Studio diary: practicing the basics!

The speakers test: https://drive.google.com/file/d/0B6wx2y ... sp=sharing I placed the speakers to form a triangle but I guess they are too close to the wall, right?

Interesting new data! Something changed... the high frequency boost is mostly gone, which makes me think that maybe you had the measurement mic too close to the tweeter, or even aimed directly at the tweeter. Now that the geometry is better, the "problem" has decreased. I'm still not seeing the high frequency roll-off that I expected, however, so that's still a mystery... Also, the level that you did those tests at is about 10 dB higher than the ones you did the other day! :shock: :!: :horse: It is very important that you do ALL your testing at the exact same level, and with the mic in the exact same place. If the mic moves even a bit (more than about half a cm), then you cannot compare the tests, since you are looking at a different location in the room. The tests are still valid: just not comparable.

I can remember reading posts from members who sounded pretty skilled here saying that they usually use a fair amount of smoothing on these EQ curves, but maybe I missed something

If you use smoothing, it can easily hide modal problems. Modes are usually very narrow, just a few Hz wide, so even mild smoothing can totally disguise them. If you can't see them, the you can't fix them! :) For example, here's a frequency response graph of the low end for your right speaker with 1/48th octave smoothing, and clearly you have major modal dips at 74 Hz, 118 Hz, 130 Hz and 172 Hz. (as well as modal peaks between) And here's the exact same graph, smoothed at 1/3rd octave: Where are the modes now? Not even the slightest hint.... Smoothing has its uses occasionally, but you need to know that if you are looking at a smoothed graph, you are only seeing a small fraction if the information, and a lot of the major important details are simply not visible. Smoothing can give you an overall impression, but it hides the very important details. If you look at an atlas of Switzerland, it's great for seeing the overall shape of the country and where the big cities are, but if I wanted to find your house, then an atlas is no use at all! I need a high resolution street map to find the details. Same for acoustics: you need the high resolution "unsmoothed" version to find the important details.

I've used the loopback calibration procedure to factor in both the mic preamp and the converters (you can see the curve in the linked measurements) so this should be taken into account already.

That's fine. That is typically what a calibration curve looks like, and entirely believable. No problem there.

EQ in the way: not at all. Just the Behringer microphone plugged into the mixer going to the converters.

No EQ set on the mixer? What mixer are you using, and are you certain that the EQ controls are set absolutely flat? Some mixers even have buttons for disabling the EQ on each channel: if yours has that, then make sure it is disabled.

The left wall is made of "Alba" or "placo-plâtre" - sorry I don't know the word in English. Not very dense, easy to drill.

Some type of gypsum board maybe? Also called plasterboard, drywall, sheetrock, and other things. something like this?

500 hZ absorbtion: could that be because I've installed too many basstraps? I've done this on the basis of the "more basstraps" kind of rule of thumb and therefore does not sound very plausible.

Actually, there is still not enough bass trapping in that room! Take a look at the waterfall plot for the low end of the spectrum: And also the RT60 graph: Clearly, there's nowhere near enough bass trapping in there. You are going to need a lot more than you have at present.

I've also recorded a bit of that song through the Behringer microphone:

Oh wow! :shock: The bass is just booming! You can hear the modal ringing going on there in the lows, rather clearly. The highs are a bit "tinny", but that might just be the speakers. Hard to say. But the lows are overboard. Here, I did a very steep low-pass filter on that, with cutoff at 120 Hz, so you can hear how the original song sounds: http://digistar.cl/sayers-forum/pask-so ... off_01.wav And also how your room sounds: http://digistar.cl/sayers-forum/pask-ro ... ff_01..wav Compare the loooong bass decay of your room against how the original was. (You will have to listen on full-range headphones, of course: you can't listen in your room, for obvious reasons...) You need more bass trapping in that room. Definitely.

I placed the speakers to form a triangle but I guess they are too close to the wall, right?

If that is the position they will be in once the room is complete, then that is the correct position for testing. And in such a small room, the speakers MUST be close to the wall. You can only move them away from the walls if the room is very large, or else you will get severe SBIR artifacts in the low-mids and lows. Also you need to set the roll-off controls on the back of your speakers to about -5 dB. However, you have the speakers lying on their side! that is not correct. They should be standing upright, with the tweeter at the top. The acoustic axis of the speaker should be 1.2m above the floor.

Is there any recommendation about the distance the microphone should be from the floor (or ceiling)?

It must be at 1.2m above the floor, the same height as the acoustic axis of the speakers, which is the same height that your ears will be when seated.

How can I know the final speakers placement as the acoustic treatment and its associated measurements are not completed yet?

That should have been decided in the design stage! :shock: :!: If you don't know where your speakers and mix position are going to be, then how did you decide on the treatment? That all goes together as part of the design process....

How do you know that I understand food analogies very well

:lol: I never met an audio engineer who didn't like good food! :) So it was sort of a rough guess... 8) The important things is that you understood the analogy! You are "tasting the cake" before you finish blending the ingredients, so you can't know yet how it will turn out. Not a good time to change the recipe, especially since you got the recipe from a Master Chef! (John Sayers)... :shot: :)

I know it sounds silly, but this room needs to be at least usable for vocal classes in about 10 days,

For recording vocals? Or just for teaching vocals? Also, I thought this was going to be a control room, so I'm not sure that I understand why you would be using it to practice vocals... Is that just temporary, or will it be used for vocal classes permanently? Anyway, it seems that the control room layout never was planned in advance, and neither was the acoustic treatment, so the best you can do for now in setting up your speakers and treatment, is to do a normal "generic" setup, modified to fit your room dimensions. So here's what I would do if that were my room, and I needed to fix it quickly to get it somewhat usable: Put each speaker about 28% of the room width away from the side walls. Your room is 4.19 m wide, so the speakers will go 117cm from the side walls (that means the MIDDLE of the front panel of the speaker will be at 117cm from the side wall, and they will be about 185 cm apart. also measured at the middle of the front panel). The room is small, so they will have to go right up against the front wall, just leaving a 10cm gap for the absorption panels to fit in. The speakers must rest on very heavy, massive stands built so that the final height of the acoustic axis is 1.2 m above the floor (note: ACOUSTIC AXIS! Not the top or bottom of the speaker box, not the tweeter, not the woofer: it is the acoustic axes we are talking about. Ask your manufacturer where that is on your speakers.). Angle the speakers inward at about 30°: Set up your chair so that your ears will be about 175cm from the front wall. If you do all that correctly, then the acoustic axes should be aiming at your ears (not your eyes), and will intersect about 15cm behind your head. Now put thick absorption panels between each speaker and the front wall, about 90cm wide, using at least 10cm thickness of medium density insulation (30 kg/m3 if you use fiberglass, or 50 kg/m3 if you use mineral wool). Put two more panels made the same way on the side walls, such that the center of each panel is 107cm from the front wall. Cover the entire rear wall with at least 15 cm of insulation, except for the corners where your superchunk bass traps should be. Build a hard-backed cloud with at least 10cm of the same insulation, and at least 16mm plywood (better: 19mm) on top. The cloud should measur 240 cm wide by 120cm deep. Hang that over the front of the room, centered 107cm from the front wall, and tilted at an angle of at least 12°. Use chains to hang it, so you can adjust the angle and height later. Put plastic on that, to prevent the insulation fibers from filtering down all over you and your equipment. Then do another REW test, and post the results here so we can see how the room is doing, and what the next step should be in treatment.

Thanks for this input.

:thu: You are very welcome! That's what this forum is all about.

I know it sounds silly, but this room needs to be at least usable for vocal classes in about 10 days,

You should be able to complete all of the above in less than ten days, and get another REW test done. I would expect a fairly big improvement in the REW results, except perhaps in the low end. That will make the room too dead, of course, but at that point you can start adding the slot walls (after calculating them!), which will return some of the liveliness. For that size room, I would aim for an overall decay time of maybe 150ms to 200ms. It would be too live if you went beyond that, I expect. Anyway, that would be my plan of action, if that was my room. Hope that helps! - Stuart -

Very, very useful Stuart - thanks a ton! Use of the room: it will primarily be used for vocal classes but also acommodate a writing/composing/producing setup. Tracking may happen but mainly for scratch vocals. Highs/no roll-off: I have no explanation for the lack of roll-off but the measurement microphone was actually pointing towards the speaker directly, so I guess this was causing the problem. Levels: I know that the levels are not consistent but I had to check the whole measurement setup again, swap microphones, use another mixer's channel, make the loopback test again, add another speaker, move the speakers, adjust their height, etc. and REW started to show those 150dB-ish measurements again. So I also needed to use the SPL calibration again, etc. All in all, I had to reset the whole thing, thus the lack of consistency. I fully understand the principle of not using smoothing for those SPL curves, no problemo and no need for a food allegory here ;-) EQ in the mixer: I'm 129% sure that there's no EQ engaged because the EQ section is simply switched off (unless my friend's console 2 first channels have serious problems). Alba: those are kind of plaster bricks, not panels. Basstrapping: I can also hear those long basses in the room. Now, I have 2 basstraps in the front and one in the back (left), all superchunks. Plus, we have small corner traps (superchunks) in the ceiling's corners. The ceiling cavities all have at least 10cm rockwool in them, as well. Practically speaking, how could I mount more basstraps than that? Quick question regarding RT60: I read on this forum that those graphs cannot be trusted from 200hz and below, is that correct? Speakers placement: OK - I was unaware that even speakers placement can be defined just looking at a SketchUp file. I will place them exactly where you explained. I know what acoustic axis is - no problem here. By the way, those ultra precise measurements leave me a bit skeptical sometimes because it would mean that if you move your head by an inch or two, the sound will not be ideal - which happens all the time in real life! You may adjust the height of your chair, sit back or lean forward, move to a side rack, etc. How do you handle those situations, then? I guess I should rephrase the title of this thread with "The joy of being a beginner" ;-) Treatment concept: I simply decided to mount as many basstraps as I could, fill those ceiling cavities and cover the rest of the walls with rockwool/slot resonators as my impression was that it was difficult to go wrong from such a basis. How wrong was I? ;-) Generic approach: (Side) panels: the side walls are already covered with 5cm Typ3 rockwool (http://www.swisspor.ch/index.php?sectio ... Page&id=77). Are you suggesting that I create additional side panels? Rear wall: it's already covered entirely with 5cm Typ3 rockwool, are you suggesting that I beef it up to 15cm? Why not create an horizontal corner trap instead? I'll try to sketchup something and post. This could serve as a big bass trap. Or maybe both? We already have a superchunk basstrap in the rear left corner, just so you know. Cloud: we already have 10cm of Typ3 rockwool in all the ceiling's cavities (see first pictures), are you suggesting to add an extra cloud panel? If so, should the plywood be facing up (towards the ceiling) or down (towards the production equipment)? I understand that you're suggesting a generic approach here but would like to clarify a bit more how to implement that in my specific case. Thanks a ton anyways!

OK - I've just Sketchupped 2 ideas to add basstrapping to the room. What about a large back wall corner trap + a beef up panel? If those ideas are valid, should I superchunk the corner trap or would an angled panel be sufficient? Also would this corner trap need to be closed or could I leave the sides open? Closing it would mean a lot more headaches, as I already have corner trap in the ceiling's corners, and where corners meet is always a difficult place to deal with (mounting-wise). About the beef up panel: would it make sense if I'd leave 5cm of air between the already installed 5cm rockwool and that 5cm thick beef up panel? Should it be thicker than 5cm? Would that make any difference? I don't really see where I could install more basstraps... I already thought that I had been very wise in mounting corner traps on the ceiling ;-)

Anyone has an idea about this SketchUpped thing?

OK, even though I am under the impression that I'm talking to myself here ;-), here are some news. The room needs to be at least usable on... Monday! But we'll have to empty the room in about 5 months anyways as I just happened to learn that the concrete floor needs to breathe for at least 6 months before I can paint it... So, this room will be a work in progress for the next months or so. I've installed the speakers as close as what Stuart explained, even though I haven't strong stands (the end user will bring his). Lights are installed and I'm almost done with the cloth. It starting to look quite good but I can't wait to add the slats! The cloud is not really possible, as there is a window on the front wall that goes up to the ceiling and it's not really possible to obstruct it. I have no answer on my questions regarding the side panels, but as we already have 5cm rock wool there, I'm assuming that it's OK? For the back wall, I don't want to waste the extra 10cm as they are needed to keep the usable surface large enough. So, we'll end up with 5cm of rockwool there instead of the recommended 15cm. Basstrapping: as I've already written, I can see that basses from 90Hz and below are sustaining too long but what would be the options to tame those frequencies? All my corners are already used for some sort of basstrapping. I've hanged a corner panel that was unused at the studio (about 160cmx100) angled at the top of the rear wall but the low-mids were more affected than the desired 90Hz and below. I'd really like to know what to do for those basses - I've heard of tuned traps like barrels with a whole, etc. but have never tried them. Anyone could point me to some examples? Now, about the mids, I guess the next step would be to make slat walls so I've spent some time today on John's Excel spreadsheet to get my head around that. Am I right to think that I should first try to work on this 500Hz bump? Apparently, I also have a serious dip between 610 and 730 Hz, anything that can be done to cure that? Thanks,

today's measurements: https://drive.google.com/file/d/0B6wx2y ... sp=sharing I think I should look into those 500Hz, right?

Quick question regarding RT60: I read on this forum that those graphs cannot be trusted from 200hz and below, is that correct?

The readings get less accurate at lower frequencies, but they are still plenty good enough to show what your room / sound system is doing. They won't be accurate enough to do a scientific research paper on the results, or write a thesis, but still entirely good for making decisions about treatment.

By the way, those ultra precise measurements leave me a bit skeptical sometimes because it would mean that if you move your head by an inch or two, the sound will not be ideal -

Exactly! :) Which is why there is no point in trying to design a room with millimetric precision. However, accuracy is important for the symmetric location of your speakers, and also for the positioning of the measurement mic every time you do a new test with REW. The mic must be in the exact same location in the room, in all 3 dimensions, precise to within a few mm. That is important. If not, then you can no longer compare the readings....

Rear wall: it's already covered entirely with 5cm Typ3 rockwool, are you suggesting that I beef it up to 15cm?

Yes. 5cm isn't much. Especially considering that very large modal or SBIR issue that is almost certainly associated with the rear wall.

Cloud: we already have 10cm of Typ3 rockwool in all the ceiling's cavities (see first pictures), are you suggesting to add an extra cloud panel?

You might need one, yes. The hard back serves several purposes, such as possibly helping to break up vertical modal response a bit, helping with RFZ, etc.

If so, should the plywood be facing up (towards the ceiling) or down (towards the production equipment)?

Facing up, towards the ceiling. Lots of insulation below it.

I understand that you're suggesting a generic approach here but would like to clarify a bit more how to implement that in my specific case.

Start with the generic, then test again with REW to see how it is going, then consider the specifics that you might also need, in addition to the generic.

I have no answer on my questions regarding the side panels, but as we already have 5cm rock wool there, I'm assuming that it's OK?

I would make it a bit thicker, if possible, especially at the first reflection points. The thick it is, the better it works, and the lower it goes in frequency.

For the back wall, I don't want to waste the extra 10cm as they are needed to keep the usable surface large enough. So, we'll end up with 5cm of rockwool there instead of the recommended 15cm.

Then there is no way that you will be able to get rid of that huge hump in the low frequency response at around 40 Hz. That is undoubtedly associated with your first axial mode, or with SBIR related to the rear wall. So the only place it can be treated, is on the rear wall...

Basstrapping: as I've already written, I can see that basses from 90Hz and below are sustaining too long but what would be the options to tame those frequencies? All my corners are already used for some sort of basstrapping.

You might want to consider a membrane trap using limp mass, but you'll still need a fair amount of depth to get down to 40 Hz. And seeing that you don't want to lose any more room depth, there aren't any more options open to you, except to just live with the distorted frequency response and time-domain response in the low end.

I'd really like to know what to do for those basses - I've heard of tuned traps like barrels with a whole, etc. but have never tried them. Anyone could point me to some examples?

Be careful of Helmholtz traps: lots of people talk about them, but they are very hard to tune correctly, and even harder to place correctly in the room. Panel traps, membrane traps and similar devices are your only realistic options, but they have to be big to get down to low frequencies. Many inches deep...

today's measurements:

One good thing! The data is now looking very believable, and very realistic. The high-end now rolls off as expected, and so does the low end. I think your earlier issues with your REW procedure are now resolved! :) I'm not sure what you did to fix it, but the graphs are now showing typical data for small rooms, more or less what one would expect.

Am I right to think that I should first try to work on this 500Hz bump? Apparently, I also have a serious dip between 610 and 730 Hz, anything that can be done to cure that?

For mid-range, just design the slot wall for general broad-band absorption. In other words, large percentage of open area, with the center frequency set to around 1 kHz and a gentle Q that covers at least an octave either side.

I think I should look into those 500Hz, right?

It's not really worth doing anything in the mid range unless you have the low end under control already. The biggest issues, by far, are under 200 Hz. I would not aim to absorb too much specifically in the 500 Hz region either, as your decay is already low in that area, as compared to around 4k. I would do general broadband across the upper-mids, both absorption and maybe some diffusion. (However, the room is rather small for diffusion, except at high frequencies.) You also need to reduce the bass roll-off on your speakers a bit: Try setting that to about -3 dB. Which speaker did you use for that test that you posted? Left or Right? You should really do 3 tests for each measurement set: one with just the left speaker, one with just the right speaker, and one with both. So what do you have in the room right now? How much of what I suggested have you done? Photos! :) - Stuart -

Hi Stuart and thanks for your feedback! I was starting to feel alone, as you're apparently the only one able to answer all the questions so many people have the forum ;-) Thanks for clarifying my above questions. What I've done so far is experiment to tame those 90Hz and below and start working on the slat walls. 2 things I've realized: closing the front window and the door at the back really changes the frequency response and the RT60 (it probably sounds evident to you but I'm amazed to see how dramatic it is). So, after a few hours of moving rockwool panels in the room, I discovered that positioning 15cm of rockwool right in the front of the desk lowers the RT60 of those 90Hz and below from 0.8 to about 0.3 seconds. I take this as good news and hope I'm right by doing so (?) ;-) I would have to package those rockwool panels in a nice way if keeping them is desired. I still have a very audible ringing in the highs and wonder if it's not that old heater located right under the front window (between the speakers) and/or the water pipes in the ceiling? I plan on completely removing the heater (or insulate it, as we don't use it anyways) and put those insulation empty "tubes" around the water pipes, to see if this helps. Any ideas around this? I've made new measurements that look better than before and will post them on Monday/Tuesday. In order to thicken up this back wall, I had sketcupped and idea - if you could give me a bit of feedback on this, that would be very helpful. Basically, the idea is to add extra layers of rockwool (and air?) from about the middle of the wall and up to the ceiling corner traps. This way, we would keep the desired space for the sofa up to about 150cm from the ground. Cloud: again, if you look at the SketchUp plot, we have the constraint of a window that opens towards the interior of the room and water pipes running at an annoying distance from the ceiling, so that leaves us with little space to play with. Side panels: I was waiting for the feedback about the 5cm being enough or not before taking action, so no progress here. I'm wondering if I should create self-contained 15cm deep "slat panels" that I could attach to the side walls? Or is any sort of diffusion a bad idea in this zone? Thanks for your help!

I was starting to feel alone, as you're apparently the only one able to answer all the questions so many people have the forum

:| Yeah, it sort of seems to be that way! I sure wish more folks would join in with responses, especially those that have already finished their rooms, or seen major benefits from advise the got here... Some people do help occasionally, when they can (Steve is one!), but more would be appreciated...

2 things I've realized: closing the front window and the door at the back really changes the frequency response and the RT60 (it probably sounds evident to you but I'm amazed to see how dramatic it is).

Darn! I probably should have mentioned that right from the start! When you do measurements in a room, it is very important to have the room all closed up, with all the doors and windows closed. If not, then you are measuring the response of the combined acoustics of the room plus the other spaces around it. Sorry about not mentioning that in the beginning! Yes, it can make a very big difference.

I discovered that positioning 15cm of rockwool right in the front of the desk lowers the RT60 of those 90Hz and below from 0.8 to about 0.3 seconds. I take this as good news and hope I'm right by doing so

That's fine, yes. The modal and SBIR issues are associated with the front/back axis of the room, so putting treatment on the front wall as well as the back wall is a good idea. And it works! :)

I still have a very audible ringing in the highs and wonder if it's not that old heater located right under the front window (between the speakers) and/or the water pipes in the ceiling?

That's a possibility, yes. That heater radiator could indeed be doing strange things to the acoustics. But is there anything else in that room that is thin, or that has a thin air gap inside it?

I plan on completely removing the heater (or insulate it, as we don't use it anyways)

Take it out, if you can. Much better to not have it there at all. And you can use that empty space for more bass trapping! :)

Basically, the idea is to add extra layers of rockwool (and air?) from about the middle of the wall and up to the ceiling corner traps. This way, we would keep the desired space for the sofa up to about 150cm from the ground.

That would work very well. I would suggest starting with just a simple frame filled with mineral wool, and see how that goes. If it doesn't do enough in the low end, then it might be necessary to make it into a membrane trap. Hopefully that won't be necessary, but keep it as an option.

Cloud: again, if you look at the SketchUp plot, we have the constraint of a window that opens towards the interior of the room and water pipes running at an annoying distance from the ceiling, so that leaves us with little space to play with.

Yeah, I noticed that: Not a lot of room there. Let's first see how things go with the rest of the treatment, and only think about using that area if you still aren't getting enough effect.

Side panels: I was waiting for the feedback about the 5cm being enough or not before taking action, so no progress here. I'm wondering if I should create self-contained 15cm deep "slat panels" that I could attach to the side walls? Or is any sort of diffusion a bad idea in this zone?

Slot walls are fine on the sides, but do try to angle them enough that you don't get reflections from the speakers coming to your ears. You might need to build them as a sort of "sawtooth" design, with two or three smaller sections that are angled greatly, rather than just one section with a small angle on it. And if you tune it to high enough frequencies (for example, the mid-highs that are causing you so much trouble!!!!), make it broadband (by varying the cavity depth, and slot and gap width), with low Q (entirely filled with insulation, right up to the front panel), then that won't cause you unnecessary issues. Some diffusion is good, and the amount created by a slot wall is fairly low level, and also occurs mostly in horizontal bands (since the slats are horizontal, which is better than having it in vertical bands or 2D bands, as caused by QRDs and Skylines. An even better option is to dived the wall into three sections vertically, with slats for the top and bottom sections, and just absorption in the middle, where your head is closest, and where the real first reflection points are. - Stuart -

Soundman2020 wrote:

Quick question regarding RT60: I read on this forum that those graphs cannot be trusted from 200hz and below, is that correct?

The readings get less accurate at lower frequencies, but they are still plenty good enough to show what your room / sound system is doing. They won't be accurate enough to do a scientific research paper on the results, or write a thesis, but still entirely good for making decisions about treatment.

OK - good to know, thanks!

By the way, those ultra precise measurements leave me a bit skeptical sometimes because it would mean that if you move your head by an inch or two, the sound will not be ideal -

Exactly! :) Which is why there is no point in trying to design a room with millimetric precision. However, accuracy is important for the symmetric location of your speakers, and also for the positioning of the measurement mic every time you do a new test with REW. The mic must be in the exact same location in the room, in all 3 dimensions, precise to within a few mm. That is important. If not, then you can no longer compare the readings....

I understand this but in my specific case, the room is already used for composing and vocal classes so I need to remove the microphone each time I make measurements. I've put marking on the ground to help, though.

I'd really like to know what to do for those basses - I've heard of tuned traps like barrels with a whole, etc. but have never tried them. Anyone could point me to some examples?

Be careful of Helmholtz traps: lots of people talk about them, but they are very hard to tune correctly, and even harder to place correctly in the room. Panel traps, membrane traps and similar devices are your only realistic options, but they have to be big to get down to low frequencies. Many inches deep...

I see, good to know!

Am I right to think that I should first try to work on this 500Hz bump? Apparently, I also have a serious dip between 610 and 730 Hz, anything that can be done to cure that?

For mid-range, just design the slot wall for general broad-band absorption. In other words, large percentage of open area, with the center frequency set to around 1 kHz and a gentle Q that covers at least an octave either side.

May I ask why those mid-range frequencies would need additional treatment? Aren't both the graph and the RT60 looking quite decent in this area?

I think I should look into those 500Hz, right?

It's not really worth doing anything in the mid range unless you have the low end under control already. The biggest issues, by far, are under 200 Hz. I would not aim to absorb too much specifically in the 500 Hz region either, as your decay is already low in that area, as compared to around 4k. I would do general broadband across the upper-mids, both absorption and maybe some diffusion. (However, the room is rather small for diffusion, except at high frequencies.)

OK - so let's focus on the basses first, then.

You also need to reduce the bass roll-off on your speakers a bit: Try setting that to about -3 dB.

Will do.

Which speaker did you use for that test that you posted? Left or Right? You should really do 3 tests for each measurement set: one with just the left speaker, one with just the right speaker, and one with both.

Oops - I've only used both. Will do L-R-LR next time.

2 things I've realized: closing the front window and the door at the back really changes the frequency response and the RT60 (it probably sounds evident to you but I'm amazed to see how dramatic it is).

Darn! I probably should have mentioned that right from the start! When you do measurements in a room, it is very important to have the room all closed up, with all the doors and windows closed. If not, then you are measuring the response of the combined acoustics of the room plus the other spaces around it. Sorry about not mentioning that in the beginning! Yes, it can make a very big difference.

No problem - it now sounds like common sense but I had cables running in and out of the room, thus the open door and window.

I discovered that positioning 15cm of rockwool right in the front of the desk lowers the RT60 of those 90Hz and below from 0.8 to about 0.3 seconds. I take this as good news and hope I'm right by doing so

That's fine, yes. The modal and SBIR issues are associated with the front/back axis of the room, so putting treatment on the front wall as well as the back wall is a good idea. And it works! :)

Should this treatment mostly be rockwool or should I plan some slat wall panels there, as well (I guess in the front only)?

I still have a very audible ringing in the highs and wonder if it's not that old heater located right under the front window (between the speakers) and/or the water pipes in the ceiling?

That's a possibility, yes. That heater radiator could indeed be doing strange things to the acoustics. But is there anything else in that room that is thin, or that has a thin air gap inside it?

Not to my knowledge, and I doubt it's the windows, as I've put rockwool in front of them but it didn't solve the issue.

I plan on completely removing the heater (or insulate it, as we don't use it anyways)

Take it out, if you can. Much better to not have it there at all. And you can use that empty space for more bass trapping! :)

I need to get the building owner's approval first but I plan on having it removed.

Basically, the idea is to add extra layers of rockwool (and air?) from about the middle of the wall and up to the ceiling corner traps. This way, we would keep the desired space for the sofa up to about 150cm from the ground.

That would work very well. I would suggest starting with just a simple frame filled with mineral wool, and see how that goes. If it doesn't do enough in the low end, then it might be necessary to make it into a membrane trap. Hopefully that won't be necessary, but keep it as an option.

I have an old rockwool panel (rockwool mounted on a thin layer of compressed artificial wood (not sure of the name)) that I move around to test options. Putting it angling downwards on the back wall had an effect on the 400-700Hz range but not lower than that. So I was more thinking of simply beefing up the backwall. A question regarding this. Should I add about 15cm (3 panels) of rockwool, end of the story? Or going 5cm panel - 5cm air - 5 cm panel would be more efficient? Also, if I go for the air gap option, should the sides be closed, or is it ok to just leave the sides open? Another alternative: I could create a self-contained slat wall built on a plywood panel? That would be plywood, a stud frame, rockwool, (thin plastic), cloth, slats. But as the plywood is hard, would that change the physical theoretical dimensions of the room?

Cloud: again, if you look at the SketchUp plot, we have the constraint of a window that opens towards the interior of the room and water pipes running at an annoying distance from the ceiling, so that leaves us with little space to play with.

Yeah, I noticed that: Not a lot of room there. Let's first see how things go with the rest of the treatment, and only think about using that area if you still aren't getting enough effect.

OK - great.

I'm wondering if I should create self-contained 15cm deep "slat panels" that I could attach to the side walls? Or is any sort of diffusion a bad idea in this zone?

Slot walls are fine on the sides, but do try to angle them enough that you don't get reflections from the speakers coming to your ears. You might need to build them as a sort of "sawtooth" design, with two or three smaller sections that are angled greatly, rather than just one section with a small angle on it. And if you tune it to high enough frequencies (for example, the mid-highs that are causing you so much trouble!!!!), make it broadband (by varying the cavity depth, and slot and gap width), with low Q (entirely filled with insulation, right up to the front panel), then that won't cause you unnecessary issues. Some diffusion is good, and the amount created by a slot wall is fairly low level, and also occurs mostly in horizontal bands (since the slats are horizontal, which is better than having it in vertical bands or 2D bands, as caused by QRDs and Skylines.

What high-mids are you talking about exactly? I've seen John's sawtooth slat panels and this would be a project in itself but I'll keep this idea for further development.

An even better option is to dived the wall into three sections vertically, with slats for the top and bottom sections, and just absorption in the middle, where your head is closest, and where the real first reflection points are.

This sounds more realistic as of now. Thanks for your very helpful inputs!

Last measurements: https://drive.google.com/file/d/0B6wx2y ... sp=sharing

Also, would such "acoustic barrels" serve as speaker stands and basstraps? http://www.thomann.de/fr/hofa_basstrap.htm

According to what we've discussed for the back wall, I've created a pretty large 88x240x17cm panel. It has 10cm of Typ3 rockwool and 7cm of air gap. I put plastic on top. So, the sandwich is: plastic, rockwool 5cm+5cm, 7cm air gap -> mounted on the 5cm rockwool insulation that is already there (with cloth). Graphs: https://drive.google.com/file/d/0B6wx2y ... sp=sharing According to the graphs, the bass frequencies RT60 is ... worse! And I can't see any benefit in the mids and highs either, or am I missing something? How come is that RT60 much worse than withOUT this panel? I can't see any logic here, I must say. Not sure what to do with that huge panel, now :-/ I'll probably hang it in the lounge or something. What do the doctors say?

So... either everybody's dead or I should upload those pics ;-) ?