Hello!

I could use some advice on what to do next to treat my room. Thank you very much to anyone who is so inclined.

If you want more details on my room here is my build thread.

Here are the basics of the room.

Width = 131" (333cm)
Height = 90" (229cm)
Length = 308" (780cm)

This is a finished corner of my basement that I built myself. Forgive my poor construction choices. I overestimated my need for isolation and then we had a baby... Oh well.

Attached are my most recent REW results.

I'm trying to get my subwoofers and speakers positioned correctly and also decide what to do with the ceiling, walls and door to the room. The room is a bit leaky (I haven't finished the door, but at the moment isolation is not a concern, and I haven't figured out the ventilation yet, so for now I'm leaving the door frame a bit leaky and the window stays open a little. I haven't passed out and I've spent a lot of time in here.

And the leaks can only help with the LF, right? .

My current plan is to get some more R30, stuff some more boxes with it and experiment with other positions.

Once I find something that works, I will think about how to install them in a way that is less... cardboard. But it's useful for testing.

Obviously, I have an issue at 79 Hz (Floor ceiling mode), which makes sense since I have barely treated the ceiling. At the moment, I'm most concerned with 150 Hz+ I assume it's all the bare walls and ceiling. I will have to cover some or all... but at what depth and with what material

Thanks!

Edit: these measurements were taken with the phase on the left sub reversed.

One of the reasons I did not add more drywall to the left wall, front wall, and rear wall with door is because when I got to this stage I tested the isolation and was pleased with the results. Probably wasn’t the smartest move as it is making speaker and subwoofer positioning more complicated since the boundaries don’t seem to reflect sound in the same way.

I will add a floor plan.

I realized that I flipped the phase on one of my subs which makes the low end way better, but contributes to that big null. So I guess phase flipping does away with the fundamental modal issues, but not the harmonics.

I think that just means I need to treat more of the ceiling and walls.

You need to hang the bass traps straddling the corners - or just lean them up and straddle the corners.

Do all four corners.

Also, straddle the ceiling to side wall where ever you can.

Then treat the side walls.

https://realtraps.com/placing_mt.htm

Check the noise figures, decay and spectogram around 120hz. Looks like a grounding buzz that you have to fix.

There is audible transformer hum coming from one of my subs. That particular hum is not a ground issue.

It isn’t that loud relative to mixing levels,
do you think it is impacting the audio and measurements?

I do have a really high end UPS that has some sort of power cleaning function. I got it for free from one of the IT guys at work because all of the batteries in it are dead, I think it has some sort of wave generator for cleaning up power issues, but the fans are so loud I haven’t used it. Might be worth replacing them if it fixes this issue, maybe I will give it a try.

Thanks!

All four corners of the room have 18-24” of uncompressed R30 and the ceiling wall corners are treated with 1” 703 + 9” of R30 all the way through the room.

I could do more on the side walls. I could also lose 8-9” of ceiling height and still have a bit of space left overhead.

Its on both of your measurements, measuring around -40db. Yeah that would bug me some personally if i were working in your place. Especially when the rest of the noise floor is low (which is a good thing congrats).

And yeah you could just get used to it and work around it, but why not just fix it?

Finding the right listening position is tough because of the dimensions. The best position i've found is around 18% from the front wall, but this is in a perfect situation where every surface is trapped perfectly( except for floor).

And even in this situation there will be a slight tilt in terms of low to hi.
(see attached pic).

Wow thanks! I think I could live with that response!

Good call on the buzz, it’s worth opening up the sub at least. I will give that a shot.

Thanks! It is quiet in there, when the dehumidifier and air purifier are off, both doors are shut and no trucks are idling nearby!

I think I made a mistake trying to make the whole front wall a trap. I figured that it would be fine to have the speakers out a couple of feet from the wall if I had the whole thing covered with 18” of R30. I’m going to leave it alone for now.

Someday, perhaps I will remove some of it to get the speakers right up against the wall... although I suppose that might change any treatment I do between now and then.

Whatever you have done is good. Better then alot of the attempts i've seen around here. The recommended NR is 10 or NR 15 max. Your space is around NR 17 which is still really quiet.

If someone were to nitpick the whole -5db null between 173hz and 308hz would be worth investigating, and some of the filtered ETC reflections on the left speaker would need to be supressed or redirected( around 500hz and 630hz).
Not to mention the early part of the ETC could be a little cleaner. This would help the clarity between 2.8khz-5.8khz.

Could the decay between 50hz-100hz around 250-300ms be suppressed more? Sure. How to accomplish with the treatment in space already?

Thanks! Having no HVAC doesn't hurt (except in July, August, January, and February). I need to build a hush box for the computer, it's very quiet but when I'm running anything GPU intensive it ramps up. I also need a new GPU.

I guess it looks a lot worse when you zoom in the graph like I have. For the most part, I am thrilled with the flatness of the Frequency Response in the low end, although the waterfall doesn't look quite as nice.


I am still learning (I have no idea) how to interpret ETC. I've attached the screen shots of the L+R measurements for 500Hz and 630Hz. I believe I have some reading to do in order to get it. Thank you for pointing out that there is something for me to see there!




Again, I will read until I can see what you are telling me.



I am probably going to move the panels that are straddling the floor/wall corners to the ceiling. They're 8" of Roxul 60. I'm sure that will help in some areas and hurt in others, but I don't think I can really leave them on the floor where they are. They take up a ton of floor space in a narrow room. I could possibly redesign them and make pseudo furniture that functions like a floor wall trap, but could also be used as a keyboard stand or equipment rack. Once they are on the ceiling, I can fix the problems that creates, hopefully without too much expense (of time, material, and floor space)

I've attached a really basic floor plan.


Thank you again for your attention and advice!

I’ve been following your build. Read the build thread the other day (fully) and saw this thread start up the other day. Keep up the great work. Your attention to detail and tenacity in the build is inspiring. Cheers.

Hey Frankie!

Thanks, bro! It’s really nice to hear you say that, especially since so many times through this process I have felt defeated by my lack of knowledge, time, or funds.

It has been a great learning experience! I’ve even gotten to use my rudimentary acoustics knowledge at my day job (AV specialist at a college) to help do some basic treatment (and avoid $5000 installs that include only foam). I mean the main thing I’ve done has been order stuff from GIK, but still. (Oh man, I love their soffit traps... so convenient and easy, and no dust or itchiness).

I have spent way more time thinking about this thing than building it or making music, but I have been able to get some money out of the studio through voiceovers and stuff so that’s cool.

My greatest fear is that as soon as its done we will have to move. Fingers crossed!

It’s been a fun read and follow. I really dig how dedicated you’ve been to it. A two year gap (roughly?) and that prob involved much thought processes during the break. Living in south jersey I understand high water tables and how bad our basements are here. Your upstate New York and “on a hill” so I’m glad it’s been two years and you haven’t had any water problems. I was concerned for bad air quality at minimum (60% humidity is the high end of normal) and basements can’t get musty. I’m glad this hasn’t been an issue for you. It would suck if you wound up moving (growing family) but you know, it could be a touch of a selling point. A private office? A gym? Plus there are many musicians in the world that may even take advantage of that space. Glad you got some money out of it so far as well. Sorry I can’t help with the acoustics, not really my thing, and your in good hands with Thrill for sure. Keep up the good work man. Can’t wait to see it finished

Congrats, studio looks like a nice fun place to create.

Ahhh!! No HVAC. That makes sense now. I was wondering how you got the AC so quiet in the basement.

Yeah no AC is a deal breaker for me in any studio i've ever worked in. The talent will never survive in the heat and so much insulation in an enclosed space traps even more heat, so it can actually feel hotter.

Instead of building a box, just change the fans for the quietest you can find.


I was thinking best or easiest case scenario of 1/3 or 1/6th resolution and the response still falls short, but not by much.




I've attached the reflections you may want to suppress for 500hz/630hz.

First of all i find it easier to study the Impulse Response and ETC with a linear Y axis (set to %FS instead of dBFS).

The Filtered ETC is primarily aimed at examining decay behaviour in different frequency bands and analysing the results per ISO 3382.

When studying the filtered ETC ideally you want the Schroeder integral curve to look like a straight line or slope with no bumps.

When you select show data panel The RT60 figures include the decay range over which they have been calculated and an "r" value, the regression coefficient, which measures how well the data corresponds to a straight line. A value of -1 would indicate a perfect fit, values lower in magnitude than -0.98 indicate the corresponding decay figure may not be reliable.

So for example if you look at at the 500hz filtered ETC data panel, EDT, T20 and T30 are all in orange. This means they aren't close enough to -1(below -.98).


Looking at 500 hz with % selected, you have a reflection that is 40% energy compared to the impulse( which is alot). It had to travel 8.69ms(9ft 9 inches) further to reach the microphone compared to the impulse itself. The impulse measurement is the distance from the speaker to the mic, so if the measurement mic is 10 ft from speaker add 9 ft 9 inches or 19ft 9 inches of travel for reflection.




If you do the entire ceiling it will hit the ceiling corners as well.

Me either!

Yeah, Thrill is helping me rearrange my understanding pretty well. Thank you!

I am lucky about the water. Good drainage. A hill and no leaves to clog the gutters on the front side of the house!

And yeah, home gym! Already has the rubber floor and everything!

Thanks Kyle! Coming from you that means a lot! I hope you’re right! I’m starting to love it!

Thrill,


that’s a lot to digest. Thank you for it.

I will dig into it tomorrow.

Re HVAC, This is just for my personal use, or helping a few friends with some projects. It isn’t too bad. If I sealed the door and window I wouldn’t be able to stand it in there. At this point I’m thinking a split ductless and an ERV/HRV somewhere down the line, unless they start making better all in one tiny home HVAC units. I almost bought a climateright but apparently they don’t make them anymore (for good reason, I’m sure).

For now I wear this cooling scarf soaked in water and turn the dehumidifier on when the hygrometer hits 55. Pretty funny, right? It really isn’t bad during the winter. The furnace is old and leaks a lot of heat into the basement. It is dry as hell down there during the winter... which is good.

I do worry a bit about radon. We had the house tested and it was fine, but after I sealed this room up, I should test again just in the studio.

I made a couple more traps today. I will hopefully measure more on Friday.

Sadly the loudest part of the computer is the graphics card. I actually really like my case. It has some dampening material in the covers and the air intakes are baffled.

Here's a link in case anyone needs a quieter case.


That is much easier to read! Thank you.

... Schroeder integral curve... yes...


"Thus, EDC (Schroeder Integral Curve) is the total amount of signal energy remaining in the reverberator impulse response at time. The EDC decays more smoothly than the impulse response itself, and so it is more useful than ordinary amplitude envelopes for estimating RT60."

So this curve measures the amount of energy at a frequency, relative to the original source signal and it should decay smoothly, which would indicate that there is less constructive/destructive interference with the original signal?

I'm fairly certain that I don't understand, but I will read about it. Also, I found ISO 3382, but I can't drop $80 on it at the moment lol. I will look at the summary again.


Wow! So the data panel shows how close or far from T20 and T30 the measurement is at a given frequency! Awesome! I always wished there was a "how far off am I?" function, and this graph has it!



Right... I remember this now from physics class, and again from trying to fix past rooms.

343m/s or 3m/ms.

I think the mic was about 5' and so that impulse went around 13-15ft. That's definitely the opposite side wall FR point. So now I know I need to add treatment to that location that is effective between 500-630.

I've probably read good explanations of that a ton of times but I needed to be walked through it (again). Thank you!

The ceiling/wall corners are straddled with 1" Armstrong Optima Ceiling tiles. (picture).
The space behind the 24" front is loosely filled with R30. I have considered moving the ceiling tile material (looks like 703 or 701) to the back of the soffit and covering the front with fabric.

Incidentally, this is a very cheap and quick way to make a decent ceiling wall corner trap that looks pretty good considering the amount of work and materials that went into it, but it does take away options for full ceiling treatment like running some wires across the ceiling filling with R30 and then covering with fabric.

Thanks again for your help!

If you want to learn about ISO 3382 watch this slide:
https://www.slideshare.net/narasimha...c-measurements

The most difficult part when analyzing measurements for octave and 1/3 filtered measurements is that the EQ's used will affect the result. That is why REW gives you 2 options: Time reversed filtering for octave measurements and zero phase filtering for 1/3 filtering.



The RT60 EDT, T20 and T30 does as well. But for anything below 200 hz you have to be careful in the results as the filtering affects the decay. For anything below 200hz, the Waterfall, Spectral decay and Spectogram gives you better and more useable results.




The challenge is how to attack the reflection without throwing off the decay at that frequency. Your RT60 Topt is relatively clean. At 500hz its a little short (.129) compared to the rest of the bands( 200hz-10khz). So redirecting the reflection to a place you have absorption already may be a better plan than absorption. But your lack of space makes it more challenging.



I only suggested if you did the entire ceiling which i always suggest for treating any small rooms or rooms with low ceiling.

By the way earlier when i modeled your space in the room simulator, i put you sitting a 1/3 from the floor which for some people is pretty low. If you have to sit near or halfway the response will change.

Thrill,

Thanks for those slides. Very concise explanation. I will have to absorb that over time I think. (Just noticed the pun... it wasn’t intentional, but let’s say it was).

How did you export that window?!

That looks incredibly useful.

I tried putting the speakers out wide, in the same plane as the subs. I still need to experiment with positioning, as the best position seems to change each time I change treatment.

I put one trap that was on a floor ceiling corner up on some light stands to see how it would do with more of the ceiling treated. It seemed to help with the 630Hz bounce, but not the 500Hz bounce.

I added 8 boxes 18x18x15 filled with uncompressed R30 at the side FR points. Seems to have improved things a bit.

I will probably have to add panels to the ceiling at the rear and center of the room as well, hopefully that will help more than it hurts.

I also added some R30 backed with 1" 703 (ceiling tiles) that I had.

The R30 was pretty gross, so I put it in trash bags (uncompressed) in the corner behind the boxes. I screwed the bags to a furring strip and propped it between the ceiling and the floor.

The waterfall looks much better in the low end now!!

I tried to match the settings of the original screenshots that Thrillfactor posted for easier comparison, but the MDat is included as well if anyone has time to look at it.

Edit:

WTF is that 300Hz null?! Floor bounce? Guess I will try different heights.

Edit:

Re sitting with ears at 33% height, that’s only 29.7” and I’m 6’2” so probably won’t work. I’m currently around 52.5” (58%).

Double edit:

I ordered a umik. I have a Nady CM100 and a weird Earthworks iPhone measurement mic with a battery. But I thought at this point it would be good to have a calibration file.

Considering getting the miniDSP 2x4 balanced as well. I’m thinking. That delay and per channel EQ could be useful on the low end if there are phase issues between the monitors and the subs.

Edit:

I guess the bass was flatter with the subs at 1/4 wavelength of the room. From the front and side walls. I suppose I could elevate them 1/4 off the ground as well and even it out even more. That might make the phase issues with the mains worse though.

A couple updates.

I realized that when I posted that measurement the microphone was not in the same position.

This brought me to try to rediscover my positioning when I took the measurement that I shared initially. I should take better notes.

But anyway. If I am trying to find the best listening position so far I have been most focused on the 30-100Hz range, since that is the hardest to treat. But I've noticed that as I move the microphone back, my low end flattens out, but the dip at 300Hz gets sharper.

So, should I leave my mic in the position where the low end is the flattest while treating, or should I be concerned with balancing the sub 100Hz and the low mids at this point?

Also, I just got a MiniDSP 2x4 Balanced and a umik. Hoping to dig into those tomorrow!

Also, just so I don't forget this. Currently settings are:

Thanks,

I'm no subwoofer expert but to me "it feels like" positioning of mains is most important and then... you add some low end as smoothly as possible?

RE your hum issues. (Sorry if this is the wrong thread for that comment?)

Noise Tips:
When dealing with Hum & Noise in a room it’s convenient to use REW RTA display. (No sweep this time)
Tweak your preamp /monitor levels, switch off LEDS, humming subs..
while simultanously watching what happens with 60 & 120Hz on the screen in Real Time!
Pic 1 RTA (Note:Example here No SPL calibration)
(IMHO, very educational to also get a feel for SPL Levels)

When you’re up and running with REW and Umik, just open the RTA window in REW. Choose: RTA 1/48 Oct, adjust the scaling for level/ frequency of interest.
Or ”Spectrum” (FFT Narrowband).

Position Tips
While still watching the RTA screen. Open REW Generator
Choose: Pink Periodic Noise. (Pink PN).
(No sweep but a periodic noise signal instead)
Set up like this (REW 5.19):



Move your microphone (or speaker) and see what happens in Real Time.
Save when y find something good. Compare with the Swept measurement.

Best

PS It's quite boring to listen to the repetetive noise so Y can band-limit to a narrow (Custom) low range.

I can’t believe it was the freaking LED... and my gain structure... that second one is more believable.

Thanks for both tips Akebrake!

I have gotten a bit spread out with my advice requests ( what I was trying to avoid when I started this thread... ������) so I appreciate your bringing me back on point. I will consolidate now that I have gotten the noise issue mostly resolved.

RE: FFT real-time analysis. I tried that long ago with my MacBook Pro and the frame rate was almost unusable and the spikes were huge. Perhaps it is better now (in this version, on windows, or there is a setting I had wrong...)

I will try this and see.

I have found optimal positions for each driver on its own (many times with and without treatment) but when I put the system all together the optimal position for each driver doesn’t really add up to the optimal position for the system. Perhaps this is due to phase, delay, different boundary rigidity issues.

Once I get these 22 batts of R38 in place I will experiment with positioning again. 11 down... 11 to go!

Thanks for your advice!

Edit: I’m sure I failed to engage smoothing when I tried to use the RTA before. I have a decent pocket RTA at work. I should probably borrow that again some time.

As soon as "personal Java" was included in REW everything went smooth.
I'm on Mac OSX 10.10.5 (late 2013) and it works fine for 5.19 but I can't open the last 20 betas on 5.20 because my OSX is getting old...

Pocket RTA? No!
Not needed because this feature is bult into REW!
REW RTA + Pink PN is complete with your calibrated Umik.
Try it! Just a couple of clicks on the REW screen. Very handy!

Pink PN gives you a stable reading with whatever resolution you like (1/48, 1/3 Oct) because send/ return are syncronised.
No averaging needed and y can save it (in a row) with "ordinary" saved REW measurement on the same screen for comparison. ("All SPL window")

Of course you'll miss the time (phase) information but see it as a quick complement to ordinary REW IR's.

A 20 times faster way to observe what happens when you move a speaker or the microphone.

Best

Hello again!

I've made some progress.

Hung R38 (almost) all around the mix position, put my little desk back in place, and messed around with speaker positioning a bit. I put a bunch of R30 boxes on a dolly that I can wheel in front of the door. I also wrapped a bunch of the 703 ceiling tiles in fabric and a bit of cardboard and put them behind the speakers, they helped when I didn't have the desk, but I will probably try removing them.

I played around with speaker positioning and got rid of some of the comb filtering.

Attached are some pictures and REW files.

I will need to play with the sub positions now that more treatment is in place to optimize them, but I'm not sure if I'm going to find a better spot than where they are.

Now I have to decide how much I have to treat the back half of the studio. Then make it not look so terrible.

Edit: I know Thrill is going to tell me to clean up the early EDT (or he will get tired of telling me that and not reply). I just don't know what is going on there. I can only imagine that it's the stands (and made worse by the desk). At least the 5ms bounce went away.

The results look pretty good to me though!

By the way, thank you very much Akebrake! This is way faster than what I was doing, and I can get rid of the comb filtering caused by my mic being 2mm out of alignment so much faster!

The reason I was thinking about the pocket RTA, is because I thought it would be hard to see the effects of changes from where I was, but with 1/12 smoothing and exponential averaging it was so much better.




EDIT:

Oh and I might as well put up these ugly ugly graphs, which clearly demonstrate the lingering issues. Some of the low end decay issues are mechanical and located elsewhere in the house... Hmm... Maybe I should put sorbothane under my fridge...

Quick question:

What's brought up the noise floor? Is it the portable air conditioner in picture?
Your NR is around 25 right now. The recommended is NR10-NR15 max. I think an NR of 20 is still passable. With an NR or 25 you will have to monitor a little hot( 85db).

Also your RT60 in the highs is 100ms which is really dry, reaching anechoic or headphones. Is this ok for you?

That’s a dehumidifier, I turn it off for tests. I may have forgotten to close the doors, now that they’re hidden behind a wall of boxes. (May have helped with LF decay... oops).

The higher NR is from the refrigerator and the air conditioners being run all the time right now. Mainly it’s the refrigerator. The condenser is right next to the load bearing wall of the house, which is the right wall of my room. All that drywall, RC-2, and green glue helps, but it’s still not quiet enough at times. I think putting a rug underneath it would go a long way towards fixing the issue. Failing that, I don’t know... sorbothane bumpers can a new fridge.

No, it’s definitely fatiguing. I am thinking that I will probably cover the rear wall with cloth walls then add some slats to the front of that.

Also, I may use different panels past the FR points instead of the R38 that’s there now.

Finally, when I move in some synths and an equipment rack or two there will be a lot more surfaces for HF to reflect off of.

Lastly, I might try my hand at making some 1/4 ply quads, not because I think they’re the right thing for the room, but because they probably won’t hurt and they would look cool.

I spent most of the day trying different arrangements:

Baseline: Pretty much where I started.

Ref 1: Hung a panel in the middle of the room. Moved some R38 to the back of the room floor wall corners. Swapped it with 4" 703 panels.

Ref 2: Staggered panels and R38 throughout the back half of the room to try to mitigate flutter echo and reclaim some HF energy. I don't think it worked very well.

Ref 3:
Moved subs from walls to just left and right of mix position.
Put panels on top of subs to combat 300Hz floor bounce from speakers.
Put R38 panels back at the front of the room around mix position.
Put cut up ceiling tiles on desk to combat desk reflections
Lowered desk (still needs to be lower, and smaller). I really just need a keyboard tray and a monitor in front of me, but it would be nice to be able to have some controllers up there as well. Maybe if I make the top of the desk severely angled it will take help.

Moved stacked Armstrong ceiling tiles (703?) from front of room behind speakers. Put in back wall corners on angle in front of corners. I figured they would bring back some HF. It did make some midrange nulls worse. I think that covering the right and left front corners with slats would be a good idea. I am not sure what spacing to use.



So far this one has the most promise. I have some sheets of pegboard that are the same size as my 2x4 bass traps. I might tape them on the front and see what that does. I doubt it will be good, but I might as well check it out.

Stupid upside down pictures...

Edit: oops baseline SPL is way off, but the bass is more even with the subs on the walls. More experimentation is probably necessary.

Hey, look at that!

Moved the mains around
Put the subs back in the front corners
Removed the floor wall traps from the front of the room.
Adjusted crossovers then EQed the subs and mains.

A low shelf boost to the mains helped clean up a big dip at 120Hz.

I think my next step is to buy or build a desk that helps with the early ETC.

2 things to keep in mind:

1) If you are going to focus alot on frequency response, you have to do more( i would recommend minimum 6-10) of measurements around the listening position and average them together to get a real picture of what's happening in terms of the frequency response.

People forget that no one mixes with their head in a vise. Its natural to move your head alot when doing a mix. Sometimes to use a piece of gear you slide left or right or behind you.

This is why you want a sweet spot to be as wide as possible.

2) As you boost your sub its adding alot of 2nd harmonic distortion around 30hz-80hz which is 55% compared to the original signal. This is a lot of distortion. Basically when you are listening to those 808 type bass and kick drums, half of what you hear right now is distortion mixed in.

Something in the gain staging is wrong or your subwoofer is only rated to a certain amount of level before distortion sets in.