The No.1 Website for Pro Audio
Secondary Mix Room Treatment Project
Old 11th July 2020
  #1
Secondary Mix Room Treatment Project

Hello, guys.
It's been a long time since I visited this [back alley portion of the] forum! I still have an untreated, control room, but it's livable and we're still too busy to shut it down for the upgrade I have in mind. But one thing that would help is fleshing out a room we are using for little more than a few vocal lessons, once a week. So we're going to transform this room into a mix station, and possibly use it for sound-for-picture in its projected future. Perhaps, once fleshed out, we can move some of the work from the control room and give that room some proper upgrade attention.

I'll start by describing the room, and let me mention that its construction is set (unchangeable). I will describe wall positions from the perspective of the mix position, roughly the center of the room and slightly forward...

PHYSICAL DIMENSIONS & MATERIALS
It's a small room, 12.0' deep x 12.25' wide x 8.25' tall. The front wall is 1 layer 5/8" drywall hung on 2x4's, against a solid, poured concrete wall with nothing but planet Earth on the other side (dirt). Side and rear walls are 2 layers 5/8" drywall, hung on 2x6's w/ insulation in the cavity, followed by 2 more layers of drywall (5/8" + 1/2"), except for the right wall, which is followed by the outdoor brick wall. Ceiling is also skinned in 2 layers of 5/8" drywall, and is spanned by 18" floor joists, 16" O.C. Floor is poured, finished, sealed concrete. Consequently, the room is highly reflective.

There are also 2 windows; the left wall has one, 3' high x 4' wide, centered, and is a bit high - starting approximately 1 ft from ceiling. The right window, 4' x 4', is also centered, also starts about 1 ft below the ceiling. The entrance is a sealed 32" storm door, glass framed in wood, located in rear wall, in left corner.


Last edited by ejbragg; 11th July 2020 at 10:37 PM.. Reason: Added Room Dim Pic
Old 11th July 2020
  #2
I have also made the following decision, based on room use & space availability, concerning placement of the [only one] subwoofer: It will be centered underneath the console, against the front wall. Therefore, all following room measurements and treatments are based on this position.

As I continue my posts, the true acoustic engineers will know that I'm no pro. I'm just a geeky jerk with a high interest in doing my own room treatments, and I am using this as both a learning tool for myself, as well as a reference to help me see what to do better in my latter control room project. My hope is that someone else may also benefit from this experiment.

I have indeed taken some measurements. I started by making my own spreadsheets, where I could manipulate the data by hand. Although I prefer Matlab, my license expired and I no longer have need to spend that kind of money. Once you guys turned me on to REW, I have appreciated this fine tool, which makes data collection much quicker than my old method.

After creating a sound card calibration file, I shot 24 positions in the room, from 12 points on the floor (2 different elevations - 2 ft & 4 ft). Across the rear of the room, left to right, were points 1-4 (I call this "Row 1". Moving forward to about 1.5 ft from center, I marked another row, points 5-8; "Row 2". About 3 ft further in, I marked a 3rd row, points 9-12; "Row 3". I did not set up a 4th row close to the front row, because this area will be taken by a desk and equipment.


Last edited by ejbragg; 11th July 2020 at 10:42 PM.. Reason: Added pic
Old 11th July 2020
  #3
I started work on the room, beginning with the lowest modal frequency problem. The lowest major (half-wave) mode in this room calculates to about 47Hz. Measurements prove it's more like 45Hz, average. The next axial mode is floor-to-ceiling: about 68Hz. I'm going to ignore this one for now, and treat the lowest axial problem, first.

Ever since I read the article "Tim's Limp Mass Absorbers", I've been intrigued and motivated to give this a try. It seems a much more efficient means to treat low frequencies than the shot-in-the-dark, or broadband bass trap approaches. Although Tim treated frequency modes with traps of small dimensions, I believe the rooms he treated were much larger than mine, so I planned to build some decent-sized boxes and try them out (understanding that smaller rooms generally have larger problems).

I and my colleague built one such absorber, 2 ft wide x 4 ft high and tested its average effect on the room. We placed it in the left front corner. Results showed a 1.0 dB reduction at the targeted frequency.

One thing we forgot to do was place absorbent material close to the membrane. So we opened it back up, installed some insulation, and sealed it back up. The result was a 1.5 dB reduction at the targeted frequency. We built a second unit and tested both. Whether we laid them on their side, stood them upright, stacked them in the same corner, or separated them, the result was the same: ~3.0 dB reduction at the targeted frequency.

Yes, I'll be posting those measurements in a bit.

Here are some pics of the units we built, as well as the room.

By the time of the picture below, we were encouraged enough to build 2 more units. The result of 4 such units offered a drastic improvement to the room....








Limp Mass Trap Computations:

Tim commented once that MLV was a bit too stiff for his taste. I found another response that someone was using rubber roofing with good results. Since it so happens that I have a supply of the thickest rubber roofing (EPDM) available on the American market, I decided to use it...

Calculate Absorber Cavity Depth
EPDM 1 layer = .5623 lb / ft2
2 layers = 1.125 lb / ft2
3 layers = 1.687 lb / ft2
Limp Mass Absorber formula: f0 = 170 / sqrt (m x d)
where m = mass in lb / ft2,
d = depth of enclosure in inches.
m = [one of the EPDM densities above]
Rearrange formula: d = 170^2 / (m x f0^2)
= 170^2 / (m x 46.5^2)
= 23 - 3/4 inch [using 1 layer]
= 11 - 7/8 inch [using 2 layers]
= 7 - 15/16 inch [using 3 layers]

(I liked the double layer result, so our boxes were based on the 11-7/8" result.)

Design Requirement:
The absorbers are tuned to f0; however, narrower cavities have a higher Q than wider ones. Because the nature of limp mass absorbers are generally low in Q, in attempt to focus on the strong mode, build suggestion with an internal partition was suggested. Box material was 3/4" MDF, as usual. (Note that partition is about 1/2" shy of the membrane, to prevent contact.)

Last edited by ejbragg; 17th July 2020 at 05:45 PM.. Reason: Added limp mass absorber computations
Old 12th July 2020
  #4
Here is a BUNCH of graphs for those of you detail-oriented folks....

ROW 1 GRAPHS















Old 12th July 2020
  #5
Old 12th July 2020
  #6
ROW 3 GRAPHS


















POINT-TO-POINT OBSERVATIONS:

On average, both peaks and nulls at and above the 45 Hz target were calmed down; however, nulls below the target freq remain relatively unchanged.

The 65 Hz peak remains relatively unchanged; however, this is a ceiling to floor mode. This mode is most noticeable in the 24" high readings; at 4ft height, this mode will be nulled out.

Specific Observations:

Comparing Same Points at Different Elevations:
1. Sharp divergence occurs at ~67 Hz: Ceiling to floor mode. (expected)

Comparing Points Across Row 1:
1. Points seem congruent until right side of room, where null
develops around 85 - 90 Hz.
2. At 48 inch height, wide null is stronger near side walls.

Comparing Points Across Row 2:
1. Wide dip around 55 Hz is more drastic near side walls at 24 in.
2. Dip is narrow and shifts higher in same manner at 48 in.

Comparing Points Across Row 3:
1. At 24 in, null at 35 Hz near side walls shifts to 27 Hz
(to a lower frequency) near center.
2. At 48 in, null still shifts lower, but less prevalent.
3. At 48 in, huge null around 85 Hz develops near center. No
such occurance is noticeable at 12 in.

Comparing Points Down Column 1:
1. In center of room, null is created close to 45Hz mode (peak)
a) Nearer the rear of the room, the null is above the peak
freq.
b) Nearer the front of the room, the null is below the peak
freq.

Comparing Points Down Column 2 AND 3 AND 4:
1. As in column 1, nulls appear on either side of the peak
mode when near center of the room.

Last edited by ejbragg; 12th July 2020 at 12:52 AM.. Reason: Added Observations
Old 12th July 2020
  #7
Instead of posting all my waterfall graphs, I decided to post one single point measurement as an example, which for the most part, represents the rest pretty well...

Fwiw, I sliced these at 20ms intervals, because I am interested in the 60 ms mark, and felt too lazy to have to count very far to see the results...

WATERFALL GRAPH, POINT 1, BEFORE & AFTER TREATMENT






General Observations:

The development of ringing at sub target freqs does not show in other graphs, except at 19 Hz. There are a few possible culprits: 2 windows, glass door in the corner closest to the mic, a small but heavy desk in the adjacent corner which I refused to remove. I do not know what causes a the 19 Hz ring - this ring shows up in every measurement. AC was turned off during testing.

Last edited by ejbragg; 15th July 2020 at 06:39 PM.. Reason: Added Observations
Old 12th July 2020
  #8
Quote:
Originally Posted by ejbragg View Post
I started work on the room, beginning with the lowest modal frequency problem. The lowest major (half-wave) mode in this room calculates to about 47Hz. Measurements prove it's more like 45Hz, average. The next axial mode is floor-to-ceiling: about 68Hz. I'm going to ignore this one for now, and treat the lowest axial problem, first.

Ever since I read the article "Tim's Limp Mass Absorbers", I've been intrigued and motivated to give this a try. It seems a much more efficient means to treat low frequencies than the shot-in-the-dark, or broadband bass trap approaches. Although Tim treated frequency modes with traps of small dimensions, I believe the rooms he treated were much larger than mine, so I planned to build some decent-sized boxes and try them out (understanding that smaller rooms generally have larger problems).

I and my colleague built one such absorber, 2 ft wide x 4 ft high and tested its average effect on the room. We placed it in the left front corner. Results showed a 1.0 dB reduction at the targeted frequency.

One thing we forgot to do was place absorbent material close to the membrane. So we opened it back up, installed some insulation, and sealed it back up. The result was a 1.5 dB reduction at the targeted frequency. We built a second unit and tested both. Whether we laid them on their side, stood them upright, stacked them in the same corner, or separated them, the result was the same: ~3.0 dB reduction at the targeted frequency.

Yes, I'll be posting those measurements in a bit.

Here are some pics of the units we built, as well as the room.

By the time of the picture below, we were encouraged enough to build 2 more units. The result of 4 such units offered a drastic improvement to the room....





Why not try at center of front and backwall?
Old 12th July 2020
  #9
Now to do some averaging and make some sense out of all this mess...

For my own education (and possibly for others, as well) we can see the characteristics of the frequency shifts as we compare the results across rows of data...

ROW AVERAGES











Old 12th July 2020
  #10
It is also helpful to do the same across the columns of data, left to right...

COLUMN AVERAGES:















Old 12th July 2020
  #11
Bird's Eye Perspective...

What I'm really trying to do here, is assess the effectiveness of the 4 panels constructed. Did I accomplish my goal? If not completely, am I heading in the right direction? It appears to me that the limp mass modules were quite effective at treating the target frequency. But I still have a ~3dB bump that I'd like to deflate a bit more.

TOTAL AVERAGE:






Further Observations:

1. 45 Hz treatment successfully calmed both peaks and nulls [in
nearby upper frequencies] with 4 panels.

2. The panels have little effect on any frequency below the
designed target.

3. The panels have little effect on the 68 Hz mode.

4. We still have a wide rise of about 3 dB close to the target frequency ~43 Hz.

5. Next problem frequency is easy to see, around 67 Hz.

6. Very interesting to me is the fact that the membrane absorber is not effective at attenuating an octave higher, as other design types do.
Old 12th July 2020
  #12
Quote:
Originally Posted by thethrillfactor View Post
Why not try at center of front and backwall?
Ahh, yes. Good thought. I have indeed checked those positions... I will post my findings on that later, as I am now searching for the next (final) position to complete the 45Hz mode treatment.

I started in the left and right corners because:
1. I wanted to hit the lowest (most difficult) mode with treatment in the corners where 3 surfaces come together.
2. I wanted to save some ceiling space along the walls for the floor-to-ceiling mode.

In any case, I did run some tests with the first 2 traps, placing them in areas beside the front (and rear) corners, but their effectiveness was not as noticeable. (And I can't really use the rear corners for permanent treatment, due to the entrance.)
Old 12th July 2020
  #13
Quote:
Originally Posted by ejbragg View Post
Ahh, yes. Good thought. I have indeed checked those positions... I will post my findings on that later, as I am now searching for the next (final) position to complete the 45Hz mode treatment.

I started in the left and right corners because:
1. I wanted to hit the lowest (most difficult) mode with treatment in the corners where 3 surfaces come together.
2. I wanted to save some ceiling space along the walls for the floor-to-ceiling mode.

In any case, I did run some tests with the first 2 traps, placing them in areas beside the front (and rear) corners, but their effectiveness was not as noticeable.
I just thought because you are trying to hit the 1-0-0 length mode, the pressure would be highest at the center of the wall.

Though if its the 1-0-0 mode the pressure will be high on the entire wall.
Old 12th July 2020
  #14
Quote:
Originally Posted by thethrillfactor View Post
I just thought because you are trying to hit the 1-0-0 length mode, the pressure would be highest at the center of the wall.

Though if its the 1-0-0 mode the pressure will be high on the entire wall.
That does make sense.
In fact, that at least lines up with my next notion / finding, which is that to attenuate the last 3 dB, the rear wall, center would be an excellent spot for the last 45Hz trap. Again, I'm about to post...
Old 12th July 2020
  #15
The following is a list of test points I conducted in search of the best spot for the final 45Hz trap(s). I am also looking for the next spot for the 68 Hz & 90 Hz traps.

These positions are all at the ceiling. I am not able to use the floor positions in any other areas of this room.

Old 12th July 2020
  #16
And here are the results of those positions with the other 4 traps in place...

S1, S2, &S3:


S12 & S4:


S11 & S5:


S10 & S6:


S7, S8, &S9:
Old 12th July 2020
  #17
And this is where I'm going to take a break and await for some input. And anyway, I have some other projects I need to get to. I appreciate you guys!
@ thethrillfactor , I will also take this time to more completely digest your suggestion... I was in abit of a hurry, this evening. Thanks for your suggestion.

Last edited by ejbragg; 12th July 2020 at 05:32 AM..
Old 12th July 2020
  #18
Quote:
Originally Posted by ejbragg View Post
And this is where I'm going to take a break and await for some input. And anyway, I have some other projects I need to get to. I appreciate you guys!
@ thethrillfactor , I will also take this time to more completely digest your suggestion... I was in abit of a hurry, this evening. Thanks for your suggestion.
Center rear seems the best location but you may have to delay it to align it properly.
Old 12th July 2020
  #19
Quote:
Originally Posted by thethrillfactor View Post
Center rear seems the best location but you may have to delay it to align it properly.
I’m not following you on the delay. What do you mean?

Edit: .... unless you’re referring to time alignment with the mains? If that’s what you mean, I’m not working on that, yet. I don’t intend to do any monitor correction at this point.... maybe never, as far as EQ correction. Phase matching /delay, yes, for sure, once I add the sub to my mains. But not for a while. Just working on low freq room modes for now, then overall bass response. Let me know if I’m misunderstanding you. Thanks.

Last edited by ejbragg; 13th July 2020 at 04:30 AM..
Old 15th July 2020
  #20
In hindsight, I have found a couple errors in my design/build:
1. When adding batting, the formula should be adjusted to reveal a shallower depth. Art Noxon, in his article wrote that a fully isothermic chamber should dictate a change in the constant from 170 to 143, due to the diminishing action of heating/cooling of the air. Since I filled approximately 1/4 of the space, I should have split the difference accordingly; 1/4 between 170 and 143 is about 163. This would have revealed a more shallow cavity.

2. On the other hand, because my batting was a bit too dense for the application, air does not freely move through the fibers. Therefore the cavity appears to be more shallow to the air flow.

Perhaps this is why, despite having built the cabinet too deep, it still seems to calm the target frequency. I’m supposing that if I were to have built the case correctly (more shallow) and used less dense batting, I’d still have landed on target and had more efficient absorption (more peak absorption) at the target frequency. And actually, if I am to be fair, it appears I actually missed the target slightly ... the peak attenuation being slightly higher than the peak room mode.

In any case, the suggestion by @ thethrillfactor agrees with the last data I assembled for the next absorber unit: I think I’ll build one for the center, rear wall/ceiling, which is almost directly in-line with the sub output.

Last edited by ejbragg; 15th July 2020 at 06:52 PM.. Reason: Added next step
Old 3rd August 2020
  #21
Ok. We have built a 5th box, tuned again to 45 Hz (ish), and placed it in the ceiling, center, opposite the subwoofer (see post 15, position S8). And indeed, the 45 Hz mode seems to have been reduced a bit more. However, a weird phenomenon has appeared. There is a new peak at 50 Hz. This did not show up until this last absorber unit, so something about it, or its effect on the room is at fault. I can't decide if the room is better off with or without this unit. Here are some shots of before and after the 5th unit ["BEFORE" referring to the room w/ 4 units in the front 2 corners]....

Last edited by ejbragg; 3rd August 2020 at 04:50 PM..
Old 3rd August 2020
  #22
Position 1, 4 Units vs. 5 Units:


Position 2, 4 Units vs. 5 Units:


Position 3, 4 Units vs. 5 Units:


Position 4, 4 Units vs. 5 Units:


Position 5, 4 Units vs. 5 Units:


Position 6, 4 Units vs. 5 Units:


Position 7, 4 Units vs. 5 Units:


Position 8, 4 Units vs. 5 Units:


Position 9, 4 Units vs. 5 Units:


Position 10, 4 Units vs. 5 Units:


Position 11, 4 Units vs. 5 Units:


Position 12, 4 Units vs. 5 Units:
Old 3rd August 2020
  #23
That's a lot to look at. I prefer to put more weight on the averages. Here's a look at the average of all points at 24 inches height...
  1. Before any treatment
  2. After 4 Units Installed
  3. After 5 Units Installed

Position 11, 4 Units vs. 5 Units:


Anyone have any ideas what might be causing this phenomenon?
Old 3rd August 2020
  #24
Quote:
Originally Posted by ejbragg View Post

Anyone have any ideas what might be causing this phenomenon?
What phenomenon are you referring to?
Old 3rd August 2020
  #25
Quote:
Originally Posted by thethrillfactor View Post
What phenomenon are you referring to?
The appearance of the 50 Hz peak. (post #21 ).
Old 5th August 2020
  #26
bmp...

Does no one have any insight on this?
Old 5th August 2020
  #27
Gear Nut
 

I don’t ej........but I do have an alternative method to share that comes from Harmon research and fluid dynamics based on wave propagation in a closed volume which makes a lot of sense when you consider the challenges faced in a small studio

So.....how this works is instead of knocking down modes and filling nulls which is incredibly complex and can eat up valuable floor and wall space, the method is to have multiple subs at multiple locations on different vertical plains. Now before you scratch your head.....we’re talking four seperate small sealed subwoofers with an F3 of 35hz....and they don’t need to be loud as the combined SPL only needs to meet reasonable levels and FM curves. Three of the subs are placed in the corners of the room and one is placed along the longer side wall midway up the vertical plain. We’re talking 1 sqft floorspace per for a 10” driver

Now the result is less obvious but if you wrap your head around fluid dynamics, it makes more sense. One propogator drives a wave with deeper nulls and modes which get amplified when combined with phase dependant reflections resulting in even deeper variance from flat......think of this like a tsunami and the two or three extremely power full waves that bash the shore over time. Now consider multiple propagations of waves created in a more chaotic method....all interacting quickly and absorbing energy from one another in a similar time domain.......the end result is exponentially more peaks and nulls with exponentially reduced amplitude. Think of this as boating in a lake or bay where there’s 2 foot waves on a windy day from all directions....but out in the ocean the result is ocean swell with long durations and wave heights of 6-8 feet!

Now it’s not a new concept....Harmon introduced it years ago and the theory is proven. Earl Geddes who did some. Very extensive research found the results to be astounding with a baseline treated space showing a response from 20-100hz plus/minus 3db with no DSP. I’m in the process of DIYing 4 subs myself of the above spec an the entire investment less than $1k including amplification and the saved space and $$$ on LF treatment?......priceless!
Old 5th August 2020
  #28
Hi, Mayhem, thanks for chiming in!
That's very interesting! If/when you start this experiment, please send me a link; I would love to see the results and gather some deeper insight. But I do get what you're talking about.

One fact about the subwoofer I'm using... it's made by Rythmik Audio and uses a powerful direct servo motor, which was designed to combat resonance from the room or other subs, so as far as subs working against each other, that may not do the same as you are explaining.... however, I can still see the same result in overall peak / null reductions. Also, it is plausible that I will have room for 2 subs, possibly even 3, but not likely more, and certainly not in any of the rear corners... more like left front, right front, center front. although that would likely make a difference, it may not be able to smooth things out as well as you are explaining. At this time, I have only the one sub. I will inevitably have to have another, but at the moment, it doesn't fit the budget.

I'm wondering where you are getting your subs? Or are you building them?
Old 6th August 2020
  #29
Gear Nut
 

Building them using 12” Aluminum cone drivers from Dayton although my room is a bit bigger than yours hence the 10” suggestion

Very familiar with Ryhmik.....might have been my first DIY effort 12 years ago with their servo driver kit. Servo tech has better control over the driver with incredible inpulse response when used in a reflex design.

One sub is better than two......so resist the temptation. Harmon suggests that three is acceptable for the system to work, four being optimal. Avoid active DSP processing unless it’s at a premium.........cheaper systems have less processing power and therefore slower in their asjustments in real time. Set and forget is where it’s at.
Old 6th August 2020
  #30
Quote:
Originally Posted by ejbragg View Post
The appearance of the 50 Hz peak. (post #21 ).
Read this paper:
http://downloads.bbc.co.uk/rd/pubs/reports/1992-10.pdf

Its one of the issues one encounters of adding helmholtz or membrane absorbers after to a structure instead of building it as part of the structure. If its not sealed correctly or if part of the box vibrates it can create peaks and resonances of their own. Also if the walls aren't stiff enough it becomes an issue as well.

Also the more and deeper you build/seal the traps, what the wavefront sees changes as the dimension of the room has changed.

Last edited by thethrillfactor; 6th August 2020 at 02:34 AM..
📝 Reply
Post Reply

Welcome to the Gearslutz Pro Audio Community!

Registration benefits include:
  • The ability to reply to and create new discussions
  • Access to members-only giveaways & competitions
  • Interact with VIP industry experts in our guest Q&As
  • Access to members-only sub forum discussions
  • Access to members-only Chat Room
  • Get INSTANT ACCESS to the world's best private pro audio Classifieds for only USD $20/year
  • Promote your eBay auctions and Reverb.com listings for free
  • Remove this message!
You need an account to post a reply. Create a username and password below and an account will be created and your post entered.


 
 
Slide to join now Processing…
🖨️ Show Printable Version
✉️ Email this Page
🔍 Search thread
🎙️ View mentioned gear
Forum Jump
Forum Jump