per octave, that still allows a fair amount of power output at 40 Hz, or 30 Hz, or even 20 Hz. And it's the vey low frequencies where vibration becomes a big issue.

In other words, if you go with a resilient mount, you need to ensure that the isolation system is tuned way lower than the bottom end of the speaker's spectrum, even if you use cross-overs, or else the speaker could still be transmitting substantial vibration into the soffit structure.

Considering the cutoff frequency of the baffle step at 447hz,

The baffle step response is not related to the speaker cut-off frequency. Baffle step is about something else entirely: it's the frequency where the front panel of the speaker stops acting like an infinite baffle, and sound can start "wrapping around" behind the speaker.. And it's not really a "cutoff" either: more like a "roll-off". But either way, the baffle step response does not influence what you are trying to accomplish with the mounting system. In simple terms, your mount must either be tuned so that it provides sufficient isolation at least an octave below the lowest frequency that your speaker will be producing in order to not transmit excessive vibration into the soffit structure, or it must be rigid, solid, and massive enough to be able to handle that energy transfer without vibrating.


- Stuart -