First measurement: 2 big problems. Any solution?

**steveo1234** · **08-22-08, 07:15 AM**

Hi all.
Ive installed my 4*AE 15 IB in my 15*12feet room. Im currently powering it with an old receiver so its probably getting around 150W or so. Im getting a Behringer ep 2500 as soon as i manage to sell it though.
Anyways. Since I have a BFD I wanted to see what im hearing (which just sounds incredibly good btw) and what I see looks like, well, not something that i would expect to sound so nice!

I ran my first set of sweeps.
This isnt looking good. Well, from 8-45 or so it looks great, but then theres a huge dip between 45 and 65. Is it even possible to do anything about this?

Second problem comes when looking at the LF Waterfall plot.
What is going on at and around 50hz? Anyone know how to solve this or what the problem could be?

Any help would be much appriciated!

**brucek** · **08-22-08, 07:37 AM**

First you need to make some changes in your graphs.

Use a vertical axis of 45dB-105dB and a horizontal axis of 15Hz-200Hz.

Note the icon that looks like a little floppy disk in the lower left corner of your graphs - click it and use it to save your graphs (at 800 wide) for posting, rather than showing the entire REW page.

Waterfall plots use the same axis as response graphs and must be set to LOGarithmic rather than LINear. Use the Freq Axis button for that.

Once you've done that, can you tell us if this is a sub only or a sub + mains measurement. The first plots you should do are sub only, with the crossover set in the receiver to the one you normally use, and the mains disconnected. That way, you can see what the sub is doing in the mix without influence.

BTW, a 10dB house curve is a bit extreme.

brucek

**steveo1234** · **08-22-08, 07:50 AM**

Hi there.
Alrighty.
Changes per request done to pictures and uploaded, see below.
These measurements were done with both sub and mains, crossed over at 80hz in my receiver. I´ll redo measurements with sub only if that help.
The 10db house curve was just a test to see if i used the correct syntax in the .txt file. The house curve is not implemented nor are any other changes. The included measurements are with no filters added to the BFD.
So, any idea whats wrong? The 50hz ringing and wide gap in the frequency respons is somewhat troublesome for me.

**brucek** · **08-22-08, 08:47 AM**

Quote:

So, any idea whats wrong?

It's impossible to tell when you have the mains in the mix. You need to disconnect the mains and do the sub only measure. This allows you to see the sub and the room by itself. Then you can reposition the sub for the best marriage with the room.

When that's complete, you add the mains and check the interaction at the crossover (where the biggest mix effect is taking place). Then usually, a small phase adjustment to the sub will clear the crossover problems.

You also need to measure from 0-200Hz (don't stop at 80Hz).

I can pretty well tell you that you won't require a BFD.

Start off by removing that house curve and running a sub only test from 0-200Hz (with the 80Hz crossover) at 75dBSPL target.

brucek

**Ricci** · **08-22-08, 01:24 PM**

Bruce,
He's got an IB.

**brucek** · **08-22-08, 01:44 PM**

Quote:

He's got an IB

hehe, fairly difficult to move that around eh?......

brucek

**Wayne A. Pflughaupt** · **08-22-08, 05:41 PM**

Quote:

steveo1234 wrote:

Since I have a BFD I wanted to see what im hearing...
Well, from 8-45 or so it looks great, but then theres a huge dip between 45 and 65. Is it even possible to do anything about this?

Since you have a BFD, I'm wondering why you would be asking this...

Quote:

Second problem comes when looking at the LF Waterfall plot.
What is going on at and around 50hz? Anyone know how to solve this or what the problem could be?

Make sure your air conditioner, refrigerator, washing machine, ceiling fan, coffee pot, lawn mower and anything else that can make noise is turned off.

Regards,
Wayne

**steveo1234** · **08-23-08, 10:53 AM**

Quote:

Wayne A. Pflughaupt wrote:

Since you have a BFD, I'm wondering why you would be asking this...

Make sure your air conditioner, refrigerator, washing machine, ceiling fan, coffee pot, lawn mower and anything else that can make noise is turned off.

Regards,
Wayne

Hia. Even though I have the BFD im still asking because I know its a bad idea to raise output using the BFD (thank you for that info). but since I have a lot of excess energy from 10-40 I was hoping that reducing that peak would somehow magically raise my 45-65 drop. Perhaps it doesnt work that way, I have no idea.

Everything is turned off. If i run a sweep without any signal to the subwoofer (only the Conrad mic connected to the computer) I get the below graph. Peaks at 50, 100 and 200Hz. With lots of ringing..

Anyways.

I redid my graphs according the Brucek´s info above and the result I got can be seen below. I also changed my soundcard setting to recognize the SPL meter as "line in". (Green graph below), Which is the correct setting?
Comments?

**steveo1234** · **08-23-08, 11:08 AM**

And just for good measure, a near mic measurement as well:

**brucek** · **08-23-08, 11:14 AM**

Quote:

I redid my graphs according the Brucek´s info above

?

* Use a vertical axis of 45dB-105dB and a horizontal axis of 15Hz-200Hz.*

Quote:

If i run a sweep without any signal to the subwoofer (only the Conrad mic connected to the computer) I get the below graph.

Use the Spectrum Analyzer tab in REW for this using the Sig Gen set for Periodic White Noise (White PN).

You'll get a graph such as this:

Quote:

With lots of ringing..

It's hard to evaluate since your target is so high. Be sure to set the target level of your measurement and the REW target line at 75dBSPL, then the graph will be exactly in the middle area of 45dB-105dB............

brucek

**clubfoot** · **08-24-08, 01:45 PM**

Do you have a calibration file for your conrad mic loaded for the measurements?

**Wayne A. Pflughaupt** · **08-24-08, 04:11 PM**

Quote:

Hia. Even though I have the BFD im still asking because I know its a bad idea to raise output using the BFD (thank you for that info). but since I have a lot of excess energy from 10-40 I was hoping that reducing that peak would somehow magically raise my 45-65 drop. Perhaps it doesnt work that way, I have no idea.

Well to start with, the 10-40 Hz area is not a "peak." It wouldn't be advisable to attempt a wholesale gain reduction across a relatively linear 2-octave span. An equalizer is not designed to be a level-matching device. Besides, it would require about a 16 dB gain reduction to get the 10-40 Hz area down in range with the 40-65 Hz area. Since the BFD's control and effect is limited below 20 Hz, you'd end up with a big honkin' boost at ~10 Hz. Probalby not a good thing...

Your best bet is simply to boost the 40-65 Hz range as needed, provided your sub system has enough headroom (which of course any equalizing requires).

Quote:

Everything is turned off. If i run a sweep without any signal to the subwoofer (only the Conrad mic connected to the computer) I get the below graph. Peaks at 50, 100 and 200Hz. With lots of ringing..

I assume you're referring to this graph?

Since it's not being generated by your system, then it is of no consequence with regards to equalizing your subwoofer.

Regards,
Wayne

**steveo1234** · **08-25-08, 03:11 PM**

Quote:

clubfoot wrote:

Do you have a calibration file for your conrad mic loaded for the measurements?

Yup!

**steveo1234** · **08-25-08, 03:14 PM**

Quote:

Wayne A. Pflughaupt wrote:

Well to start with, the 10-40 Hz area is not a "peak." It wouldn't be advisable to attempt a wholesale gain reduction across a relatively linear 2-octave span. An equalizer is not designed to be a level-matching device. Besides, it would require about a 16 dB gain reduction to get the 10-40 Hz area down in range with the 40-65 Hz area. Since the BFD's control and effect is limited below 20 Hz, you'd end up with a big honkin' boost at ~10 Hz. Probalby not a good thing...

Your best bet is simply to boost the 40-65 Hz range as needed, provided your sub system has enough headroom (which of course any equalizing requires).

I assume you're referring to this graph?

Since it's not being generated by your system, then it is of no consequence with regards to equalizing your subwoofer.

Regards,
Wayne

Hi.
Well, its not technically a peak I guess. But, it is very easy to hear that the IB is "bottom heavy" which is a bit distracting. I will try boosting to see what happens.
Im currently having a lot of problems getting measurements. My mic either refuses to send a signal to my computer or REW is registering (an impossible) 240dB.
I need to play around a bit more to find the source of the error...

**steveo1234** · **01-03-09, 10:28 AM**

Hi all.
I decided to give REW a new try today. I spent some 2 hours trying to get resonable levels on the input and output as well as the SPL-meter and managed to get this respons:
To me it looks kind of hopeless to fix with the BFD?

**Wayne A. Pflughaupt** · **01-03-09, 11:22 AM**

Well, it certainly is a challenge. Did you ever try boosting the depressed area with the BFD?

Regards,
Wayne

**steveo1234** · **01-04-09, 05:04 AM**

Not yet. Ive been playing around with the filters in REW trying to get a good curve and I ended up with this:

http://img291.imageshack.us/img291/7...rrectedoh4.jpg

That will take 6 and 7 db boost at 53 and 60hz. Not sure how it will work out but im going to feed the BFD with this and see what happens. I have a lot of headroom with subs and amps so it´s possible it might work.

**steveo1234** · **01-05-09, 04:06 AM**

Ok.
With 5 filters, most of them with a gain of -4 or -5 I got the following result:

This looks pretty good to me. I decided not to touch the "bump" from 10-30hz as I enjoy the extra infra energy.
On the downside: I had to add a 55hz filter with a wide gain of 10dB. This isnt something I expected and im not sure I should keep it, but, it DOES give a good curve.
Any comments?

(Added original measurement for reference)

**Wayne A. Pflughaupt** · **01-05-09, 10:08 AM**

Quote:

steveo1234 wrote:

On the downside: I had to add a 55hz filter with a wide gain of 10dB. This isnt something I expected and im not sure I should keep it, but, it DOES give a good curve.

Considering how small your room is, and how huge your sub system is, I'll be surprised if you have any problems. In any event, this post should make you feel better.

Regards,
Wayne

Advertisement