Project studio/music room - construction phase

[kominak]

Hello,

after a year-long pause I got back to work on my small studio room and I decided to write a diary as I progress to keep me on track.
Design has been discussed here and final sketchup version is here:
https://www.dropbox.com/s/q711hjdtw98ug ... 0.zip?dl=0

Stuart helped immensely with the design and he's the main reason that the sound isolation and the HVAC works to my satisfaction - many thanks Stuart!

The current status is:
- 2 leaf sound isolation structure is done (had to remove the windows and replace the door since the pictures in design thread), and everything works fine - I measured >50db TL.
- HVAC is partly done - silencers are in place and are working great, but AC conditioning is not yet installed.

Pictures of the room:
Back(stale air out + silencers. Picture is stitched from multiple shots - the reason for black areas.):

studio_back.jpg

Front (fresh air in):

studio_front.jpg

Martin

[kominak]

Wood for the soffit structure arrived...

IMG_6139.JPG

... and after some fun with circular saw:

IMG_6144.JPG

[Gregwor]

Glad to see you're back at it!

What out of this world door hinges did you use that allowed you to get away with using only 2 on your heavy door!?????

Also, do you have any silencer box install pictures to share? I'd like to see how you implemented them in an existing build. Moreso how to did the penetrations and connections!

Excited to see more progress!

Greg

[kominak]

Hey Greg,
I don't have the pictures of the insides of the silencers, but I updated my first post to include the link to the sketchup design - the silencers were made exactly like Stuart designed them.
The ducting in the machine room is not yet finished so I didn't post pictures yet.

The door hinges I used are these:
https://www.sfsintec.biz/en/web/technol ... _3d_1.html

The door is made out of solid spruce so they are not that heavy and the hinges are working fine so far (they are rated up to 100kg per pair).

Hopefully after long dreaming I'll make some real progress at last

Martin

[Soundman2020]

Also, do you have any silencer box install pictures to share? I'd like to see how you implemented them in an existing build. Moreso how to did the penetrations and connections!

That was a rather complicated design, due to the restrictions imposed by the existing structure, and also the climate.... Note where Martin lives... It took me quite a while to get that right... Basically, here are sleeves going through the wall to the utility room behind. Top box is supply, bottom one is return, with two ducts along the edges of the ceiling, leading to the front, where the return registers will be, over top of the speaker soffits.

Nice work. Martin! And I'm glad it's working as designed.



- Stuart -

[kominak]

Hello,

does anybody know the optimal weight load per square inch of 50 duro sorbothane in 1/4" thickness (or similar, suitable for soffit mounted speakers) that will get the correct 10-20% compression? Before I start with the math formulas in sorbothane design guide I'm hoping for an easy answer

I'm thinking of using this: Sorbothane strip 36" x 2" x 1/4 50Duro"

I plan to use it the way that Stuart described in this post.

I was trying to contact the company for the answer, but they are trying to sell me the hemispheres instead .


Another question - the central part of the soffit structure. I plan to put big LCD scren there - is it better to build it into the wall (so that the screen is flush with the soffit wall):

13122007539.jpg

Or should I not complicate it, make the whole front wall solid and without cutouts and put the LCD screen in front of that.

My instincts tell me it's better to do it like in the first picture, but they have been wrong before

Many thanks!
Martin

[kominak]

Logging some small progress - the soffit frame is almost done and the speaker boxes are finished. I'm waiting for the speakers to arrive (should be any day now). I'll post the first measurements of the room once they will be at home.

Studio3_stitch_smallsize.jpg

speaker_boxes_stitch_smallsize.jpg

[kominak]

Speakers finally arrrived

Here they are temporarily installed to make some initial audio measurements of the room:

studio4_stiched_smallsize2.jpg

Anybody got some clever trick to make the sorbothane less sticky? I used sheets of paper to slide the speakers into the boxes more easily, but I don't want the paper stuck there for the permanent installation.

The soffit cavities are filled with mineral wool (30kg/m3 density). I took measurement before and after putting the hanging absorbers in the bottom part of the soffit (with both speakers playing).

Before:

without_hangers_both.jpg

After:

with_hangers_both.jpg

To tell the truth I was expecting some obvious improvement from the hanging absorbers, but they actually made the situation worse, if I'm reading the graphs correctly?
Is it possible I didn't construct them the right way? They are structured as follows: 30mm mineral wool - 12mm birch ply - 30mm mineral wool. The mineral wool density is 30kg/m3 and the plywood is on the heavier side.

I'm about to fill the back wall under the silencers with mineral wool (30cm thick) and will measure the room again.

[Soundman2020]

Anybody got some clever trick to make the sorbothane less sticky? I used sheets of paper to slide the speakers into the boxes more easily, but I don't want the paper stuck there for the permanent installation.

Then you didn't build the enclosure boxes correctly! You'll find it next to impossible to slide anything on Sorbothane, and then find it even more impossible to adjust the pads for the correct balance and deflection with the speaker already inside the enclosure.

The correct way to do it is to start by place the speaker on the pads resting ONLY on the bottom panel of the enclosure, OUTSIDE of the soffit (resting on a flat, level workbench or table, for example) to get the balance and deflection precisely correct (important!), THEN build the rest of the box around that, adding in the side and top panels of the box and the other sorbothane pads as you go (also with correct deflection). If you don't get the Sorbothane deflection and balance exactly correct, then the speaker won't "float", and will couple to the enclosure box at some frequencies (or all frequencies!).

By the way, your enclosure boxes don't seem to be massive enough. It looks like only one layer of MDF there: probably not enough, for those speakers.

I took measurement before and after putting the hanging absorbers in the bottom part of the soffit (with both speakers playing)

You need to measure ALL combinations of speakers, not just both! You need to measure them independently and also together. If not, then you cannot see how the mid range and high end are doing, and you cannot see how the speaker balance is doing (left vs. right). Measuring both together is fine for seeing how the low end is working out, but it tells you nothing more than that.

In addition, you have not calibrated your REW setup correctly, so your data isn't much use anyway. You should calibrate then repeat all the tests.

I'm not sure if you have seen this thread: http://www.johnlsayers.com/phpBB2/viewt ... =3&t=21122 that explains how to set up and calibrate REW, and how to take the measurements correctly.

To tell the truth I was expecting some obvious improvement from the hanging absorbers, but they actually made the situation worse, if I'm reading the graphs correctly?

Impossible to say, since you only posted the full-range frequency response graphs, which are not important at all for evaluating bass trap response... We would need to see all of the other data to determine if the hangers are working or not. Frequency response plots are not much use for that, and in fact are the least useful of all the REW graphs in general, for determine the full picture of acoustic response in the room. You also need to examine the data at the correct scale, and with the correct filtering, to see what is really happening.

They are structured as follows: 30mm mineral wool - 12mm birch ply - 30mm mineral wool. The mineral wool density is 30kg/m3 and the plywood is on the heavier side.

Birch ply? You used high density PLYWOOD for the hanger cores? Ummmm!!!! That's about TWICE the density you need, and much too rigid. You should be using Homasote for that. Homasote is a low density light-weight, soft fiber-board. Nothing at all like plywood. It's the type of material used as a backing behind dartboards, and for notice boards in offices, schools, churches, etc. to pin paper messages. That's what you need for the core of acoustic hangers. Plywood is much too dense, and much too hard (rigid).

I'm about to fill the back wall under the silencers with mineral wool (30cm thick) and will measure the room again.

Before you do that (and I'm sure you are not going to like what I'm about to say! ), I'd suggest that you take ALL of that out again, mount the speakers correctly inside their enclosure boxes, correctly balanced and floated on their Sorbothane pads, do the complete REW calibration procedure and the full "baseline - empty room" test with REW, exactly as outlined in the thread I linked above, with NO treatment at all in there: just the speakers, nothing else. Then put JUST the hangers in (including the insulation on the back, side, and bottom of the hanger enclosure), and test again like that. THEN add the other insulation, around the speaker enclosure and above it, and do a third test. Only then will it be possible to see if the treatment is working as it should.

Also, it looks like you have enough space to get one extra hanger in there: if you bring those four a little closer to each other, and slide them all over a bit, I think there's enough space to fit a fifth hanger in there. Highly recommended!

And finally: Are you sure your speakers are set at the correct height? They look rather low to me...


- Stuart -

[kominak]

Oh Stuart, you really know how to spoil the fun
Just kidding - you know how much I value your input!
I was expecting much worse, actually - going through your post it seems like I did at least some of the things right, it just wasn't very obvious from my post.
I'll try to make it more clear - please let me know if I understand you correctly (but don't be afraid to ruin my optimism, if it's needed )

Then you didn't build the enclosure boxes correctly! You'll find it next to impossible to slide anything on Sorbothane...

Exactly! it was impossible to slide the speakers in, but after I covered the sorbothane pads with sheets of paper, it was possible - the paper got attached to the naturally sticky sorbothane and the speakers were sliding on the paper. I'm pretty sure I got the sorbothane deflection right - the results of the calculation were confirmed with actual measurements in place. For these REW measurements I used 2 rectangular pads of 50 duro sorbothane. Here's the calculation for one pad bearing half of the speaker weight:

Sorbothane_pads.png

The weight is 50lbs per speaker. To calculate the height of the box I took speaker height, added 2x height of the sorbothane pad and deduct 2x deflection. Same for the width.
I used two pads on the bottom and two pads on the top - after I found the balanced position on the bottom I glued them in and mirrored their exact position on the top with another two pads. I didn't worry about the sides just yet.
I really didn't want to leave the speakers permanently in the soffits just yet - I'm too afraid I'm gona scratch them when building things around them - that's why I'm looking for a way to slide them in and out.

For permanent installation I'm gonna use thin strips around the perimeter of the box, like you suggested in this post.

Calculation for one strip/one side of the box/one speaker (perimeter + strategically placed small pads/offcuts from this strip):

Sorbothane_strip.png

By the way, your enclosure boxes don't seem to be massive enough. It looks like only one layer of MDF there: probably not enough, for those speakers

It's two layers of 22mm MDF (with the exception of bottom, where there's only one layer, that's screwed to another MDF in the soffit structure) - it feels pretty rigid (and heavy!).

... you have not calibrated your REW setup correctly, so your data isn't much use anyway. You should calibrate then repeat all the tests.

Hm, why do you think so? I did actually calibrate my measurement system: I'm using miniDSP UMIK usb measurement mic, which is natively recognised by REW software and it comes with it's own unique calibration file from manufacturer. It's possible I did something wrong, but the setup seemed pretty straightforward. Here are the setup/measurement steps I followed: https://www.minidsp.com/applications/ac ... p-with-rew. With the exception of the desired sound level (75dB vs 80dB) it looks identical to your guide. I included the calibration file with the measurements below.

Birch ply? You used high density PLYWOOD for the hanger cores? Ummm!!!!! That's about TWICE the density you need, and much too rigid. You should be using Homasote for that. Homasote is a low density light-weight, soft fiber-board

You got me on this one Thank you for your explanation! I was searching for Homasote, but it's not available around here... do you perhaps know of some substitute material? Would cork work? Or very lightweight soft thin plywood (balsa)?

...since you only posted the full-range frequency response graphs, which are not important at all for evaluating bass trap response...

...do the complete REW calibration procedure and the full "baseline - empty room" test with REW, exactly as outlined in the thread I linked above, with NO treatment at all in there: just the speakers, nothing else

By some lucky chance I do have measurements of the empty room, taken some time ago (calibrated for 75dB sound level, too): Baseline_empty_room. It's not the same speakers, but I hope that it could still be usefull, given that we're solving 40dB hills and dips here, not 5db differences between speakers...

Here are the measurements of the speakers with soffit stuffed with insulation and hangers in place: Soffit_hangers

And here is the measurement with insulation filled soffits but without hangers: Soffit_without_hangers

Also, it looks like you have enough space to get one extra hanger in there: if you bring those four a little closer to each other, and slide them all over a bit, I think there's enough space to fit a fifth hanger in there. Highly recommended!

Will definitely do this when rebuilding the hangers. Thanks for the recommendation!

And finally: Are you sure your speakers are set at the correct height? They look rather low to me...

The acoustic axis of the speaker (center of the middle driver) is 125cm above the ground, as per your design - the picture is stiched from multiple shots so the perspective may be a little misleading...


Many thanks for your input, Stuart - I really appreciate it!

As you can probably feel, I'd REALLY hate to have to remove all the insulation, break the speaker boxes and start again . But will do it, if absolutely necessary.
I'll definitely rebuild the hangers and post measurements then...

Thank you once again!
Martin

[Soundman2020]

What speakers are those? (Make, model)

Are you planning to use a sub as well?

You have your resonant frequency rather high for the rubber pad calculations: 31 Hz implies that you won't get decoupling until about 50 Hz, and good isolation starts at about 90 Hz. Assuming you won't be using a sub, that's too high. You should be shooting for a resonant frequency much lower than that. There's a LOT of energy in the low end, between 35 and 100 Hz, in modern music. Try to get resonance down to less than 20 Hz. 15 if possible. Yes, it's hard to do that!

You also have your excitation frequency set too high: more realistic would be at the lowest fundamental frequency that any of your instruments produce: maybe 35 Hz or so, which is in the range of the bass, and even keyboards.

With the exception of the desired sound level (75dB vs 80dB)

86 dBC, with all speakers running (80 dBC individually, for two speakers). That's the "standard" calibration level for movie houses and studios, and is a typical monitoring level while mixing. If you don't run the system hot enough while testing, you won't necessarily trigger all the potentially bad modes.

The acoustic axis of the speaker (center of the middle driver) is 125cm above the ground, as per your design - the picture is stiched from multiple shots so the perspective may be a little misleading...

OK, that should be about right. It's probably just the image that makes it look too low.

I don't have time to analyze your REW data right now, but I will when I can.

- Stuart -

[Soundman2020]

Added later: There's a major issue with all of your measurements. You have very severe mains hum interference:

Martin-REW--WF-50hz-hum.png

That huge spike at 50 Hz that never dies away is electrical mains hum. It's only 10 dB down from the calibration level (65 dB vs 75 dB), so it must be VERY audible! (Unless it is a faulty mic or mic cable, in which case you would not hear it).

It's in all of your measurements (except the original ones that you said were done with different speakers). The strange thing is that it seems to be phase inverted between the speakers, so I'd check all your cabling too. Perhaps you have an XLR cable that has pin 2 and 3 swapped at one end? I'm not 100% sure if that's the case, but it looks like it.

So you'll need to solve that issue first, then repeat all of your measurements, and we can take a look at the data.

but I hope that it could still be usefull, given that we're solving 40dB hills and dips here, not 5db differences between speakers...

Actually, you don't have 5 dB difference between speakers: you have 26 dB difference between speakers! Here's the plot:

Martin--REW--Speaker-Difference-Comparison.png

26.04 dB difference, to be more precise. That's VERY large! Something strange going on there. So yes, it seems that we ARE solving rather large hills and dips between speakers!

- Stuart -

[kominak]

What speakers are those? (Make, model)

ATC HTS40 - they are from their HiFi line, but ATC uses the same drivers/crossovers in both their Pro and HiFi speakers (HTS40 uses the same drivers as SCM25 and SCM45).
These are sealed enclosure, passive speakers, with crossover optimised for near wall placement (so the baffle step EQ won't be needed).

Are you planning to use a sub as well?

No, these should go reasonably low.

Try to get resonance down to less than 20 Hz. 15 if possible. Yes, it's hard to do that!

I played with sorbothane calculator and to get under 20Hz I will need to use thicker sorbothane (1/2" instead of 1/4").

I want to ask - how critical it is to seal the speaker within the box with the sorbothane strip around the perimeter? Versus using only pads?


Measurements:

That huge spike at 50 Hz that never dies away is electrical mains hum. It's only 10 dB down from the calibration level (65 dB vs 75 dB), so it must be VERY audible! (Unless it is a faulty mic or mic cable, in which case you would not hear it).

Good catch! It seems that the USB mic cable was faulty. Thank you!

86 dBC, with all speakers running (80 dBC individually, for two speakers). That's the "standard" calibration level for movie houses and studios, and is a typical monitoring level while mixing. If you don't run the system hot enough while testing, you won't necessarily trigger all the potentially bad modes.

I took a new set of measurements - should be calibrated correctly and I also hope I solved the 50Hz hum problem - could you please do a quick check?
Here's the MDAt file: soffit_with_hangers_calibrated

Today I found the material similar to Homasote - soft cellulose fiber board with the density around 280kg/m3. I will rebuild the hangers tomorrow.

Actually, you don't have 5 dB difference between speakers: you have 26 dB difference between speakers!

I suppose you are talking about the difference between baseline measurements (old speakers) and measurements with new speakers? Or left vs right speaker?

I remembered the baseline measurement was done in really empty room - without the silencers . As they are quite massive, could they be responsible for the differences in old and new measurements?

To honor the effort you put into this I finally decided to do it right and follow your advice - I'll take all the insulation plus hangers out of the picture and do a new baseline measurement - hopefully tomorrow.

Many thanks!

Martin

[Soundman2020]

What speakers are those? (Make, model)

ATC HTS40 -

That's what I thought. OK, specs say they go down to 32 Hz, and REW agrees: It's showing -6 dB at 32 Hz, which is exactly on spec. So that's what you should set your excitation frequency to when figuring your pads: 32 Hz. That's the lowest frequency that they can reproduce comfortably, and it can be reached by a 6-string bass, or an organ, or a piano, or most keyboards, and even some electric guitars.

These are sealed enclosure, passive speakers, with crossover optimised for near wall placement (so the baffle step EQ won't be needed).

That's fine. You don't have your baffles on yet, so that explains why the level is slightly low. But that can be fixed, if need be, with digital tuning.

No, these should go reasonably low.

Even more reason why you should be looking to isolate them down to low frequencies! If you did have a sub, then there would be a crossover that rolls off the low end of the mains below about 80 Hz, give or take, so it wouldn't have been quite as big an issue. But with no sub, your mains are covering the entire spectrum, so they do need to be isolated down as far as they go.

I played with sorbothane calculator and to get under 20Hz I will need to use thicker sorbothane (1/2" instead of 1/4").

Right! At least 1/2". Sometimes I have to use thicker stuff.

I want to ask - how critical it is to seal the speaker within the box with the sorbothane strip around the perimeter? Versus using only pads?

Your speakers are sealed, not ported, so there won't be too much over-pressure "pumping" inside the soffit. Thus, it's not critical in your case. It matters a lot more for folks who have rear-ported speakers.

I took a new set of measurements - should be calibrated correctly and I also hope I solved the 50Hz hum problem - could you please do a quick check?

Looks god from here! No more mains hum, and the levels seem fine. Make sure you keep those exact levels for all future measurements!

Today I found the material similar to Homasote - soft cellulose fiber board with the density around 280kg/m3. I will rebuild the hangers tomorrow.

Perfect!

I suppose you are talking about the difference between baseline measurements (old speakers) and measurements with new speakers? Or left vs right speaker?

Nope! That's the difference between your left and right speaker, in the most recent measurement set you did until yesterday. And it has not improved in today's version. In fact, it is worse:

Martin--REW--Speaker-Difference-Comparison--V2.png

Probably due to the higher level after calibration. But clearly, that's a big problem. Specs call for +/-3 dB, worst case, and preferably +/-1 dB.

I'll take all the insulation plus hangers out of the picture and do a new baseline measurement - hopefully tomorrow.

But I would also get the speaker mount fixed first: you do need to be sure that you aren't transferring any vibrations between the speakers and soffit structure (both directions: to and from).


- Stuart -

[kominak]

Rebuilding the hangers took longer than expected, so no new measurements today...

The hangers are much lighter in weight and somewhat more "springy". I think I get it now, why it should work much better.

...specs say they go down to 32 Hz, and REW agrees: It's showing -6 dB at 32 Hz, which is exactly on spec. So that's what you should set your excitation frequency to when figuring your pads: 32 Hz.

Looks like these 4 disc pads should do the trick? (50 duro, 0,5" thick, 1" diameter):

Sorbothane_discs.png