EqualizingTwo Subs

[Nordo]

I've done a search but I can't find any info on this.

I have two "identical" diy subs each with two 10" drivers and two 4" tuned ports.
I have REW, a RS SPL meter, laptop, and BFD Pro.

How do I EQ two subs?

Also is there any recommended House Curves for my dedicated Home Theatre? (Subs and Mains). The room is 6.054m(19'10") x 4.25m(13'11"), with an average height of 2.89m(9'6") (The room has a 12" step acroos the middle). The room is used almost exclusively for movies. The walls and ceiling are heavily treated at the first reflection points. The floor has heavy carpet, and there are bass traps in the front corners and along the rear wall.

Thanks

[Wayne A. Pflughaupt]

If they’re both in the same location, you’d EQ them the same as one.

If they’re separated, things get complicated. You’d have to EQ them each separately, then check combined response.

Also is there any recommended House Curves for my dedicated Home Theatre? (Subs and Mains).

Part 1 of our house curve article gives instructions on how to determine a house curve for your sub. Part 2 covers mains + subs together.

Regards,
Wayne

[Blaser]

Wellcome Nordo!

I have 2 subs as well. If they are symetrically placed in your room and equidistant from you, you may just equalize them as one. That is what I am doing and itworks fine, but their response is identical at LP to begin with.

If that is not and cannot be your case, we may discuss it later, but it isa bit more tricky.
I believe the thread title should be equalizing 2 subs, not tuning which is quite different.

House curve depends on your taste and your subs capabilities. For me I do 0 at 80 and 10 db at 35 Hz, but I play a lot...

[Nordo]

Thanks Wayne.

Currently I have one sub centred against the rear wall (behind a couch), and the other centred about 5' in front of the front seats.

I was thinking of keeping the rear sub in it's position, then, without any filters, move the front sub around until I get the smoothest SPL from the combined subs, then start my EQing from there. Is that a logical approach?

I'm new to REW, and I've been trying to find and read all your articles.

The reason I started to look at EQing the subs is because one of my sub drivers is bottoming during heavy passages. So I am hoping to incorporate a subsonic filter in my EQing with the BFD. Is this possible?

Thanks

[Nordo]

blaser wrote:

Wellcome Nordo!

I have 2 subs as well. If they are symetrically placed in your room and equidistant from you, you may just equalize them as one. That is what I am doing and itworks fine, but their response is identical at LP to begin with.

If that is not and cannot be your case, we may discuss it later, but it isa bit more tricky.
I believe the thread title should be equalizing 2 subs, not tuning which is quite different.

House curve depends on your taste and your subs capabilities. For me I do 0 at 80 and 10 db at 35 Hz, but I play a lot...

I tried a sub against the rear wall, and a sub against the front wall. This placed them symetrically, but at different distances from my main listening position. Also I didn't like the sound.

A simple REW test of each sub shows a lot of output below 20Hz. This should be where the port does all the work, but the drivers are really flapping about, and one is making loud popping sounds (bottoming?)
The attachments show my rear sub (blue) and my front sub (green)

[Nordo]

Wayne
Can you change the thread topic as blaser suggests?

Got to go to work. Won't be back online for about 12 hours

[brucek]

The reason I started to look at EQing the subs is because one of my sub drivers is bottoming during heavy passages. So I am hoping to incorporate a subsonic filter in my Eking with the BFD. Is this possible?

No, you can't fashion a high pass filter with the BFD.

A simple REW test of each sub shows a lot of output below 20Hz.

They both show usable output to about 15Hz. The signal below that is noise. REW tends to compensate down low since you have the C-Weight box checked (as the C-Weight graph drops, the noise signal rises. Uncheck it, since you have a cal file loaded.

It would help out if you used the standard axis of vertical 45dB-105dB and horizontal 15Hz-200Hz. It makes it easier for everyone to read. Also uncheck the soundcard cal and meter cal to remove the lines from the graph.

brucek

[Wayne A. Pflughaupt]

Nordo wrote:

The reason I started to look at EQing the subs is because one of my sub drivers is bottoming during heavy passages.

I’ll bet one of them measures not as loud as the other with the SPL meter, so you’ve turned it up to compensate? I had the same problem when I tried separate placement of my subs, because one location was very inadequate.

As brucek has noted on other threads, the main reason to go with separated subs is if it gets you a favorable response curve that you aren’t able to achieve from a single location. There can be other reasons to separate them – perhaps blaser can give his – but typically if you do, it only works well in or near the corners of a perfectly symmetrical room – one with “shoebox” dimensions, for example.

So I am hoping to incorporate a subsonic filter in my EQing with the BFD. Is this possible?

The BFD doesn’t have a subsonic filter. People have accomplished “creating” a low pass filter by stacking multiple filters, effectively the same thing as what you’re wanting, although a high pass (aka subsonic) is more problematic. I modeled a crude one in REW using several filters, shown in this graph:

As you can see, the best you can do is attenuate the lows only to a certain level, not an infinite slope like you’d have with a real high pass. Still, it might be enough to do the trick for you. Eliminating the last filter (#9) and tweaking the others would move the cut-off point up to ~25 Hz. It wouldn’t be hard to move the cut-off up higher, if you needed to.

The caveat is that you’ll only be left with a single channel to do your subwoofer equalizing, so if you want to EQ your subs separately and create a high-pass for one, you might need a second BFD.

Overall, unless you have a compelling reason to keep them separated, I’d put the subs together and EQ as one.

When you re-do your graphs to the correct scale as brucek suggested, I’d like to see one with combined response as well as the two separately

Wayne
Can you change the thread topic as blaser suggests?

Done.

Regards,
Wayne

[Sonnie]

For the record... level match the two subs and equalize them together. It is next to impossible to equalize two subs individually... it just does not work and I have yet to see anyone do it successfully. You listen to them both at the same time... if you measure them separately, you will not be measuring what you are really listening to and the response will not be the same - no way it can be.

I have three subs and when I did equalize them, I equalized them all together... works wonderfully. When I had two subs, whether symmetrical or not, I equalized them together (and yes I moved them religiously and equalized them - I was a testing fool to say the least)... it worked wonderfully and is very simple to do. No reason to make it complicated.

Level match, measure both, equalize both and it is flawless. I have done it over and over and over in numerous setups and it works every time. If you think about it, it makes perfect sense why it works.

[SteveCallas]

Sonnie wrote:

For the record... level match the two subs and equalize them together. It is next to impossible to equalize two subs individually... it just does not work and I have yet to see anyone do it successfully. You listen to them both at the same time... if you measure them separately, you will not be measuring what you are really listening to and the response will not be the same - no way it can be.

If it's worth anything, I agree. The sound waves created from the subs will interact with each other on their way to reaching your ears, creating cancellations and spikes in the response. Only way to take that into account is by measuring both subs at the same time. Before even touching the EQ though, experiment with moving the two subs around a little bit, even if only 1-2' this way or that, it can make a huge difference.

With that said, I would still go ahead and at least measure each sub individually - if you find one has a great natural response and the other one horrible, I would move the horrible one next to the great one if possible, then use those locations as your starting point.

[jerome]

You have probably already discovered that but it will save you a lot of time to get some help when equalizing you subwoofers.
When you find a frequency range to adjust, the other person only needs to sit in the sweep spot and tell you when the SPL goes up (or down) when you're modifying the levels/phases or moving the sub(s) around. This way, you don't need to run REW for every configuration. You only do it for the best configuration.

I have used this method a few times and it went really fast. It's a bit tricky at first but works very well when you're used to it

[Nordo]

brucek wrote:

No, you can't fashion a high pass filter with the BFD.


They both show usable output to about 15Hz. The signal below that is noise. REW tends to compensate down low since you have the C-Weight box checked (as the C-Weight graph drops, the noise signal rises. Uncheck it, since you have a cal file loaded.

It would help out if you used the standard axis of vertical 45dB-105dB and horizontal 15Hz-200Hz. It makes it easier for everyone to read. Also uncheck the soundcard cal and meter cal to remove the lines from the graph.

brucek

Sorry about the axises. I thought John had set the defaults in v 4.0 to suit the forum. I didn't realise I had to limit the axises for sub readings.
I also thought the Cal file allowed for the C-weighting. If I uncheck the C-Weight box, should I also turn C-Weight off on the meter?

[Nordo]

Wayne A. Pflughaupt wrote:

I’ll bet one of them measures not as loud as the other with the SPL meter, so you’ve turned it up to compensate? I had the same problem when I tried separate placement of my subs, because one location was very inadequate.
No, both volume controls on the sub amps (Parapix 2x50w) are set about 1/4 using SPL meter and 75db.
As brucek has noted on other threads, the main reason to go with separated subs is if it gets you a favorable response curve that you aren’t able to achieve from a single location. There can be other reasons to separate them – perhaps blaser can give his – but typically if you do, it only works well in or near the corners of a perfectly symmetrical room – one with “shoebox” dimensions, for example.
I have a perfectly symetrical dedicated HT room, but I don't mind stacking the subs on top of each other if that simplifies things (they are side firing).


The BFD doesn’t have a subsonic filter. People have accomplished “creating” a low pass filter by stacking multiple filters, effectively the same thing as what you’re wanting, although a high pass (aka subsonic) is more problematic. I modeled a crude one in REW using several filters, shown in this graph:
I've got a lot to learn about applying filters (I do know that fewer is good). A few years ago when I first got my BFD, I got one sub to look pretty good on the SPL graphs, but since reading posts in this forum, I realise I may have produced a poorier sound audibly.

As you can see, the best you can do is attenuate the lows only to a certain level, not an infinite slope like you’d have with a real high pass. Still, it might be enough to do the trick for you. Eliminating the last filter (#9) and tweaking the others would move the cut-off point up to ~25 Hz. It wouldn’t be hard to move the cut-off up higher, if you needed to.
Are you saying that 1 to 8 are all used to achieve the dip below 20Hz?
The caveat is that you’ll only be left with a single channel to do your subwoofer equalizing, so if you want to EQ your subs separately and create a high-pass for one, you might need a second BFD.

Overall, unless you have a compelling reason to keep them separated, I’d put the subs together and EQ as one.
That's not a problem as they are only each about 15" high (37" x 24" in plan).

When you re-do your graphs to the correct scale as brucek suggested, I’d like to see one with combined response as well as the two separately
I have a combined graph, but I'll adjust the scales before posting.



Done.

Regards,
Wayne

Thanks

[Nordo]

Sonnie wrote:

For the record... level match the two subs and equalize them together. It is next to impossible to equalize two subs individually... it just does not work and I have yet to see anyone do it successfully. You listen to them both at the same time... if you measure them separately, you will not be measuring what you are really listening to and the response will not be the same - no way it can be.

I have three subs and when I did equalize them, I equalized them all together... works wonderfully. When I had two subs, whether symmetrical or not, I equalized them together (and yes I moved them religiously and equalized them - I was a testing fool to say the least)... it worked wonderfully and is very simple to do. No reason to make it complicated.

Level match, measure both, equalize both and it is flawless. I have done it over and over and over in numerous setups and it works every time. If you think about it, it makes perfect sense why it works.

Sonnie
When you say to "level match the two subs", are you saying to set them both individually to, say 75db (which I did), or are you saying to stack them on top of each other?
I'm happy to EQ them at the same time, but maybe move the front one around til I get the best graph before I start to EQ?

[Nordo]

OK, this is the graph of my two subs measured together (correct scales) un-EQed and in their current positions.

[JohnM]

Nordo wrote:

If I uncheck the C-Weight box, should I also turn C-Weight off on the meter?

No, the cal file compensates for the effect of the meter's C weighting.

[brucek]

OK, this is the graph of my two subs measured together

Are you sure you need to equalize?

[Blaser]

Hey, the graph looks very good already. The dip agt 56 Hz won't propably be noticeable. Try with the speakers and also try playing with phase and see if you can get rid of it.

[Nordo]

blaser wrote:

Hey, the graph looks very good already. The dip agt 56 Hz won't propably be noticeable. Try with the speakers and also try playing with phase and see if you can get rid of it.

Even though the combined curve is quite good, I still want to try and suppress the output below, say 25Hz. In very heavy passages one of the drivers makes extremely loud "popping" or "snapping" noises, and it seems to be moving around +20mm. The Xmax is 8mm. The port is supposed to be doing all the work low down - not the driver ! I guess it won't hurt to play with some virtual filters to see what I can achieve.

I can't alter the phase in my sub amps. Audyssey (in my Onkyo 805 receiver), does phase alignment, but I don't think I have any way of changing the phase of the subs relative to each other. However, I can move them relative to each other. Currently I have tried to have them equidistant from the main listening position (this centres them across the width, but puts them at odd distances within the length of the room). When I get a bit of time I will see what happens if I move one by small amounts before I start any EQing.

[Sonnie]

Nordo wrote:

Sonnie
When you say to "level match the two subs", are you saying to set them both individually to, say 75db...

Correct...

I would also try stacking them as well as moving them around. I found having one in front and one in rear to work out very good in my particular room. It works well enough I need no equalization.

[Wayne A. Pflughaupt]

Nordo wrote:

Are you saying that 1 to 8 are all used to achieve the dip below 20Hz?

Yup. Note that the #1 filter is very broad with a deep cut. It’s the one that’s causing the signal reduction below 22 Hz. The other filters are basically used to boost the signal above 22 Hz back up to roughly trace the Target Curve. You can create the same thing on-screen in REW by using the same filter values in the EQ Filters panel (which is why I showed them on-screen for you). Then you can play with the filter settings to see what each one is accomplishing.

I have a perfectly symetrical dedicated HT room, but I don't mind stacking the subs on top of each other if that simplifies things (they are side firing).

I’ll bet you’ll find that either co-locating them in a single corner, or splitting them between the two front corners, will get you better output. That would allow you to reduce your sub volume, which would solve your bottoming out problem.

Sonnie wrote:

For the record... level match the two subs and equalize them together.

SteveCallas wrote:

If it's worth anything, I agree.

Probably not a problem in a shoebox room like yours, Sonnie, or Nordo’s. Would never work in a room like mine (you’ve seen it). ’Course, I only have one good location for a sub anyway, so it would be foolish to try to equalize from a poor position.

Regards,
Wayne

[Nordo]

Wayne A. Pflughaupt wrote:

Yup. Note that the #1 filter is very broad with a deep cut. It’s the one that’s causing the signal reduction below 22 Hz. The other filters are basically used to boost the signal above 22 Hz back up to roughly trace the Target Curve. You can create the same thing on-screen in REW by using the same filter values in the EQ Filters panel (which is why I showed them on-screen for you). Then you can play with the filter settings to see what each one is accomplishing.

I’ll bet you’ll find that either co-locating them in a single corner, or splitting them between the two front corners, will get you better output. That would allow you to reduce your sub volume, which would solve your bottoming out problem.


Probably not a problem in a shoebox room like yours, Sonnie, or Nordo’s. Would never work in a room like mine (you’ve seen it). ’Course, I only have one good location for a sub anyway, so it would be foolish to try to equalize from a poor position.

Regards,
Wayne

Obviously solving sub problems by re-positioning instead of EQing is preferable. And I have the ability to position my subs in my shoebox room anywhere I like.
I first tried one centred on the rear wall, and one centred on the front wall (as recommended through testing by Todd Welti from Harmon Inter.). This didn't sound too good, and also meant the subs were not equi-distant from the main listening area.
Chris from Audyssey Labs recommended that to EQ the subs with Audyssey, they had to be equi-distant from the main listening area. And I think everyone here is saying the same thing if using REW and BFD.
Locating them in the front corners as you suggest, would put them equi-distant from the two main front seats (and equi-distant, but further away, from the rear couch). I have triangular shaped traps in these two corners, but the subs (firing each end), are long enough to run across in front of the traps. I have tried to do a quick and dirty sketch of my room below.

Whether I put the subs in the front corners, or whether I stack them, I'm going to have to run some more wiring. Should I stick with my current positions, or should I try the other suggested positions (with temp wiring), and see what results I get?

I guess to obvious answer is "give the other positions a go".

[plundstrom]

brucek wrote:

REW tends to compensate down low since you have the C-Weight box checked (as the C-Weight graph drops, the noise signal rises. Uncheck it, since you have a cal file loaded.

Hey brucek and others!
I have question about this C-weight box. In REW help file it says clearly that C-weigth box should be checked when using SPL meter and uncheck when using real mic&pre-amp. This guy was using RS SPL and you asked him to uncheck this box?

Whcich one is the correct way? Manual or your advise?

[brucek]

Which one is the correct way?

The meter or mic calibration file over-rides the C-Weight compensation over the frequency end points of the files. So, for example, lets assume a meter files extends from 10hz to 200Hz.

If the C-Weight box is checked, then below 10Hz and above 200Hz, the calibration extends using the last values for those frequencies. So, if the last value was -5dB at 10hz, a straight line would extend down to 0Hz at -5dB. The same would apply above 200Hz for its value.

If the C-Weight box is checked, the slope of the C-Weight curve is extends beyond these end points, so that the meter calibration file only substitutes the values over its range.

Seems like a good idea to check the C-Weight box then. But here's the rub. When the C-Weight box is checked and the resulting calibration continues lower and lower following the C-Weight curve (below 10Hz in this example), it compensates by adding that inverse value to a subwoofer signal that is now in the noise and unusable. the plotted response looks like a rising signal at low frequencies. I can't even count the number of people that post graphs down to 2Hz and say look at how great my sub works down to 2Hz.

By unchecking the box when measuring subs, this problem is removed. In fact, the new version of REW (soon to be released) remedies this situation.

Here's a pictorial of what I mean in the following jpgs.

One has the C-Weight box checked and the other doesn't. See the wacky rising signal in the checked example. It isn't the best example in the world, but it's all I had... You get the idea.
For the full range measure there's no problem, but you should not be using an RS meter for full range anyway. In addition, a full range ECM has its cal file go to 20K, so unchecking the C-Weight is still a good idea if you use a cal file (in this version anyway).........





brucek

[plundstrom]

Thanks brucek. I see your point with really low frequencies but I'm still not sure about others. If we are interested only about normal usable frequencies say 10Hz to 200Hz and RS SPL calibration file is used, then does it make any difference to results if C box is ticked or not?

ps. I think REW online manual should be "corrected" with this mater.