How am I doing with my BFD? (1 Viewer)

[Rick Radford]

I'd appreciate some feedback from you folks who've done this before. I feel a bit uncertain setting the bandwidth parameter.
I find the peak of the freq. (dB), then see how wide the peak is (Hz), then convert from Hz to fractional octaves for the width.
For instance: a peak at 46 hz with the valley to the left bottoming at 40 Hz.. I figure +/- 6 Hz... or 8/60ths octave. ((8/60)*46). Is this the way you guys figure the bandwidth?
Anyway, using the right front corner since it gave the flattest response, I set up 3 filters and fumbled my way to this:

If the graph doesn't come up, try this link.
At this point, understanding that I'm expecting an improved sound, I am pleased with my first efforts. I was hoping to get less of a sawtooth effect, and probably can once I get comfy with adjusting the parameters of the BFD.
Next up: eq for the LF corner.. and then decide which corner sounds best.
Edit: corrected the corner placement to the RIGHT (not left) front corner. sheesh.

 

[ThomasW]

Can't see your graphs.

The bandwidth is expressed as a fraction, 20/60= 1/3 octave, 40/60= 2/3 octave, 10/60= 1/6 octave, etc....

 

[Rick Radford]

Can't see your graphs.

Dang it. I had problems with that too until I went to GeoCities with the same browser and looked at the uploaded graphs there. Then they showed up in the msg here just fine.

I don't know what I'm doing wrong with the graphs.

I understand that the bandwidth is expressed as a fractional octave, but I need some way to relate to the graph so I can tell how wide to set the filter.. so it seemed reasonable to determine how wide (Hz) I wanted to filter, then convert that to a fractional octave for the BFD.

IE, if a peak is at 40 Hz, then 6/60 octave = 4 Hz, right? So if I set 6/60 in the BFD, do I get +/- 4 Hz.. or +/- 2 Hz at whatever gain I select?

Am I confused as to how to set this up?

 

[Jeffrey Noel]

To see the images, right click on the "broken" pics and choose properties. Then copy the entire picture address and paste it in your browser, then hit enter, and viola! Now you can see!

 

[Rick Radford]

Anyone have an idea what causes a broken pic? I don't understand why the img function works erratically for me.

 

[Brian Fellmeth]

Jeffery's trick works, but all of a sudden the pic appeared in the thread also. Anyway,

IE, if a peak is at 40 Hz, then 6/60 octave = 4 Hz, right? So if I set 6/60 in the BFD, do I get +/- 4 Hz.. or +/- 2 Hz at whatever gain I select?

I wasn't sure of this myself, so I fired up Behringer's simulater and looked at a single -16db filter centered at 100 Hz, 60/60 bandwidth (a full octave). Your question is- would this filter between 50 and 200 Hz or between 75 and 150 Hz. There is no hard and fast point where the filter dies out- it slowly rolls off. If you define the boundrys of the filter as the place where it is 1/3 stregnth then that is between 50 and 200. Or, plus or minus 4 Hz in the context of your original question. So it would seem that the bandwidth of the filter in fractional octaves is the distance from the center to the edge, not the distance from one edge to the other.

One last remark- I suggest including main speakers in the next round of tweaking. You could spend a day getting that graph perfect with 8 or 9 filters only to find out that phase cancillations and augmentations with the mains playing create possibly large humps/valleys around the crossover frequency.

 

[Wayne A. Pflughaupt]

Rick,

Post your un-equalized 1/6-octave readings here and I’ll tell you how to set your BFD.

If you haven’t already, put your sub in a corner, the one with the longest uninterrupted wall length in both directions. This will give the highest output and the smoothest response, and should have minimal “sawtooth” readings.

Regards,

Wayne A. Pflughaupt

 

[Rick Radford]

Brian, thanks for the feedback.
Regarding the graphs.. they would not come up at all for me today.. even if I clicked on the link. But when I browsed to the graph on my geoCities site and left it up while I came back here, then the msg showd the graph. What's up with that? Anyone know? (I only tried with Opera 6.0 today).

I suggest including main speakers in the next round of tweaking

And that is something I meant to do.. but forgot.
Oh well.. another day. I have lots of learning to do yet anyhoo. And now that I've discovered the autorun ETF feature, I can create filters and watch the result on the fly.. and barely have to think about what I'm doing.
The toughest part of using the BFD and ETF is trying to predict what the bandwidth adjustment will do. But now I can crank in a freq, select the cut I want, then vary the BW on the fly and watch the result on the screen.
If anyone else is reading this and stumbling through the learning process with me. Stay tuned. I'm comfy in my test dummy job.
A quick refresher: here's my uncorrected RF and LF graphs (LF spl adjusted to match RF at 20 Hz). You can see the reason I have to cut the trim by 6 dB when the sub is in the LF corner.

Now, here's a graph of both corners after applying filters (2 for the LF and 3 for the RF corner). It (LF) stays within +/- 3dB until ~67 Hz (~1/3 octave of 80Hz.. crossover). But the RF corner stays within +/- 3 dB until well after the 80 Hz crossover freq. It just requires 6 dB more trim than the LF when calibrating with Avia/VE.

 

[Rick Radford]

Wayne,
I completely missed your post. I musta been cranking out my epic response to Brian and just missed seeing yours. Sorry!

Post your un-equalized 1/6-octave readings here and I’ll tell you how to set your BFD.

Sure wish I knew how to get those numbers for you. I am using ETF and it generates its own MLS tones. If you're familiar with ETF, can you direct me how to furnish what you're asking for? Also, the ETF RS calibration file doesn't go below 315 Hz or so, I believe.. so my graphs don't show any mic comp for the RS spl meter errors.
If you'll note in my response to Brian, I posted a graph of my unequalized Low Freq Response graph. My LF corner has the longest uninterrupted wall; it also gives me about 6 dB more headroom, but does not provide the flattest response.
Note the following diagram for my room layout. This shows the sub in the LF position. I've got about 3 feet open either side of the armoire...just room for the main and the sub.

Thanks to Sonnie's post in the HW section, I discovered I'd not used the chart to setup the filter freqs correctly.
OTOH, the autorun feature of ETF made it simple to make changes to the BFD and almost instantly see the effect onscreen.

 

[Wayne A. Pflughaupt]

Rick,

I don’t know if it’s my computer, but I can’t get any of your diagrams to open. Will the ETF program let you save a response plot to a document that you could e-mail to me?

Regards,

Wayne A. Pflughaupt

 

[Rick Radford]

Will the ETF program let you save a response plot to a document that you could e-mail to me?

I can do a screen print, paste in MS Paint, and save as a jpg file. Will that do?

 

[Wayne A. Pflughaupt]

Rick,

That will be fine. Bring it on!

Regards,

Wayne A. Pflughaupt