(many)room acoustic questions

[bpape]

It might.

First of all, there is always some overlap above and below the xover point so you might be getting some additive effects.

Secondly, the peak at 120Hz might be a harmonic of a lower peak that can be treated in the subwoofer range. For instance, if you have a peak at 40Hz, 120 would be the 2nd harmonic (40, 80, 120, 160, etc.) or it could be the first harmonic of a peak at 60Hz.

 

[Sonnie Parker]

The BFD will handle pretty much any frequency your sub can produce... it doesn't care what your crossover point is. However, in your case it depends on what you are measuring. If you are measuring your sub only and you have a peak at 120hz, the BFD will have no problems eliminating that peak. If you are measuring your sub + mains then the peak could be coming from your mains, which would be impossible for the BFD to tame, since your mains are not in the BFD loop.

 

[Bevan]

appreciate the help both of you.

i ran the sub by itself and got 'peak' of minus 4db bellow flat at 116hz.
then i ran the mains by themselves and got plus 6db at 116hz.

seems like they are both contributing. that the sub is outputting 66db of a 70db note should make it easy to chop i'm guessing.

but i'm surprised that i dont have a fundamental at 56(1/2 of 116), or 39hz(1/3 of 116) like you suggested.

i dont have a graph to show you unfortuately, but appart from the peak at 116 i have

102hz = -9
94hz = +4
82hz = +8
75hz = +10
60hz = +8
56hz = +4
36hz = -7

from 56-94 the curve doesnt drop bellow +4

i got my bfd this week but still need to get an rca to 1/4" adaptor before i can set it up.

i spent all day re-arranging my living room to find a final setup that is livable before i get to work with the bfd.

i have chosen to sit at 1/2 the lenth of the room, in a massive node inorder to help my subs amp out as much as possible. figure i can flatten the 56-94hz hump, and the 116hz peak, maybe boost the 102hz dip by 5db(what you think???) and live with the dip at 36hz.

again, thanks for any imput. i'll post my results here when i get my bfd and room wizzard running.

cheers

b

 

[Bevan]

brian, sonny, only now just found out about the private message thing, sorry for ignoring you.

see you at the bfd forum one of these days.(wishing i wasnt a mac user at the moment as i cant get rrom wizzard up and running, need to 'borrow' girlfriends laptop...)

 

[bpape]

IMO, it's quite likely that the main speaker placement is the culprit for the 116Hz hump. Distance from the side and front wall can be adjusted to compensate for this. Also, you can play with treating directly behind and beside them with some absorbtion that will help tame this boundary effect.

Putting your seating right in the middle of the room to help the sub is backward to me. Put your seat in a position where it is in a minimum of peaks and nulls and then move the sub around to smooth the response below the xover. Then add in the mains and adjust them position wise to complement the overlap and general FR while still maintaining good imaging.

 

[Sonnie Parker]

Bevan...

Download the Excel Workbook from the BFD GUIDE (link in sig) and input your sub only measurements into the workbook. Then we will have a graph. If you don't have Excel then post your measurements here.

Make sure you use 1/6 octave measurements using the Radio Shack SPL Meter so we have correction values, unless you have a more accurate mic with other correction values. As you input your raw measurements into the Excel Workbook, it will automatically calculate the corrections and give you a graph of the corrected response for the RS meter.

Bpape has a good point about absorbtion panels, unless you have an auto-eq feature built into your pre-pro or receiver. Make sure you get absorbtion panels that are specifically for the frequencies you wish to absorb. Most of the panels available for wall placement near your speaker first and second order reflections will absorb down to about 100-125hz... but usually only a few db. It may help some with your mains.

The BFD will deal nicely with any problems below your crossover point. Typically it's best to place you sub as far into a front corner as possible. This usually will be the place where you'll have less dips to worry about, although it may create some peaks, this is where the BFD shines, it will eliminate just about any peak.

 

[Bevan]

thanks for that.

got my sub in the corner sonnie. more output here, and it didnt seem any less flat to me. musical accuracy is my main concern, not max spl's, and even though i was sure before i got the sub that i wouldnt put it in the corner that is where it seemed to work best.


bpape, i had my listning possition at 1/3 of the way into the room before yesterday. gave the flattest response before i had the sub(fewest peaks anyway)

but with the sub in place i have very limited output in the 40's and 30's. and i'm a bit scared to turn up my subs gain more than i have to.

now that i sit in the middle of the room these low frequencies are only -4db bellow flat. the curve is now not as flat as before admittedly, but the main problem now is the 50-90hz peak which i should be able to treat with the bfd?

as i understand it sitting in the middle of the room is meant to increase general spl levels as well increase the peaks. but can it also accentualte the dip?? if not, i'm not following why it would be a bad idea to sit here(provided one had eq)

cheers

bevan

 

[Bevan]

woahoo! got the bdf up and running!

easier than i had expected.

put one filter in, a -6db cut centred at 75hz. bandwidth = 20. pretty wide but maybe not too wide as my himp spanned over 30h?

big difference. i like, but think i can do better. definitely can do flatter but might end up with a house curve of some sort i expect.

one things i didnt expect was that i was down a few db through the 30's and 40's. can a cut 75 effect further down the frequency range? or maybe my bandwith was too wide?

and when i then used the AVIA disk to ballance mains vs sub, i had to crank the sub up a few db. i would have expected this as the 75hz hump was obviously bringing up the average, but this is not too different from just boosting the 30-40hz range to start with, which i didnt want to do.

and even stranger, the dip i had around 100hz, i tried to boost it 8db(just to see), and it seemed to make the dip even more severe obviously a room node that doesnt want to be messed with.

no more time to fiddle today, more tomorrow

cheers

b

 

[Sonnie Parker]

Yes... a bandwidth of 20 can definitely effect a wide range of frequencies. Very seldom do we ever use a bw that wide. Try starting small with 2 or 4 and working up if needed. 20/60 is 1/3 octave wide. The bw settings on the BFD are setup to effect response above and below the actual bw range.

Bevan... what measuring instrument (mic) are you using?

 

[bpape]

Corner mounting will give you the most output from the sub but also the least smooth response curve due to exciting all of the room modes in all 3 axis since you're at the end of all 3 in the floor/wall/wall corner.

Turning up the sub is not a bad thing as long as you stay within it's parameters.

My point with playing with the positions of the mains is that as you approach the front and side walls, you'll DELIBERATELY introduce peaks and nulls into the response. The trick is to find the place where you're introducing peaks to compensate for your nulls and visa-versa.

Don't be afraid to use a lot of different filters on the BFD - that's what makes it so handy. As Sonnny said, find the narrow peaks and tame them first. You'll be surprised at what other things will change with those. Then work to the less major ones that come in groups so you can work with them together. Then you can use the level overall to balance things out.

Before you start, make sure that the phase is set correctly on your sub or you'll be fighting battles that don't need to be fought.

 

[Bevan]

sonnie, i using the radioshack analogue meter. am roughly allowing for the errors in my head, adding about two to anything in the 40's and 3 to my 32hz peak. i'm taking so many measurements that i dont have time to graph anything. would like to see a visual, but once youve watched the needle for the 100oth time you begin to for a good picture in your head of what the curve looks like.

brian, wish you hadnt suggested moving my mains again, not after i promised my girlfriend that if she let me get the bfd she could put the speakers anywhere she wanted had them 6' out but now they are 3' from the front wall and 5'from the sides. not too bad for imaging, and i like the fact that with my new arrangement my tv has got about 50% bigger for no extra cost. so they might have to stay as is.

got to within +- 4db bellow 100hz using 6 filters. sounding very clean and tight. very happy.

was wondering if there are any rule of thumbs when guessing bandwidth and centring. if one had a uniform hump from 50-80, with its peak at 65, where would one centre the filter and what bandwidth would you guess at. i'm thinking that the filter should not be at the halfway point(65) but before it?

one last question, i read that making cuts does not neccessarily increase headroom. why not? wouldnt a cut give you the same extra headroom that a boot of the same amount would take away? and wouldnt a single cut at 40hz with a single boost at 50hz(of the same db and bw) leave you better off in terms of dynamic headroom?

thanks again guys

b.

 

[Sonnie Parker]

sonnie, i using the radioshack analogue meter. am roughly allowing for the errors in my head, adding about two to anything in the 40's and 3 to my 32hz peak. i'm taking so many measurements that i dont have time to graph anything. would like to see a visual, but once youve watched the needle for the 100oth time you begin to for a good picture in your head of what the curve looks like.

Bevan... I hate to spoil your party, but you are not getting anywhere allowing for errors in your head and not using a graph.

Here are the correction values for the RS Meter:

Now let me show you the difference in a non-corrected and corrected response:

Now here's a look at how a chart can help you determine what to adjust and how it was adjusted to obtain a very nice house curve:

If you want to get your sub right, you might want to reconsider roughing it. Even the best technical wizards I know have not been able to achieve an acceptable response curve without some sort of graph.

 

[Bevan]

i was thinking that if i ever get a bigger sub i'll get a microphone or something more accurate than the radioshack meter, but i thought that since my sub only goes to 30hz i could wing it as the errors are not more than 4db and i'me only aiming for flat within +-4db. for the moment anyway- now i just want to sit back and enjoy music for a few days. then i'll get back to my tweaking with the aim of getting something like your curve.

will try to get another flattish curve with my speaker plugs out(see my other thread), and will try some house curves. your house curve of about 15db seems a lot to my eye though as my flattish curve sounds pretty good to my ear. but i'm looking forward to experimenting.

thanks for the help

b

 

[Bevan]

say, do you know what kind of sweep tone the AVIA dvd used i.e 1/6th octave, 1/12 octave, linears etc?

its a bit too fine, and quick, for me to write down for the purpose of graphing. but this degree of resolution would be nice if i could get room wizzard to do the graphing for me...

 

[bpape]

Check out Ethan Winer's site. You can download test tones in 1Hz increments. SnapBug also has some for download I believe.

 

[Sonnie Parker]

Hey Bevan... did you ever get things going with your sub again?

Sweeps are continuous... usually running from 10hz to 200hz or similar.

Btw bpape... SnapBug is the BFD GUIDE.

1hz sinewaves and sweeps can be downloaded from http://sinewave.bfdguide.ws

 

[edwelly]

Guys - I don't have a BFD but I gotta say there was some GREAT info in his thread. I am now adding a BFD to my list of things to get this spring.

Also, Sonnie: GREAT looking home theater room you got there!!!