Limp mass membrane material

Posted: Wed Jul 09, 2008 9:09 pm

So I got the mic, got the software, convinced it to work Wink

set the mic up, calibrated everything...... and finally hit the test button

This is what I got......

SURELY this can't be right? Sad

(would explain alot though! Wink

)

Good news is....

The hole @100hz has gone Wink

Posted: Thu Jul 10, 2008 6:55 am

Yeah, I hope that's not right. Hopefully someone will have some measurement advice for you.

If it is, you have a pretty repetitive comb filter happening which is probably a single reflection that you are dealing with. Is there a way you can find where the dips are exactly? With that information you could narrow it down to a reflection. They seem to be all multiples of about 1100Hz which would be about 12" for a full period, 6" for a 180 phase shift. If the frequency response is correct, you are looking for a difference in reflection distance of 6". My guess would be side reflection and ceiling reflections are interfering. So to fix it you would need high frequency absorbers or diffuser in those locations.

Can someone else confirm this or tell me I'm full of it please?

Posted: Thu Jul 10, 2008 11:40 am

Thanks for the help Gecko, I appreciate it.

I think what I should have done was to be a little less impatient and wait for the mic to turn up before making the modifications to the traps. At least that way I'd know if I'd made any headway.

It looks like the membranes have helped the lowend but introduced alot of reflection problems. First port of call is to turn the traps around so the membranes are on the back, that should return some of the hi/mid absorption.

It's been hard to generate any interest in this topic, in several threads across the net and it feels like I'm p*ssing in the dark here. It's hard when you're trying to get your head around new concepts and I would appreciate some advice.

Thanks

Posted: Thu Jul 10, 2008 12:55 pm

scoobz

please explain exactly how you are performing your test - details matter - so don't be brief

Rod

Posted: Thu Jul 10, 2008 1:08 pm

Hi Rod,

I'm using the behringer ecm8000 on a microphone stand with the front of capsule where my ears usually are. I'm calibrating the level using an spl meter, I've tried using one and both speakers (in the case of a single speaker I will point the mic at it) Using both speakers gave a horrible result, so i'm guessing that's a no no.
I've tried the pink noise and swept sine generators, always 5 seconds. I'm always well out of the way usually behind the mic a few feet. Can't think of anything else specifically so if i've missed anything let me know.

Posted: Thu Jul 10, 2008 1:17 pm

GeckoMusic wrote:

Yeah, I hope that's not right. Hopefully someone will have some measurement advice for you.

If it is, you have a pretty repetitive comb filter happening which is probably a single reflection that you are dealing with. Is there a way you can find where the dips are exactly? With that information you could narrow it down to a reflection. They seem to be all multiples of about 1100Hz which would be about 12" for a full period, 6" for a 180 phase shift. If the frequency response is correct, you are looking for a difference in reflection distance of 6". My guess would be side reflection and ceiling reflections are interfering. So to fix it you would need high frequency absorbers or diffuser in those locations.

Can someone else confirm this or tell me I'm full of it please?

This gels with the HUGE resonance in the low 30s and mid 60s shown on the waterfall.

Posted: Thu Jul 10, 2008 4:13 pm

What I don't understand is why I'm getting such heavy comb filtering.

I've done everything I can, within the boundaries of my knowledge. Here is a little run down of the facts

Room :
L 4.63m
W 4.25m
H 2m

Speaker face is 0.9m from rear wall (speaker end)
Listening position is 1.9m from rear wall
Ear height 0.73m from ceiling 1.27 from floor
Speaker size 393mm (h) 250 (w) 290 (d)
Top of monitor 0.63m from ceiling 1.38m from floor

On the side reflection points I have 4" of 60kg/m3 rockwool covered with thin fabric hung 4" from the wall
I have a ceiling cloud which contains a single layer of rockwool covering the bottom (again covered with fabric)
I thought these were the main culprits of comb reflections

I have a similar panel running the width of the speakers (+0.3 either side ish) on the back wall to stop radiating frequencies.

I have corner traps 100mm thick which now have a thin membrane spray mounted (on the back turned them around as an experiment). These are mounted across the corners with a further piece of 100mm rockwool standing behind them.

I have redistributed about 5 panel's worth of rockwool into bags which are now lining every wall/floor corner all the way around.

On the back wall I have two skyline diffusers, a bookshelf, some convex mirror tiles and a leather sofa.

Please tell me what I'm missing because either I'm measuring wrong or my room is horrible

Posted: Thu Jul 10, 2008 6:04 pm

If the microphone is a little off center it could cause that comb effect with two speakers. It could be your sweep generator as well. Some of them step the frequencies, so if it is a fast sweep it might step by some.

Not really relevant, but it might make you feel better about your room. This is what I started with:

this is a 20-200Hz sweep. That's right: about 30dB trough to peak. It's still a work in progress.

Posted: Fri Jul 11, 2008 9:36 am

Thanks again gecko, you're right, your post did make me feel a bit better Wink

I'd like to know if there's a flaw in my measuring with etf (like I said I had to convince it to work)
I have a motu2408mkII lightpiped into a d8b, so i was doing all the routing required in the mackie (left channel back to the program, right channel to the speakers>microphone>program) etf couldn't use my motu mixer and kept wanting to use the virus ti as a soundcard (I don't use it as such) So I was feeding it the right thing at the right input and it seemed to be working but I don't know if the round trip of having to go thru the mackie and back would mess with it's timing and give the skewed response seen above

I demoed eaw set it up on quite a fine resolution (maybe 12th oct, don't remember) matched the levels and used the difference to make an eq curve.
What I also did was read up on my sub. It turns out the LF roll-off curve is exactly what I want (it starts a@50hz and goes down to zero) so I engaged the -6db switches, giving about -4db@30-40hz, turned the sub down a bit more.
That tweak along with an eq with -3db@35hz +2(ish)db @ 100, 300 and 2k (which I tweaked by ear to about 4.5k) gives a full sound.

To be honest I was expecting the top end to be smooth, given that I've addressed the relfections but I was in there last night and my ears told me the tweaks were a big improvement over what I had. For example, the proper lows (30-60) really seem full now and provide a solid bedrock for the frequency spectrum, still haven't got all the 100hz I'm looking for but I'm going to have to accept that unless I move room I'm probably not going to get much better.

Please let me know of another way i can test the room (btw I don't have IE on my work machine so Room Wizard EQ won't work) just to eliminate the etf theory.
If I made a sine sweep in an audio editor and recorded that and compared the two files would that work (is that what you did gecko?)

I want to know I've done all I can for this problem before going back to making some music.
So Rod, was my measuring practice ok?
Cucco, I'm going to remeasure today but you're saying if I can control that area (30-40, which I think I have gained some ground on) that would affect the entire frequency range?
I just thought the top end would be alot smoother.

Posted: Fri Jul 11, 2008 10:38 am

Why does your ref channel have 3bd troughs in it?
I have no clue, but that looks nasty.

Posted: Fri Jul 11, 2008 10:55 am

scoobz wrote:

Thanks again gecko, you're right, your post did make me feel a bit better Wink

no problem. My pain your gain

scoobz wrote:

If I made a sine sweep in an audio editor and recorded that and compared the two files would that work (is that what you did gecko?)

That's pretty much what I did. I used a program called sweepGen to make the sound, and recorded in Audacity. SweepGen steps 500ms per frequency, and makes 60 second long sweeps, so to see it the troughs are really there you may have to just sweep a small piece, say 500Hz to 1500Hz?

Posted: Fri Jul 11, 2008 11:57 am

Nice one, just got sweepgen, although I can't register it to save the generated file. I might just use wavelab. Will let you know the results

Posted: Sat Jul 12, 2008 2:37 pm

Ditched sweepgen, used wavelab instead. Then went back to etf! lol
I think you were right gecko, with the comb filtering. I've done a few experiments since then, including trying different speaker locations. All resulted in the same top end but just introduced nodes @ different bass frequencies.
I also tried bypassing my sub and that gives a smoother response that doesn't extend as far, but had no connection to the comb problem.
I tried putting 4" tiles imbetween the lights and the side wall, also a strip of 2x4 tiles (all 4") down the centre of the room. That helped alot, but didn't solve the comb problem.
Now the interesting bit, I tried laying more foam on the mixer, this gave a MUCH better reading, the comb filtering reduced by a couple of db in both directions. This was improved a little more by holding another piece against the ceiling trap.
This would suggest to me that the ceiling trap isn't doing it's job. Maybe the material is too thick and pulled to taut, which is causing hi frequency reflection.
I'm still going to superchunk the corners behind the traps, completely cover the rear wall in a broadband trap 4 inches from the wall and get some top corners looked at.
As for the ceiling trap, I guess I'm gonna have to take it down and recover it in a thinner material?

Posted: Tue Jul 29, 2008 10:19 pm

I'm finally back!!! It's taken some doing but I think I've made some headway, in fact my ears are telling me it's the best my studio has ever sounded.

These are like for like comparisons

AFTER :

BEFORE :

Far fewer peaks and troughs

Bass end

AFTER :

BEFORE :

Notice the much longer decay times lower down the scale

And this is a little comparsion I made, both are the same scale

The only things I'm not particularly happy about is a -6db dip @ 100hz (Probably down to speaker placement) and a canyon @ 400hz (again down to postioning I think, i never liked 400hz anyway Wink

)

Please let me know if you think this is an improvement and if I'm reading it right

S

Posted: Wed Jul 30, 2008 2:20 am

What looks like 70hz changed from a trough to a peak on your second test. Did you change the mic placement or anything with the monitors?

This topic sponsored by: