Hey Thomas, Sorry about the long delay in getting back to you! It's been real hectic here for a while... the same day you posted, my daughter was admitted to hospital for major surgery... she's OK now, but that sort of sidelined me from things here a bit. Then Christmas, New Year... But that's all gone now, and I'm getting back to normal... and catching up on the forum!
So....
If there’s a way to safe some of my sketchup and optimize the room for this mixed format approach, your input would be highly appreciated, sine I’ve only a vague idea of what to change...
Basically, the idea is to eliminate reflections from the rear speakers getting back to your ears at the mix position. You can do that by re-angling the hard, reflective surfaces at the front of the room and also slightly adjsuting the rear speaker angles, until you get the situation where there are no serious reflections heading your way. You can also do it by replacing some reflective surfaces with absorptive ones, or even with rounded diffusive ones: it is possible to create treatment that the main L and R speakers will "see" mostly as solid reflective surfaces, (to help with RFZ) but that the rear surrounds will "see" as rounded slats and mostly absorptive. It takes a lot of careful figuring to do that, but it can be done.
Or you can use combinations of the above. As long as you achieve the end goal of not having reflections from the rear surrounds to the mix position, while still having good reflection control for the L and R, then that's fine.
I’ve used the PDF from “The Recording Academy's Producers & Engineers Wing - Recommendations For Surround Sound Production”. There’s a drawing on page 30.
Right. That's one of the standards, but your layout still doesn't seem to match that: As you mentioned, you speakers are not on the "circle", so yes, you will need to have delays on all of the other speakers, to match the time-of-flight to the one that is furthest away.
Also, I have a beef with that diagram: it places the sub in the center of the front wall: that's not a good location for it, as I'm sure you know. It should be off-center, as I mentioned before. The center of the front wall is a modal null for several frequencies in any room, so those frequencies would be somewhat attenuated in the entire room. Subs should be off-center to the left or right. And the best solution is a pair of subs, o even a set of four subs. But that's for a separate discussion...
I’m sitting at approximately 40% if you do the math, but I had to start somewhere. So, as you also say, I only use it as a start to base all other Ideas on. Because only then I was able to place the speakers and check how far they have to be away from the walls and check the dimensions, etc.
That's fine. Many engineers prefer to sit a bit forward of the famous "38%" spot, but a bit behind it is fine too. As long as your predictions show that you will get a good stereo image and broad, clean soundstage, then all is well!
At the moment, they are in a perfect 60° triangle. I’ve used this angel, because with it I was able to use speaker soffits (in my case: the whole front wall) were the speaker is not in the middle of the soffit, but in a 38-62% ratio to the next plates that have a different angle.
I've written about this before, but I think I should do a complete article on it, and posit it here permanently: There is no need have a "perfect 60° triangle". It's nice if you can get it, but not necessary. Nothing will happen if the intercept angle happens to be 58° or 62°. In other words, the speaker toe-in angle does not need to be 30°. In fact, anything between about 25° and 40° is fine, and it is even possible to go out to 20° and 45° under certain circumstances, with careful design. There are many reasons why you might need to do that, such as an unusually wide or narrow room, or window, doors, pillars, in the room that need to be taken into account, or a need to fit in an unusually large or small console, or other furniture. There is no acosutic reason to only ever use 30° toe-in. In fact, if you look into psycho-acoustics, human hearing is most sensitive about 40° off the median plane, so purely form that point of view, it would make sense to spread the speakers wider than the "perfect triangle" allows.
I could go into a lot more detail on that, but that's the basics: 30° toe-in° (60° triangle) is not written in stone, and you can vary the angle a bit, if needed.
the speaker is not in the middle of the soffit, but in a 38-62% ratio to the next plates that have a different angle. This will help to keep the frequency response flat I’ve read.
That's correct, yes. The "rule of thumb" that I use is to offset the speaker 3/5 across the baffle: so your layout is very close to that. It's not so much frequency response that suffers, but rather intensity response too. There can be "lobing" of certain frequencies if you have the speaker perfectly centered in the soffit. So yes, you are doing the right thing.
because the Eve Audio SC 208 have a bass reflex port that faces backwards, which could change its intended tuning frequency if the speaker is too close to a wall.
There are ways of dealing with that. It is possible to damp the reflex port, acoustically, so it is not a problem any more.
Take a look at this thread:
viewtopic.php?f=8&t=14 Those are Eve SC-407 speakers, soffit mounted, vertically! Against the rules. When I designed that room many years ago, I found that I would need to set them up like that for many reasons, so I contacted Eve headquarters, and discussed this with their chief engineer. A very nice guy, and very helpful!. I told him what I wanted to do, and why: he agreed that it was the best option for that room, and we came up with a plan. Those SC-407's also have bass reflex ports on the rear, and there was very little space inside the soffit. So I damped the reflex ports, basically killing them, with the blessing of the chief engineer himself at Eve Audio. He even gave m some tips on the best way to do that with the SC-407. And I also mounted them vertically, instead of the more recommended horizontally. You can see the results on the customer's own website.
So it is possible to deal with bass reflex speakers in soffits... or close to walls. The wall will do that the reflex port was trying to do out in the open, and the wall will do a better job, so there's no problem with damping the reflex port. Some people don't understand this, and think that you can't flush-mount a rear-ported speaker, but in fact it is possible, if you do it right.
So, it would be possible for you to get your SC-208 closer to the front wall if you need to... but do also be careful about cooling airflow. Leave enough space to ensure that air can circulate around the speaker freely.
Mhh, okay. The PDF I’ve shared the Link above says it should be like that. Could you explain this a little further?
If the sub is in the middle of the front wall, it is sitting in a null for every possible first-order axial mode in both the 1.0.0 and 0.1.0 directions, and thus also every tangential mode associated with the 1.1.0 axis, as well as all possible oblique modes. At the same time, it is also sitting in the modal peak for all of the second order modes of those same axes. Thus, it can potentially excite all of those second-order modes with high efficiency, and also fail to produce much sound at the frequencies of the first-order modes. In other words, it distorts the frequency response of the entire room, across the entire spectrum. (The frequency response curve of a room is highly dependent on the modal response, as caused by the speakers). If you offset the sub to one side or the other, then you no longer have that problem: you move the speaker out of the modal points, into areas where modes are less affected: it smooths out the response somewhat.
A pair of subs would be possible to build (but I will need two amplifiers with DSP, which will be expensive)
Unless you use active subs. I really like the Neumann KH-805 for this: it goes down really low, and a pair of those can produce excellent results, even in a small room.
but I think I won’t need that much amount of bass energy in this small room...It could lead to much more problems I fear
It's not about how much power the sub can produce, but rather about how you use that power. It's about placing the speakers carefully, and tuning the carefully, to deal with some of the room problems, in addition to just playing music.
those Eve Audio SC 208 speakers go down to 36 Hz.
The KH-805 goes down one entire octave more! Yes, it goes down to 16 Hz. The spec sheet says 18 Hz, but they are being conservative there: In several places where I have used that in studios, it consistently puts out plenty of energy way down at 16 Hz,... and even lower. In fact, it goes down to about 12 Hz. Here's the final spectrogram for a small control room I designed a few years ago. similar to yours, where I used a KH-805 to clean up the low end:
That's pretty darn amazing, as you can see. Perfectly clean, smooth, even, flat, balanced bass all the way down to 12 Hz.
To put that in perspective, here's the original first spectrogram from the same room, before it was treated and tuned:
So that's the level of improvement you can get when a room is properly designed, properly treated, and properly tuned, including the placement and adjustment of a sub to deal with the low end.
I only intend to use the Sub for Video postproduction exclusively. Maybe check the “club character” of a song with the 2.1 config but that’s it.
That would be a mistake...
That's not what subs are meant for. Contrary to popular belief, the purpose of a sub is not to "make some stuff sound good if it has a lot of bass". The real purpose of a sub in a studio, is acoustic, not music. It is there to even out the low end, smoothing away issues such as SBIR and modal problems that still remain after the room has been treated. By smoothing out the low-end response, it helps flatten the frequency response curve. If you turn it off, then it can no longer do that, and your entire room response is messed up!
This is not at all intuitive, and many people don't understand how a sub can fix problems in the mid range, or higher, but it does.
It's a little complex, but basically it has to do with the modal response of the room. As I mentioned before, the low frequency modal response "shapes" the frequency response curve for the entire room, and the entire spectrum. If a room could have no modes at all (in other words: no walls and no ceiling), then the frequency response would be flat. As soon as you have one single mode in the room, that puts a "kink" in the response curve, all the way up... because you can't just have one single mode! If you have a mode at, for example, 30 Hz, then you also MUST have one at 60 Hz, ...and another at 90... and another at 120... and so on, all the way up the spectrum, at every multiple of 30 Hz. Because modes are resonant, they are also harmonic. Every single harmonic of that fundamental mode is present in the spectrum. So just from having one single modal resonance in the room, that messes up the entire spectrum: the full frequency response curve is no longer flat, all the way up to 20,000 Hz, just because of one mode. Now, if there were only some way to deal with that one single mode at 30 Hz, if you could flatten it out somehow, then you would also flatten out every single harmonic... and the full frequency response curve would be flat again. If only there were some way to deal with that fundamental mode, way down at 30 Hz. If only..... If only you had a sub! With a sub, you can indeed deal with that mode. And with several others as well. By carefully placing and tuning the sub, you can flatten out the entire frequency spectrum again, by using the sub. Then, every time that any sound came out of the sound system at any of the frequencies that the sub covers, the sub would produce that tone in the right way, and at the right location, to keep things flat around the rest of the room. But if you don't have a sub, or if you turn it off, then it can't do that, and you mess up the frequency response again....
So that's what studio designers and acousticians use subs for: not just to make the music "sound nice", but rather to deal with acoustic issues in the room that only a sub can fix.
That's how I got the response of that room so flat, and clean: by using a sub that goes down low enough to do the job.
Yes, there are still some artifacts visible in that "final" graph: certainly. So by rolling off the very low end on the sub (everything below 17 Hz), well below the musical spectrum, I improved things even more, to this:
So that's what a sub will do for you, in a well design room, properly treated, and properly tuned.
The Eves have their own DSP with some filters.
Yes they do, but they do not have the precision needed to tune a room. They are great speakers, and I have used Eves in several rooms, but I just set them mostly flat with a slight bass roll-off, then I do the actual tuning in hardware.
If they don’t match the problems I will use the control room functions of my DAW, which allow me to put VST plugins in the monitor way.
It might seem that this will work, but in reality it wont. Once again, it does not have the precision of a dedicated hardware DSP, and there's another major issue: it only works when you use your DAW! You wont have any tuning at all if you just play back something from your console, or anything else that does direct to the speakers. Even if you play music on the computer itself without running the DAW, then you don't have any correction.... This is not a good solution for a pro studio, and I don't know of any major studio that works like that.
For the Sub, I wanted to use a modular system from sure electronics (
http://store.sure-electronics.com/). They sell amplifiers, DSP programmers, DSP break-out circuits and DSP boards separately. You can combine them as needed for the task. The ADAU 1701 DSP board they offer is a bit older but performs well @48 kHz and has a simple graphical programming interface, where you can combine different audio tasks like delay, iir filters, fir filters, etc.
I'm sure it is possible to string together a bunch of stuff and get a good result, but I find it a whole lot easier to just buy something like a KH-805 and an SP-2600, and I'm done! That's all I need. The 805 goes down way lower than anything else I'm aware of in the price range, and works very well. The 2600 gives me a really good range of filters with reasonably good precision, to do the tuning I need to do on all the speakers. And I use REW for all the rest.
The flush mounting baffle, which I’ve described in my second post of this thread, is designed to guide most of the mid range wavelength (the speaker are narrowing as higher the frequency gets) above my head.
It looks like you are trying to do a CID concept design, but you are not following the full CID principles. You should read the actual paper on that, to find out how to do it. It's old, but you can probably find it on-line. I'll see if I have it somewhere. However, I'm not aware of many studios that were ever built using CID (apart from those few at the BBC), and it is certainly not a commonly used design concept. It has several drawbacks, one of which is that it leaves the room overly live, and it also only works for a single location in the room. In the years since CID was first proposed, better concepts and designs have appeared. That's why I suggested using a hard-backed, angled cloud. Because you an do more with it than you can with a fixed ceiling panel. A hard-backed cloud can do several things at once, but a CID ceiling cannot.
With a short gap for the HVAC ventilation pipe (the air should be sucked out at this point of the room),
Yes, but not through a pipe! It should go out through a silencer box, then into the actual HVAC duct.
However, I’ve tried that before and it seems that it is really hard to get a good working reflection free zone that way if I only use an angled cloud instead of the two angled plates that are connected to the flush mounting baffle.
It is ALWAYS hard to produce a good RFZ in most small rooms!
But it's a lot easier if you have a cloud that you can adjust the angle of, to tune the room as needed. In the room for the graphs that I showed above, I initially go the angle wrong, and needed to adjust it. I was getting unexpected reflections (I still don't know where they came from), but changing the angle of the cloud fixed it. And it also fixed some other issues.
Clouds can be used to do a lot more than just establish an RFZ: The can also be used to smooth out bass issues, for example. Sticking to the example of the same room, here's the "before" and "after" waterfall plots regarding the cloud:
BEFORE CLOUD:
AFTER CLOUD INSTALLED:
You can clearly see the huge improvement in the low end.
Clouds can do lots of things: CID panels can't.
For my understanding, it is even more critical to have two different angels in my case,
You mean like this?
That's a cloud I designed for a very small room, in two sections at two different angles.
because the near field sound waves radiate in a spherical sector, like you said, and therefore the approximation to a ray do not reflect the reality exact enough.
It is the mid and low frequency sounds that expand spherically, and not just in the near field. You can probably figure out why I did some of the things you can see in those two clouds.
Or is it not as critical as I think it is? Please advice in this case, because I’m kind of lost here. And because the approach for a surround configuration is different,
Right. It is important, and getting the angle right is also important. For a mult-channel room, you would check that by revers-ray tracing: from the rear speakers back towards the front of the room, to see what they hit, and where they would be reflected.
That’s good to hear, since it will save me a lot of time and money if I don’t have to build them! But another reason why I choose this design was the fact that I did not get a good reflection free zone in the horizontal position without those walls, so I thought it will not harm to do “two-in-one” side walls. If you could take a look at the following ray tracing pictures, you can see what I meant. If I don’t want to cover the window, or change the angle so that the window is not convered, a reflection free zone does not work, if the ray tracing is trustworthy.
I think you maybe missed the point I was making: you do need the "wing" sections between the actual soffit face and the wall, and it does need to be angled, yes. But the side wall itself does not need to be angled. If you have your mix position in the best location, and the soffit and wings correctly designed, then there should be no need to have the side walls angled. Unless the room is very narrow, in which case there are other ways of dealing with that...
Okay, understood. So the only way to avoid this (which should be possible if I understood you right) is to build without the outer framework and just build the front with the reflection free zone baffle (freestanding because of the decoupled floor impact protection) and treat the room for acoustically. And check how the reflection free zone baffle has to be altered to prevent problems with the rear speakers.
That's the basic idea, yes!
Regarding the cloud: it is possible to hang it on the ceiling directly, because it is not touching anything.
Absolutely! That's the way I normally do it. I assume your ceiling is able to support the weight. Just use good strong chain, and closed-loop hooks to hang it. Make sure the chain and hooks can support ten times the weight of the cloud, for safety. So if the cloud weighs (for example) 30 kg, then make sure that the chains and hooks can support 300 kg.
I will look for a good paint in this case and use ordinary MDF. So, what do you recommend to use to combine those two layers for the baffle? Ordinary wood glue? Greenglue (which is hard to get and therefore also very expensive)?
Wood glue is fine. Gree Glue is not glue at all! Despite the name it is NOT adhesive, and CANNOT be used to stick things together. I wish they would have chosen a better name... It's a great product, and really, really good at what it does: in fact, it is without equal for that. It is a constrained layer damping compound, for use between two layers of drywall in isolation construction, and it excels for that. But it is not glue. I didn't watch the video yet, but my guess would be that the people who have "differing opinions" about it, are not using it the way it is supposed to be used, for the purpose it is designed for. It has been extremely well tested in independent acoustic testing laboratories, and the test results have been widely published: It does exactly what the manufacturer claims it does. If you try to use it for another purpose, then it probably won't work well. But when used correctly, for the design purpose, it is amazing stuff.
But anyway: just use wood glue, or even better don't glue it at all: just screw the panels together around the edges (from the back).
Could you explain a bit further why it needs to be this heavy? When I’m developing Hifi Speakers e.g. I can test different baffle designs using cardboard and except of the frequencies less than 200 Hz, the measurements are nearly the same when the speaker is made out of MDF.
The baffle is a very large surface: many times the surface area of the typical baffle on a studio speaker. It is subject to much larger forces, because of the acoustic loading imposed on it by the room, and also the loading that it imposes on the room. It must not vibrate, so it has to be massive, and rigid. And it is precisely the frequencies below 200 Hz that it must deal with. In fact, it deals with everything down to the lowest frequency present in the room, and it must have the mass and rigidity needed to handle that, without flexing or vibrating.
Allright! This is the answer of a question that I’ve asked above! So, no big baffle, just a stripe of reflective surface for the speakers to work out the refelction free zone. But how much bigger than the speaker does it have to be?
I think you misunderstood: You DO need a very large baffle around the speaker: as large as possible. But at the same time, you can't cover the zone too far down towards the floor, or too far up towards the ceiling, because there are other things going on up there, that need to be considered as well. Ideally, the baffle would go floor to ceiling, but in practical reality it cannot do that, because there are other conflicting needs. So the baffle has to be a bit shorter, in order that the other things can work.
Understood. I will not have a console, I have a “reflection free desk” (expanded metal with cloth over it to let the lower frequencies through)
Excellent!
But do I really need a ventilation slot directly above the speaker (like it is shown in the concrete soffit you have posted earlier)?
You need to ventilate somewhere! I used to do it like that, with a slot in the front of the baffle, copying a really old concept from somewhere else, but now I do it differently. I try to keep it concealed, in the top panel of the soffit enclosure, so it ventilates up through the hangers, and out the front and side of the soffit module.
I don’t want to connect the floor with the walls because I will lose the intended acoustic insulation from the decoupled floor
Decouple the speaker itself, not the soffit! The old-school method for soffit-mounting speakers was to grip them tightly in a very, very massive, rigid box inside a very, very massive, rigid structure. But better ways have been developed over the years, and most up-to-date designers don't do it that way any more. Instead, we completely "float" the speakers inside the soffit, so that they are free to vibrate as much as they want, but cannot transmit any vibration into the soffit structure itself. Thus, the soffit framing and baffle do not get any vibration from the speaker itself, since it is fully decoupled from the structure.
I understand that you will keep some of your secrets! No problem. Just to get this right: Is it critical to build a bigger cabinet with all the insulation that is sealed? Or can I just build a frame, mount the soffits and just build fake walls above and below? Like I said, I want to pack all the insulation in plastic sheets, so that the fibres can’t get into the air. I’m planning to use plastic sheets that are only 10 µm thick and therefore are very light. Is this a problem?
As I hinted at above, there are two basic methods for soffit-mounting speakers. The "old" way, where the speaker is mounted tightly in a very rigid, very massive enclosure, that is then rigidly mounted to the framing and baffle of the soffit, which is also very rigid and massive. That's not a good method, and seldom done these days by designers who are up to date on acoustics and technology. Since the wooden enclosure box must fit tightly around the speaker cabinet, it can damage the cabinet as you try to force the speaker in and out, and it prevents proper ventilation of the cabinet too: it prevents air flowing over the sides, top and bottom, which is important for some speakers, as you know yourself. It also changes the characteristics of the speaker, as you also know, because it changes the density and weight of that cabinet.
The other, more modern (and more sensible method) is to "float" the speaker on resilient pads. You still have a massive, rigid enclosure around the speaker, but it does not fit tightly: it is held on place by a few strategically placed rubber pads, carefully located to absorb the movement of the speaker as it vibrates, and carefully calculated in terms of size, thickness, deflection, and resilience, so that the speaker does indeed "float". The secret to getting this right is the calculation of the pads. It's not that hard, but it does take time and effort, and care. Basically, the speaker and the enclosure box form an MSM resonant system, so you just need to calculate the resonant frequency for that, and make sure that it is at least one octave lower than the lowest frequency that the speaker will b producing. It's that simple. If you know the resilient characteristics of the pads, the weight of the speaker, and the frequency, then it's not hard to do the math and figure out how big to make them, and where to put them. The complexity comes in when you realize that you have to do this on ALL sides of the speaker (not just under it), and that speakers are not well balanced: they are heavier at teh front, or back, or to one side... so you have to adjust the sizes of the pads to compensate for that. Then also, the pads on top of the speaker are putting pressure on it, pressing it down, so the pads under it are loaded not just with the weight of the speaker but also with the pressure from the top pads.... Yeah, it gets to be complicated. That's why I wrote a program to do all the calculations for me, which makes it much faster, and simpler....
With this second method, there is now space around the sides, top and bottom of the speaker cabinet, inside the enclosure box, so it can still have air flowing around it, as the designer intended. It is also held gently, not tightly, on the resilient mounts, so the cabinet is not damaged from grinding it against the enclosure box, and the original cabinet is still there, not in direct contact with the enclosure box, so it behaves as the original speaker designer intended. So this method is far, far better than the old method... but it involves math that some folks are not prepared to do (in case they get it wrong, I suppose...), so the only option they have is to use the old, outdated methods. That's a pity, because it really isn't that hard to float it, and the results are so much better! For example, taking the room associated with all the graphs above again: I floated those speakers like this, and you can see the results. And the Eve SC-407 speakers in the thread I linked you to above, Studio Three Productions, are also floated in the same way: you can see the results their too. I could show you many others where I have gotten similar results, so it's pretty clear that it works very, very well when done right. I have heard some people (who don't seem to understand the physics of floating systems) claim that this is not a good way to soffit-mount speakers, and it does not work, and it produces artifacts, and they don't like it... but the truth is that they are wrong: It does work, really really well. Much better than the old "super-massive cabinet crusher" method.
I can't say any more about how I do it, as it's taken me years to develop my method. I probably said too much here already. But I do offer a "soffit design service" if you are interested, where I design just the soffits for any give room and speaker. I have a standard design for doing that, and it's fairly easy to adapt to different rooms shapes and sizes, and different speakers. It just takes a few hours to do, for most speakers and rooms. Of course, all the designs I do like this look similar from the outside, since its the same basic design, but inside they are all different, adapted specifically for each client. If you don't feel up to the task of doing all that design work yourself for your soffits, then that service is available. PM me. You give me the dimensions of your room, and the make and model of speaker, and I send you back the complete dimensioned 3D plans, tailored specifically for your room, and your speakers.
Good to know, but that’s your secret, isn’t it?
Yep!
Oh Boy, I hope it won’t be necessary to replace the speakers this often in my case. But I get the point. Question is, am I able to build something similar in my room?
Certainly! That option isn't in my standard soffit design, but I can easily add it for specific cases.
In fact, I originally came up with the concept for Studio Three Productions (see link above), as they owner thought he might want to upgrade his speakers one day. He originally had some old Genelec's in there, and that's what I design the soffit mounts (and the entire room) for. Then about a year after the studio was completed, one of his Genelec's developed a problem, so he decided to replace them, with the Eve SC-407's. It was simple to build new internals for the new speakers, then slide out the old modules, slide in the new ones, and within a few hours he was up and running with the new speakers. We also had to re-tune the room, of course, since the Eve SC-407's are rather different form the Genelec's, sonically, and he also wanted to add a second sub at the same time. But the actual swap of the speakers was down in a few hours, with very little down-time for the studio. It is not possible to do that with the old "cabinet crusher" method of soffit mounting, since everything is tightly bolted and nailed together...
Thank you for this advice. Like I said, I will change the design accordingly and leave the slat resonators out if it.
- Stuart -
Translate This Page
Wow! That's fantastic! You have an entire facility at your disposal, and your own room as well! Lucky guy....
).
