Mixing & mastering room with vaulted ceilings

[maxlor]

Hi everybody and thanks for all the great info on this site!

I just moved into a new house and I'm trying to decide the best way to acoustically treat my mixing / mastering room. I am renting the place so I can't make any big structural changes. My budget is between $2000 - $3000. Loudness is not an issue as there are no neighbors nearby. I listen at a pretty consistent 80 to 85 db (c-weighted, slow).

MDAT for untreated room

MDAT with mineral wool panels in corners and at first reflection points (not ceiling)

Room Dimensions:
2.87m W, 5.97m L, 3.2m H at peak, 2.49m H at speakers, 2.7m H At LP, 2.7m at the back of the room.

There is a door in the center of the back wall leading to a utility room. There is a door in the rear left wall of the room (approx 40 cm from the back wall) leading to the rest of the house. There are two air conditioning vents on the left wall about 2.8m high. There is a sketch of the side walls attached (not to scale).

The floor is ceramic tile on a concrete slab. The walls and ceiling are drywall. The ceiling is vaulted. The peak of the vaulted ceiling is closer to the back of the room. Apologies for the lack of construction details as I did not build the place. I can find out more from the landlord if needed.

I've done some preliminary testing / speaker / LP positioning using pink noise / etc. But I'm happy to change these if you have any better ideas. Here's what I landed on:

Mic / Listening position: 179cm from front wall, 146cm from left side, 115.5cm high.

Speakers 48cm back from front wall. 59.5cm from each side wall. Tweeter at ear height 115.5cm

Subs sounded best just outside the speakers, equidistant from the mic. This filled in the 100 hz area nicely.

What is the best way for me to treat this room?

In the past I've always just hit all the reflection points and corners with 2" 4" or 6" mineral wool. But this room is bigger (longer / taller) than any I've treated before. So I'm curious whether I should be using pink insulation, reflective treatments, etc.

Any help is much appreciated!

Thanks
Max

[Soundman2020]

Hi Max, and Welcome!

I am renting the place so I can't make any big structural changes.

OK, but I guess you can still hang things on the walls, right? I've heard of some places that don't even allow that.... Hopefully that's not the case with you.

Room Dimensions:
2.87m W, 5.97m L, 3.2m H at peak, 2.49m H at speakers, 2.7m H At LP, 2.7m at the back of the room.

Complicated shape! Especially for a mastering room. The floor area is reasonable, and the ceiling heights are not bad, but the shape is complex.

Mic / Listening position: 179cm from front wall, 146cm from left side, 115.5cm high.

Perhaps try a little further back: Maybe around 190-210. For each listening position, rotate the speakers to point at the correct new aim point (more on that below).

Speakers 48cm back from front wall. 59.5cm from each side wall. Tweeter at ear height 115.5cm

Try raising your speakers a little: they are probably a little too low "standard" height is 120cm, and that's for the acoustic axis, not the tweeter. The axis will be below the tweeter, not commonly at the tweeter. I generally prefer to have the speakers a little higher that that anyway, for other reasons, such as desk reflections.

You could probably spread them a little further apart (closer to the side walls): try somewhere around 62-65, and once again, with each change rotate them to point at the correct aim spot.

About that aim spot: Forget what you have heard about the infamous "equilateral triangle": I could give you a long explanation about why that isn't the best solution, but for now, think of this: The best sound from your speakers is on-axis, obviously: that's what they are designed for, and that's where they are measured: to be heard on-axis. But if you set them up in the equilateral triangle, then they will be aiming at your EYES, not your ears! In order to get them aimed where they should be, you need to rotate them a little, so that they are both pointing at a spot about 30-60cm behind your head. If you do that, then axis of each speaker will be just grazing the tip of your ear pinna, or even better, just a bit beyond the pinna, and that's where you want it. So each time you move the speakers or the mix position, rotate the speakers so that they are pointing at that spot behind your head, and not AT your head.

Subs sounded best just outside the speakers, equidistant from the mic. This filled in the 100 hz area nicely.

Right, but that doesn't seem to be the best spot: from the REW data, it looks like there's some interaction going on with walls, mains, and between them. I would suggest that you try the "plane wave" setup, and see if that helps.

What are your subs and mains? (brand, model) How do you have the cross-over set? There's a big difference between the empty room and the treated room: the overall level dropped by about 1 dB, but the mid range dropped by about 3 dB. That's hard to explain for just the treatment that you did. It seems that something else changed as well, which makes comparisons rather hard...

Also, are you using some type of bass management system, or low shelf EQ? The reason I ask is because there's practically no difference with and without subs, in terms of frequency response:

MAXLOR--REW--FR--18..1k--L-vs-LS.png

Blue is left speaker alone, green is left plus sub. The level goes up just a little (about 3 dB), yes, but the shape of the curve hardly changes at all. The -3 dB point only move from about 32.9 Hz to 30.2 Hz. Not much of a change, for having added subs to the system. So I suspect that either the subs don't do much at all (the don't go low enough), or there's some type of EQ involved here that is rolling off the low end.

What is the best way for me to treat this room?

There's some modal stuff going on in the low end (especially 32 Hz, which seems to be your primary axial length mode), so I would suggest deep bass trapping. Your overall decay times are very long: around 360 ms, which is too much for that room. It should be around half that: maybe 180 to 210 ms. So, you will need some general absorption spread around evenly to get there, but taking care not to take out too much in the 200-300 Hz range, which is a common issue with small rooms. Most of the overshoot is in the high mids and low part of the highs (about 1k to 6k), so I would start by targeting that region exclusively, then see how it looks, and figure out what to do next.

In the past I've always just hit all the reflection points and corners with 2" 4" or 6" mineral wool. But this room is bigger (longer / taller) than any I've treated before. So I'm curious whether I should be using pink insulation, reflective treatments, etc.

In general, the room isn't too bad, especially considering the strange shape, but there are many issues that do need dealing with. Diffusion would help, and your room is long enough to be able to do that, so you could consider putting deep absorption across the entire rear wall, then adding large diffusers just in front of that, tuned to that 1k-6k range. It's going to be a challenge, because of the shape of the room, and also because you want it for mastering, which generally means a higher standard than just tracking/mixing.

- Stuart -

[maxlor]

Thanks so much for the quick and informative response, Stuart!

I am renting the place so I can't make any big structural changes.

OK, but I guess you can still hang things on the walls, right? I've heard of some places that don't even allow that.... Hopefully that's not the case with you.

Yes, the owner actually designed the room to track drums in it, so he's familiar with acoustic treatment and OK with me hanging things to my heart's content. However the treatment I've used so far in this room has been just stacked against the wall (I haven't made any holes yet).

Mic / Listening position: 179cm from front wall, 146cm from left side, 115.5cm high.

Perhaps try a little further back: Maybe around 190-210. For each listening position, rotate the speakers to point at the correct new aim point (more on that below).

Speakers 48cm back from front wall. 59.5cm from each side wall. Tweeter at ear height 115.5cm

Try raising your speakers a little: they are probably a little too low "standard" height is 120cm, and that's for the acoustic axis, not the tweeter. The axis will be below the tweeter, not commonly at the tweeter. I generally prefer to have the speakers a little higher that that anyway, for other reasons, such as desk reflections.

In my last place I had them a bit higher where my ear was between the tweeter and the midrange driver. I'll move them up!

You could probably spread them a little further apart (closer to the side walls): try somewhere around 62-65, and once again, with each change rotate them to point at the correct aim spot.

About that aim spot: Forget what you have heard about the infamous "equilateral triangle": I could give you a long explanation about why that isn't the best solution, but for now, think of this: The best sound from your speakers is on-axis, obviously: that's what they are designed for, and that's where they are measured: to be heard on-axis. But if you set them up in the equilateral triangle, then they will be aiming at your EYES, not your ears! In order to get them aimed where they should be, you need to rotate them a little, so that they are both pointing at a spot about 30-60cm behind your head. If you do that, then axis of each speaker will be just grazing the tip of your ear pinna, or even better, just a bit beyond the pinna, and that's where you want it. So each time you move the speakers or the mix position, rotate the speakers so that they are pointing at that spot behind your head, and not AT your head.

Ah yes I agree with this. Currently speakers are about 171cm apart and I'm sitting about 122cm from the speaker plane. I will try moving them closer to the wall as you suggested. I've read that it is not advisable to place speakers equidistant from the front wall and side wall. I fear that if I move them wider they will be in this range. Is this something I shouldn't worry about?

Subs sounded best just outside the speakers, equidistant from the mic. This filled in the 100 hz area nicely.

Right, but that doesn't seem to be the best spot: from the REW data, it looks like there's some interaction going on with walls, mains, and between them. I would suggest that you try the "plane wave" setup, and see if that helps.

What are your subs and mains? (brand, model) How do you have the cross-over set? There's a big difference between the empty room and the treated room: the overall level dropped by about 1 dB, but the mid range dropped by about 3 dB. That's hard to explain for just the treatment that you did. It seems that something else changed as well, which makes comparisons rather hard...

Also, are you using some type of bass management system, or low shelf EQ? The reason I ask is because there's practically no difference with and without subs, in terms of frequency response:

MAXLOR--REW--FR--18..1k--L-vs-LS.png

Blue is left speaker alone, green is left plus sub. The level goes up just a little (about 3 dB), yes, but the shape of the curve hardly changes at all. The -3 dB point only move from about 32.9 Hz to 30.2 Hz. Not much of a change, for having added subs to the system. So I suspect that either the subs don't do much at all (the don't go low enough), or there's some type of EQ involved here that is rolling off the low end.

I'm not sure why the mid-range dropped so much. I was very careful not to move the speakers or mic while adding the treatment, but of course it's possible I may have bumped something.

The speakers are Adam S3x-V and the subs are NHT b-20 which have a built in crossover unit. The subwoofer LP is set at 95hz I believe. There is no HP applied to the mains.

I set up the subwoofer position / crossover with the treatment in place. So if you look at the "Treated" L vs L-S you will see that it filled in a hole in the 100hz area. Because the speakers are almost flush with the front wall, they are very full sounding already. So the subwoofers aren't needed to boost the low end much, I was just hoping they would activate a different part of the room and fill in some holes.

I did a little reading about the "Plane Wave" setup, but I'm not sure I understand it completely. Does this mean I should orient both the subwoofers on the same plane, parallel with the front wall?

What is the best way for me to treat this room?

There's some modal stuff going on in the low end (especially 32 Hz, which seems to be your primary axial length mode), so I would suggest deep bass trapping. Your overall decay times are very long: around 360 ms, which is too much for that room. It should be around half that: maybe 180 to 210 ms. So, you will need some general absorption spread around evenly to get there, but taking care not to take out too much in the 200-300 Hz range, which is a common issue with small rooms. Most of the overshoot is in the high mids and low part of the highs (about 1k to 6k), so I would start by targeting that region exclusively, then see how it looks, and figure out what to do next.

Thanks, this is great info. Do you have any recommendations on how to spread absorption evenly without taking out too much 200 hz? I noticed in my treated vs non-treated graph 200 hz already took quite a hit.

I'll move the speakers / LP / Sub around a bit and make some more measurements ASAP.

[Soundman2020]

In my last place I had them a bit higher where my ear was between the tweeter and the midrange driver. I'll move them up!

For this, and for everything else you do, run a REW test after each change, and make notes in it about what you did just before the test. That will give you a complete record of what you tried and what the results were. For example, for choosing the height, do a series of tests in small increments of maybe 1cm at a time, and see what that does. Then you can cycle through those incremental tests and spot trends fairly easily. It's actually very interesting to do that, and see what a difference such small changes can make. Ditto when you move the speakers and mix position: Do that in small steps of maybe 5 to 10 cm, and do a test at each position, then cycle through them quickly to spot trends.

I will try moving them closer to the wall as you suggested. I've read that it is not advisable to place speakers equidistant from the front wall and side wall. I fear that if I move them wider they will be in this range. Is this something I shouldn't worry about?

As with most generalities, it's not always true. Sometimes you might actually WANT to do that, to get a phase cancellation at a particular modal frequency. If the speakers are in a null for a mode, then they won't be able to excite it, and in a square room, for example (not your case, but hypothetically) it might be good to put them in the null in both directions. Or perhaps you get lucky, and in a rectangular room the same distance would put you in one modal nulls for the length axis and another modal null for the width.

That said, you also don't want to put the speakers too far away from the front wall, as that would move the first SBIR dip down too low.

In other words, it's a juggling game! You have to play with all of those parameters to find the best overall compromise that doesn't make any one factor too bad, or too good at the expense of others. You can do the math to figure this out, then test it... or you can just move things around in small steps, to see those trends, and maybe use the trends to predict a better spot. For example, if you see that the low-end response is getting flatter around 34 Hz as you move in one direction, then keep going in that direction! That will have negative consequences in other places, of course, but it might still be worthwhile. Try not to focus too much on one single thing, at rather look at the bog picture as you nudge things around: pretty much always one or two things get better, but others get worse.... it's all about compromise.

Of course, if you soffit mount (flush mount) your speakers, then you can eliminate many of the artifacts caused by just having the speakers in the room, but that's another discussion....

I'm not sure why the mid-range dropped so much. I was very careful not to move the speakers or mic while adding the treatment, but of course it's possible I may have bumped something.

Maybe something changed on your DAW? Perhaps you used a different setting on your mic pre-amp, or in the software some place?

The speakers are Adam S3x-V

Nice!

and the subs are NHT b-20

I found the manual for those on-line, and they only go down to 29 Hz, so that would probably explain what I saw with the 30 Hz roll-off. The S3x-V goes down to 32 Hz by itself, so the subs aren't really adding a lot, and that's also what I was seeing in your data.

I set up the subwoofer position / crossover with the treatment in place. So if you look at the "Treated" L vs L-S you will see that it filled in a hole in the 100hz area.

I did see that, yes, but if you look at the final LR and LRS in the current setup, with the room treated, you'll see that it didn't actually accomplish what you were hoping for:

MAXLOR--REW--FR--12..500--LR-vs-LRS--treated-V1.png

Purple is the LR result, and light blue is LRS. The dip is still there, except that it moved down a few Hz, from about 98 to 91, but there's now an eve larger dip at 200 Hz: that took a nose dive by more than 10 dB! The reason for that 200 Hz dip is also clear, when you look at the phase overlaid on top of the FR:

MAXLOR--REW--FR--12..500--LRS--treated-V1-plus-phase.png

You have a phase flip, right there at 200 Hz! It spins through 360° over a range of just a couple of Hz. That is almost certainly a phase cancellation caused by the something in the room, such as SBIR for example, since it is also there on the LR test, as well as the LRS, but is NOT there on the S-- test alone... so it is not related to the subs directly, only to the mains. But when you add the subs, then multiply the problem due to their location in the room, and the extra energy. Try flipping the phase on both subs, and see what you get...

Because the speakers are almost flush with the front wall, they are very full sounding already. So the subwoofers aren't needed to boost the low end much, I was just hoping they would activate a different part of the room and fill in some holes.

One question: from the manual I found on-line, it's not clear id the subs are set up as stereo, or if they are both summed to mono? In other words, is each sub getting the exact same summed mono signal, or are the each getting different signals, from the L and R channels? It might be an idea to do a REW test of the subs individually: first just the left sub by itself, with nothing else at all, then just the right sub by itself.

I did a little reading about the "Plane Wave" setup, but I'm not sure I understand it completely. Does this mean I should orient both the subwoofers on the same plane, parallel with the front wall?

Not really, sort of, yes, no!

This is usually done with several subs, but it can also be accomplished with just 2, to a certain extent. The idea is to have one set of subs on one wall (often the front of the room) and another set on the opposite wall, phase inverted. So the ones at the front are "pushing" a wave down the room, all in phase, and the ones at the back are pushing the inverted copy of that wave in the opposite direction... and sometimes this can be considered from another point of view: the ones at the back are "sucking up" the wave produced at the front. So that wave is produce on one wall, runs down the room, and is cancelled at the other wall. It can also be done across the room: it does not necessarily need to be done lengthwise. Width-wise works OK too. Often, this is set up by four subs on one wall (eg, front), at the 1/4 wave points of that wall, and another four on the rear wall, in the corresponding locations. All of them acting together produce a plane wave that travels down / across the room. A plane wave is called that because the wave-front is flat, not curved: all of it would strike the other wall at once, with each part of the wave hitting the wall at the same time, rather than each part hitting at a slightly different time if the wave-front where curved. That has acoustic benefits that are rather complex to get into, but one of those is that you can cancel the entire wave at the other end of the room all at once, if you produce an identical wave that is phase inverted. That doesn't happen with a spherical wave-front.

Why does it work, and what does it do? It greatly reduces modal issues, because a mode that is triggered on the one wall, is cancelled on the other wall! It gets "sucked up", so to speak. So it doesn't get to resonate, like it would otherwise. That's the theory.... Exactly as the mode arrives at the opposite wall, the other speaker produces a mirror image, hopefully getting 100% phase cancellation, so the mode is stopped dead in its tracks. For a large room, you usually need a delay on one sub that is equal to the time distance across the room, ensuring total cancellation, but for small rooms and across the width, sometimes you can get away without that.

So, in your case, since your room is not symmetrical front-to-back (the ceiling rises and falls), that's not going to work so well. But it CAN work sideways, because your room is symmetrical that way. You would need to set up one sub o the floor, in the middle of the left side wall, and the other in the same place in the right side wall. Flip the phase 180° on one of them, set them both to the exact same level, send them both the exact same signal (in other words, the summed base signal), and you should get a reasonable facsimile of a plane wave like that. It should give you decent smooth bass coverage through the entire room. You might need to nudge things around quite a bit to get it to work, because there's always some non-symmetrical stuff going on, and you might need to add a delay as well (as mentioned above).

One drawback is that having the subs out to the sides might make them "localizable", meaning that you can hear the direction where the bass is coming from: that's why it is better (when possible) to do it front-to-back. It might be possible to reduce the localization effect by sliding both subs forward, closer to the front wall. You'll lose a bit of the plane-wave effect like that, but hopefully not enough to be noticeable.

Once again, it's a juggling game: playing to give one aspect priority over another, if something becomes objectionable.

If it works well, you get very even bass coverage throughout the room, and you minimize modal problems. It smooths things out.

However, furniture and non-symmetrical things in the room can mess it up, so try to minimize that. Keep your desk as low profile and acoustically invisible as possible. Ditto your chair, and the client couch at the back (if you have one). Minimalist is the key here!

It's very time consuming and rather boring to set up, but if you can get it to work, it's a really, really, really good thing!


- Stuart -

[maxlor]

I've taken your advice and tweaked the speaker positioning / height. I tried many different configurations and landed on this:

Speakers 43cm from the back and side wall. That makes them 2.3m wide. I've moved the LP slightly closer at 174cm. The tweeters are now at 121cm but my microphone height has stayed at 115cm as this is comfortable and tests well.

This new positioning has cleaned up some of the problems in the 30hz and 100-200hz range. But it has caused new problems in the 300 hz range.

New measurements:
Untreated MDAT
Treated MDAT

Thank you for the wealth of information on the concept of planar wave subwoofer placement. These NHT subs are indeed fully stereo, but have the option to be summed to mono as well. It also includes 4 different phase positions -180, -45, 0, 45, 180. However, with the mains in the corners, you're right that the subs are not adding much as the bass is already sounding full. So, for the purposes of this initial setup and testing, I'm going to simplify the process by leaving out the subs.

Of course, if you soffit mount (flush mount) your speakers, then you can eliminate many of the artifacts caused by just having the speakers in the room, but that's another discussion....

I would love to soffit mount, but I think I need to wait until I have a more permanent room before embarking on that journey.

For the "treated" example, I have take much more care in panel choice and placement. I have purposely aimed to tame 1k - 4k without taming too much 200-300 hz. And the RT60 is looking much more balanced across the frequency spectrum.

With this speaker placement, I do not have much room behind the speaker for treatment. Is this preferable to a slightly less-optimal speaker placement which would allow deep front corner traps?

Do you have any recommendations on optimal thickness / positioning of ceiling treatment? Or is that impossible to calculate with this type of ceiling?

I want to cover the whole back wall in 12" thick trapping. From my reading about density, depth, etc. it seems like something less dense like pink fluffy insulation would do better than rigid mineral wool to absorb bass at a 12" thickness. Is that misguided or am I on the right track?

Thanks again for giving me so much great info. Any amount of help is appreciated.
Max

[DanDan]

Are your speakers 43cm from the wall behind them? (I would call that the Front Wall, but never mind)
Are you measuring from the back of the speaker or the front cone side?
Most of the speaker manufacturers and much technical writing recommends they be as close to the FW as possible, OR as far away as possible.
Genelec used to write no further than 20cm OR no closer than 2.2M
Optimising speaker and listener positions is like 3D Chess. Add subs and it goes 4D.
Whether you start there or end up there though, do try the speakers almost touching the FW. Mine have little rubber bumpers, they are that close.
I would try Speaker Wadding or other fibre in the ADAM ports, or even block then entirely. They are pretty chuffy down at 30Hz. Let the subs do their work.

DD

[maxlor]

Are your speakers 43cm from the wall behind them? (I would call that the Front Wall, but never mind)
Are you measuring from the back of the speaker or the front cone side?
Most of the speaker manufacturers and much technical writing recommends they be as close to the FW as possible, OR as far away as possible.
Genelec used to write no further than 20cm OR no closer than 2.2M
Optimising speaker and listener positions is like 3D Chess. Add subs and it goes 4D.
Whether you start there or end up there though, do try the speakers almost touching the FW. Mine have little rubber bumpers, they are that close.
I would try Speaker Wadding or other fibre in the ADAM ports, or even block then entirely. They are pretty chuffy down at 30Hz. Let the subs do their work.

DD

Hi Dan,

Thanks for chiming in. The front face of the speakers are 43cm from the front and side wall. (That was a typo, I call it the front wall too!) The back side of the speaker is about 10 cm from the walls.

Due to the shape of the speaker and stand I was unable to go much closer, but I could rig up a way to test them touching the corner.

So, it sounds like you're saying it's likely better to have the speaker touching the corner drywall than allowing any treatment behind?

Thanks!
Max

[maxlor]

Hi all,

I took DanDan's advice and tested with the speakers as far in the corners as possible (32cm from front and side wall). The results were very favorable in the 200-300 hz range. But below 200 hz, the old position (speaker 43cm from front and side walls) looks better to me.

MDAT 32cmx32cm Untreated

MDAT 43cmx43cm Untreated

I tried many different speaker positions in between these two, and these two tested best.

I also tested several different listening positions forward and back. The listening position of 174 cm from the front wall seems to be the sweet spot. And playing music while listening from different positions has reinforced this as well.

I'm curious what you think and whether you would agree that 43cm looks better than 32cm. There are pros and cons to each and I wonder how you weigh these.

Thanks!
Max