BFD EQ Help (1 Viewer)

[Citizen87645]

I would appreciate some advice on how to proceed with equalization of my sub. I only have two filters applied - one at 36 Hz and the other at 47 Hz. As you can see it looks as though I made the dip at 63 Hz worse. I've tried various filters on the 63 Hz frequency but nothing seems to affect it. I assume that means the other filters are the cause, but both are very narrow (setting of "5" for bandwidth). Thanks in advance.

 

[Wayne A. Pflughaupt]

It looks like you’re setting your filters too wide, Cameron.

Start with the 47Hz filter. It should probably be about 1/4-octave. Start with it at 1/4, and boost about 8dB, then go back and take new readings at 40,45,50 and 56Hz. (You may have to get new baseline readings before you start, unless you have some kind a pre-set level for this.) Pay special attention to the outer frequencies, 40Hz and 56Hz. You ultimately want the widest bandwidth that does not change those readings. So, if you see they’ve risen, tighten up the filter.

If they haven’t changed, make the filter a little wider, and keep increasing and taking new readings until you begin to see a rise at 40Hz and 56Hz. When that happens, decrease the filter to the last setting that didn’t show a rise. Make sense?

You have to keep in mind that the filter will “spread out” if you increase amount of boost, so be prepared to readjust the bandwidth. Just the opposite if you decrease the boost – the filter will “shrink,” so be go back and open it up as needed.

After you get the optimal 47Hz filter, you might want to start over with your 36Hz filter. It should retain constant readings at about 28 and 40Hz.

If nothing affects the 63Hz problem, it may be a null. If that’s the case, there’s not much you can do with it.

Hope this helps.

Regards,
Wayne A. Pflughaupt

 

[WayneO]

I'd cut the peak at 74hz down 6-7 dbs next and see what happens.

 

[Citizen87645]

If I have the bandwidth set to "5" doesn't that mean it's 1/12 of an octave? Should I make it narrower than that or just try the 1/4 octave (a setting of "15").

Thanks for mentioning the null. I hadn't thought of that.

 

[PaulT]

If you want to check the 'theoretical' results of changes to the filters in the BFD I would suggest this great excel spreadsheet by Anthony Gomez:

http://www.pvconsultants.com/audio/eq/peq.htm

I used it in conjunction with the BFD spreadsheet you are currently using and I was able to tune in the sub with less measurements.

I would agree with Wayne that the 63Hz dip is probably from the room. Can you move the sub half a foot one way or another to test that ?

The 'x' value for Bandwidth is x/60th of an Octave, so 5 is 1/12th.

 

[Brian Fellmeth]

Can you move the sub ? Might help with that 63 hz dip that will be difficult to equalize out.

 

[Citizen87645]

That other spreadsheet is great. I know it will not duplicate exactly in the real world, but it's good to see what the filters are doing, especially for a visual person.

I'm pretty stuck with the sub location, at least at my present level of motivation. I would have to do some major reorganization and I already did a round of that recently (fortunately not because of the sub, 'cause then I'd be really annoyed.)

 

[Wayne A. Pflughaupt]

I don’t have a BFD, so I can’t interpret their number/bandwidth settings. But a 1/12-octave filter is pretty narrow. That’s what you would use to tackle that null (assuming it could be EQ’d).

I do know that the BFD has a tendency to “spread” a lot wider than its bandwidth settings indicate when the boost or cut gets substantial. This is a common problem with cheap equalizers. So who know, a “5” could be a so-called 1/12-octave filter.

The good news, though, is you can adjust the bandwidth to where you need it, no matter what the thing “says” it is. Just try the method I outlined above and you can’t go wrong – although it may be time consuming.

Regards,
Wayne A. Pflughaupt

 

[Citizen87645]

Well, I spent another evening making adjustments. I've noticed it's hard to judge things after an extended period of hearing test tones. Everything just sounded weird last night but much better this morning after letting my ears rest a bit. Here is the latest graph.



I tried to EQ the dips at 63 and 100 but nothing moved them. I'm pretty pleased with the results from 40 to 56, the area that seemed the most in need of adjustment (besides the apparent null areas). I tried making additional filters for the smaller peaks and dips below 40 Hz, to no effect. I'm guessing this may be the wrong approach and it would be better to adjust the existing filters.

My main question now is what my expectation should be. Should I keep tweaking to improve it or is this pretty good? I do like the more even response, but am not sure if I should be concerned about those smaller peaks and valleys in the 20 - 40 range.

 

[Citizen87645]

I have filters at 47, 38 and 25 Hz.

 

[Richard Greene]

The goal with a BFD should be bass in a +/-5dB range.
Sometimes the best you can do is +/-6dB
You seem to be close to +/-5dB.
The major problem getting to +/-5dB is not the peaks
(the BFD can handle them) -- it's one or two nulls.

You can't affect standing wave nulls with an equalizer.

And you can't simply move your seat (or speakers) to get away from all nulls in a room.

But you can often move your ears far enough away from the deepest portion of nulls so that all you have left is one or two nulls in the -5 to -10dB range.

It's most accurate to measure the sound in two locations roughly eight inches apart and then average the results to represent what you hear through your left and right ears.
This makes little difference with peaks, but can make a significant difference when measuring nulls.

The standing wave null roughly half way between the left and right walls, for one example, is so narrow that it's possible to measure -20B at the trough of the null and
-10dB when measuring just one foot to the right or left of center.

A null of 63Hz. would occur roughly halfway between opposing surfaces 565/63 = 9 feet apart or
one-quarter (or three-quarters) of the way between
opposing surfaces about 18 feet apart

Opposing surfaces are two walls or the floor & ceiling.

Your frequency response is pretty good except for the peak above 100 Hz. that should be eliminated.

Subs sound best when they have little output above roughly 80Hz. so you can't hear male voices through them while all other speakers are turned off.

 

[Lee Carbray]

Cameron were your measurements taken with the mains on or off?

 

[Citizen87645]

Thanks for the info Richard. I'll take a whack at the peak at 111.

I disconnected all my speakers during the measurements.

 

[Lee Carbray]

You might want to do the measurements with the speakers on. That is, after all, how you listen to the system. It looks form the plot that your crossover is probably at 80Hz? You will find that the mains will probabaly affect the dip at 63, we will have to wait and see it its good bad or otherwise.

You will also be able to see how you the sub is interacting with your mains. How well the levels are matched and how well the phase is set etc. You can also then smooth out the response out past 80Hz.

What are you running any way? We may be ablet o suggest some alternate settings if we know you speakers and receiver.

 

[Citizen87645]

I tried cutting down the peak at 111 Hz and the only thing that seemed to put a dent in it was a filter at 100 Hz cranked down to -48 dB. Didn't sound very good though, so I turned that filter off and have it back as it was with the first graph. Here are the results with the 4th filter:



I'll go ahead and take some measurements with the speakers connected.

I am using:

-Pioneer DV-563A
-Outlaw ICBM with crossovers set to 80 Hz for mains and surrounds; 100 Hz for center channel
-Onkyo 575x receiver
-JBL S26 mains/surrounds and S-Center

 

[Richard Greene]

I'm not sure why you would not be able to reduce the peak at 111Hz. unless your bandwidth is too narrow and the center frequency is wrong.

Most peaks are addressed with 1/6 octave (10/60)
to 1/4 octave (15/60) bandwidths.

Once in a while there are two adjacent peaks that need
1/3 octave (20/60) bandwidth.

I'd try a 1/4 (15/60) bandwidth at 110Hz. with a -20dB setting to start with. That should make an audible reduction in the upper bass.

I'm not sure what your crossover frequency is but I'll assume 80Hz. That means you need little output over 100Hz. so could even experiment using a 1/2 octave bandwidth at 125Hz.

These relatively broad bandwidth cuts to the upper bass frequencies should be easily audible with you standing right next to the BFD twisting the dial.

 

[Citizen87645]

If it is another null at 100 Hz wouldn't that explain the ineffectiveness of the filters?

I think I tried just about every possibility. I tried different frequencies (from 100 to 125 Hz), increasing the bandwidth (as high as 1/2 octave). I think what finally had an effect was a filter at 100 Hz, 1/2 octave and -48 dB. But like I mentioned, it didn't make the overall sound quality better.

 

[Citizen87645]

I took measurements with speakers connected and disconnected and remeasured all frequencies with the BFD with filters and without. Here's a composite of the two graphs.



Thoughts?

 

[Lee Carbray]

The new dip at 89HZ could be a phase issue. Try adjusting the phase to see if that helps befor trying to boost it.

Notice the peak at 111Hz is not an issue with the mains on.

 

[Lee Carbray]

As an experiment, since you have the ability, is to try your crossover for the mains surrounds at 60hz. Just to see what it does.