My BFD results. Need advice, please. (1 Viewer)

[Troy_K]

In this bottom graph, the sub only is in blue.

 

[Brian Fellmeth]

Under your banwidth column, are those units 60th of octave ? If so, your filters are real tall and spiky- makes me wonder what is going on between your data points. Suggest confirm with slow sweep that you don't have a sawtooth hiding under that apparently perfect curve. If you pass that test, looks like as close to perfect as is possible.

 

[Troy_K]

After about 6 hours of failing, I figured that it was easiest to use narrow filters of large quantity than fewer, broader filters. With the broader filters, my valleys grew deeper, and I didn't want to boost unless absolutely necessary. I have tones for every frequency 10-100 hz, and I do have some very narrow (1-2 hz) peaks and valleys still that aren't apparent by my charting, but only a few.

Should I be concerned by all the heavy attenuations I have done, especially in the upper bass regions? I haven't cranked it up yet to see if I lost any headroom, but my amp is not very large that I'm using with my SVS 20-39 CS+.

Does anybody have any suggestions on how to level out my FR based on my non-eq'd levels that I can try with a different preset? My non-eq'd graph shows no peaks, but only valleys, so this was challenging to work out. I hope I did it right.

Thanks,
Troy

 

[Darrin_R]

Your not going to lose headroom from making cuts.

Presets 8, 9, and 10 might be a little aggressive but I think it looks like a nice house curve.

How much did you have to increase the gain on the amp when you recalibrated the sub?

 

[brucek]

Troy_K,
Your end result certainly looks good, and it seems to integrate with your mains rather well.
Your BW's do look a little thin, but when I entered them into the BFD design software, the resultant curve (which represents the voltage waveform that you are sending from the BFD after the filters), seems OK. It's a good idea to downoad this software and enter the filters into it and look at the graph of what your filters looks like to give you a better idea of what is going on. I suspect you could have combined a few filters and ended up with the same result with a bit less spikey graph, but it's not too bad. If you've done perhaps the 1/12th octaves and a few sweeps and you don't notice anything funny, then I'd leave it alone.
You can download the design software from Behringer and enter your filters to see what your filter curve looks like. It's useful to understand the width and interaction between your filters, and also the resultant phase changes you're ending up with. Using it can sometimes show you that two filters are really fighting each other and you can just use one and accomplish the same result. Getting a feel for the phase changes you have introduced with your filters is also fairly interesting. It can cause some unwanted cancellations around the crossover overlap with your mains.
The web site is below:
http://www.behringer.com/05_support/...mid=4&lang=eng
Select FEEDBACK DESTROYER DSP1100P Design and download the zip file.
Or download the latest version that works with XP from here: Link Removed
The latest version has a smaller screen than the original, so I prefer the original.....
You have to play around and get used to the software, but it's useful once you get the hang of it just to see how filters work. When you first start this program it shows a MIDI port error. Click OK and the program will start. The error is because you haven't set the MIDI port parameters yet. You can set them once your in the program and it won't do that error the next time.
brucek

 

[Wayne A. Pflughaupt]

Troy,
That’s an awful lot of filters for unequalized response that isn’t all that bad to begin with. You might want to look at this link to get a handle on how filters and bandwidth work in the “real world”:
http://www.hometheaterforum.com/htfo...407#post708407
You might want to check out some of the other posts on this thread as well. After it was said and done, the “owner” of the thread was able the achieve the same response with 4 filters that he had previously accomplished with 12 filters.

... I didn't want to boost unless absolutely necessary.

It’s fine to boost when you need to. In fact, look at your filter chart: Ten of the eleven filters are cutting. You didn’t do yourself any favors by doing this – the overall gain is dragged down so far you will have to turn the sub way up. Well, there goes all the “headroom” you thought you were conserving.
In fact, many of those valleys you initiated were caused by unnecessary frequency – like below 36Hz for instance. If the goal is flat response, it makes no sense to EQ a 1-octave area that is almost perfect to begin with - practically half of the sub’s entire frequency range in this case.
Looking at your unEQ’d response, your 32Hz and 100Hz points are pretty much at locations that will make a nice house curve. Don’t mess with anything below 32Hz – it’s flat and it’s fine. Work on sloping down from that point to put 100Hz at 89-90dB (or 5-6dB below 32Hz). This will give you a nice house curve, and you will probably find that the sub’s gain setting will be close to where it was before EQing – maybe even a little less.
So it’s entirely possible you will end up with a lower gain setting on the sub despite boosting the 40Hz area considerably. This is certainly better than before, EQing overall gain down so far that the sub has to be turned way up.
Regards,
Wayne A. Pflughaupt

 

[brucek]

Troy,

I think Wayne is quite correct in his advice. You result is good, but you can do it another way with less filters.

Sometimes dips or valleys won't respond to equalization. But when they will, it's smarter to attack one dip than it is to cut all other frequencies, and then compensate by increasing the overall sub gain. The multiple filters will each introduce its own phase alteration. Fewer filters means fewer phase problems. Go ahead as I suggested, and enter those filters into the design software and check your resultant phase curve - a bit messy for sure..

Either way, the frequency where the dip is located is going to end up being boosted either by a single gain filter or by increasing the overall sub gain to compensate for all the cuts at all other frequencies. You will lose headroom at the dip if it's equalized in any fashion, but as long as you don't push too much voltage to that frequency, so that you've eaten all your headroom up, you'll be fine.

You could do as Wayne suggests and just use the minimum number of filters required to remove the dip and end up with a fairly flat response with the natural drop you seem to have already, and then use a single filter in the higher frequency range to install increased house curve if you still think you require it. Generally if you raised the 40Hz dip and reduced the 56Hz area, you'd have a good response.

brucek

 

[Brian Fellmeth]

The multiple filters will each introduce its own phase alteration.

I'm not sure that this is true since the BFD does its filtering in the digital domain. Analog filters introduce phase shifts, but I'm not sure that the number crunching of digital filters has this problem.

 

[brucek]

Brian,

I'm not sure that this is true since the BFD does its filtering in the digital domain. Analog filters introduce phase shifts, but I'm not sure that the number crunching of digital filters has this problem.

Yep, absolutely. Interestingly, the design software I gave the download link for is quite useful in this regard. It displays two graphs. One graph shows the effect of a filter and the other graph shows the resultant phase shift caused by that filter. When you enter all your filters you see the overall response and phase change that will be introduced to the BFD's incoming signal.. This is quite useful in ensuring that you aren't changing the phase around the crossover point..

brucek

 

[Troy_K]

Bruce/Wayne,

Put those two names together, and we have Batman!

So what you are saying is I should try this: To get a flat curve, I could simply boost up 40 and 71 hz and take 56 down a bit?

What do you suggest I start with for Q? I'm a bit worried about addding too much gain at 40, as I anticipate it will need at least 12-15 db gain which really seems to rattle things in areas other than my listening position. How much of a gain is safe to add?

Reading some of Vince's posts, I was led to believe that it was better to bring all other levels down to your valleys rather than boosting up your valleys.

FWIW, prior to having a BFD, I had my Denon's sub volume at around -10 with my sub amp cranked up. This was about 3-4 db hot according the S&V setup DVD. After hooking up the BFD and EQing, I still have the Denon at -10 to -8 for a 3-4 db hotness. Interesting considering all the attenuation I dialed in.

Thanks for all your input so far, this board has been very helpful to many. I look forward to further advice. I'm so glad the BFD has many presets to dinker with!

Troy

 

[Wayne A. Pflughaupt]

What do you suggest I start with for Q? I'm a bit worried about addding too much gain at 40, as I anticipate it will need at least 12-15 db gain which really seems to rattle things in areas other than my listening position. How much of a gain is safe to add?

I’d start with a Q of 11.5 (1/8-octave) for the 40Hz problem. The idea is to accomplish your needed boost and not change 36Hz. If you see 36Hz boosting, squeeze the Q a little tighter.

To reduce 56Hz, start with a Q of 4.3 (1/3-octave). Since it won’t be cut much, you might need to open up the Q a little.

For 71Hz start with a Q of 11.5 (1/8-octave again).

Regards,
Wayne A. Pflughaupt

 

[Troy_K]

Wayne,

Thanks for the suggestions. I'll try that later when I have more time to EQ.

Troy

 

[Troy_K]

woops..

 

[Sonnie Parker]

Wayne,
Just a line to let you know, I added the above HTF link concerning bandwidth to the BFD Guide. I also mentioned your name, hope you don't mind.
You all look like you've got this stuff hammered down pretty well here. Awful nice of all of you to dedicate the time and effort to help others.
Troy,
Your graphs look good. Glad you like the sound. As they have suggested, if you can get the same sound with less filters, I would go for it.