Converted cowshed, cathedral ceiling, One room studio build

[astro_mat]

Apologies if this post is long, I just want to give all the detail. I think this may be an interesting project to some as there are elements of my design that don’t seem to be addressed elsewhere. I hope I have followed the rules correctly. Although I have lurked on this forum (and others) for a long time, I have never posted so, here goes.

Overview:

My project is a converted stone cow-shed in rural west of Ireland. It is a standard rectangular shape, gable roof structure. The building has approximate internal dimensions of 7.9m Long, 3.9m Wide with approximately 0.5m stone walls. The walls are 2.6m high and on this sits a gable roof consisting of 150mm rafters and a 12/12 pitch. A stone wall 4.9m from the back wall divides the floor area into two rooms (Live space and workshop area) the live performance space being approx 5x3.9m. On top of this wall and the workshop area, sits a Mezzanine area that will be the control room 2.2m above the ground level and extending 1.58m over the live area. This elevated area will contain steps up to the upper floor and a a treated acoustic space described later.

Design Points to note:

This will be a “One-Room” design meaning that all activities will be in the same acoustic space. No areas will be isolated from the others. This will hopefully make construction simpler/cheaper as I will not be trying to isolate any rooms. Also, treatment should be easier as bass frequencies have more space to expand into (Hopefully).

As this is in a rural area, sound isolation is of moderate concern there are no houses nearby and I am willing to work with outside sounds leaking in. The only Isolation will come from sealing the space as much as I can from air leaks, the 0.5m thick stone walls and the foam insulation in the roof space. Adding Isolation would reduce the size of the already small roof space (control room) too much to be practical. The added benefit is that bass frequencies are free to leave the structure rather than bounce around within it which should help with treatment.

Main treatment is from covering the entire roof with 50mm RW3 for mid and high attenuation. There are currently bass traps in the front corners which are built in and approximately 75cm deep at the deepest point. They are faced with 100mm RW3 and filled with uncompressed Knauf earthwool fluffy loft insulation. Further bass trapping is provided by filling in the space where the roof meets the upper floor with a 50mm RW3 face and the same light fluffy infill. This is approximately 30cm deep and will continue around the whole structure (See top view image).

The control room floor is 2 layers of 18mm ply, 2 layers of plasterboard between the joists, light fluffy infill, 1 layer plasterboard. This is only half finished at the moment (no in-fill or last layer yet)

A dead (ish) recording space will be under the mezzanine. The Walls here will be lined with 50-100mm RW3, fabric and then slats (if needed after measurement). the front of this area will have a movable, treated gobo/workbench (not pictured). This will give me a seperate recording space that is more acoustically dead than the rest. The gobo can be stored under the mezzanine when being used as a bench. Pulled out to give a little vocal room or guitar amps or moved to the side for full band performance.

The stone walls will be pointed with lime with deep (80mm) gaps between the stone. Hopefully this will give me some nice diffusion for drum sounds? I will add panels and traps following measurement of the space if needed.

The roof construction is 100mm Foam insulation between rafters (50mm air gap behind). There is then 50mm of Rockwool RW3 as acoustic treatment. There will then be fabric and finally slats. Slat/slot width and placement to be determined after measurement to achieve the desired RT60

The odd size and dimensions of the space makes calculations hard. Because of this, I have took a “build-measure-build” approach.

Slat/slot width on the roof and in dead space to be determined after measurement to achieve the desired RT60

There is a large living area adjacent to the construction. Tie lines will allow me to use this area for larger ensembles recording guitar amps etc.

Goals:
To have a multifunction recording space where everything is set up, ready to go.
Too finish the control room part so I can set up my equipment there as I wait and save for the rest of the build.
To have as balanced sounding control room as I can given the limitations.


Where in the process:
Currently in the project, I am working on the roof. I have completed covering it in 50mm RW3 and the bass traps are in place. My room measurements seem to be good but any further experiments with panel placement to treat anomalies seem to make the response worse. I am trying to decide if this is as good as I can get it given my limitations before I cover with fabric and experiment with slats. It is worth noting that placement of speakers is limited to a max distance between speakers of __cm and this obviously places them very close to the angled side walls. I am using a pair of Neumann KH80’s with the Neumann KH805 Subwoofer. I had originally intended to mount the KH80’s on the wall with 100mm RW3 behind in order to reduce SBIR but I seemed to get better results with them spaced 59cm away from the front stud wall. As the speakers are bass managed, this gives seems to give me some lee-way as to placement away from the wall. The sub does seem to make the floor resonate at about 20Hz but I am not sure there is much more I can do about this other than completing the damping under the floor. The sub has been mounted on an isolation platform consisting of two semi inflated inner tubes sandwiched between two concrete slabs. There seem to be dips caused by side wall reflections in the response. I have tried deepening the absorption here. I have also tried angled solid reflectors but all efforts seem to make it worse.

BUDGET:
I am time rich but cash poor so am willing to wait in order to save the cash to build this the way I want. That being said, I am also very frugal so would hope to do the rest of the bare-bones construction within €2000.

QUESTION:

Given the limitations I have, can I consider my control room REW results as good as I can get? I have tried various angled reflection panels and absorption panels but I am guessing that the proximity to the wall and the consequential shallow angle of incidence means that absorption panels are not very effective and reflect the waves more than absorb?. I am open to suggestions?

The link to an .mdat file is here

I include some screenshots as well.

Many thanks in advance.

[DanDan]

Hi Matthew, welcome. That is a very comprehensive post, great sketch upping, and it looks like a wonderful place.
But it is hard for me/us to see the wood from the trees.
I am unclear as to how much has been done and what is proposed.
What are you asking?
Could you isolate your questions into punchy one liners? Perhaps just four to begin with.


DD

[Gregwor]

Can you post some real life pictures?

Greg

[astro_mat]

Many thanks for the reply. I will attach some images to show the stage that I am at now. I omitted them before as I find i quite difficult to take pictures (With my tablet) that give a good representation of the space and felt that the sketch up files showed it better.

I guess my main question at this stage is;

Can I consider my results good enough at this stage to proceed with covering with fabric and then adding slats?

I don't think I can improve on my results given the limitations I have ie: speakers close to slanting side wall and low ceiling. I have experimented with speaker/listener position, adding angled reflection panels and extra absorption to the sides and roof but all attempts seem to make the results worse. I am, of course, open to your suggestions.

Mat

[astro_mat]

Here are some images. I hope they are clear enough.

[Soundman2020]

I don't think I can improve on my results given the limitations I have ie: speakers close to slanting side wall and low ceiling. I have experimented with speaker/listener position, adding angled reflection panels and extra absorption to the sides and roof but all attempts seem to make the results worse. I am, of course, open to your suggestions.

I would suggest that you seriously consider flush-mounting ("soffit mounting") your speakers. With all those crazy angled surfaces so close to the speakers, and to your head, I would try to eliminate as many artifacts as possible, and the easiest way of doing that is to soffit mount your speakers. It's a lot of work, yes, but the results would be worth it. Soffit mounting is the single best thing yuo can do for your room acoustics in any room, but it makes even more sense in a complex room like that.

I took a look at your REW data, and there's several issues. First, the room is very dead. Way too dead. That can be recovered with your slats, hopefully, so it's not too much of an issue at present, but you'll have to tune it carefully. Here's the situation:

Astro-Mat-_REW--RT--40-12k-b.png

Most of the spectrum is way below 100 ms, which is way too dry, but below about 500 Hz it starts improving, and below about 300 Hz it is decent, so you need to take care that you only liven up the high-mids and highs, with a bit of attention to the low mids, while not doing too much at all to the lows. Thus: careful tuning of your treatment.

There's also the issue of the coupling to the large room down below: that is very evident in the ETC data:

Astro-Mat-_REW--IR-200ms-b.png

Basically, it's a huge mess! The Schroeder integral tells the real story here. There's two wildly different slopes (at least two: perhaps more), and the overall decay is anything but smooth. The room itself dies after about 60 ms, then the rest of the studio kicks in and provides the tail end, but mostly in the lows. Here's what it should look like, for a similar sized control room with proper treatment:

FCLD-sample--REW-IR--250ms.png

The Schroeder integral should be a straight line, angled downwards at the target decay rate for the room. Yours can never be flat, since you have two starkly different coupled acoustic spaces, but at least it should be cleaned up a little!

You also have a serious imbalance between your speakers: the right speaker is several dB louder than the left:

Astro-Mat-_REW--FR--12--24k--Speaker-Difference-Compare.png

That graph shows the DIFFERENCE in level between your left and right speakers, where the "0" line across the middle marks the level where both speakers are the same. Wherever the curve rises above zero, your left speaker is louder at that frequency. Wherever the curve falls below the line, your right speaker is louder at that frequency. As you can see, for the entire top half of the spectrum (everything above 1 kHz) your right speaker is louder, and from 3 k upwards it is considerably louder: about 5 dB to 7 dB louder, in fact! So there's some problem with your signal chain there, ir perhaps it's the speakers themselves not being matched. But there's a big issue there: That's not acoustic. The room cannot do that to your levels. That's a speaker or signal chain problem, and it needs to be found and fixed.

I don't really have more time to get into greater detail, but those are the issues I see at first glance.

- Stuart -

[astro_mat]

Many thanks for taking the time to analysis my problem, it is much appreciated.

I would suggest that you seriously consider flush-mounting ("soffit mounting") your speakers. With all those crazy angled surfaces so close to the speakers, and to your head, I would try to eliminate as many artifacts as possible, and the easiest way of doing that is to soffit mount your speakers. It's a lot of work, yes, but the results would be worth it. Soffit mounting is the single best thing yuo can do for your room acoustics in any room, but it makes even more sense in a complex room like that.

I had considered flush mounting and would be happy to put the time and effort in. I ruled it out for several reasons:

1) I figured that the main benefit of flush mounting is eliminating SBIR effects from the front wall by limiting the radiation from the speaker to the front hemisphere. I was hoping that this could be reduced enough by positioning the speaker very close to the wall and using absorbtion behind it (As recommended by dandan and others).
2) Unwanted secondary sound radiation from the monitor cabinet's edges is already dealt with pretty well by the cabinet design of the Neumann KH80 speakers.
3) The Neumann KH80's do not lend themselves particularly well to flush mounting (Curved front bezel, amp situated at back requiring ventilation).
4) Due to the strange angles and hard to predict room involved, I thought that it would be better to be able to re-position the speaker to find the best place rather than having it fixed in a wall.
5) I was worried that the increased power needed would mean that the speakers I have would not be adequate.

I could sell the speakers and find a more suitable pair but I am really fond of them. If there is no other way around it, I would be happy to take the effort to flush-mount but I'm not sure that with all the other limitations I have, It would be worth it. I am certainly keeping it in mind as a future option.

I took a look at your REW data, and there's several issues. First, the room is very dead. Way too dead. That can be recovered with your slats, hopefully, so it's not too much of an issue at present, but you'll have to tune it carefully.

Most of the spectrum is way below 100 ms, which is way too dry, but below about 500 Hz it starts improving, and below about 300 Hz it is decent, so you need to take care that you only liven up the high-mids and highs, with a bit of attention to the low mids, while not doing too much at all to the lows. Thus: careful tuning of your treatment.

This is great to hear, I had suspected it to be too dead and it certainly sounds it! I will be adding slats to the wall to bring the lively-ness back. I am thinking now that I may experiment with different arrangements and slat-spacing before applying the fabric but assume that this will not effect the lowend. For now, I am just seeing if I am going in the right direction.

Basically, it's a huge mess! The Schroeder integral tells the real story here. There's two wildly different slopes (at least two: perhaps more), and the overall decay is anything but smooth. The room itself dies after about 60 ms, then the rest of the studio kicks in and provides the tail end, but mostly in the lows. Here's what it should look like, for a similar sized control room with proper treatment:

FCLD-sample--REW-IR--250ms.png

The Schroeder integral should be a straight line, angled downwards at the target decay rate for the room. Yours can never be flat, since you have two starkly different coupled acoustic spaces, but at least it should be cleaned up a little!

This is fascinating stuff! Would this situation be improved by adding absorption to the large live space? I plan to fit two large bass-traps in the far corners and the whole area under the mezzanine will be absorptive. could this help this situation?

You also have a serious imbalance between your speakers: the right speaker is several dB louder than the left:

Astro-Mat-_REW--FR--12--24k--Speaker-Difference-Compare.png

That graph shows the DIFFERENCE in level between your left and right speakers, where the "0" line across the middle marks the level where both speakers are the same. Wherever the curve rises above zero, your left speaker is louder at that frequency. Wherever the curve falls below the line, your right speaker is louder at that frequency. As you can see, for the entire top half of the spectrum (everything above 1 kHz) your right speaker is louder, and from 3 k upwards it is considerably louder: about 5 dB to 7 dB louder, in fact! So there's some problem with your signal chain there, ir perhaps it's the speakers themselves not being matched. But there's a big issue there: That's not acoustic. The room cannot do that to your levels. That's a speaker or signal chain problem, and it needs to be found and fixed.

I hadn't noticed this. It could be that I had not set the levels correctly between the two speakers (there is a gain knob at the back). as this is a technical issue I am not as concerned about this as I am about the acoustics for now (Although I will investigate when it comes time to take more measurements).

Another possibility that I have been a bit worried about is that the Neumann KH805 sub (Non DSP) feeds into the DSP controlled Neumann KH80's meaning that there will be some latency (and thus phase differences) between the two. This can be somewhat mitigated by the sub's built in phase switches but not perfectly. I am not sure weather this could be causing the issue. How do you set REW up to compare the phase between the left and right speakers? That is very handy and would help me to test if I have it right next time.

Mat

[Soundman2020]

1) I figured that the main benefit of flush mounting is eliminating SBIR effects from the front wall by limiting the radiation from the speaker to the front hemisphere.

That's just one of the benefits, but not the only one by any means! Others include: extending the low frequency response, correcting the power imbalance, eliminating edge diffraction artifacts, eliminating comb filtering and other phase-related artifacts, improving focus, improving clarity, tighter bass, etc. etc.

I was hoping that this could be reduced enough by positioning the speaker very close to the wall and using absorbtion behind it

That reduces SBIR yes, but does not eliminate it. It pushes the first dip up the spectrum, into the low mid-range, where it is less noticeable and easier to treat, but it does not eliminate it.

2) Unwanted secondary sound radiation from the monitor cabinet's edges is already dealt with pretty well by the cabinet design of the Neumann KH80 speakers.

Wellll.... Let me put it this way: there is marketing hype, then there is technical reality. I love Neumann speakers, and consider them among the best: I have used them in several studio designs. But I don't fall for their marketing hype. They are great speakers, but they cannot defy the laws of physics. They KH80 has slightly rounded edges (small radius) and a curved front face. That's all. In the very best case, that will slightly decrease edge-diffraction artifacts, but it does not eliminate them. It would have to have extremely rounded surfaces to have a significant effect. The best possible shape for a speaker, is a sphere shape, or egg shape, where the entire cabinet is curved back on itself. In theory, that's the shape that produces the least edge diffraction, but even then it's not perfect. And the KH80 is a far, far distance away from that.

Here's a graph published by Sound on Sound magazine a few years ago, that shows the frequency response for a number of possible speaker cabinet shapes. "A" is clearly the best, but even then it still does not get flat response. The KH 80 is more like "K". Or maybe a combination of "K" and "I":

Speaker-response-curves-for-various-cabinet-shapes--03.jpg

So no, despite their wonderful marketing, and the great technical design team at Neumann, the KH80 has not eliminated edge-diffraction, nor any of the other artifacts related to putting speakers in a small box. This is not about how the box is shaped, but simply because the box is too small. The ONLY way to eliminate edge diffraction completely, is to make the box large enough that the longest wavelengths that it can produce are still small when compared to the size of the speaker. And since the KH80 can produce wavelengths that are 21 feet long (6.5 meters), the front baffle of the speaker would need to be 21 feet wide, at least, in order to fully eliminate all edge diffraction. In fact, theoretically, to fully eliminate it, it would have to be an "infinite baffle", many times wider than the lowest wavelength... Which is impossible, of course.

That said, soffit mounting actually DOES create an infinite baffle, if you do it right, because the entire wall surface of the room becomes the baffle... But even if you can't achieve that, you can still get a large improvement, way beyond what the cabinet itself can do. No matter how well they shape the edges the fact remains that the KH80 is only 154 mm wide, and therefore has a baffle step response that commences at 1524 Hz and continues all the way down to 381 Hz, with the mid point of the response being at 762 Hz. If you put that in a soffit that is 1m wide, then you force the baffle step response way down: it now STARTS at 116 Hz, the mid point is 58 Hz, and the bottom end is at 29 Hz. Simple physics. Thus, a soffit will eliminate the problem completely, because the mid point of the curve is already around the cut-off frequency for the speaker. And even if you can't make the baffle 1m wide, whatever width you can achieve is still going to be a great improvement.

3) The Neumann KH80's do not lend themselves particularly well to flush mounting (Curved front bezel, amp situated at back requiring ventilation).

Neither of those is a problem. Not in the least. ALL speaker soffits need rear-panel ventilation, regardless of where the amp is (even if it is located in a box that is external to the soffit). That's a basic part of soffit design: leaving a clear air path up the rear of the speaker. Pretty much all active studio monitors these days have amps on the back panel, and it just is not a problem. I do this all the time for the studios I design. Here's an example: thread about Studio Three Productions' studio That one has rather large Eve Audio SC-407's, which have huge amps in the back. They are also rear-ported, which introduces additional issues that you won't have to deal with, with your KH80's. And look at the results from that room. Here's another example, currently under construction: thread about Steve's high-end control room in New Orleans And yet another, recently completed: thread about Martin's small but perfect control room in Slovakia There are many more examples on the forum of studios with soffit-mounted speakers.

The curved front face makes no difference either: if it is possible to soffit-mount a Genelec 8050, which is far more curved than your KH80, then it is possible to soffit-mount yours:

JKB-DNMK--Genelec-8050--Soffited-S47.jpg

4) Due to the strange angles and hard to predict room involved, I thought that it would be better to be able to re-position the speaker to find the best place rather than having it fixed in a wall.

The best place is IN the wall! And there's only one single variable that needs to be optimized for soffit-mount speakers: how far across the baffle. The height is already fixed, carved in stone. The distance into the room is also fixed: carved in stone. The only thing you can change is the location across the width of the baffle, and that is easy to calculate.

5) I was worried that the increased power needed would mean that the speakers I have would not be adequate.

??? What "increased power needed" are you talking about?!?!? You seem to be badly misinformed! By soffit mounting a speaker, you REDUCE the power needed, because you ELIMINATE the power imbalance that is inherent to having a speaker in the room! You will, in fact, have to TURN DOWN the level of the woofer by several decibels: if not, it will be too loud. The reason is very simple: when you have a speaker radiating into a room, half of the power goes towards you, and the other half goes backwards: away from you. When you put it in a soffit, ALL of the power goes towards you, and NONE goes away from you. The soffit baffle prevents the "wrap around", where half of the power is lost due to going the wrong way. And since you also no longer waste power in the SBIR and other phase cancellations losses, you just don't need as much power.

This is very simple to illustrate: When you need to shout really loud, you automatically cup your hands around your mouth... because that prevents some of your shouted energy from wrapping back behind your head, and instead directs more of it in the direction you are facing! Exact same principle as a soffit baffle. It also improves the impedance mismatch between your mouth and the air around you (just like a soffit or waveguide does).

So whoever told you that you need MORE power if you soffit-mount, is unfortunately very wrong, and very ignorant of speakers and acoustics. You need LESS power to get the SAME level, when you have your speakers mounted in soffits. You also get more power in the low end (better low frequency extension), because the power imbalance has been restored.

Technically, a speaker that is not soffit mounted radiates into 4-pi space, spreading all it's energy into a sphere. A speaker that is soffit mounted radiates into 2-pi space, spreading all of its energy into just half a sphere, and therefore the level INCREASES by 6 dB.

I could sell the speakers and find a more suitable pair but I am really fond of them.

There's no reason to get rid of them; coupled with the kh805, that's a very nice setup for a small studio.

I would be happy to take the effort to flush-mount but I'm not sure that with all the other limitations I have, It would be worth it. I am certainly keeping it in mind as a future option.

If you don't do it now, it's hard to do later. That would mean re-doing the entire front end of the studio. The treatment for a studio that has the speakers on stands is different from the treatment for a studio with soffit-mounted speakers.

I am thinking now that I may experiment with different arrangements and slat-spacing

There's no need to experiment! Just do the math! Calculate what slat widths and spacing and open area you will need to deal with the problems you are seeing in your REW data. It would take you a long, long, long time to arrive at the right answer by just cutting wood at random and spacing it randomly, then cutting more wood and spacing it randomly again... You'd waste a lot of wood like that, too! It's much better to do the calculations, and make sure you are cutting each piece of wood to the correct width, and placing it correctly, with the right spacing on each side... Tune the room, using the prediction tools.

with different arrangements and slat-spacing before applying the fabric

The fabric goes on FIRST, before you put up the slats! You can't put the fabric on after the slats are already nailed in place! If your slats are tuned as Helmholtz resonators (a "slot wall"), then the fabric is part of the system, since it damps the oscillation of the air slug in the slot. But if you are not going to make them into Helmholtz resonators, then the fabric is still needed, simply to block the light from getting inside, and making the ugly details of your wall interior, visible to everyone in the room! So, in both cases you need to put the fabric up first, before the slats go on.

...experiment with different arrangements and slat-spacing before applying the fabric but assume that this will not effect the lowend.

Slot walls can be tuned quite low, if you want them to be. Or if you don't tune them, then they can still have significant low-end absorption if the underlying insulation is of the correct type and thickness, and the coverage is correct. The fabric can have a very significant effect if you do decide to make slot walls, tuned as Helmholtz resonators, for the reason I mentioned above: the fabric is part of the Helmholtz mechanism, as it damps the motion of the air slug, and removes a lot of the resonant energy. So it all depends on how you design your walls, and what techniques you use.

This is fascinating stuff! Would this situation be improved by adding absorption to the large live space? I plan to fit two large bass-traps in the far corners and the whole area under the mezzanine will be absorptive. could this help this situation?

It will make a difference, yes. But "how much difference" depends on many factors.

Also, since the area downstairs is going to be your live room, you don't want to make it dead! A "dead live room", is a contradiction... You want to play drums down there it seems, and also record other instruments, so you certainly do not want to put too much absorption down there! That would be a mistake. However, you also don't want to make it too reverberant: You are showing three solid stone walls down there, which is good, but might be overkill for that room: it's going to sound like a reverb chamber like that. I would suggest balancing the acoustics better, to be more usable for rehearsing and recording. The ceiling treatment, for example, doesn't seem to fit with the overall purpose of the studio.

I hadn't noticed this. It could be that I had not set the levels correctly between the two speakers

Then you didn't do the calibration correctly! Here's the procedure for doing it right: How to calibrate and use REW to test and tune your room acoustics

I am not as concerned about this as I am about the acoustics for now (Although I will investigate when it comes time to take more measurements).

You cannot get valid REW data for making decisions about how to tune your room, if you are not making the measurements correctly! You need to fix the problem with your speakers first, then make the measurements that you will rely on when calculating the treatment.

Another possibility that I have been a bit worried about is that the Neumann KH805 sub (Non DSP)

The 805 actually does have a DSP. The only way to bypass the DSP in the 805 is to use the right channel input ONLY, and set the rear panel channel input selector switch to "Right". If you are feeding both left and right inputs to your 805, and then going on to your KH80s as satellites, then you ARE using the DSP inside the 805. To understand this, take a look at page 29 of the manual. It is impossible to avoid using the DSP if you are using the left input...

feeds into the DSP controlled Neumann KH80's meaning that there will be some latency (and thus phase differences) between the two.

Irrelevant. There will be a much, much larger delay simply due to the difference in path length between the mains and your ears, and the sub and your ears. There's also the simple fact that you will pretty much always see a group delay issue in the low end, in any system, simply because of the way low frequency sound propagates, as compared to high frequency sound. DSP latency is maybe half a millisecond or so, probably less. Time-of-flight delays can be many times that. The phase controls are designed for that, not for DSP latency.

This can be somewhat mitigated by the sub's built in phase switches but not perfectly.

No it cannot, and that is not the purpose of those switches anyway! The phase rotation switches are there to compensate for the physical distance between the mains and the sub, and also to correct for phase issues caused by the room itself and the location of the sub and mains in the room. For example, I often use a sub to fill in the floor bounce which is inevitable in all rooms. If I have two (or more subs) then I often set them up as a plane wave bass array, to kill some of the modal issues, and I also place them in the room so that they reduce the SBIR issue from the rear wall, or the front wall, or the side walls. That is what the phase switches are used for. Not to try to "fix" the non-existent problem caused by latency in the DSPs. The phase switch is an acoustic compensation control, not a signal latency control. The phase issues caused by the physical presence of the sub in the room, and by the room itself, and by the furniture in the room, are far, far larger than any issues caused by signal latency.

How do you set REW up to compare the phase between the left and right speakers? That is very handy and would help me to test if I have it right next time.

Not really, but here's what it looks like anyway:

Astro-Mat-_REW--PH--15--21k--Speaker-Difference-Compare.png

Astro-Mat-_REW--PH--15--500--Speaker-L-and-R.png

The first graph shows the DIFFERENCE in phase between the two speakers (not considering the sub) across the entire spectrum, and the second graph shows the two individual phase plots (up to 500 Hz): Purple is the left speaker, green is the right. As you can see, it is impossible to fix that with the phase switches on the sub, since the problem occurs outside of the range that the sub covers. And even if it was in the right range, it would still be a mistake to try to correct it with the phase switches... because they apply ONLY to the sub, not to the individual speakers!

You can generate those graphs yourself using the "overlays" button at the top of the screen, then selecting "phase". But there's nothing you can do to fix the differences between the left and right channels using the phase rotation controls on the subs. And as I explained above, that would be the wrong use of the phase rotation in any case. Even if you could use it for what you want, you would then be unable to use it for the REAL purpose for which it is needed.

Hopefully all of this helps to get you in the right track!


- Stuart -

[astro_mat]

Thanks for your detailed and in depth response. You have been very kind to spend so much time on this and your comments have been very helpful.

Before responding to some of your points, I just want to pull it back a bit here. In summary, you have identified the following problems;

Problem 1) The control room is too dead
Problem 2) The control room being coupled to the live room causes a "huge mess" of an ETC
Problem 3) There is a speaker imbalance.

And this solution;

Solution a) soffit/flush mount my speakers

Clearly Soffit/flush mounting is the BEST option, but, I’m not sure which of the problems listed above it will address?
I know my room has a long way to go, but it would be useful if you could comment on some of the other aspects of the REW measurements? perhaps the frequency response...or the waterfall plot? are the dips and peaks within an acceptable range? as a beginner in this field, it is important to be able to see what a realistic aim would be for my space. Your graphs are very impressive but, given my limitations, mine will never look like yours so I am keen to know what the best is that I can shoot for?

I have the following remarks and questions:

Problem 1) The control room is too dead

Calculate what slat widths and spacing and open area you will need to deal with the problems you are seeing in your REW data.

Which problems do you mean?

It would take you a long, long, long time to arrive at the right answer by just cutting wood at random and spacing it randomly, then cutting more wood and spacing it randomly again... You'd waste a lot of wood like that, too! It's much better to do the calculations, and make sure you are cutting each piece of wood to the correct width, and placing it correctly, with the right spacing on each side... Tune the room, using the prediction tools.

Sorry for the confusion, I didn't mean that I would just "randomly cut up wood". Since there are no Helmhotz resonators (and thus, cannot use the equation) I was thinking that i would choose a suitable absorption/reflection ratio and use this to decide the slat pattern. Is this correct? What calculations are you referring to? what prediction tools?

Problem 2) The control room being coupled to the live room causes a "huge mess" of an ETC

Also, since the area downstairs is going to be your live room, you don't want to make it dead! A "dead live room", is a contradiction... You want to play drums down there it seems, and also record other instruments, so you certainly do not want to put too much absorption down there! That would be a mistake. However, you also don't want to make it too reverberant: You are showing three solid stone walls down there, which is good, but might be overkill for that room: it's going to sound like a reverb chamber like that. I would suggest balancing the acoustics better, to be more usable for rehearsing and recording. The ceiling treatment, for example, doesn't seem to fit with the overall purpose of the studio.

Quite right, I do not want to make the Live room too dead but it can certainly be "deadened" up more than it currently is. There are 4 stone walls down there. Approximately 25 - 30% of this will be covered by 100mm RW3 (The dead space under the mezzanine and the corner absorbers) Which I am hoping will tame it but not too much (as you say). This again, can be livened up by the addition of slats to adjust the ratio of absorbtion/reflection as needed. It is also worth noting that although they are solid stone, they prove to be quite diffuse due to the deep (>100mm) grooves between the stones. I would not say that it approaches the sound of a reverb chamber in it's current state.
So....,am I correct in saying at this point that by livening up the control room and adding a bit more absorption to the live room, this could (With care) balance out the ETC to something closer to the nice slope we want? What would you suggest for the ceiling treatment?
Problem 3) There is a speaker imbalance.

I hadn't noticed this. It could be that I weathad not set the levels correctly between the two speakers

Then you didn't do the calibration correctly! Here's the procedure for doing it right: How to calibrate and use REW to test and tune your room acoustics

They are very good instructions and I did try to follow them before posting. There was a lot of messing around at that stage and it is possible that an adjustment was made and not returned. I will investigate the next time I do measurements. Unfortunately, in the time since posting original, I have decided to get things moving by re-pointing the stone in the live area. This has meant that all equipment has been moved to the control room area and so measurements at this stage would be confused by a lot of equipment that was not there before.

I am not as concerned about this as I am about the acoustics for now (Although I will investigate when it comes time to take more measurements).

You cannot get valid REW data for making decisions about how to tune your room, if you are not making the measurements correctly! You need to fix the problem with your speakers first, then make the measurements that you will rely on when calculating the treatment.

Agreed, but as the situation has changed (see above) I can hopefully use the measurements I have as a guide until i can re-measure and fix the problem.
Solution a) soffit/flush mount my speakers

2) Unwanted secondary sound radiation from the monitor cabinet's edges is already dealt with pretty well by the cabinet design of the Neumann KH80 speakers.

Wellll.... Let me put it this way: there is marketing hype, then there is technical reality. I love Neumann speakers, and consider them among the best: I have used them in several studio designs. But I don't fall for their marketing hype. They are great speakers, but they cannot defy the laws of physics.

So no, despite their wonderful marketing, and the great technical design team at Neumann, the KH80 has not eliminated edge-diffraction, nor any of the other artifacts related to putting speakers in a small box. This is not about how the box is shaped, but simply because the box is too small. The ONLY way to eliminate edge diffraction completely, is to make the box large enough that the longest wavelengths that it can produce are still small when compared to the size of the speaker. And since the KH80 can produce wavelengths that are 21 feet long (6.5 meters), the front baffle of the speaker would need to be 21 feet wide, at least, in order to fully eliminate all edge diffraction. In fact, theoretically, to fully eliminate it, it would have to be an "infinite baffle", many times wider than the lowest wavelength... Which is impossible, of course.

I'm sure you're right and that i'm no expert in this matter but if the effects of edge diffraction manifest themselves in a frequency response graph, then the published frequency response graph by Neumann for the KH80 seems pretty good to me and certainly better than the ones in the Sound on Sound graphic?

KH80_response_90dB_510.gif

3) The Neumann KH80's do not lend themselves particularly well to flush mounting (Curved front bezel, amp situated at back requiring ventilation).

Neither of those is a problem. Not in the least.

I hadn't realised that it would be so straight forward to ventilate the back of the amp. I must admit that it has been a while since I researched (And decided against) sofit/flush mounting the speakers. I am certainly re-considering now.

The curved front face makes no difference either: if it is possible to soffit-mount a Genelec 8050, which is far more curved than your KH80, then it is possible to soffit-mount yours:

JKB-DNMK--Genelec-8050--Soffited-S47.jpg

That looks great. I realise that it is difficult, but not impossible to shape a baffle to the curve. Do you have more information of the project that image refers to?

4) Due to the strange angles and hard to predict room involved, I thought that it would be better to be able to re-position the speaker to find the best place rather than having it fixed in a wall.

The best place is IN the wall! And there's only one single variable that needs to be optimized for soffit-mount speakers: how far across the baffle. The height is already fixed, carved in stone. The distance into the room is also fixed: carved in stone. The only thing you can change is the location across the width of the baffle, and that is easy to calculate.

Again, getting more convinced here. Just what is "location across the width of the baffle" and how would you calculate this?

5) I was worried that the increased power needed would mean that the speakers I have would not be adequate.

??? What "increased power needed" are you talking about?!?!? You seem to be badly misinformed!

Again, my bad, it has been a while since I researched (And decided against) soffit/flush mounting the speakers. I didn't really think before writing this and of course you are right.

I would be happy to take the effort to flush-mount but I'm not sure that with all the other limitations I have, It would be worth it. I am certainly keeping it in mind as a future option.

If you don't do it now, it's hard to do later. That would mean re-doing the entire front end of the studio. The treatment for a studio that has the speakers on stands is different from the treatment for a studio with soffit-mounted speakers.

Sorry, I wasn't too clear here. What I meant by "future option is to continue with my experiments for now such as more absorption in the live end and addition of slats etc and then address whether or not to soffit/flush mount after further measurements. It would not be any harder to Soffit/flush mount because nothing would have changed at this section of the build.

Sorry if this is long, but this all very useful stuff

-Mat-

[Soundman2020]

Hi again, Mat! I'm glad to see that you are re-thinking things, and working your way through this..

Problem 1) The control room is too dead
Problem 2) The control room being coupled to the live room causes a "huge mess" of an ETC
Problem 3) There is a speaker imbalance.

And this solution;

Solution a) soffit/flush mount my speakers

Clearly Soffit/flush mounting is the BEST option, but, I’m not sure which of the problems listed above it will address?

Actually, soffit mounting won't fix any of those directly, except maybe the speaker imbalance. Indirectly, it will help with livening up the control room, but only to a small extent. The soffits are mostly solid wood, so they are reflective surfaces and will therefore add some liveliness to the room.

This gets a bit complicated, as you have to sort of switch your mental position to being a speaker inside a soffit, instead of being an engineer at the mix position. Try to imagine being a speaker mounted in a soffit. What do you see in front of you? Those are the surfaces that your sound will reach, and those are the ones that will reflect best. And you'll notice that, as a speaker in a soffit, you don't see any of your own soffit at all! You see mostly the mix position, straight ahead of you, and the side walls, ceiling and floor, and a bit of the other soffit. But you can't see your own soffit, because you are inside it, flush with the front face. So the sound you produce won't be reflected off your own soffit: it can only be reflected by the other surfaces that you can see. Why is that important? Because each speaker only "sees" half of the front end of the room! So its energy can, at best, only be reflected by half of the soffits. What it sees most is the absorptive surfaces on the ceiling, side walls, and at the rear of the room. Those are the surfaces that tune the liveness of the room, FROM THE SPEAKERS POINT OF VIEW. And since they are all designed to eliminate specular reflections, and only return a low level diffuse field, what goes out from you, as a speaker, doesn't come back. It's gone. To you AS A SPEAKER, the room looks pretty dead, except for the diffuse reverberant field.

OK, now forget that you are a speaker, and go back to being an engineer at the mix position: What do you see in front of you now? You see the entire front end of the room, which is pretty much all reflective surfaces! So as you talk, or people around you talk, everything you say will be reflected back to you by the front surfaces. So the room sounds rather live FROM YOUR POINT OF VIEW AS AN ENGINEER. In fact, its very nicely live, and you can hear yourself and people around you rather clearly. But the sound from the speakers goes right past you, and doesn't come back, except as a low level diffuse field.

The purpose of this mental exercise is to point out that the exact same room can sound very different when viewed from two different but key locations. The speaker and the engineer "see" the room very differently.

So, if the soffits don't make the room sound a lot more live as seen by the speakers, what do they do?

Here's a partial list:

1) They correct the POWER imbalance inherent in any stand-alone speaker. For high frequencies, the sound leaves the speaker mostly like rays or cones, focused directly in the direction the speaker is facing. So all of the mid and high power goes straight out toward the mix position. For low frequencies, the sound leaves the speaker like a balloon inflating around it: equally in all directions. So only HALF of the low frequency sound goes towards the mix position: the other half goes the opposite way. Thus, there is a power imbalance. If you are a speaker designer, you need to pump twice as much power into the woofer as you do into the tweeter, just to get them both at the same level at the mix position: Have you ever noticed why speakers have much larger amps for the woofer, than for the mids and tweeters? This is part of the reason why: to compensate for the power imbalance, the woofer has to work twice as hard as the tweeter. A soffit fixes that. By putting a very large front panel around the speaker, ALL of the low frequency energy is now going forward to, into the room, towards the mix position. The soffit forces that to happen. So you can turn down the power of the woofer by half, and get correctly balanced power distribution. That's why speaker manufacturers put the bass roll-off controls on the rear panel of their speakers: so you can turn down the bass amp if your speaker is radiating into half-space, instead of full space. The soffit does naturally what the speaker designer had to fake electronically. So you can turn of the faker, and let the soffit do the job, naturally.

2) They extend the low end. Because the front baffle of a typical speaker is small, it can only produce sound cleanly down to a certain frequency. When you put a much larger baffle around it, it can now produce sound cleanly at lower frequencies, which it could not reach before.

3) They eliminate SBIR from the front wall. Some of the sound that went off behind a speaker that is not soffit mounted, will hit the wall and bounce back, where it will then interfere with itself, causing a major phase cancellation at an exact frequency that is easy to calculate. The only parameter you need to know is the distance from the speaker to the wall. That creates the famous "SBIR dip" in the frequency response. A soffit eliminates that completely, because there is no sound bouncing off the front wall and coming back. The speaker is flush with the wall, so it is impossible to have any SBIR from that.

4) They eliminate comb filtering and other phase issues. The SBIR dip is also associated with a pattern of cancellations all the way up the spectrum, right to the top end. That too disappears, as do all other artifacts related to phase problems on the front wall.

5) They eliminate edge diffraction. Relate to point #1: at high frequencies, all of the sound goes forward, partly because the waves are much smaller than the size of the cabinet. At low frequencies, the waves "wrap around" behind the speaker, because they are much larger than the size of the cabinet. And at some point in between, the waves are similar in size to the cabinet, and are therefore diffracted at the edges of the cabinet, causing "hash" in the frequency and phase response. This happens at the point where the wavelength matches the width of the cabinet, and for quite a range around that. Soffit-mounting eliminates this, because the waves are ALWYAS larger than the "cabinet", at all frequencies because the soffits blend into the walls and each other: there is no "edge" to a soffit, so the waves cannot diffract around the "edge".

6) They enable design concepts such as RFZ, NER, and CID: soffit-mounting is a key part of these studio design concepts, since they use the soffits as part of the geometric shaping of the front of the room, which is a critical part of forcing the reflections past the mix position.

7) Etc. There are other reasons too, but any one of those is already plenty good enough reason to do it. All of them in combination make it a no-brainer... Soffit mounting is the single best thing that can be done to a studio, to improve the acoustic response. There's other things you can do to a room, of course, and you'll still need to do some of those, but no other single treatment has such a huge and positive effect on so many aspects at once.

I know my room has a long way to go, but it would be useful if you could comment on some of the other aspects of the REW measurements? perhaps the frequency response...or the waterfall plot? are the dips and peaks within an acceptable range?

I would love too, but I don't have the time right now to do that. Maybe next week. But basically, you can find all those answers in ITU BS-1116-3, chapters 7 and 8. The document is easy to find on-line. Quick summary: the ideal goal is to have flat frequency response, within +/- 3dB across the entire spectrum. That's very, very hard to do. Most pro studios get maybe +/- 6 dB. Most home studio builders are very happy if they can get +/- 10 dB. That's very respectable for a home studio, and a good goal to aim for. But more important is the time-domain response. The overall decay rate is calculated for the room size. Then your goal is to ensure that the decay rate for each individual 1/3 octave band is within a range of +/- 50ms of that overal number, for the range 200 Hz to 4 kHz, +/- 100 ms above 4 kHz, and rising to as much as +300 ms down to the bottom end. That's really hard to do. Most pro studios get close, but don't achieve it. Some do. Home studios are lucky to get +/- 100 ms across the mid range, and rising to maybe 600 ms or more in the low end. There's more in that document, but those are the two key points: flat frequency response, and flat time-domain response.

Your graphs are very impressive but, given my limitations, mine will never look like yours so I am keen to know what the best is that I can shoot for?

See above: +/-10 dB frequency response is good for a home studio. +/- 100 ms across the spectrum is also good.

Which problems do you mean?

On your waterfall plot, the tall spikes that die out slowly are resonances. They mess with the room response, both frequency and time domain. You can likely tune some of your slats to deal with those, reducing the intensity and length of the resonance. You can fix others with absorption in the right pace in the room, at the right frequency.

Since there are no Helmhotz resonators (and thus, cannot use the equation)

Why not? Is there a reason why you don't want to tune your slot wall? There's some good things you can do with tuned slot walls. It's not that hard to do, if you aim for broad-band.

I was thinking that i would choose a suitable absorption/reflection ratio

You could, yes, but that would also tune the Helmholtz resonance of your slots... "Percentage open area" is one way of looking at the tuning, if you are tuning it broad-band. Of course, if you prefer to go for a low open area, then you can tune each slot individually, rather than tuning the wall as a whole.

What calculations are you referring to? what prediction tools?

Here's a good one that I use quite often: http://www.acousticmodelling.com/ Especially the multi-layer one. It's pretty good. There are others, but that's a good start. And it's FREE! REW also has a built-in room simulator, for example.

Quite right, I do not want to make the Live room too dead but it can certainly be "deadened" up more than it currently is. There are 4 stone walls down there. Approximately 25 - 30% of this will be covered by 100mm RW3 (The dead space under the mezzanine and the corner absorbers) Which I am hoping will tame it but not too much (as you say). This again, can be livened up by the addition of slats to adjust the ratio of absorbtion/reflection as needed.

Have you considered making variable acoustic devices, so you can change the sound and response of the room as needed? It's not that hard to do.

So....,am I correct in saying at this point that by livening up the control room and adding a bit more absorption to the live room, this could (With care) balance out the ETC to something closer to the nice slope we want? What would you suggest for the ceiling treatment?

Yur two rooms are directly coupled, since the rear end of the control room is fully open to the live room. So there will ALWAYS be two distinct decay rates on your Schroeder integral: one for the initial decay in the CR, followed by the lower one for the LR. That's not a problem, really. But what is a proboem is the uneven bumps and dips on both parts of the curve. That needs to be evened out.

I'm sure you're right and that i'm no expert in this matter but if the effects of edge diffraction manifest themselves in a frequency response graph, then the published frequency response graph by Neumann for the KH80 seems pretty good to me and certainly better than the ones in the Sound on Sound graphic?

The graph you show is how the speaker performs in an anechoic chamber: that's a special room, very large, that is very carefully treated to produce no reflections at all. Zero. Nothing. It is deader than dead. Basically the room is not there: it is invisible to the speaker. The sound leaves the speaker going in any direction, and never comes back. Not even a little bit. That's the only real way that you CAN test a speaker, because if you test it in any other type of room, then the room interacts with the speaker, and you end up getting a graph that shows how both react with each other! So that graphs shows ONLY how the speaker reacts when there is no room around it. Not even a floor: the speaker sits on a sand that sits on a thin wire mesh suspended over several meters of empty space. So the graph shows the true response if the speaker, all by itself.

But when you put the speaker in a room, that graph goes out the window. The room dominates big-time over the speaker. The acoustic loading and modal response of the room shapes how the speaker will act, and it is rather different. The steady-state speaker response in a typical normal room, is very different form the steady state speaker response in an anechoic chamber. You can't compare the two, unfortunately. There are other measurements that you can do in the anechoic chamber to predict how the speaker will react in a room, but the speaker manufactures usually don't do those tests (they cost money!), or they don't publish the results, because they always look ugly, and the vast majority of speaker buyers wouldn0't even know how to interpret them. So there's no point in publishing them. Thus, it's hard to figure out in advance how a speaker and a room will react together: the best you can do is to just test it, with REW or something similar. The SOS article shows the predicted response for the speaker in a typical room.

I hadn't realised that it would be so straight forward to ventilate the back of the amp. I must admit that it has been a while since I researched (And decided against) sofit/flush mounting the speakers. I am certainly re-considering now

Here's how I normally do it:

soffited.speaker-with-wire-mesh-chimney---SAFE--01-SML-ENH.jpg

The wire mesh "chimney" provides a clear path up the rear, and prevents the insulation from blocking that path. In this photo the insulation isn't there yet, but normally there's insulation around the speaker enclosure box. There's a large slot in the shelf that the box is sitting on, and another slot in the top panel above the speaker soffit, or in the front baffle. That provides a clear path for air to flow past the rear of the speaker, by convection.

That looks great. I realise that it is difficult, but not impossible to shape a baffle to the curve. Do you have more information of the project that image refers to?

Unfortunately, I can't give you more than that. That comes from a project I designed for one of my clients. I'm sure you can understand that I can't share all the details of a client's studio.

Again, getting more convinced here. Just what is "location across the width of the baffle" and how would you calculate this?

That just means where on the baffle you cut the hole for the speaker to poke through. It's not a good idea to center it in the middle of the baffle, as that can create unwanted frequency response issues, so it should be offset from the center, towards one side or the other. A good spot, theoretically, is 3/5 of the soffit panel width. So if the panel is 5 feet wide, for example, then cutting the hole so it is centered 3 feet from one edge and 2 feet from the other edge, is a good spot.

Sorry, I wasn't too clear here. What I meant by "future option is to continue with my experiments for now such as more absorption in the live end and addition of slats etc and then address whether or not to soffit/flush mount after further measurements. It would not be any harder to Soffit/flush mount because nothing would have changed at this section of the build.

There's a different approach to treating a room that has soffit-mounted speakers. If you have already treated most of the room, then decide to soffit mount, you would have to change some/all of that treatment you already did. If you soffit mount, you would normally then use the ful RFZ concept for the design of the rest of the room, so that's the way you would treat it. But if you don't soffit mount, then that would not be an option, as RFZ only really works if you do soffit mount. So you'd need to do a different form of treatment.


- Stuart -

[astro_mat]

Wow! this is all great, your explanations have really helped to clear things up for me. Sorry about the delay in replying. Life etc....

I know my room has a long way to go, but it would be useful if you could comment on some of the other aspects of the REW measurements? perhaps the frequency response...or the waterfall plot? are the dips and peaks within an acceptable range?

I would love too, but I don't have the time right now to do that. Maybe next week. But basically, you can find all those answers in ITU BS-1116-3, chapters 7 and 8. The document is easy to find on-line. Quick summary: the ideal goal is to have flat frequency response, within +/- 3dB across the entire spectrum. That's very, very hard to do. Most pro studios get maybe +/- 6 dB. Most home studio builders are very happy if they can get +/- 10 dB. That's very respectable for a home studio, and a good goal to aim for. But more important is the time-domain response. The overall decay rate is calculated for the room size. Then your goal is to ensure that the decay rate for each individual 1/3 octave band is within a range of +/- 50ms of that overal number, for the range 200 Hz to 4 kHz, +/- 100 ms above 4 kHz, and rising to as much as +300 ms down to the bottom end. That's really hard to do. Most pro studios get close, but don't achieve it. Some do. Home studios are lucky to get +/- 100 ms across the mid range, and rising to maybe 600 ms or more in the low end. There's more in that document, but those are the two key points: flat frequency response, and flat time-domain response.

Your graphs are very impressive but, given my limitations, mine will never look like yours so I am keen to know what the best is that I can shoot for?

See above: +/-10 dB frequency response is good for a home studio. +/- 100 ms across the spectrum is also good.

Which problems do you mean?

On your waterfall plot, the tall spikes that die out slowly are resonances. They mess with the room response, both frequency and time domain. You can likely tune some of your slats to deal with those, reducing the intensity and length of the resonance. You can fix others with absorption in the right pace in the room, at the right frequency.

Thanks for the link to the document. This is all at the root of my original question. It is hard for me to determine from the very technical nature of that specification document, My response seems to fit what can be considered a good room at the moment? If "Measured in 1/3 octave bands" means 1/3 octave smoothing, it is within a range of 7.6dB (+/- 3.8dB)? And if taken within the limits of Figure 2, it seems to be +/- 2.3dB between 250 an 2000Hz? This would be further improved by using the equalisation within the KH80. I realise that it's not all down to the frequency response but I guess I am looking for some positives at this stage. Can you give me any?

Since there are no Helmhotz resonators (and thus, cannot use the equation)

Why not? Is there a reason why you don't want to tune your slot wall? There's some good things you can do with tuned slot walls. It's not that hard to do, if you aim for broad-band.

I was thinking that i would choose a suitable absorption/reflection ratio

You could, yes, but that would also tune the Helmholtz resonance of your slots... "Percentage open area" is one way of looking at the tuning, if you are tuning it broad-band. Of course, if you prefer to go for a low open area, then you can tune each slot individually, rather than tuning the wall as a whole.

As I understand, I cannot tune the wall. There is only 50mm depth of rockwool before a non-solid, non-airtight space. Do you mean that I can use the helmholtz equation in this case? How do I "aim for broadband"? Could you please expand on this?

What calculations are you referring to? what prediction tools?

Here's a good one that I use quite often: http://www.acousticmodelling.com/ Especially the multi-layer one. It's pretty good. There are others, but that's a good start. And it's FREE! REW also has a built-in room simulator, for example.

I have used the pourous absorber calculator but had never notice the multi layer one! I will definitely have a play with it.

Quite right, I do not want to make the Live room too dead but it can certainly be "deadened" up more than it currently is. There are 4 stone walls down there. Approximately 25 - 30% of this will be covered by 100mm RW3 (The dead space under the mezzanine and the corner absorbers) Which I am hoping will tame it but not too much (as you say). This again, can be livened up by the addition of slats to adjust the ratio of absorbtion/reflection as needed.

Have you considered making variable acoustic devices, so you can change the sound and response of the room as needed? It's not that hard to do.

I have thought of variable acoustics but did not want to disturb the look of the stone walls. Perhaps I could use heavy drapes on the walls. Another option is to separate the acoustic spaces with an absorbent wall (non-solid, where the banister is). This could either be a full wall or perhaps just making the aperture between the two spaces smaller by adding absorbers to the railing? Or perhaps a simple heavy drape a his point?

So....,am I correct in saying at this point that by livening up the control room and adding a bit more absorption to the live room, this could (With care) balance out the ETC to something closer to the nice slope we want? What would you suggest for the ceiling treatment?

Yur two rooms are directly coupled, since the rear end of the control room is fully open to the live room. So there will ALWAYS be two distinct decay rates on your Schroeder integral: one for the initial decay in the CR, followed by the lower one for the LR. That's not a problem, really. But what is a proboem is the uneven bumps and dips on both parts of the curve. That needs to be evened out.

But how would I even this out?

I'm sure you're right and that i'm no expert in this matter but if the effects of edge diffraction manifest themselves in a frequency response graph, then the published frequency response graph by Neumann for the KH80 seems pretty good to me and certainly better than the ones in the Sound on Sound graphic?

The graph you show is how the speaker performs in an anechoic chamber: that's a special room, very large, that is very carefully treated to produce no reflections at all. Zero. Nothing. It is deader than dead. Basically the room is not there: it is invisible to the speaker. The sound leaves the speaker going in any direction, and never comes back. Not even a little bit. That's the only real way that you CAN test a speaker, because if you test it in any other type of room, then the room interacts with the speaker, and you end up getting a graph that shows how both react with each other! So that graphs shows ONLY how the speaker reacts when there is no room around it. Not even a floor: the speaker sits on a sand that sits on a thin wire mesh suspended over several meters of empty space. So the graph shows the true response if the speaker, all by itself.

But when you put the speaker in a room, that graph goes out the window. The room dominates big-time over the speaker. The acoustic loading and modal response of the room shapes how the speaker will act, and it is rather different. The steady-state speaker response in a typical normal room, is very different form the steady state speaker response in an anechoic chamber. You can't compare the two, unfortunately. There are other measurements that you can do in the anechoic chamber to predict how the speaker will react in a room, but the speaker manufactures usually don't do those tests (they cost money!), or they don't publish the results, because they always look ugly, and the vast majority of speaker buyers wouldn0't even know how to interpret them. So there's no point in publishing them. Thus, it's hard to figure out in advance how a speaker and a room will react together: the best you can do is to just test it, with REW or something similar. The SOS article shows the predicted response for the speaker in a typical room.

Understood.

I hadn't realised that it would be so straight forward to ventilate the back of the amp. I must admit that it has been a while since I researched (And decided against) sofit/flush mounting the speakers. I am certainly re-considering now

Here's how I normally do it:

soffited.speaker-with-wire-mesh-chimney---SAFE--01-SML-ENH.jpg

The wire mesh "chimney" provides a clear path up the rear, and prevents the insulation from blocking that path. In this photo the insulation isn't there yet, but normally there's insulation around the speaker enclosure box. There's a large slot in the shelf that the box is sitting on, and another slot in the top panel above the speaker soffit, or in the front baffle. That provides a clear path for air to flow past the rear of the speaker, by convection.

Thats a great example. Thanks!

That looks great. I realise that it is difficult, but not impossible to shape a baffle to the curve. Do you have more information of the project that image refers to?

Unfortunately, I can't give you more than that. That comes from a project I designed for one of my clients. I'm sure you can understand that I can't share all the details of a client's studio.

Understood, I had to ask

Again, getting more convinced here. Just what is "location across the width of the baffle" and how would you calculate this?

That just means where on the baffle you cut the hole for the speaker to poke through. It's not a good idea to center it in the middle of the baffle, as that can create unwanted frequency response issues, so it should be offset from the center, towards one side or the other. A good spot, theoretically, is 3/5 of the soffit panel width. So if the panel is 5 feet wide, for example, then cutting the hole so it is centered 3 feet from one edge and 2 feet from the other edge, is a good spot.

I was worried that there would not be enough space to build big enough baffle?

Sorry, I wasn't too clear here. What I meant by "future option is to continue with my experiments for now such as more absorption in the live end and addition of slats etc and then address whether or not to soffit/flush mount after further measurements. It would not be any harder to Soffit/flush mount because nothing would have changed at this section of the build.

There's a different approach to treating a room that has soffit-mounted speakers. If you have already treated most of the room, then decide to soffit mount, you would have to change some/all of that treatment you already did. If you soffit mount, you would normally then use the ful RFZ concept for the design of the rest of the room, so that's the way you would treat it. But if you don't soffit mount, then that would not be an option, as RFZ only really works if you do soffit mount. So you'd need to do a different form of treatment.

Also worried that there is not enough space for full rfz?

[DanDan]

Matt asked me to give the thread a push to see what/who else comes in.
Some checking.....
Is the .mdat current?
What Mic? Pointed which direction?
Are the positions of the speakers and listener and sub optimised?
What is that huge 20Hz thing? Is that the sub or is it a motorway nearby?
How does it sound? Can you hear the reverb of the room below and is this 'nice' or no?

Some observations....
Lots of 50mm could be dodgy, I have made that mistake myself. But roofs can be like Limp Bags to LF so your absorption spectrum may actually be pretty balanced.
Your Waterfalls do suggest that.
Obviously it is very dead. That wouldn't bother me personally. Lots of lath would be lovely, but hard to predict the spectrum. Light wooden laths can suck out even more warmth. Outside of the listening Zone Withour Early Reflections, one could use very wide laths. There is an option or an additional possibility. A pair of rear/side speakers and a Lexicon...
This could provide a 'bridge' to somewhat linearise the Schroeder slope. Or just controlled liveness, envelopment.
NorthSlutz member jim1961 has a wonderful long thread called something like 'My Listening Room'.
He has a LOT of broad spectrum absorption. He tried all sorts of Kickers and Diffusors but eventually added a quite affordable Lex.
Such treatment usually results in an almost anechoic flat response from good speakers. This sounds unnaturally bright compared to normal speaker in normal rooms. This needs to be tailored towards a Target such as the Bruel and Kjaer Curve.
There is a fair amount of wobbling in the mid range. Take a look at the first 30mS of the ETC.
A table for keyboard, mouse or, or mixer, might block some of the floor reflection.
Those speakers sound good to me. A very competitive price and performance zone. Sub has the advantage of optimum LF location.
BUT, the KH310 is a vastly superior speaker. Effortless and with no reflex port noise down to below 30Hz. A dome mid which moves you into a different class of speaker, real 3 way. Waveguides and a sweet tweeter. Even at twice the price I don't think you will find anything equal or better.
BTW they are made here in Ireland.
DD

[astro_mat]

Thanks for commenting.

Is the .mdat current?

It is current to when the pictures were taken. In the time since, I have made some changes while await what to do with the control room (Roof). I have started chipping away the old render in the live area and will start to repoint. I guess this would add more diffusion to the area? To do this, I have moved all equipment upstairs so it is fairly crowded up there. I have not done any recent measurements but I may do this out of interest and to perhaps fix the speaker imbalance pointed out by stuart.

What Mic? Pointed which direction?

It is (wait for it.......) a peavey pvr-1 omnidirectional electret. Not a usual choice but it seems like the correct mic to use and I already had it. The direction is as described in SOUNDMAN-2020 PROCEDURE FOR CALIBRATING AND USING R.E.W. "facing straight forwards and slanted a bit upwards, angled around 60° "

Are the positions of the speakers and listener and sub optimised?

As far as I can tell and within my limitations, the position is optimised. This position seems to yield best results.

What is that huge 20Hz thing? Is that the sub or is it a motorway nearby?

I think that is caused by the floor vibrating that the sub is sat on. This must be it's resonant frequency. I have yet to fill in all the joists with fibre-glass (As described in the 1st post here) but I am guessing that this is something I may have to just live with?

How does it sound? Can you hear the reverb of the room below and is this 'nice' or no?

It sounds great to me. I cannot detect any of the reverb from the room below.

Lots of 50mm could be dodgy, I have made that mistake myself. But roofs can be like Limp Bags to LF so your absorption spectrum may actually be pretty balanced.
Your Waterfalls do suggest that.

This is what I was thinking, are the waterfalls OK?

Obviously it is very dead. That wouldn't bother me personally. Lots of lath would be lovely, but hard to predict the spectrum. Light wooden laths can suck out even more warmth. Outside of the listening Zone Withour Early Reflections, one could use very wide laths.

What dimension would be very wide laths be?

There is an option or an additional possibility. A pair of rear/side speakers and a Lexicon...
This could provide a 'bridge' to somewhat linearise the Schroeder slope. Or just controlled liveness, envelopment.

Very interesting, I do have a Lexicon MPX1 and I plan to have PA speakers in the live room on the back wall anyway but I would much rather get it right without this.

NorthSlutz member jim1961 has a wonderful long thread called something like 'My Listening Room'.
He has a LOT of broad spectrum absorption. He tried all sorts of Kickers and Diffusors but eventually added a quite affordable Lex.
Such treatment usually results in an almost anechoic flat response from good speakers. This sounds unnaturally bright compared to normal speaker in normal rooms. This needs to be tailored towards a Target such as the Bruel and Kjaer Curve.

That is what I found and have read your views on the Bruel and Kjaer Curve with interest. The DSP in the KH80 does allow for quite good manual EQ adjustment which I have played around with.

There is a fair amount of wobbling in the mid range. Take a look at the first 30mS of the ETC.
A table for keyboard, mouse or, or mixer, might block some of the floor reflection.

Do you think this is floor bounce? I have a table up there now so this can be tested.

Those speakers sound good to me. A very competitive price and performance zone. Sub has the advantage of optimum LF location.
BUT, the KH310 is a vastly superior speaker. Effortless and with no reflex port noise down to below 30Hz. A dome mid which moves you into a different class of speaker, real 3 way. Waveguides and a sweet tweeter. Even at twice the price I don't think you will find anything equal or better.
BTW they are made here in Ireland.
DD

Believe me, I would LOVE to go for those speakers. Are you suggesting to use them without the sub? would changing to these speakers improve matters ie: room response etc.

Many thanks for putting so much thought into this.

Mat

[DanDan]

For inspiration regarding laths this is my go to. Rod Gervias design https://thebunkerstudio.com
I would probably have little or no Lath on the ceilings or walls in the ZWER, Zone Without Early Reflections.
But behind you, why not go for plenty of lovely looking lath. If you are not recording back there, or doing much listening, you could perhaps liven things up a bit by going for sheets of wood or plasterboard rather than lath. On the ceiling at least. I'd like to hear what Stuart thinks of that notion.
That Mic has a pretty smooth looking FR, but they don't say was this on axis. I presume it is. In any case, without a Cal file I wouldn't trust the HF absolutely. Go by ear.
I thought you were using KH120, although they did look a little small in the pics.
The DSP Manual Eq is a great facility to have. BUT, I Master here, totally ITB, (you may have noticed Greenshine almost permanently on the Radio here!) and I very rarely settle on Linear Phase Eq.
Could be paranoia about side effects, but despite the added seeming transparency I just always shy away from it.
From a quick read I gather there is a Minimum Phase option on your KH80's. I would try that.
That aside, I have no doubt that the 310 is a vastly superior speaker. Which you could put under DSP control via Dirac Live.
3 Way with a Dome Mid and Sealed LF is just all grown up, adult.
I don't have a sub, but most engineers who visit think I must have one hidden somewhere.
In your bigger and probably deader room, it would be worth trying the KH310s full range, no Hi Pass, PLUS your sub.
I don't have much personal experience of Subs, but I would point out that placing them in a null is a thing, say at half height or even under seat! or your work surface...... an open mind anyway.
Their polar pattern is controlled by waveguides, but I can't imagine them having any influence on room tone.
Indeed, what room tone?
In rooms such as this I think it best to go for anechoic in the first 20-30mS.
A great pal, Mastering Engineer Ruairi O Flaherty, now in the USA, coined the phrase..... Big Headphones.

If you suspect floor bounce, or any other reflection, move the mic a couple of feet towards the mirror spot. If you are correct the reflection will move to earlier in the ETC graph.
DD

[astro_mat]

For inspiration regarding laths this is my go to. Rod Gervias design https://thebunkerstudio.com

You may have guessed that Rod's treatments were a major influence on the design of my room. I have been studying the construction photos of that studio, as well as others, to try to see just what is going on. I have noticed some interesting features which I could perhaps bring up in another post? It is a shame that Rod does not go into more detail about his "signature" lath treatments but, as he has mentioned in a gearslutz post, he is not in the business of selling treatments.

I would probably have little or no Lath on the ceilings or walls in the ZWER, Zone Without Early Reflections.
But behind you, why not go for plenty of lovely looking lath. If you are not recording back there, or doing much listening, you could perhaps liven things up a bit by going for sheets of wood or plasterboard rather than lath. On the ceiling at least. I'd like to hear what Stuart thinks of that notion.

Agreed about the ZWER and this is what I have done in the design. When you say “behind you” do you mean in the live area or in the area just behind the listening zone (before the end of the mezzanine). Are you suggesting the sheets in order to maximise the increased reflection?

That Mic has a pretty smooth looking FR, but they don't say was this on axis. I presume it is. In any case, without a Cal file I wouldn't trust the HF absolutely. Go by ear.

Will do. I may upgrade the mic later but it will do for now.

I thought you were using KH120, although they did look a little small in the pics.

Thats ok. Would you say that the KH120 would be a better choice?

The DSP Manual Eq is a great facility to have. BUT, I Master here, totally ITB, (you may have noticed Greenshine almost permanently on the Radio here!) and I very rarely settle on Linear Phase Eq.
Could be paranoia about side effects, but despite the added seeming transparency I just always shy away from it.
From a quick read I gather there is a Minimum Phase option on your KH80's. I would try that.

I didn’t realise that. I will try it.

That aside, I have no doubt that the 310 is a vastly superior speaker. Which you could put under DSP control via Dirac Live.
3 Way with a Dome Mid and Sealed LF is just all grown up, adult.
I don't have a sub, but most engineers who visit think I must have one hidden somewhere.
In your bigger and probably deader room, it would be worth trying the KH310s full range, no Hi Pass, PLUS your sub.

I would love to go for this option but I am afraid that it will have to wait until the studio makes me some money.

I don't have much personal experience of Subs, but I would point out that placing them in a null is a thing, say at half height or even under seat! or your work surface...... an open mind anyway.
Their polar pattern is controlled by waveguides, but I can't imagine them having any influence on room tone.
Indeed, what room tone?

What would placing them in a null do?

In rooms such as this I think it best to go for anechoic in the first 20-30mS.
A great pal, Mastering Engineer Ruairi O Flaherty, now in the USA, coined the phrase..... Big Headphones.

Good suggestion.

If you suspect floor bounce, or any other reflection, move the mic a couple of feet towards the mirror spot. If you are correct the reflection will move to earlier in the ETC graph.
DD

I will definitely investigate with this when the next phase of measurements comes along.

A few questions at this point;

What do you think about soundman2020’s suggestion of full flush mounting? For some reason, I am finding it hard to get behind. I would value your opinion.
What would be your opinion about the reverb tail coming from the live room into the control room? Is it something to worry about? Could acoustic curtains across the mezzanine be a solution? Do I need to close off the mezzanine opening?

Many thanks

Mat