Ran REW, now what?

[sud98]

So I need to try it with the mic 3 feet from the back wall, also I plan to use to bass traps behind the two tower speakers and check the effects. The bass traps are around 8 inches and 4 feet by 3 feet, so will put in the front two corners with an air gap of around 12 inches.

Let me know I need to test anything else.

 

[elangoas]

I did some maths, assuming total surface area reqd for bass traps..

To absorb a 37Hz frequency, requires a 91inch thick bass trap (per attached PDF).. So assuming it is 1' x 1'.. L = 12 inch, W = 12 inch, D/H = 91 inch..

Total Surface Area of the cuboid = 2 (lw + wh + hl).. In this case is 4,656 sqinch..

The bass traps are around 8 inches and 4 feet by 3 feet, so will put in the front two corners with an air gap of around 12 inches.

Here L = 48 inch, W = 36 inch, D/H = 8 inch ..

Total Surface Area of the cuboid = 2 (lw + wh + hl).. In this case is 4,800 sqinch..

I have no idea, if this is correct (or) wrong.. If this calc is wrong, plz correct it..

 

[sud98]

I ran REW multiple times in the last week. Initially to test the mic position at 3 feet, but then ran it at the default 1 feet and yet the 37db peak is not prominent and instead there is now a dip at 55hz.

I tried with different things to understand the issue but cant find the root cause of the change.

Here is the mic with various distances from the back wall
https://drive.google.com/open?id=16X9oCOgoktPMSEBVch2w5I0GsCBvQ_1h

The speakers with mic at 1 feet from back wall
https://drive.google.com/open?id=16X9oCOgoktPMSEBVch2w5I0GsCBvQ_1h

The mdat file with the current drawings
https://drive.google.com/open?id=1zXXi3LwX220_akGskpPQH-5FfOZC41CO

Not sure of the cause and how to handle these dips and whether this is the right measurement.

 

[Hari Iyer]

My suggestion is this,
- Turn the right speaker completely off.
- Start with the left speaker against the wall.
- Plat some good recording that has some very g9od lows, mids and highs.
- Start moving the left speaker into the room, till the bass kicks in. You can sit in your sweet spot when you ask someone else to move the left speaker for you. Move gradually the left speaker as the point where the bass kicks in will be very small.
- repeat the process for the right speaker.

- Rake the speaker slightly towards back till you find the vocalist in front of your ear level.
- Toe in the speaker till you can size the vocalist properly.
- keep REW away for the time being and trust your ears more.

 

[elangoas]

I ran REW multiple times in the last week. Initially to test the mic position at 3 feet, but then ran it at the default 1 feet and yet the 37db peak is not prominent and instead there is now a dip at 55hz.

I tried with different things to understand the issue but cant find the root cause of the change.

Here is the mic with various distances from the back wall
https://drive.google.com/open?id=16X9oCOgoktPMSEBVch2w5I0GsCBvQ_1h

The speakers with mic at 1 feet from back wall
https://drive.google.com/open?id=16X9oCOgoktPMSEBVch2w5I0GsCBvQ_1h

The mdat file with the current drawings
https://drive.google.com/open?id=1zXXi3LwX220_akGskpPQH-5FfOZC41CO

Not sure of the cause and how to handle these dips and whether this is the right measurement.

Both the links have the same picture.. Did you miss to update the link for an other measurement ?..

Are the measurements posted here, only for FS (or) FS + sub?

Surprising how the measurement has changed from the first pic you posted..

 

[sud98]

That was a spurious result as the settings were incorrect.

 

[elangoas]

That was a spurious result as the settings were incorrect.

You planning to get miniDSP to correct it (or) try other methods?..

 

[sud98]

Moved it from the other thread

A while since I updated this thread. I looked at options for me. First was to buy a minidsp and experiment with filters, then realised its a narrow solution and needs room acoustics correction before dsp.

SO did my research and with the help of FM ELango, decided on building broadband absorbers to the room. First intention was to improve the absorption coefficient of the room as my walls were thick concrete.

To build the thickest possible trap, decided to first build corner traps. Got inspired by Santy's build to start with a triangular trap. Ran into issues with stability as I was planning 8 feet high traps and the slabs kept falling down.

For the absorbent material, decided on Rockwool. While I initially was planning a very dense 96kg/m3 slab, but read a lot more reseach which pointed one of the keys to a successful broadband absorber is to have low gas flow resistivity. This is to ensure that the waves actually go through the layers than getting reflected. This is especially true if you are building traps > 4 inches.
Source -

So bought the following materials

3 bundles of Rockwool from Rockwool India 48 kg/m3 - Manal Rockwool - 950 Rs. per bundle of 14 slabs of 2 inch thickness

1 staple gun - miles TP10
A box of staples

Steel Mesh for panels besides the corner traps Rs. 35/sqft

Plywood - 2 sheets
L clamps to clamp it to the wall
6 kgs (18mts) of Spandex lycra to double wrap the traps - 3200 (Courtesy FM Kannan of Tirupur)

Sharp Knife
Gloves (used gardening rubber ones)

I had around 15 inches of corner space next to the screen, so decided to build a square pillar of 1 foot by 1 foot rockwool slabs with the ply adding around an inch. Wanted to build it till the ceiling but was difficult to clamp it as well didnt want to do a joint in the ply, so restricted to 8 feet.

Step 1
Cut the Rockwool slabs into squares
You will notice that unlike the higher density slabs, these disintegrate easily, so mask, gloves, socks and full sleeves are essential

https://drive.google.com/open?id=13E0aqFDM9anxqC1MRis_unbw7W7G3dow
https://drive.google.com/open?id=1bYxsqK1nxJf-qK7Kk5OLXHqzmz9K5_Rm

Step 2
Build the frame
https://drive.google.com/open?id=1gA_gWA1d8vO_ljAZFSp4KMQ99TzDQDK_

Step 3
Stack the rockwool slabs

Step 4
Cut the cloth, since its stretchable by around 10% so its nice to wrap around. However, at that height its still a pain to stretch and staple, so had to take the help of my better half.
https://drive.google.com/open?id=1DvJjlGJT1ISELu4rl66S42xxek3HG2CN

Step 5
Install in the right slot and clamp using an L clamp just to make sure that it doesnt get pulled down
https://drive.google.com/open?id=1i7Os6vIZD1lUqCKD-QBy_CW1iVprgveV

To build a panel (either horizintal or vertical)
Step 1
Build a frame of 4 inches depth and 1*0.6mts from inside which is the measurement of the slab
https://drive.google.com/open?id=1i7Os6vIZD1lUqCKD-QBy_CW1iVprgveV

Step 2
Cover one side with Steel Mesh and staple

Step 3
Fill two slabs and cover with another steel mesh and staple
https://drive.google.com/open?id=1yxuKlnZzMZHkzLc9OHB2Rhoi9jpH17hK

Step 4
Cover with cloth and staple used around 2.5 mts to cover the full frame
https://drive.google.com/open?id=1noYAc6oQWV3WxiTPlyHV4qXYplwY8VHF

Tried to also make triangular for the rear wall, however, somehow the structure wasnt stable and the square one was taking too much space behind the reclines, so kept that for a later day.

Ran out of mesh for the other panels, to be put at the first reflection points.

Total cost - 8k for ply and labour, 3k for cloth, 3k for rockwool, 1k for the steel mesh. Rest are usable for other applications

 

[elangoas]

Tried to also make triangular for the rear wall, however, somehow the structure wasnt stable and the square one was taking too much space behind the reclines, so kept that for a later day.

Very neatly done @sud98 .. Looks great.. Hope you complete your triangular traps for rear wall too soon and check for improvement thru measurement..

 

[sud98]

Yes, will cleanup and do a measurement this week. Fingers crossed

 

[sud98]

The Before



and the after

 

[sud98]

From the measurements

and the waterfall

There is an improvement in the mid bass and the highs. The nulls are lesser. The big room mode at 37 Hz has had a very minor effect as expected. Need to handle that through other means. Not sure if I need to add more absorbents at the first reflection and other areas.

 

[elangoas]

From the measurements and the waterfall

There is an improvement in the mid bass and the highs. The nulls are lesser. The big room mode at 37 Hz has had a very minor effect as expected. Need to handle that through other means. Not sure if I need to add more absorbents at the first reflection and other areas.

If you can generate a waterfall measurement before adding bass traps, comparison of pre & post traps will give you rough idea of how much has been cleaned up.. ( Pre waterfall above post -trap waterfall in images)..

 

[Hari Iyer]

The CSD plot with a different rise time could give more information on the waterfall plots imo.

 

[sud98]

The old waterfall

The old spectrogram

The new spectrogram

 

[Hari Iyer]

Spectrogram shows single frequency boom around 35 Hz. The higher octaves at 70 Hz and 105 Hz are virtually absent. If you convert 35 Hz to wave length that gives one side of room as around 32 feet. This is the length that you need to tame imo

 

[sud98]

Spectrogram shows single frequency boom around 35 Hz. The higher octaves at 70 Hz and 105 Hz are virtually absent. If you convert 35 Hz to wave length that gives one side of room as around 32 feet. This is the length that you need to tame imo

At that frequency which is the room mode the option would be to build a resonator or a dsp to handle it or try antimodes. While I have the dsp, wanted to check other options first.

 

[Hari Iyer]

At that frequency which is the room mode the option would be to build a resonator or a dsp to handle it or try antimodes. While I have the dsp, wanted to check other options first.

If there is a peak in the minimum phase group delay, then DSP cannot be used to tame the response. You can try adding more than 2" high density coir or absorbent to check if that flattens the boom. Better if you have a window on the other end try keeping it open and measure again to check. This will lead you to which way you should proceed.

 

[elangoas]

The old waterfall

If you could resize the waterfall charts alone from 20Hz - 300 Hz alone for pre & post traps comparison, you will most likely see the improvement..

Also you have taken the measurement of Front FS alone.. If you can move the single sub to the center of the room and measure sub + L & R, then that 35Hz room mode and 70Hz mode should be minimised..

You can try adding more than 2" high density coir or absorbent to check if that flattens the boom.

Are you referring to taming the 35Hz room mode with 2inch high density panel?..

 

[Hari Iyer]

Can't say if it can be tamed unless tried. But the entire wall needs this treatment as the peak at 35 Hz is too high compared to the higher harmonics. If there is a window at the other end, it will be wiser to open them only for measuring.