Ceiling Treatment for Listening Room / Acoustic Guitar Playi

[philhjobim]

I am adding thermal insulation to a very cold room which is a converted stable with thick stone walls. Half of the exterior walls are also underground.

The ceiling plasterboard will be taken down and 75mm PIR foam sheet placed up into the cavities between the 5inch joists. The joists are under a bedroom with chipboard floor.

I would welcome advice on how to finish the ceiling surface, for the room to sound good for hifi listening and playing acoustic guitar. I guess I'm looking for a balanced result not dry. I may occasionally record my Martin and Taylor guitars with my TLM103 but recording is not the primary room use so I am not looking for a dry acoustic.

The floor will be concrete>30 mm PIR foam sheet>22mm chip flooring>thick carpet
Three walls are thick stone>25mm unfilled air gap>concrete block tanking wall>50mm PIR foam sheet> surface finish will be decided last to allow adjustment to the acoustic.
One end wall is exposed original rough stone blockwork.

Room dimensions 7.0m long X 3.7m wide X 2.1m high

Budget in hundreds rather than thousands preferred xD

Thanks, Phil

[Soundman2020]

Hi there Phil, and Welcome!

The ceiling plasterboard will be taken down and 75mm PIR foam sheet placed up into the cavities between the 5inch joists

PIR is great thermally, but pretty useless acoustically, because it is closed-cell foam. To have usable acoustic properties, the insulation you use mist be open-cell, meaning that it consists of numerous tiny air spaces that are interconnected with each other, within some form of "web". Fibrous porous absorbers, such as mineral wool and fiberglass insulation, are the best acoustically, and also pretty good thermally. I'd suggest you switch to those if you want your room to be usable acoustically, in addition to being well insulated for heat.

The joists are under a bedroom with chipboard floor.

Do you need acoustic isolation up there? In other words, do you need to keep things quite by reducing the amount of sound that leaves your room and gets into the bedroom, up through that floor?

I would welcome advice on how to finish the ceiling surface, for the room to sound good for hifi listening and playing acoustic guitar.

Making the room sound good for critical listening and rehearsal has very little to do with how the surface of the ceiling is finished: it has everything to do with how the room is treated acoustically. With a room that size, here is the situation:

philhjobim-simulation-01.jpg

That's the predicted response for your room, untreated, in the low frequency part of the spectrum, which is the most critical. The final finish that you put on the wall and ceiling surfaces won't do much to fix that. What WILL do a lot to fix it, is the materials you use to treat the room before you put the finish on.

I guess I'm looking for a balanced result not dry.

In a small room like that with a very low ceiling, your options are rather limited. If you don't treat the low end sufficiently it will be boomy. Treating the low end enough, automatically implies that you are way over-treating the high end, which will make it muffled, muddy, thick, dead. You can deal with that to a certain extent with correct use of diffusive and reflective surfaces, but then you could mess up the mid range. Getting a balanced acoustic response out of a room like that is really hard to do.

The floor will be concrete>30 mm PIR foam sheet>22mm chip flooring>thick carpet

... which would automatically trash the high frequency response of the room, while also reducing the headroom even more, to give you an even lower ceiling. Carpet is pretty much the worst possible surface treatment that you can put into a small room. If you really want to make a good room sound bad, carpet the floor. And if you want to make a bad room sound terrible, carpet the floor with thick carpet.

Carpet is a selective acoustic absorber: it is really good at absorbing high frequencies, (which you do NOT need to absorb in a small room), it is measly and unpredictable at absorbing the mid range (which needs to be absorbed in a carefully controlled manner in a small room), and it does nothing at all for low frequencies, which need massive absorption in small rooms. In other words, carpet does the exact opposite of what is needed. Take a look at photos of professional studios from all around the world, or even good home studios as you see all over this forum: how many of them have carpeted floors? Answer: around zero. You pretty much never see carpeted floors in studios, for a good reason: it kills the acoustics. At best, there might be some carefully positioned carpet in a few key locations in the room, but you practically never see fully carpeted floors. It's a huge mistake. (The only place you see that is in the "YouTube" studios, built by people who don't have a clue about acoustics, and are the same ones who are likely to put egg-boxes on the ceiling and try to float their floor on tennis balls or rubber pads.)

Three walls are thick stone>25mm unfilled air gap>concrete block tanking wall>50mm PIR foam sheet> surface finish will be decided last to allow adjustment to the acoustic.

Once again, the PIR foam is useless acoustically.

What I would suggest is that you first run a simple acoustic test in that room, using the free REW software, and post the results here on the forum for us to take a look at. Based on that we can suggest specific treatment options, and specific locations where they would need to go, in order to give the room good acoustic response that would be somewhat usable for what you want. The room is too small and the ceiling is too low to be able to get really good response out of it, but it can still be pleasant in there, if it is treated correctly.

Basically, you would need to forget the PIR completely, and switch to something that is acoustically useful, such as fiberglass or mineral wool, for all of your insulation. That would do both things at once that you need: good thermal insulation, and good acoustics. Put that in your ceiling joist bays, frame up your walls with furring and do the same there, and use a semi-rigid fiberglass under your floor with proper underlay and laminated flooring on top. The wall framing might need angling in places, both to reduce flutter echo and also to provide a base for broad-band treatment devices, such as slot walls, and the ceiling pretty much has to be mostly fabric, perhaps with a few key reflective surfaces embedded in the insulation at angles that would help with diffusion, to a minor extent. That's about all you can do up there. With a normal height ceiling the correct approach would be to hang several small hard-backed clouds at angles, but you don't have enough height to do that.

So that would be my suggestion.


- Stuart -

[philhjobim]

Hi there Phil, and Welcome!

Wow! So much detailed information, thank you so much for your welcome and your time and advice.

Insulation: OK can switch to a fibre insulation I will do some reading up.

Ceiling: I already have enough PIR here to do the ceiling. Hopefully it will be OK to install that and then place fibre insulation underneath for the acoustic damping? I find the joists are 6 inches now I have checked so there is another 3 inches to play with.

The joists are under a bedroom with chipboard floor.

Do you need acoustic isolation up there?

No, it’s a spare bedroom also in my property and there is no conflict of activity times.

With a room that size, here is the situation:

philhjobim-simulation-01.jpg

Eeek look at that spike/notch. I knew from the moment I fired up my hifi after moving in 3 years ago, that the space was tricky!

The floor will be concrete>30 mm PIR foam sheet>22mm chip flooring>thick carpet... which would automatically trash the high frequency response of the room

OK so carpet doesn’t go back in, some version of wood flooring over underlay. I have been looking for an excuse to get some oak flooring xD

Three walls are thick stone>25mm unfilled air gap>concrete block tanking wall>50mm PIR foam sheet> surface finish will be decided last to allow adjustment to the acoustic.
Once again, the PIR foam is useless acoustically.

What I would suggest is that you first run a simple acoustic test in that room, using the free REW software, and post the results here on the forum for us to take a look at.

OK bear with me for a few days please, there is a stub wall dividing the room at present which is being taken out and when it has been removed I will have a go with REW and post the result. Many thanks for your amazing help, I'm quite excited about doing something to improve the room sound now. I have nice sounding hifi equipment and monitors and instruments and I do notice when the room doesnt sound good, so its really nice to be able to improve the outcome and do the sounds justice.

Phil

[Soundman2020]

Wow! So much detailed information, thank you so much for your welcome and your time and advice.

That's what we're here for!

Ceiling: I already have enough PIR here to do the ceiling. Hopefully it will be OK to install that and then place fibre insulation underneath for the acoustic damping? I find the joists are 6 inches now I have checked so there is another 3 inches to play with.

If you already have some PIR, then I would use it on one of the walls, not on the ceiling. Ceilings are critical, yours is already very low, and I'd suggest using every last fraction of an inch for acoustic treatment up there. Also, since that is a habitable space up there (not an outside surface), then there's no need for massive thermal treatment. I'd go with several inches of fiberglass or mineral wool. Fill those bays completely, but perhaps adding some angled massive surfaces first, to help with diffusion and maybe even a bit withe modal response.

Eeek look at that spike/notch.

Yup! Scary, isn't it BUt the good news is that the terrifying curve can be smoothed out quite a bit, with good treatment AND ALSO good room layout. What speakers will be using? That also matters....

OK so carpet doesn’t go back in, some version of wood flooring over underlay. I have been looking for an excuse to get some oak flooring xD

Try to get it as thin as possible, in order to not reduce your headroom even further. In other words, if you have the option of going with 8mm laminate that looks a lot like oak, or 2" of real oak, then go with the laminate! Also, don't do any type of flooring that would create an air cavity under it. Some types of sold wood flooring do need that... Beware!

- Stuart -

[philhjobim]

If you already have some PIR, then I would use it on one of the walls, not on the ceiling. Ceilings are critical, yours is already very low, and I'd suggest using every last fraction of an inch for acoustic treatment up there. Also, since that is a habitable space up there (not an outside surface), then there's no need for massive thermal treatment. I'd go with several inches of fiberglass or mineral wool. Fill those bays completely, but perhaps adding some angled massive surfaces first, to help with diffusion and maybe even a bit withe modal response.

OK fibre insulation for the whole ceiling depth then. I'll ask more about the angled surfaces when the room testing is done

Yup! Scary, isn't it BUt the good news is that the terrifying curve can be smoothed out quite a bit, with good treatment AND ALSO good room layout. What speakers will be using? That also matters....

The hifi speakers are 2001 B&W CDM1SE "shelf" speakers, front ported as they used to be in a much smaller! room in central London, and an ACI Titan II Kit Active Sub. No crossover, the ACI has twin freq adjustable rolloff filters which are adjusted to blend it to the CDM freqs and the CDM port plugs are left in. Driven from Copland CSA-28 hybrid valve/transistor amplifier/Transparent cables.

Studio/PC monitors are Adam A5/Sub 7.


I have a UMIK on order and the stub wall has been removed so I should be able to attempt some tests at the weekend.

Thanks,
Phil

[philhjobim]

I haven't used room measurement software before so please correct me if I have not got this right.

Freq plots for:
Green: CDM1SE single speaker raised 1 meter on keyboard stand, 1 metre from walls port plug in. Does it really do 85dB at 20 Hz?
Purple: CDM1SE 1m from floor and Titan Sub on floor. Sub settings left as they are usually set by ear, both filters set for 12db/octave 50Hz rolloff. Phase set to 0, gain at 40%.
Red: Adam A5 1 m from floor single monitor.
Blue: Adam A5 and Adam sub 7 placed 1m from floor on mount next to monitor. Crossover (Sub 7 controls this) 85Hz.

Measurements with UMIK1 aimed vertically, placed 0.5m high on low side table on centre axis of room 4 metres from speaker.

barnuntreated0915.jpg

The measurement files are here: http://www.hickslondon.com/pond/barnuntreated0915.mdat

North side of the room has a staircase coming down. There is a cottage style door at the top of the stairs.

barnnorth1200.JPG

I plan to open the side of the staircase up to the ceiling to increase the visual space in the room.

South side has a porch door and french doors. Both are being replaced with full length glass doors to increase light.

barsouth1200.JPG

There is one more door in the east wall as shown in the picture at the top of the thread.

Thanks,
Phil

[Soundman2020]

Unfortunately, there was a glitch with the forum server yesterday, and your photos seem to have disappeared, with no trace! Please could you re-post the photos. Sorry about that!

speaker raised 1 meter on keyboard stand, 1 metre from walls

If you have it 1m from the wall and also 1m from the floor, you are multiplying the SBIR artefacts. The correct height for speakers is to have the acoustic axis about 120cm to 140cm above the floor, which is the same height as the ears of most people when they are seated. Please not that this refers to the ACOUSTIC AXIS of the speaker, not the top or bottom of the cabinet. The manufacturer can tell you where the acoustic axis of your speaker is. You should then set up the speaker so that the distance from the side walls is not the same as the height, and also the distance form the front wall is not the same as either of those. For your room, the acoustic axis of the speakers should be about 103 cm from the side walls, and the speakers should be tight up against the front wall, with just a small gap of about 10cm behind them.

Measurements with UMIK1 aimed vertically, placed 0.5m high on low side table

The mic should also be placed at 120cm above the floor, and should not be on a table, desk or anything like that. The surface of that able is likely causing severe artifacts for the mic. Set up the mic on a proper mic stand, so that it is about 120cm above the floor.

The measurement files are here:

Even though you didn't set up the speakers and mic correctly, there is still some useful information in the MDAT file. As you can see, the measurements are a fairly close match for the predicted response. There are differences, of course, and there are reasons for those, such as the staircase and the doorways you mentioned.

Overall, the room is way to live right now! Very boomy, with a lot of ringing in the mid range too. Treatment will make a BIG difference here, for sure!

Does it really do 85dB at 20 Hz?

No, it certainly does not! That is probably noise coming from elsewhere. Where you inside the room when you did the measurements? Were the doors open? Next time you test, use the "Start Delay" feature of REW. That creates a pause between the time you hit the "Start Measuring" button, and the actual start of the test signals. Set that to several seconds, so that you have plenty of time to get out of the room, close the doors, and wait another few seconds for things to calm down inside there.

The good news is that you can improve things greatly with treatment!


- Stuart -

[philhjobim]

OK so I have tried to follow the instruction to the letter. I emptied the room as much as possible, set up the PC in the adjacent room as it has noisy graphics card fans, got the acoustic axis position from B&W and mounted the speaker with acoustic axis 130cm from floor. I placed the speaker axis 103 cm from the side wall plane, although the wall there is a french window so it is further away, and 12 cm from the "front" wall, but there is a fireplace which stands out 36cm from the wall so the speaker is 12 cm from that but partly in front of an alcove which normally contains a bookcase. The UMIK is placed at the height of the acoustic axis and I took four measurements, moving the UMIK 50cm across the room centre line, and 30 cm further back to get measurements at the corners of a rectangle which is kinda centred in the wider area of the room.

barnsouthwall.jpg

Measurements were taken with the doors closed. The noise meter read about -50dBZ, as the building is on a country hillside mainly birds tweeting and the odd sheep! The bass response graphs look like this:

barnuntreatedbassresponse041215.jpg

The datafile is available at http://www.hickslondon.com/pond/barnunt ... 5data.mdat

Views in other directions:

barnwestnorthwalls.jpg

barnstairs.jpg

barnnortheastwalls.jpg

As I mentioned I plan to open up the side of the stairs and move the stairs door to the top of the stairs.
Thanks, Phil

[philhjobim]

More details:
Floor Plan

barnfloorplan1200.jpg

Local building supplies offers 25mm Knauf Acoustic Floor Slab
http://www.knaufinsulation.co.uk/en-gb/ ... -plus.aspx
My default floor option would be concrete>Acoustic Floor Slab>Floating Interlocking Engineered Oak Boards, so no gaps between boards, no void underneath.
Does the acoustic slab meet the underlay requirement or are additional layers needed?

Ceiling Plan - actual joist positions.

barnceilingplan1200.jpg

Local building supplies offer Knauf Rocksilk RS45. 2 x 75 can be placed up into the joist gaps. Then cover with light plastic to eliminate falling dust / fibres, and jute fabric ie burlap/hessian? You talked about adding deflection surfaces, I am imagining you are describing something like this:

ceilingdetail1200.jpg

Is plasterboard sufficient or is chipboard required?
Should slab be placed behind and in front of the deflection?
Any particular pattern of placements?

Staircase modification:

barnnorthwallelevation1200.jpg

Shows proposed alteration of stair boxing in. There is also a space under the stairs currently cupboard which could be left open or used for trapping volume.

Wall treatment:
Propose to batten out and fit 50mm Knauf Rocksilk RS45 between battens on the three external walls (ie not the east exposed stone wall).
The surface finish as advised.

I have about 10 sq m of books which could be strategically placed to cover wall areas if that helps.

Many thanks for your advice.

[philhjobim]

It would be great to get some more input, I need to buy materials now. Many thanks for your advice.

[philhjobim]

Did I exceed my question quota?

I have filled between the rafters with 6 inches of RS45. No deflection panels as I was hoping for more detail on that first.

End walls and two thirds of side walls battened out with 2 inch RS45.
Floor materials ordered, 1 inch of 100kg rockwool batt, and 20mm thick engineered oak tongue and groove planks for the floor, which will be glued together.

Boxing in around stairs has been removed.
I will do another room measurement when the wall and floor is installed.

Placing the speakers at the end wall with absorption behind them, is a revelation! I remember those speakers sounding like that in the demo room when I bought them

Q1) Please see questions above about deflection in the ceiling spaces
"Is plasterboard sufficient or is chipboard required?
Should slab be placed behind and in front of the deflection?
Any particular pattern of placements?"

Q2) Can I add plastic sheet to contain the RS45 dust?
Any particular specification?

Thanks

[Soundman2020]

Sorry! I must have missed your reply...

I downloaded your MDAT file to look at it, but frankly it doesn't make a lot of sense! The data in there looks like it comes from a much larger room. Did you have the doors open to another room, maybe? Or the windows? Or something else?

Also, the REW data says that you did the tests at a level of 120 dB!!! That's unlikely... That implies that you did not calibrate REW using a proper hand-held sound level meter before you did the tests. So please calibrate REW, then repeat the tests.

Your basic treatment plan looks good, though. It should work, but the room is going to need more treatment than you mentioned.


- Stuart -

[philhjobim]

Sorry! I must have missed your reply...

I downloaded your MDAT file to look at it, but frankly it doesn't make a lot of sense! The data in there looks like it comes from a much larger room. Did you have the doors open to another room, maybe? Or the windows? Or something else?

Also, the REW data says that you did the tests at a level of 120 dB!!! That's unlikely... That implies that you did not calibrate REW using a proper hand-held sound level meter before you did the tests. So please calibrate REW, then repeat the tests.

Your basic treatment plan looks good, though. It should work, but the room is going to need more treatment than you mentioned.

- Stuart -

Thanks for your reply.

The doors and windows were closed in the previous test. I can only suppose that some of the surfaces are thin enough to allow some leakage. The doors are very thin tongue and groove board and rail cottage doors and dont close firmly in the frame. The external French doors have paper thin lower panels. The ceiling was at the time plasterboard sections with no skim plaster, and above that there is some free pathway to the ceiling joists of the adjacent small kitchen. Maybe some of that is altering it. Anyhow I have retested in the room with the basic treatment installed.

The previous test was done loud as I was interested in seeing maximum waterfall decay. The SPL is from the UMIK calibration file so should be correct. I did the new tests at about 80dB. The single CDM1 speaker was placed 4 inches from the back wall and 2 inches to the left of the fireplace side wall. (pictures below show the arrangement and the current room treatment). The acoustic axis was placed at 100cm from the floor, as that is my ear height on my low sofa. The stairs wall and understairs cupboard are now removed, and all basic treatment installed, but not the engineered oak floor, the floor is currently unpainted concrete. 3 tests were done with the UMIK at acoustic axis height, 1 [green] positioned at the centre of the sofa (which is centred to the fireplace) 2 [orange] above the left seat 3 [purple] above the back of the left seat. A fourth test [blue] was done to see if fireplace wall reflections were altered by moving the speaker forward to 25cm from back wall

barnbasetreatment161115mdatbass.jpg

The mdat file is available at http://www.hickslondon.com/pond/barnbas ... 61115.mdat

Looking in 4 directions around the room....

barnbasetreatmentsw.jpg

barnbasetreatmentnw.jpg

barnbasetreatmentne.jpg

barnbasetreatmentse.jpg

And a view looking up the stairs. The stairs ceiling is sloped but a bit off parallel to the stairs stringer. The whole side wall of the stairs has been treated to damp the flutter echo which was audible in that space. Around the corner is a small landing space leading off to upstairs rooms (doors closed).

barnbasetreatmentstairs.jpg

Looking forward to your comments, and copying down the outstanding questions for when they can be answered.

Q1) Please see questions above about deflection in the ceiling spaces
"Is plasterboard sufficient or is chipboard required?
Should slab be placed behind and in front of the deflection?
Any particular pattern of placements?"

Q2) Can I add plastic sheet to contain the RS45 dust?
Any particular specification?

Many Thanks, I really appreciate your help.
Phil

[Soundman2020]

The doors and windows were closed in the previous test. I can only suppose that some of the surfaces are thin enough to allow some leakage

I think I figured it out: the extra "tail" is probably fro the staircase and whatever is up there above.

The ceiling was at the time plasterboard sections with no skim plaster, and above that there is some free pathway to the ceiling joists of the adjacent small kitchen. Maybe some of that is altering it.

That might be it as well. It's hard to say, except that to note the the decay time is for a two-part room that is bigger than what is in the photos

The doors are very thin tongue and groove board and rail cottage doors and dont close firmly in the frame. The external French doors have paper thin lower panels.

Those will need to be beefed up with mass.

The SPL is from the UMIK calibration file so should be correct.

The calibration file is not for SPL: all that it does is to show the variations in the frequency response of the mic. In addition to that, you have calibrate REW itself, so that it knows how loud the level is in the room. If you don't tell it, there is no way that it can figure that out for itself! It has no idea what gain settings you have on your mic pre-amp, console, interface, or anything else, nor what settings you have on your amplifiers, speakers, cross-overs or anything else. You need to tell REW what the actual level is in the room, when REW is playing a calibration signal out to the speakers. You do that by looking at your own hand-held sound level meter (not an app on your iPhone! You need a REAL sound level meter), then tying that reading into the calibration panel in REW. Only then can you do your tests accurately.

After you do that, DO NOT CHANGE ANYTHING! Keep all of the settings on all of your gear exactly the same for all future tests.

The mdat file is available

That's a lot better, and clearly the second position for the speakers is the best, and don't sit in the left seat for critical listening!

There's still some modal stuff that could benefit from further treatment, and there's a lot of early reflections coming from somewhere, but that's a big improvement! You could probably add some reflective surfaces again, in a limited manner )widely spaced slats, for example), to help bring back some life to the room: it is overly "dead" right now, and could be improved.

It would be good to see some individual readings from the left and right speakers, instead of just both at once, as I suspect there's going to be some large symmetry problems due to the staircase. IS there any chance you could replace that with something more open? Maybe a metal grill staircase, or anything that has a lot less large flat reflective surfaces on it?

"Is plasterboard sufficient or is chipboard required?

Either would work fine.

Any particular pattern of placements?"

Random will do fine. Some facing forward, some backward, some left, some right. You could try to do that in a numeric sequence, such as the ones that are used for building Schroeder diffusers, but I doubt it would be worth it, since your scale is so much larger, and you are using a different principle of operation.

Can I add plastic sheet to contain the RS45 dust?

Sure! No problem. Except on the first reflection points: For those, just use a very fine weave, thin fabric.

Any particular specification?

Anything from very thin plastic, such as painter's drop cloth, to very thick plastic, such as is used under house foundation slabs. That affects the frequency range that is reflected, of course; the thicker the plastic, the greater the range (extending downwards). since you have fairly even decay times right now, use a combination of thick and thin. Perhaps use the thinner stuff up near the top of the panels, thick in the middle, and medium near the bottom. Or any variation that you like. Do half the room like that, then test again with REW to see how it is going, and make decisions about how to do the rest based on that.

Many Thanks, I really appreciate your help.

- Stuart -

[philhjobim]

The SPL is from the UMIK calibration file so should be correct.

The calibration file is not for SPL: all that it does is to show the variations in the frequency response of the mic. In addition to that, you have calibrate REW itself, so that it knows how loud the level is in the room. If you don't tell it, there is no way that it can figure that out for itself! It has no idea what gain settings you have on your mic pre-amp, console, interface, or anything else, nor what settings you have on your amplifiers, speakers, cross-overs or anything else. You need to tell REW what the actual level is in the room, when REW is playing a calibration signal out to the speakers. You do that by looking at your own hand-held sound level meter (not an app on your iPhone! You need a REAL sound level meter), then tying that reading into the calibration panel in REW. Only then can you do your tests accurately.

Ummmm. There are no preamp settings because UMIK is a USB mike.

As I understand it since V5 of REW the software designers worked with the mike manufacturer so that REW imports the calibration file and uses that to interpret absolute spl.

"The UMIK is automatically calibrated by REW for sound level (this information is in the calibration file)."
https://www.minidsp.com/applications/ac ... p-with-rew

If you click on calibrate in the level meter a pop up confirms "Calibration is not required. The USB mike is calibrated based on the sensitivity data in the calibration file".

If I have not understood this please correct me.

Phil