BFD plot, after 2 weekends I'm finally getting somewhere. [Archive] - Audio & Video Forums

PDA

View Full Version : BFD plot, after 2 weekends I'm finally getting somewhere.



kexodusc
06-19-2005, 02:05 PM
I've had this thing since mid-March, but last weekend (June 11) was the first time I could hook it up and fiddle around with it.

I have to say, this thing is worth every penny, I no longer have a big room induced suckout and bottom end boom, which has allowed me to turn the sub up without the annoying bass peaks. I had a difference of over 24 dB between my biggest peak and dip. Let's say a response +/- 12 dB...pretty bad. Now it's about +/- 5 dB. The graph doesn't include all actual measurements, there' s a 3 points in particular where my response still varies by 5 dB but I just never bothered to modify the spreadsheet.

On that note, my in room results are pretty odd....most on-line node/mode calculator's I've seen predict different behavior. I attribute this to the 2 doors, 2 windows, and half-foot countersunk walls along 1/2 the room...maybe sub position as well...anyway.

The BFD is as frustrating to learn at first as everyone mentions. But half an hour or so everynight and a few afternoons on weekends goes a long way.

My reference level was 64.5 dB, though I had originally done these at 80 and 70 dB. As I'd hoped, the results were pretty much linear except the dips were a bit lower relative reference at higher SPL. I chose the lower volume because there were too many buzzes and rattles adding a dB or two at some frequencies.

I would like to lower this curve overall by 1 or 2 dB, but it's not as simple as the knob on my sub amp or the receiver settings for some reason.

I'm definitely going to aim for the "house-curve, boosting frequencies below 40 Hz or by about 6 dB per octave, my hearing definitely drops off a bit down that low. I suspect I'll have to start all over again - it's unreal how boosting or cutting low frequencies affects the higher ones. I'll basically have to increase the gain on my amp and cut the frequencies above 40 Hz I think.

I'm also going to have to try all this again with the Radio Shack SPL meter if I ever get it back from my brother.. My Galaxy Audio SPL meter is much more accurate above 40 Hz, I'm not so sure about below that point though, and it'd be nice to double check.

All in all, this could be the best purchase in my home theater system to date. You really can't explain how much bass you are missing (or adding) once you tighten up the in-room response. I've also increased my crossover from 60 Hz back up to 80Hz because, well, it sounds alot better, the 15" sub doesn't betray it's location too much, and more consistent response is well worth the trade-off. Surprisingly this benefits multi-channel audio more than movies (probably because of bass management issues).

At $100 if you're still on the fence, go ahead and get one of these. It's a damn cheap upgrade that dramatically improves 2 or 3 octaves of info, and lets you squeeze the most out of your sub.

The blue curve is the measured in room response without the parametric eq, the pink curve is with parametric eq, the yellow is the reference level.

hmmmm
06-19-2005, 03:48 PM
Well, since there is no such thing as a stupid question...do you have a software program that works with the BFD that prints out the graph? I still haven't purchased one because it does seem like a pain to set up.
Also, have you noticed when you play DVD-A and SACD, your system doesn't play as loud? Since we have about the same setup I was curious to know.

kexodusc
06-19-2005, 03:54 PM
There is software you can download, however, it really isn't that bad. If you can build a speaker and sub, you can figure this thing out. Not to mention there's 47 years of experience with this unit lurking here on the forums, that's all the help you'd ever need.

If you have a Radio Shack SPL meter, download some test tones and check out the response of your sub between 20 and 90 Hz. You'll be surprised what room acoustics do. If your response curve is as bad as mine was (or heaven forbid, worse), you owe it to yourself.

I do lose a couple of dB's from my cheapo Toshiba DVD-A/SACD player. That's normal though, probably just a different output sensitivity (ie: voltage). My Arcam CD player is also a bit louder than most of my other units. So is my X-Box.

Don't worry about turning up your receiver any, the numbers in the display are arbitrary and don't necessarily mean that the amplifiers are at max.

Geoffcin
06-19-2005, 03:57 PM
You had a 20dB slope over ~30 hz?! That is one heck of a mis-match, no wonder you were disapointed with dual subs. Mongo like BFD, Mongo want's BFD now!

kexodusc
06-19-2005, 04:16 PM
Well, that 20 dB is kinda misleading...there's a couple small narrow peaks not accounted for in the graph on the way down that slope, but yeah, an aweful lot of info lost. .
The old PW-2200 didn't go near as low despite Paradigm's specs, but I'll give that sloppy ported bastard credit, it, because it rolled off the overall slopiness wasn't quite as dramatic. I think my room has a few troublesome nodes that are sucking bass out....and I get that double peaking in the low 30's and mid 20's.
My ceiling isn't flat either, maybe this is messing things up...

Sounds awesome now!!! I'm hearing all the notes. I think the house-slope will be even better, wonder if anyone around's tried that?

hmmmm
06-19-2005, 05:38 PM
There is software you can download, however, it really isn't that bad. If you can build a speaker and sub, you can figure this thing out. Not to mention there's 47 years of experience with this unit lurking here on the forums, that's all the help you'd ever need.

If you have a Radio Shack SPL meter, download some test tones and check out the response of your sub between 20 and 90 Hz. You'll be surprised what room acoustics do. If your response curve is as bad as mine was (or heaven forbid, worse), you owe it to yourself.

I do lose a couple of dB's from my cheapo Toshiba DVD-A/SACD player. That's normal though, probably just a different output sensitivity (ie: voltage). My Arcam CD player is also a bit louder than most of my other units. So is my X-Box.

Don't worry about turning up your receiver any, the numbers in the display are arbitrary and don't necessarily mean that the amplifiers are at max.

I have the radio shack spl meter. I'll test it out. I do have a pretty normal "rectangle" room with a couple of entrances that can be closed off with doors and one hallway that has an extremely thick curtain that I pull closed to seal the room as good as I can. I have a few bass traps that made a big difference also. I'll let you know what happens when I get a chance to test the fifteen incher:)

PAT.P
06-19-2005, 06:32 PM
Kexodusc Can this BFD work when Im using 2 sub? What about the cross over setting on receiver?Pat.P

kexodusc
06-20-2005, 03:36 AM
Kexodusc Can this BFD work when Im using 2 sub? What about the cross over setting on receiver?Pat.P
Yes, the BFD lets you program different parametric eq filters into each sub separately if need be. Up to 12 each (which is more than you'll ever need). I use 8 filters on my sub, I think I could actually cut it down to 6 with a bit more experience

I'm going to add a second sub eventually, I imagine I'll have to start all over again.

The crossover on the receiver really just affects when the bass is rolled-off the sub and transferred to the speakers. I had mine at 60Hz, but now after customizing the response curve, it sounds much better all the way to 80 Hz in my setup. I get a flatter response over a wider range of frequencies this way. Your setup might produce different results. Many people cross their subs over even higher after incorporating the BFD. I guess experimentation is required. I'm happy with 80 Hz for now.

agtpunx40
06-20-2005, 06:30 AM
I also recommend the bdf. While my original response wasn't quite as bad, it was pretty bad, and is alot better now. There is one very narrow huge dip about -8 or -10, and one that's about -5 that I can't seem to get rid of no matter where I move the sub. Well, I have a few more places to check, but I haven't had the time lately, and while they are both very large dips, they are also very narrow, and aren't as noticable as you might think. One is at 45hz and the other's at 56hz. Just to give you an idea, at 40, 50 and 60 hz, the response is all back to normal, and is actually pretty straight other than these 2 dips. I'm going to start working on my house curve this weekend, and I'm going to try giving each of these a narrow boost of at least a few dbs. Did you add dbs anywhere, or mostly just drop some and move the sub around? In my last place, I found that worked much much better (as most people on these boards reccommend), although the dips were much wider.

hmmmm
06-20-2005, 06:35 AM
Where is a good place to download the test tones? I don't recall my Avia disk having the 20-90 test tones in increments. A friend of mine has it at the moment.

kexodusc
06-20-2005, 07:05 AM
I also recommend the bdf. While my original response wasn't quite as bad, it was pretty bad, and is alot better now. There is one very narrow huge dip about -8 or -10, and one that's about -5 that I can't seem to get rid of no matter where I move the sub. Well, I have a few more places to check, but I haven't had the time lately, and while they are both very large dips, they are also very narrow, and aren't as noticable as you might think. One is at 45hz and the other's at 56hz. Just to give you an idea, at 40, 50 and 60 hz, the response is all back to normal, and is actually pretty straight other than these 2 dips. I'm going to start working on my house curve this weekend, and I'm going to try giving each of these a narrow boost of at least a few dbs. Did you add dbs anywhere, or mostly just drop some and move the sub around? In my last place, I found that worked much much better (as most people on these boards reccommend), although the dips were much wider.

I have a 2 boosts and 6 cuts. The 1 boost was narrow and was necessary only because of the cuts on either side of it. The other boost is in that big dip you see, and contrary to the advice I also receive it did help. It was the last thing I tried and I think I boosted +4 dB between 50 and 60 Hz somewhere. You'll notice that my pink curve still shares a lot of movement in the same general areas as the blue curve with the biggest exception being around 50 Hz, I think that's where the one troublesome boost was required.

I ended up increasing the sub level by several dB's from my pre-BFD setting. Once I started cutting frequencies, I noticed my average response was well below my speaker's reference level of 65 dB or so. I think my sub's gain knob is around the 12 o'clock position now instead of the 10 o'clock it was at. I also boosted the receiver gain to the sub when following the BFD comprehensive setup guide.

Hmmm, this site has everything you need to know about the BFD and links to test-tones.

http://www.snapbug.ws/bfd.htm

agtpunx40
06-20-2005, 07:46 AM
yea, that site on snapbug is really helpful. I guess I'll be trying some boosts this weekend, and see how it works. I also notices that I was turning up the sub volume after eqing. The reason is just that you no longer have those huge humps in the response which are really loud compared to everything else. This was even worse in my case because I had a big hump at the top and at the bottom. Especially when watching tv shows on regular cable, this was simply awful in some cases. Deep voices had their low end raised alot, there was little bass in the middle, and very deep bass was alot louder. Even with those big dips I was describing, which are noticable but not as bad as without the bfd, the response is much more even and natural in most circumstances.

kexodusc
06-20-2005, 07:57 AM
That's just it, you tend to compromise your listening by either setting the sub's output to the highest peak, in which case all other frequencies are artificially diminished, or a somewhere lower where the swings in SPL are volatile. This can give the impression of boominess and poor response.

Personally, I think everyone should have one of these...I'm finally seeing more manufacturers offer at least one or two filters with their subs now.

agtpunx40
06-20-2005, 08:19 AM
couldn't agree more. I can't believe people spend as much as they do without the $100 or so to get one of these, and the relatively small amount of time to set it up. Almost as much as I can't believe my experience this weekend buying a new tv. Neither best buy, nor circuit city sell any setup disks for tvs. I can't imagine someone spending thousands on some ultra expensive plasma, and then not bothering to spend the small amount of time and money to set it up properly.

kexodusc
06-20-2005, 11:11 AM
One thing I'll add about the BFD...it's even uglier than you'd expect. Those blinking LED's might serve a purpose, but they're definitely not made with home decor in mind...The rack mount brackets are detachable, but there's no substitute brackets to protect the control panel against dust etc. And I suspect many people will require a 3-prong-to-2-prong adapter for the power cable to prevent ground loop. Still, you can't complain for $100.

Woochifer
06-20-2005, 02:46 PM
Looking at that before and after curve, it seems that you've done a fine job with your first set of parametric filters! Seeing that huge contrast right in the middle of your original response curve would single-handedly ruin whatever aspirations to accurate bass you might have had.

Your testimonial just adds to the increasing list of "best $100 I've spent on my system" praises for that Behringer Feedback Destroyer. Even without additional tweaking, just getting rid of those two large peaks and that big dip in your bass response alone would be enough to vastly improve the quality of your bass.

One thing I would add. The BFD adds 1-2 ms of delay onto the signal, so you might want to decrease the distance of your subwoofer on the setup menu from the receiver accordingly. The flatter and more even frequency response has already worked wonders on my bass. Since my older receiver locks everything in at Yamaha's old 90 Hz crossover point, and does not allow for any delay compensation on the subwoofer, I think those are really the only adjustments remaining that I need to make.

kexodusc
06-20-2005, 03:08 PM
Wooch - I had a lot of help from a few people by e-mail with way more experience than me...but I still have at least 14 hours of tinkering on this thing. As I mentioned, there's still a few problematic areas that just don't show up on the plot because of the 1/6 octave intervals in data entry, so this curve looks better than it really is (I should be marketing speakers) but yeah, the difference is incredible.

My curve is really more of a semi-eliptical shape. I'd like to try the so-called house-curve to compensate for my hearing (and add some fun-factor to movies).
My room is over 500 square feet in an odd sort of shape, but rectangular for the most part. The sloped ceiling at one end probably adds some trouble as well, the living room at my last place didn't produce nearly as problematic a response curve.
As for delay, you're about the 14th person that's mentioned that to me. The BFD has a huge on-line fan-club, another benefit that really helps with the learning curve.

Wooch, couldn't you compensate for the lack of delay on your sub by reducing delay on the other 5 speakers? Isn't it all relative? Just a shot.

Woochifer
06-20-2005, 04:05 PM
Wooch, couldn't you compensate for the lack of delay on your sub by reducing delay on the other 5 speakers? Isn't it all relative? Just a shot.

Can't do it because my receiver does not alter the timing for the main channels. Only the center and surround speakers have adjustable delay. The subwoofer is unfortunately locked in with the mains.

The ability to tweak with the settings and create a house curve is one of the fun aspects of having up to 12 parametric filters to play around with. I tweak with my system every few months and experiment with other settings to see what other bass response curves float my boat.

kexodusc
06-20-2005, 04:16 PM
Can't do it because my receiver does not alter the timing for the main channels. Only the center and surround speakers have adjustable delay. The subwoofer is unfortunately locked in with the mains.

The ability to tweak with the settings and create a house curve is one of the fun aspects of having up to 12 parametric filters to play around with. I tweak with my system every few months and experiment with other settings to see what other bass response curves float my boat.

Your receiver sounds like my old RX-V795a all over again. Though I do miss the knob controls that unit had, much more personal than digital settings.

Any recommendations for attacking the house curve? I'm thinking that increasing the sub-amp's gain and cutting frequencies with filters at 6 db/octave or so is the safest bet, but I have enjoyed some success with boosting some frequencies. It be nice if I could just undo some of the cuts below 40 Hz and let it go. I'll use program 5 for that.

Woochifer
06-20-2005, 04:34 PM
Your receiver sounds like my old RX-V795a all over again. Though I do miss the knob controls that unit had, much more personal than digital settings.

Mine's the RX-V800, basically the successor to the 795, except with Yamaha's then new digital Top-ART design, which put a lot more setup options into the on-screen setup menus.


Any recommendations for attacking the house curve? I'm thinking that increasing the sub-amp's gain and cutting frequencies with filters at 6 db/octave or so is the safest bet, but I have enjoyed some success with boosting some frequencies. It be nice if I could just undo some of the cuts below 40 Hz and let it go. I'll use program 5 for that.

Generally, the advice is to use the BFD for attenuation only. Especially with low level signals, the BFD's known for adding a layer of noise and coloration to the signal. Might want to play around with the gain and just use the filters for cutting the signals, and create your curve that way. Isn't this fun?

I guess what I like the most about using the BFD is how well it illustrates the link between what the technical measurements say you're supposed to hear, and what you actually hear.

kexodusc
06-21-2005, 03:55 AM
I guess what I like the most about using the BFD is how well it illustrates the link between what the technical measurements say you're supposed to hear, and what you actually hear.

Ain't that the truth. Too bad all our living rooms weren't anechoic chambers and our sweet spots 2 meters from the speakers.
I spent 3 hours yesterday spinning Pink Floyd's DSOTM SACD, Beethoven's 7th SACD (Kleiber), Porcupine Tree "Up The Downstair", Alisson Krauss & Union Station "New Favorite" SACD (wicked bluegrass), a few Widespread Panic DVD-A's, and Crystal Method's "Legion of Boom" DVD-A. It's like a buying a brand new subwoofer. I can't wait to build the 2nd one.

Crystal Method should be a compulsory listen for everyone with a subwoofer. I'm not sure if I prefer the DVD-A track or the DTS ES track. There's definitely an added spacial element in the DTS ES version.

I'll probably fiddle around some more trying to flatten out the remaining +/- 5 dB peaks and dips a bit more.

kingdaddykeith
06-22-2005, 03:29 PM
They do work great, been using one for years. However, do be careful with narrow Q boosting, it will cause severe ringing and that's worse then a deep narrow dip sound wise.

mixadude
06-22-2005, 11:53 PM
Some interesting stuff here.

I've been using much better parametric EQs as a professional for decades. The best thing I can suggest is that less can be more. Try to follow the inverse of your (general) room response without going hog wild. On a sub, 4 filters is what I'd consider max or 2 might be better! Three wider filters is most likely gonna give better overall long term results than 6-8 narrow ones. Keep in mind what filters do to phase coherancy!

Also cut only is prolly the way to go. Boost can cause some gnarly problems.

I'm surprised Smaart (http://www.siasoft.com/) hasn't made it into the "high end" HiFi/HT arena. Real time Windoze computer based FFT. Check it out, I been using it since ver 1.0 ;)

kexodusc
06-23-2005, 03:52 AM
Some interesting stuff here.

I've been using much better parametric EQs as a professional for decades. The best thing I can suggest is that less can be more. Try to follow the inverse of your (general) room response without going hog wild. On a sub, 4 filters is what I'd consider max or 2 might be better! Three wider filters is most likely gonna give better overall long term results than 6-8 narrow ones. Keep in mind what filters do to phase coherancy!

Also cut only is prolly the way to go. Boost can cause some gnarly problems.

I'm surprised Smaart (http://www.siasoft.com/) hasn't made it into the "high end" HiFi/HT arena. Real time Windoze computer based FFT. Check it out, I been using it since ver 1.0 ;)

That's some good advice, and is pretty much consistent with what others are telling me.
I'm currently using 8 filters. Two of these were very narrow cuts attempting to eliminate isolated +3 peaks. I can kill those easily enough as they'd lie in my +/-3 dB tolerance range. You're suggesting I could be adding undesireables into the equation that would overshadow the benefits of the flatter response. Guess I'll have to try that, there's 9 more memory slots available.

After checking things out again, I'm pretty confident I can cut that down to 4 easy enough. Incidentally, I have not heard any negatives out of the slight boost I have applied, I'm wondering if the fact our hearing is so insensitive to the sub 80 Hz frequencies that I just can't tell? What exactly would "ringing" sound like below 60 Hz?

I've seen others acheive even better results with 2 or 3 well implemented filters. As I get more experienced with this unit I hope to minimize the processing while maintaining or improving the results...in the meantime, the improvement is literally 100%. I'm hearing things I never could before.

kingdaddykeith
06-23-2005, 05:48 AM
Ringing is when the signal stops and the speaker still plays a harmonic for a few ms. It's easy to hear, play a test tone un filtered and listen to the tonal characteristic, then add a narrow Q filter with about 10db of boost, then listen to that tone again, it will not sound like a solid tone. You should notice that it will have some sustain and harmonics that make the tone sound like…..Ringing. In addition, make sure you are using very high resolution in you test tone measurements, I play tones with 1Hz spacing, you could do it every 2 Hz, but I wouldn’t recommend any less rez then this if you really want to know what’s going on in the room.

kexodusc
06-23-2005, 05:58 AM
Ringing is when the signal stops and the speaker still plays a harmonic for a few ms. It's easy to hear, play a test tone un filtered and listen to the tonal characteristic, then add a narrow Q filter with about 10db of boost, then listen to that tone again, it will not sound like a solid tone. You should notice that it will have some sustain and harmonics that make the tone sound like…..Ringing. In addition, make sure you are using very high resolution in you test tone measurements, I play tones with 1Hz spacing, you could do it every 2 Hz, but I wouldn’t recommend any less rez then this if you really want to know what’s going on in the room.

Thanks for the description. Am I correct in assuming then that the ringing would occur at harmonics only above (and in multiples) of the fundamental? And, narrow Q cuts won't result in any ringing? Or would they also apply cut effects and possibly diminish at harmonics?

I play sine waves test-tones in intervals of 1Hz. Takes forever, but you're right, it's the only way to know what's going on. I'm also at the mercy of my 3 SPL meters accuracy. Correction values for the standard Rat Shack meters are available only at 1/6 octave increments. My Galaxy Audio meter is highly praised, we use it at our shows when setting up, but I'm not sure of its accuracy below 50 Hz or so.

kingdaddykeith
06-23-2005, 06:27 AM
Thanks for the description. Am I correct in assuming then that the ringing would occur at harmonics only above (and in multiples) of the fundamental? And, narrow Q cuts won't result in any ringing? Or would they also apply cut effects and possibly diminish at harmonics?.

They are harmonics but I'm not sure about the details, I only know there is nothing you can do to avoid this side effect, I have a narrow 12db dip at 53 Hz and could get it moved out of the pass-band with a lot of EQ, but it sounded worse then leaving the dip there because of the ringing. Cutting frequencies don’t have this side effect, only boosting. The only way to properly fix a dip is broad-band absorption, IOW, Room Treatment.


I play sine waves test-tones in intervals of 1Hz. Takes forever, but you're right, it's the only way to know what's going on. I'm also at the mercy of my 3 SPL meters accuracy. Correction values for the standard Rat Shack meters are available only at 1/6 octave increments. My Galaxy Audio meter is highly praised, we use it at our shows when setting up, but I'm not sure of its accuracy below 50 Hz or so.

Eric Wallin has posted a method of modifying the RS SPL meter that makes it very accurate from 20-20K, I did it and it works great.

http://mysite.verizon.net/tammie_eric/audio/rsmeter/33-2055/33-2055.html

Andrew Pratt
09-28-2005, 07:00 AM
I thought I'd let you know that Sonnie has set up a forum to help us BFD owners. Its located here http://www.nextlevelav.com/forum/forum.asp?FORUM_ID=39

Andrew