DIY flush mount speakers

Forum for all aspects of speakers and speaker design.

Moderator: Aaronw

Eric Best
Senior Member
Posts: 311
Joined: Fri Feb 14, 2003 1:51 am
Location: Lansing, MI USA
Contact:

Re: DIY flush mount speakers

Post by Eric Best »

Here are the waterfalls from the left and right speakers from 1M. Don't think the problem is there.
right speaker.jpg
left speaker.jpg
"It don't get no better than this"
Eric Best
Senior Member
Posts: 311
Joined: Fri Feb 14, 2003 1:51 am
Location: Lansing, MI USA
Contact:

Re: DIY flush mount speakers

Post by Eric Best »

spekaer with calibration.jpg
I turned off the calibration file on the previous two measurements, then turned it back on.
"It don't get no better than this"
Eric Best
Senior Member
Posts: 311
Joined: Fri Feb 14, 2003 1:51 am
Location: Lansing, MI USA
Contact:

Re: DIY flush mount speakers

Post by Eric Best »

Now it has come back and I can't get rid of it.
"It don't get no better than this"
Eric Best
Senior Member
Posts: 311
Joined: Fri Feb 14, 2003 1:51 am
Location: Lansing, MI USA
Contact:

Re: DIY flush mount speakers

Post by Eric Best »

I know, not another post from Eric!!!

I completely reset my system and did not install the sound card calibration file. It is gone.
"It don't get no better than this"
Soundman2020
Site Admin
Posts: 11938
Joined: Thu Aug 21, 2008 10:17 am
Location: Santiago, Chile
Contact:

Re: DIY flush mount speakers

Post by Soundman2020 »

I completely reset my system and did not install the sound card calibration file. It is gone.
Ok, that would make sense. It was that ringing (feedback loop) that I mentioned previously, on your soundcard calibration. Probably something was messed up with the calibration procedure, as I doubt that your card actually rings like that!

Try repeating the calibration process carefully, and checking that you have no feedback loops anywhere. It might not be a physical loop, but rather a logical loop, hidden somewhere in the internal digital routing path of your signal chain.

OK, on to the next tests! "Walking mic, going up and down".... (Maybe this one should be the "hopping mic"...)


- Stuart -
Eric Best
Senior Member
Posts: 311
Joined: Fri Feb 14, 2003 1:51 am
Location: Lansing, MI USA
Contact:

Re: DIY flush mount speakers

Post by Eric Best »

It has been coming and going. It is ringing from the soundcard, and I can't isolate it. I am doing measurements using my headphone out from my laptop with a USB mic and they look ok but everything drops off over 10k, so I hope i can find the problem in the soundcard!
computer soundcard.jpg
"It don't get no better than this"
Soundman2020
Site Admin
Posts: 11938
Joined: Thu Aug 21, 2008 10:17 am
Location: Santiago, Chile
Contact:

Re: DIY flush mount speakers

Post by Soundman2020 »

It's interesting you say that. I've come across a lot of people trying to use USB mics, and ending up with problems. It seems the technology for USB mics just isn't there yet. I really, really prefer to stick to a good external interface (which can be USB, as that technology does work), and a good XLR connected, phantom powered acoustic measurement mic. You actually an calibrate that setup, and it works. So that's what I'm recommending from now on.

So this month I have two new additions to my "Don't do this" list:

- Do not, ever, under any circumstances, buy a Behringer ECM8000 acoustic mic.
- Do not, ever, under any circumstances, buy a USB connected acoustic measurement mic.

I have a feeling the list might grow...

Anyway, I hope you figure out a solution!

- Stuart -
Eric Best
Senior Member
Posts: 311
Joined: Fri Feb 14, 2003 1:51 am
Location: Lansing, MI USA
Contact:

Re: DIY flush mount speakers

Post by Eric Best »

I needed to select the ASIO drivers in REW for my sound card to work correctly. I am all calibrated, and one week later, ready to function correctly :) So be ready for a whole lot of data to be posted!
with asio drivers.jpg
"It don't get no better than this"
Eric Best
Senior Member
Posts: 311
Joined: Fri Feb 14, 2003 1:51 am
Location: Lansing, MI USA
Contact:

Re: DIY flush mount speakers

Post by Eric Best »

OK, on to the next tests! "Walking mic, going up and down".... (Maybe this one should be the "hopping mic"...)
With a caveat. There was no accurate SPL on these. I won't have them until tomorrow.

https://www.dropbox.com/s/pg0neh7pceecq ... .mdat?dl=0
"It don't get no better than this"
Soundman2020
Site Admin
Posts: 11938
Joined: Thu Aug 21, 2008 10:17 am
Location: Santiago, Chile
Contact:

Re: DIY flush mount speakers

Post by Soundman2020 »

With a caveat. There was no accurate SPL on these.
No problem. For this test, SPL accuracy isn't so necessary.

... and my suspicion is confirmed! your 80 Hz dip is related to the vertical axis. It's SBIR. You are probably going to need a ceiling cloud that extends over the mix position to deal with that. I'd suggest making it hard-backed, and hang it an angle that compliments your CID layout.

To see the issue for yourself, select the "SPL & Phase" tab in REW, set the limits to Left=18, Right=500, Top = 110 / 1440, and bottom = 35 / -1440. Hit "Apply". Then hold down the left "Alt" key on your keyboard, and use the up and down arrow keys to cycle through the various height graphs fast. You'll see the SBIR dip "moving" from left to right (changing frequency). That "Alt-arrow" trick is very useful for cycling through a bunch of similar graphs, to see what is what. Since SBIR is a phase issue, and is related to distances, it will appear at a different frequency for each different distance as you move the mic around the room. And since modal activity is only relate to room dimensions, modal issues do not change frequency as you move the mic. They only change in amplitude. Also, modal issues "ring" and decay slowly, whereas SBIR does not. It is pure phase cancellation, so there's no resonance involved. That was the purpose of the first "walking mic" set: it clearly shows which problems are modal, and which are SBIR (among other things). Sometimes you can get lucky and find a spot where an SBIR problem just happens to cancel out a modal problem, but it doesn't happen too often. And it doesn't "fix" the problem: it just disguises it, an only works for that exact spot in the room (sort of like badly done "room correction EQ").

The large, broad dip at 80 Hz is related to the distance between the ceiling and the floor, and the smaller, sharper mobile dip is related to the mic position within that ceiling-floor dip. An angled, hard-backed ceiling cloud should help to "blur" that dip more, making it lower intensity and broader band.

You can see more SBIR issues happening at around 140 Hz and 420 Hz, plus a couple of others. Those will all improve with a cloud.

Quick test: If you have a full panel of 4" thick 703 available, try putting that on the floor where the mic stand is, then set the mic stand on top of that again, in it's normal position, with the height adjusted back to mix position height again (above the floor, not above the 703), and do a test like that. You should see another small reduction in that dip. If that works, then a cloud would work even better.

Also, don't forget to try what I suggested yesterday: move your mix position a few inches forward. It should be a little smoother there. That's something you can also see on your "walking mic" test, by cycling through the measurements. But don't go more than about 4 or 5" forward, or you'll start getting slightly off axis from your speakers.

- Stuart -
Eric Best
Senior Member
Posts: 311
Joined: Fri Feb 14, 2003 1:51 am
Location: Lansing, MI USA
Contact:

Re: DIY flush mount speakers

Post by Eric Best »

Soundman2020 wrote:Also, don't forget to try what I suggested yesterday: move your mix position a few inches forward. It should be a little smoother there. That's something you can also see on your "walking mic" test, by cycling through the measurements. But don't go more than about 4 or 5" forward, or you'll start getting slightly off axis from your speakers.
We did this, it minimized the 80hz dip by a couple db.

Quick test: If you have a full panel of 4" thick 703 available, try putting that on the floor where the mic stand is, then set the mic stand on top of that again, in it's normal position, with the height adjusted back to mix position height again (above the floor, not above the 703), and do a test like that. You should see another small reduction in that dip. If that works, then a cloud would work even better.

- Stuart -
Unfortunately, this didn't budge anything around 80hz. It did flatten out the low mids a little more though, so I have decided that I'm going to put more absorption there anyway where the bright green is in this picture. That area is 12" thick, so it will be a super chunk style.
additional absorber.jpg
We did a test with the SPL meter, and there was a 19dB difference at 80Hz between the mix position and the corners in the ceiling, so more bass trapping won't hurt and it might help. This will probably be a couple day project.
"It don't get no better than this"
Soundman2020
Site Admin
Posts: 11938
Joined: Thu Aug 21, 2008 10:17 am
Location: Santiago, Chile
Contact:

Re: DIY flush mount speakers

Post by Soundman2020 »

It's not just absorption that you need. It's a proper cloud, with a thick, massive hard back (I normally use 3/4" MDF), angled. And it needs to be angled lower towards the speakers, higher over the mix position. Absorption will help, but you really need a massive "reflector" that works down at low enough frequencies (ie, it needs to be big), and has a decent angle on it, to not just reflect but also "smear" the wavefront.

Here's some REW graphs from a room I did a few years back. This is still the early stages of treatment: the "before" graphs are the response of the room with only superchunks and other bass trapping, and the "after" graphs are the next stage in the process: after installing the large, angled, massive, hard-backed cloud:

Frequency response before:
RDMOUS-for-eric-no-cloud-FR-20-500.jpg
Frequency response after:
RDMOUS-for-eric-WITH-cloud-FR-20-500.jpg

Waterfall before:
RDMOUS-for-eric-no-cloud-WF-20-500.jpg
Waterfall after:
RDMOUS-for-eric-WITH-cloud-WF-20-500.jpg

You can see how the SBIR stuff around 100 Hz cleaned up nicely.

After that I still did a lot more treatment in that room, to iron out the other stuff, but you can at least see what a difference a cloud can make.


- Stuart -
Soundman2020
Site Admin
Posts: 11938
Joined: Thu Aug 21, 2008 10:17 am
Location: Santiago, Chile
Contact:

Re: DIY flush mount speakers

Post by Soundman2020 »

It's not just absorption that you need. It's a proper cloud, with a thick, massive hard back (I normally use 3/4" MDF), angled. And it needs to be angled lower towards the speakers, higher over the mix position. Absorption will help, but you really need a massive "reflector" that works down at low enough frequencies (ie, it needs to be big), and has a decent angle on it, to not just reflect but also "smear" the wavefront.

Here's some REW graphs from a room I did a few years back. This is still the early stages of treatment: the "before" graphs are the response of the room with only superchunks and other bass trapping, and the "after" graphs are the next stage in the process: after installing the large, angled, massive, hard-backed cloud:

Frequency response before:
RDMOUS-for-eric-no-cloud-FR-20-500.jpg


Frequency response after:
RDMOUS-for-eric-WITH-cloud-FR-20-500.jpg


Waterfall before:
RDMOUS-for-eric-no-cloud-WF-20-500.jpg
Waterfall after:
RDMOUS-for-eric-WITH-cloud-WF-20-500.jpg


You can see how the SBIR stuff around 100 Hz cleaned up nicely.

After that I still did a lot more treatment in that room, to iron out the other stuff, but you can at least see what a difference a cloud can make.


- Stuart -
Eric Best
Senior Member
Posts: 311
Joined: Fri Feb 14, 2003 1:51 am
Location: Lansing, MI USA
Contact:

Re: DIY flush mount speakers

Post by Eric Best »

Like this?
cloud.jpg
"It don't get no better than this"
Eric Best
Senior Member
Posts: 311
Joined: Fri Feb 14, 2003 1:51 am
Location: Lansing, MI USA
Contact:

Re: DIY flush mount speakers

Post by Eric Best »

I rigged up a cloud. with 3/4" mdf and a 4" 2'x4' panel of 703.
20170514_142901.jpg
Didn't get any change at 80hz.
with and without cloud.jpg
So I decided to use my rig to fill the upper corner with 8" of 703.
20170514_151033.jpg
I didn't have enough to do both, but it made a significant change.
spl base and one corner filled.jpg
here are the waterfalls. First, no corner filled.
baseline waterfall.jpg
Here with waterfall with one corner filled.
waterfall no cloud corner filled.jpg
"It don't get no better than this"
Post Reply