74hz dip in a 4x3m room

[diogodasilva]

Hi guys.

I've finalized my build (http://www.johnlsayers.com/phpBB2/viewt ... =2&t=20355)

Construction build to understand the ceiling:

IMG_3480.JPG

IMG_3502.JPG

IMG_3503.JPG

Here are some pics from the final build

Photo 24-08-16 15 18 34_web.jpg

Photo 24-08-16 15 18 13_web.jpg

Photo 24-08-16 15 17 44_web.jpg

REW FILE:
https://www.dropbox.com/s/hb6rw3mdimo4w ... .mdat?dl=0

REW SCREEN SHOTS:

4.jpg

5.jpg

6.jpg

Now, as you can see I have a huge DIP at around 74hz.
My room dimensions are 4x3m and ceiling height of around 2.7 average.

Treatments:
6x 20cm deep bass traps. 60kg/m3 rock wool
4x front corner
2x back right corner
6x 10cm deep panels. 32kg/m3 rock wool
3 each reflection points on the sides
ceiling fully treated with 10cm deep 32kg/m3 rock wool on the concrete combs.
acoustic ceiling membrane (pvc) micro dots (0,3mm), whole ceiling
qrd diffuser back wall
5x broadband panels, 5cm deep 32kg/m3 rockwool

MIC:
Sonarworks XREF 20 calibrated mic
Interface: Apogee DUET
How to properly address it ? I also see some SBIR on the freq chart, hould that have anything to do with the fact that the computer display is closer to the listener that the speakers are? EDIT* I guess not, I've made some measurements without it and it is still there. so it seems a back wall interaction from the room modes. 35hz peak, 75hz null, 110hz peak.
The 35hz peak has been tamed with the EQ on the Opal which has settings for that. but the 75 is missing some treatment.

I've run some tests on that specific frequency of 74hz.
It is more related to the length of the room than the height or width.

Moving within the room with that freq playing there is a DIP on the listening position at around 1,10 to 1,20 from the front wall.
Moving backwards to 1,60 from the front wall the dip shifts from the 74hz to the 90hz mark with an 11 DB cut.
REW here:
https://www.dropbox.com/s/637wur6794q4t ... .mdat?dl=0

REW SHOTS:

1.jpg

2.jpg

3.jpg

The speakers are as close to the front wall as possible, due to my limited length it seemed like the best idea to place them there and move the SBIR to as high of a fq as possible.

Is it the case to build more specific tuned bass trap devices such as a limp membrane and replace the diffusers with that instead?
Or should I stuff all of the back wall with 30cm deep porous absorption?

Always appreciate the input guys,
thanks a lot!

[Soundman2020]

Diego, please post all images right here on the forum, where they can be seen always. If you post them on hosting sites, they will not be available in the future to other people who might be following your build.

For example, right now your link to "cloud.candelaluz.com.br" produces nothing! Just a pure white screen, with no images or text. Not even an error message.

So unfortunately, I can't help you, because I cannot what you are referring to.

The same applies to your Dropbox images: they might be there today, but they will eventually expire, then your thread will be no help to anyone else.

What you can put on dropbox, is the actual MDAT file from REW! That is too big to put on the forum, so it is OK to put in on Dropbox, where I can download it and analyze it properly.

Also, in your original thread, you didn't provide any updates in almost six months, so we did not see each stage of the build, or get to offer you suggestions and advice that might have prevented this situation you have now. In the photos, I can see several problems with the way the room is right now, all of which could have been avoided if you would have let us know your plans at each stage.

So, to summarize: please post the photos and REW images here on the forum, and post the actual REW MDAT file on DropBox, with a link here.


- Stuart -

[diogodasilva]

Diego, please post all images right here on the forum, where they can be seen always. If you post them on hosting sites, they will not be available in the future to other people who might be following your build.

For example, right now your link to "cloud.candelaluz.com.br" produces nothing! Just a pure white screen, with no images or text. Not even an error message.

So unfortunately, I can't help you, because I cannot what you are referring to.

The same applies to your Dropbox images: they might be there today, but they will eventually expire, then your thread will be no help to anyone else.

What you can put on dropbox, is the actual MDAT file from REW! That is too big to put on the forum, so it is OK to put in on Dropbox, where I can download it and analyze it properly.

Also, in your original thread, you didn't provide any updates in almost six months, so we did not see each stage of the build, or get to offer you suggestions and advice that might have prevented this situation you have now. In the photos, I can see several problems with the way the room is right now, all of which could have been avoided if you would have let us know your plans at each stage.

So, to summarize: please post the photos and REW images here on the forum, and post the actual REW MDAT file on DropBox, with a link here.


- Stuart -

Sorry!!
I fixed and added more information!
thanks

[diogodasilva]

SOme more extra tests after repositioning the speakers.
I removed the stands so I could place the speakers and desk closer to the front wall.
Then my new listening position with this configuration is 80cm from the front wall.

From this position I measured full response without any speaker EQ, reponse with speaker EQ on and finally response with EQ on and Sonarworks room EQ on.

EQ DIP 40HZ OFF.jpg

EQ DIP 40HZ ON.jpg

ROOM EQ ON.jpg

Rew file:
https://www.dropbox.com/s/xluh2tbv2n74g ... .mdat?dl=0

[Soundman2020]

REW FILE:
http://cloud.candelaluz.com.br/190C1j0Z450l

Result:

diego-blank.jpg

... And

REW here:
https://www.dropbox.com/s/637wur6794...4_59.mdat?dl=0

Produces this:

diego-404.jpg

- Stuart -

[diogodasilva]

Here is the rew measurements with increments of 10cm from the front wall to the back.

As you can see the freq does shift upwards so it is most likely related to the rear wall reflection?

Shall I remove the diffuser and build a frame to put 10cm of rockwoll on the front and leave a 20cm gap to the brick wall? totalling 30cm deep trap? from floor to ceiling I can build almost 7sqm (will have do discount the door)
https://www.dropbox.com/s/u89xrxry9vyz3 ... .mdat?dl=0

[diogodasilva]

REW FILE:
http://cloud.candelaluz.com.br/190C1j0Z450l

Result:

diego-blank.jpg

... And

REW here:
https://www.dropbox.com/s/637wur6794...4_59.mdat?dl=0

Produces this:

diego-404.jpg

- Stuart -

Replaced the links

[Soundman2020]

The final link works (the REW with the 10cm steps between readings).

However, the data does not seem valid.

I'm wondering why the signal level drops nearly 1 dB between measurements. The level drops by nearly 8 dB over a distance of 1.4m, which is not explainable by any acoustic reasons. There should be a drop off as you move way from the speakers, yes, but it's just too big in your case. It should also be exponential (3db per distance doubling), but yours seems to be linear. There is no reason why each 10cm move of the mic should produce a 0.5 dB drop in level. The room is not that big, so all of them should be at about the same level with only slight differences.

What happened?

Your data is not valid like that. You cannot draw any conclusions from suspect data.

You need to re-calibrate your system carefully, find out why the level is dropping so much as you move the mic, then repeat the measurements.

- Stuart -

[diogodasilva]

The final link works (the REW with the 10cm steps between readings).

However, the data does not seem valid.

I'm wondering why the signal level drops nearly 1 dB between measurements. The level drops by nearly 8 dB over a distance of 1.4m, which is not explainable by any acoustic reasons. There should be a drop off as you move way from the speakers, yes, but it's just too big in your case. It should also be exponential (3db per distance doubling), but yours seems to be linear. There is no reason why each 10cm move of the mic should produce a 0.5 dB drop in level. The room is not that big, so all of them should be at about the same level with only slight differences.

What happened?

Your data is not valid like that. You cannot draw any conclusions from suspect data.

You need to re-calibrate your system carefully, find out why the level is dropping so much as you move the mic, then repeat the measurements.

- Stuart -

Hi stuart.
I don't know! But even in person the level drop is noticeable.
I guess the room being all trapped is contributing to the level drop?
Note that I'm measuring with only one speaker at a time.
Level is calibrated to 85db on first measurement point with spl meter.
maybe I need to re-calibrate after each step?

[diogodasilva]

I have identified a low level 60hz HUM probably due to the mains not grounded properly.
I will have to verify that.

If I decrese the MIC input and increase the speaker amp power (higher SPL) to improve the SNR will the results still be valid?

Screen Shot 2016-08-25 at 13.19.33.png

[Soundman2020]

Assuming the REW data is reasonably good, let's analyze that:

Now, as you can see I have a huge DIP at around 74hz.

I don't see that at all! There's no dip at 74 Hz. :

diego-fr.jpg

There's a dip at around 54 Hz and another at 62 Hz, but nothing unusual at 74 Hz. What is the issue you are hearing at 74 Hz?

4x front corner
2x back right corner

Maybe it's just the angle of the photos, but to me it doesn't look like you have those absorbers angled at 45! across the corners. They seem to be mostly up against the side walls, with just a slight angle.

6x 10cm deep panels. 32kg/m3 rock wool - 3 each reflection points on the sides

The top and bottom one on each side don't need to be there. They can be better used on the rear wall. You only need panels at head height, so re-arrange just one panel on each side, centered on the exact first reflection point.

ceiling fully treated with 10cm deep 32kg/m3 rock wool on the concrete combs.

acoustic ceiling membrane (pvc) micro dots (0,3mm), whole ceiling

Why? What frequency is that tuned to?

qrd diffuser back wall

Get rid if that. The room is too small to be able to use numerically tuned diffusion. That might be the cause of some of the frequency and phase shifting issues that are visible in the mids and highs.

I also see some SBIR on the freq chart, hould that have anything to do with the fact that the computer display is closer to the listener that the speakers are? EDIT* I guess not, I've made some measurements without it and it is still there. so it seems a back wall interaction from the room modes. 35hz peak, 75hz null, 110hz peak.

No, the SBIR issue is not related to the screens. It is related to the rear wall. There is no null at 75 Hz. And the peaks are 35 Hz and 110 Hz are modal, not SBIR.

The 35hz peak has been tamed with the EQ on the Opal which has settings for that. but the 75 is missing some treatment.

You cannot fix room modes or SBIR issues with EQ. Modes are time-domain problems, not frequency problems, so EQ can only reduce the intensity of the problem slightly, and only for one specific location in the room. SBIR is due to phase cancellation, so there is nothing at all that EQ can do for that. If you try to boost the level in the SBIR dip, you just make it worse, not better, since there is more energy available to cancel itself out.

I've run some tests on that specific frequency of 74hz.

I still do not see any problem at all at 74 Hz, so I'm not sure what you are talking about here!

The cursor on the FR graph above is centered on 74 Hz. What do you see there that is causing you worry?

Same for the spectrogram and waterfall (below):

diego-sp.jpg

diego-wf.jpg

I just don't see anything at all wrong with 74 Hz!!! Are you SURE you posted the right REW file?

The speakers are as close to the front wall as possible,

Right, but do you have absorption between them and the wall? If not, then you need to do that. You need at least 10cm of absorption between the rear of each speaker and the actual wall of the room.

Also, you have your speakers laying down on their sides: that's no good. Stand them up vertically. If you lay them down, you get this problem:

Speakers-mounted-vertically-and-horizontally-standing-up-laying-down-on-side.jpg

From the photos, it also looks like your measurement mic is facing directly forwards, pointing straight at the front wall: That's no use. It needs to be angled upwards at about 60°

Note that I'm measuring with only one speaker at a time.

That's not clear from your REW file at all! There are no comments or indications in there. You should ALWYAS label EACH test to show which combination of speakers you used. I use a series of three letter at the start of each name, to show that: "LRS", for Left Right Sub". If one of the speakers was off (or not present), I replace that letter with a dash. So "L-S" means the test was done with just the left speaker and the sub, "--S" means just the sub, "-R-" means just the right speaker, and "LR-" means left and right but no sub.

Here's what you need to do:

- Remove the diffusers,
- Take off the six side absorbers, use four of them on the back wall (more below) and center the remaining two at the exact first reflection point.
- Take the four free absorbers, and set them up in the middle of the back wall, covering as much surface as you can, and separated from the wall by about 6 inches (15cm). If they won't all fit, then double-up: put the extra one in front of the others, so you have extra thickness
- Angle the absorbers that you have in the room corners so they are 45° diagonally across the corners.
- Set up the speakers correctly: vertically, the correct distance from the side walls, and angled inwards at the correct angle. (if you don't know what the correct distance and angles are, give me the dimensions of the room and I will tell you)
- Set up more absorbers between the speakers and the front wall. At least 10cm thick.
- Set up the mic at the correct location for the mix position, and angle it upwards 60° (if you don't know where the correct location is for the mix position, give me the dimensions of the room and I will tell you)
- Turn off all equalization, and set the speaker controls to about -3dB roll-off for the bass, but everything else flat.

Now repeat all of your tests. For each location of the mic, do one test with just the left speaker ("L--"), one with just the right speaker "-R-", and one with both speakers ("LR-").

Do not adjust anything in the system between tests! Leave all of the controls exactly the same.

Take some photos of how the room looks when you do that testing.

I have identified a low level 60hz HUM probably due to the mains not grounded properly.

I noticed that. It's probably not a mains problem, but rather a problem with your cables or equipment, or a wiring problem in the room. Did you get install with star-grounding?


- Stuart -

[diogodasilva]

Hi Stuart.

I've tried some re-positioning of the traps and speaker position.

I've removed some panels from the first reflection points on the sides and put two covering the diffusers and two behind the speakers on the front.
Also removed one of the membranes from the ceiling just for the sake of testing.

Apparently not much benefit.
Here is the file and pictures.

Speakers centered on the room 132cm apart from each other.
Mic position 130cm from front wall.
Mic height 120cm from floor.

I'll run some more tests with open door to see if it is of any help.

https://www.dropbox.com/s/di1fsjocdhl81 ... .mdat?dl=0

[Soundman2020]

I've tried some re-positioning of the traps and speaker position.

The traps seem to be good, but your speakers are still not set up correctly: the angle is clearly too large, which is probably the cause of the big dip in the high-mids, and your mix position is too close to the front wall.

According to your original thread, the room is 240 cm wide there. Assuming that's still correct, set up the speakers 70 cm from the side walls, which means they will be 100 cm apart. If the room is a different width, then tell me what it is and I'll tell you where to put the speakers.

Also, your speakers are still sitting on shelves on top of the desk. You'll never get good response like that. Put them on stands behind the desk, not touching it. Pull the desk forward into the room just enough so you can do that. The stands need to be massively heavy, not flimsy. Some people use hollow metal stands filled with dry sand, others stack up cement blocks or bricks.

Now for the angle: Your original thread does not show the length of the finished room, but at a guess I'd say about 480 cm. So set up a mic stand 226 cm from the front wall, on the room center line, and set it up vertically at that point. That's the aim point for your speakers. Set the angle of each speaker so that it is pointing exactly at that mic stand. I'm talking about the acoustic axis of the speakers: that's what needs to point at the stand.

Set up your listening position at 176 cm from the front wall (30cm in front of the mic stand). That's where your head should be, and it's where you should put the measurement mic when you do all future REW tests. The mic must be angled upwards at about 60° for all measurements.

If the room is not actually 480 long and 240 wide, then let me know the real dimensions, and I'll give you new locations. From the photos, I'm not convinced that those are the real dimensions. It looks like you built the room different from what the original plan shows.

Speakers centered on the room 132cm apart from each other.
Mic position 130cm from front wall.

Assuming the dimensions I gave above are correct, your speakers are too far apart, angled too greatly, and your mix position is too far forward. It's very nearly in the quarter-wave null / peak for that room.

I'll run some more tests with open door to see if it is of any help.

Why? That doesn't make much sense. You won't be mixing with the door open, and it will add a second, much longer tail to the decay times, since you will have both rooms acoustically coupled. Keep the door closed.

You also need to turn up the level of your right speaker by about 1 dB. It's not at the same level as your left speaker. Actually, it's a bit less than 1 dB. About 0.8 dB. Do it very carefully: it's just a small touch.

Your 40 Hz roll-off is helping a bit to reduce the modal ringing there, but it is also killing your entire bass end. From about 120 Hz down, there's a big hole in your low end. Turn it off for all future measurements, and make sure that all the controls on the speakers, console and everywhere else are set for flat response.

Your latest set of tests are also incorrect: they only start at 35 Hz, so there's no information in the very low end. Run ALL of your tests from 17 Hz to 22,000 Hz.

So please correct all of the above, then do another set of REW tests.

- Stuart -

[diogodasilva]

I've tried some re-positioning of the traps and speaker position.

If the room is not actually 480 long and 240 wide, then let me know the real dimensions, and I'll give you new locations. From the photos, I'm not convinced that those are the real dimensions. It looks like you built the room different from what the original plan shows.

- Stuart -

Hi Stu!
The room is 300cm wide and 400cm depth
please let me know the correct positions.

Thanks again!

[Soundman2020]

The room is 300cm wide and 400cm depth

Then the speakers would need to be about 85cm from the side walls, and therefore about 130 cm apart. The mix position (your head) would need to be on the room center-line (left/right), and about 160 cm from the front wall. The speakers should be aimed at a spot directly behind your head, about 190cm to 200 cm form the front wall.

- Stuart -