Repositioned sub and started to re-EQ

**Malice** · **01-31-07, 04:29 PM**

Well, having lived with the sub in one position for over a year and having got continual gripes from 'er indoors about the "boominess" (even though I had used the BFD to flatten the response), I decided to move the sub to the front next to the TV and inside the front speakers. I thought it may integrate better and also means that the wife is not sat right next to it when watching a film!!

I shan't bore you with the many sweeps and graphs of different combinations of crossover, PEQ on the amp and phase settings on the sub. But suffice it to say I was nearly disappearing up my own **** with all the data sets. What I will say however is that I could not believe that for a different XO on the amp, the optimum phase on the sub had to change. On a 60Hz XO to get the best overall response of mains and sub together around the XO point the phase on the sub had to be 0 deg. (I swept for 0, 45, 90, 135 and 180 and compared the output). When I switched to an 80 Hz XO, the optimum phase on the sub was found to be 180 deg. That was a surprise as I assumed the amp would have been consistent in its output across all XO.

Anyway, to the point.

I wanted to sweep the mains and sub with the BFD in the loop but with no filters active, so that if there was a delay/phase shift through the BFD I could "optimise" the phase from the outset.

As far as I am concerned, this is the best I achieved when having the amp set to "Normal" PEQ (for the upper frequency EQ), a manual check of levels using the SPL meter and the Denon test tones, but leaving the distances as automatically calculated by the amp on auto set up.

(See attached JPG)

This is the response WITH NO EQ AT ALL!!! Note I get in room response of 15Hz at the target levle of 75db!

Question is, should I try to tame the peak at 25Hz? Is it too much?

Thanks

Bob

EDIT I've added the SUB ONLY graph and MAINS ONLY graph.

**brucek** · **01-31-07, 04:51 PM**

Quote:

Question is, should I try to tame the peak at 25Hz? Is it too much?

Yes........................

brucek

**Malice** · **01-31-07, 05:53 PM**

Well, I've created two BFD presets:

(1) Flat (ish)
(2) House curve with 20 Hz + 6db and 50 Hz at 0 db

I'll live with both for a while and see what's what!

But I am really pleased with the end result as these are the final combined responses of mains and sub.

**cyberbri** · **01-31-07, 06:03 PM**

Maybe cut 25Hz 10~12dB...

**Malice** · **01-31-07, 06:06 PM**

I have to assume it is an anomoly in the calibration file and sound calibration because I cannot believe my system has this response....

**cyberbri** · **01-31-07, 06:18 PM**

Quote:

Malice wrote:

I have to assume it is an anomoly in the calibration file and sound calibration because I cannot believe my system has this response....

Yeah, don't look at anything below 10Hz or so.

**DrWho** · **01-31-07, 06:38 PM**

Are your mains set to large or small?

**brucek** · **01-31-07, 06:43 PM**

Quote:

I have to assume it is an anomoly in the calibration file and sound calibration because I cannot believe my system has this response

No, that's a result of the meter cal file stopping at 10Hz and then the C-weight carrying on down very low. When that curve reaches the noise, it compensates just as a standard calibration file does to any signal, by applying the inverse gain to it.

As said, ignore anything below 10Hz.........

brucek

**Malice** · **02-01-07, 06:28 AM**

Quote:

DrWho wrote:

Are your mains set to large or small?

Small. If they were large the MAINS ONLY SWEEP would look like very similar to the SUB down to just below 30Hz. My MAINS go low with room gain.

The top line is the MAINS with no processing (i.e. LARGE) and then 40, 60, 80, 100 XO's.

**Malice** · **02-01-07, 06:47 AM**

I think what is now clear to me is that I have accidentally stumbled upon a reasonable position for my mains and sub. The MAINS in their current position get the life sucked out of them at about 68Hz, and although there is a similar dip in the sub at the same frequency, I presume its position / phase setting has minimised this to an extent where the mains fall off has been all but compensated by the Sub. More or less.

I good few evenings' work I'd say!

**DrWho** · **02-01-07, 03:16 PM**

Something isn't right if you've got over 15dB of gain at 25Hz...that's on the order of what you would get in a car - I doubt your room is that small. In fact, your speakers are rated at -6dB at 30Hz...so call it -9dB at 25Hz and you're looking at over 24dB. This is not normal at all.

What kind of mic are you using? Is everything calibrated correctly and all that? What's the size of your room?

**Malice** · **02-02-07, 06:16 AM**

Hi Mike,

I'm using an old style (square) Ratshack SPL meter with the appropriate calibration file. The REW and PC calibrate fine with internally generated tones (both speaker pink noise and sub pink noise) and this validated OK when using the Denon's test tones (once the Denon had auto EQ'd @ 75db using the Denon Mic) Everything seems to align with respect to calibration and levels.

Room is not entirely rectangular!

**DrWho** · **02-02-07, 09:41 AM**

Room gain for your room shouldn't start any sooner than 30Hz, so there is no acoustical explanation for the insane boost you're seeing. And then it would be an engineers dream come true if your speakers were exhibiting this natural behavior - but I highly doubt any marketing firm would allow underated specs for hte frequency response...

Assuming your entire process is correct, then the only explanation is that the calibration file for your meter isn't accurate - basically your meter is exhibiting wierd behavior.

What do your ears tell you?

**Malice** · **02-02-07, 10:10 AM**

Mike,

Before I knew about calibration files for the ratshack meter and before I knew about REW, I used Sonnies test tones and burned a disk and manually plotted a response in Excel. This was when the SPL meter was new out of the box. I had been aiming for a target level of 80db!

As far as my ears are concerned, "untamed" feels too boomy at the lower levels. Tamed and the sweep levels feel about even, but perhaps a little flat at the lower end.

EDIT, just as a note, I remember when I did the manual 1hz test tones, the sub was significantly louder to my ears over the 20 - 35 range, and borne out by the SPL readings.

**JohnM** · **02-02-07, 12:59 PM**

That's not room gain at the bottom end, it is the first length modal resonance at around 26Hz and the first width mode at or just over 30Hz. Can easily get modal gain of 15dB or more at those frequencies.

**brucek** · **02-02-07, 01:39 PM**

Quote:

it is the first length modal resonance

brucek

**Malice** · **02-02-07, 01:53 PM**

Me being a complete half wit with the finer technical nuances in this thought room gain and modal gain were one and the same thing.

JohnM, your use of the word "resonance" perfectly describes what I appear to hear near the frequencies that Brucek has kindly chipped in with; with single test tones that I did over a year ago, the room seemed to take on a life of its own and nearly took off! People in the other part of the house were complaining about the loudness of it. TBH when I first set out to manually measure and EQ the sub, I tried to achieve a target level of 85db, but as I approached (with hindsight) the modal frequencies, particularly the one about 26Hz, the SPL meter was registering 115dB. No wonder there were complaints!

**DrWho** · **02-02-07, 05:30 PM**

Serves me right for rushing numbers trying to catch the bus - there's quite a difference between 11 and 21 feet....

I would like to take a look at the energy time curves...if it's really entirely the result of a standing wave, shouldn't you see a series of spikes with 38ms intervals? And then wouldn't it take at least 32 equal magnitude reflections to achieve 15dB of gain?

I ask because I don't know and am skeptical of the possibility for 15dB gain on a standing wave...especially considering that a good portion of the energy will be transmitted through the walls/windows. Where is the energy coming from?

On a side note, if you reduce the response of the subwoofer to flatten out the peak, then aren't you also reducing the sub in compensation for the response of the mains? If so, then aren't you compromising the LFE channel that is only being sent to the sub? If you set the LFE to Sub + Mains, does it still engage the highpass filter to the mains?

**Malice** · **02-02-07, 07:54 PM**

Quote:

I would like to take a look at the energy time curves...if it's really entirely the result of a standing wave, shouldn't you see a series of spikes with 38ms intervals? And then wouldn't it take at least 32 equal magnitude reflections to achieve 15dB of gain?

I've attached the mdat for the sub on its own with the 80 XO.
NOTE: In order to upload the file I had to zip it and then rename it to mdat as at 750.2 Kb it exceeded the upload limit of 750kb! So "save as", rename to a "zip" extension, unzip and there you have my raw data. Make of it what you will coz I certainly I am not on your wavelength (sic) at all with the level of detail you are talking about! This is obviously an area of expertise for you!

If you want the mdat of the mains on their own with no XO in play, let me know as well!

Quote:

I ask because I don't know and am skeptical of the possibility for 15dB gain on a standing wave...especially considering that a good portion of the energy will be transmitted through the walls/windows. Where is the energy coming from?

Is there any way there is an unusual contribution from the fact that I have a bay area with walls at 45 deg and an entrance at a 45 deg angle as well?

Quote:

On a side note, if you reduce the response of the subwoofer to flatten out the peak, then aren't you also reducing the sub in compensation for the response of the mains? If so, then aren't you compromising the LFE channel that is only being sent to the sub? If you set the LFE to Sub + Mains, does it still engage the highpass filter to the mains?

In order to get the option to have LFE + Mains I have to set the speaker config to set the MAINS as "Large" in which case this bypasses the crossover filter and would make EQ'ing too difficult or impossible perhaps! As it is setting all the speakers to SMALL will route MAINS below the XO and the LFE to the sub.

If you look at the first post you will see that the curve for the combined MAINS and SUB is almost identical to the sub on its own up to about 65 Hz, so I assume if I tame the sub to a flatish response then any LFE will also be tamed, but not to the extent of it being non-existent.

Bob

**DrWho** · **02-03-07, 02:26 AM**

Hey Bob,

I downloaded your file, converted to .zip and then unpacked and tried to load the measurement in REW and I get an error saying that it's not compatible with REW4. I wonder if something got messed up during the whole process?

Maybe try sending it via email? mebentz2 -at- gmail.com
(better yet, you could send me your impulse response - which you can obtain via the export option).

Btw, I'm by no means an expert - I know just enough to get myself in trouble

I'm sure John is right, but I'm just not seeing it...I've always been told that the rule of thumb was 6dB for standing waves?

**Chrisbee** · **02-03-07, 05:22 AM**

The two slanting windows behind the listening position seem to focus optically slightly in front of the hot seat.

One could well imagine that long soundwaves would also be focussed (or concentrated) just there.

I can't save any images in *** Paint.NET or I'd post a picture to show what I mean.

**Malice** · **02-03-07, 06:19 AM**

I know what you mean Chris, no need to stretch your Painting skillz for me!

Whether or not I have made any difference to this potential issue is not clear, but behind the seating position I have some very heavy "velvet" curtains that go all around the bay with about 400mm of the bay wall showing top and bottom.

I doubt they will be much of a barrier for the LF though!

MIKE

I've emailed the MDAT. The WAV export didn't appear to work for me and the text export was bigger than the MDAT!

FWIW, I just downloaded the file to another directory, renamed the extension to zip, unzipped the mdat and it loaded fine into REW with no issues for me.

Just to whet your appetite, this is the Energy time graph

JOHNM: I could not get the Y axis to scale above +20. It appears to be a max?

**MACCA350** · **02-03-07, 07:22 AM**

Quote:

Malice wrote:

I could not get the Y axis to scale above +20. It appears to be a max?

Select the 'Plot Response Normalized' option down the bottom, this will fix it.

cheers

**Malice** · **02-03-07, 07:37 AM**

I had tried that MACCA350, but as it puts the highest peak at 0 you have to do mental maths to calculate how many dbs any peak is over the target response. Having relative dbs below the highest peak could be useful in other situations, but I think it would help me better to understand the mechanics at play and how and by how much I appear to be getting up to nearly 20db boost with room modes at certain frequencies.

Unless I am misreading the normalised graph?

**MACCA350** · **02-03-07, 09:37 AM**

Quote:

Malice wrote:

I had tried that MACCA350, but as it puts the highest peak at 0 you have to do mental maths to calculate how many dbs any peak is over the target response. Having relative dbs below the highest peak could be useful in other situations, but I think it would help me better to understand the mechanics at play and how and by how much I appear to be getting up to nearly 20db boost with room modes at certain frequencies.

Unless I am misreading the normalised graph?

As I understand it, the Energy-Time graph helps you determine the Reverberation Time Decay RT60(time required for a sound in a room to decay by 60 dB) which in the average home theater is normally between 300-600ms. To do this you will need to normalize the peak to 0db so you can see where the -60db point is easily.

cheers

Advertisement