my bass response curve...what to do next (1 Viewer)

[PaulDA]

Here is a plot of my bass response curve. I used the Snapbug Excel worksheet with the Radio Shack corrections. The blue line plots my response from the listening position, vol. set to 0dB (the automatic reference volume of my Integra receiver--my speakers are calibrated to 75dB (also the recommended setting for my receiver as per its manual). I ran the test tones from my laptop to my receiver via analogue stereo phono plug to rca plug cable, with full output from the laptop volume setting.

The sub is a Boston Acoustics PV900. Crossover was set in the receiver at 80hz and the crossover bypass input was used in the sub. Phase was set to 0 degrees and gain was set at about 40% (no numbers on the dial).

Edit:

Can't find way to post the graph so here are the corrected numbers


FREQRAW SPLSPL (comp)
166879.5
1879.587.5
208895.5
229298.5
2597102.0
28101105.0
31.5100103.0
36101103.5
40102104.5
459698.0
509495.5
5691.593.0
6367.569.0
719091.5
8090.592.0
898687.5
1008385.0
1117778.0
12575.576.0
142.565.566.0
16059.559.0

Any suggestions?

 

[JohnSmith]

A good tip would NOT to output tones at 100dB, unless you want to toast your sub's voice coils!

Here's your graph..
http://nathansilly.012webpages.com/graph99.jpg
copy link to IE then press enter.

 

[ChrisWiggles]

It looks terrible, as do most response curves. You probably need to work on placement, and bass trapping in the room, coupled perhaps with an EQ to top things off.

I don't know how much the sub's response is aggravating things either.

 

[PaulDA]

John,
I did not know that such output could damage the sub. Thanks for the tip. I'll dial it down for future tests.

Chris,
The primary problem is weak bass from the instrument (bass guitar/stand up bass) as the dip is at a fairly important frequency range.

 

[oliverLim]

I had the exact same dip at about 61hz +- 2. Moving the sub to different locations only helped a little. The dip was mainly contributed by the listening position and is about 25db down from peak and 15db down from the average. Will lowering the average to 65 +- 5db result in any loss of efficiency?

Oliver

 

[Max F]

I wonder, since you ran the test soo loud (hope you had ear plugs), if you were causing distortion or shaking things in the house which cause the high decible readings at 25-45 hz.

Basically, your running your sub too hot. You need to turn down your sub and bring down the levels that are below 80hz so that they are similar to your front speakers (the values above 80 hz).

 

[PaulDA]

The test tones only last about 8 sec., and they didn't seem that loud (and I'm not loud volume kind of guy, movies are listened to at -20dB from reference and music often lower than that). That said, I'll rerun the tests at a lower output, but I don't think things will change very much, on relative terms. Also, only the sub was on for the readings (even above the xover point) as I'd disconnected the other speakers.

 

[ScottCHI]

although the sub's response alone can be informative, what you're really interested in seeing and tuning is your combined sub+fronts response curve. that is, after all, what you will be using and what you want to be flat, especially through the xover region. so,if you're going to spend the time trying to smooth curves, then smooth the curve that represents the way you'll be using the system, in it's entirety. don't focus too much on the sub's curve, alone. yes, there will still be a problem ~60Hz, but you might as well be tuning the entire system's curve and not just the sub's while you're doing this exercise.

 

[Wayne A. Pflughaupt]

Specifically, the crossover can introduce phase-related problems in response an octave or so above and below the crossover point. Often these problems can be corrected with equalization.

Regards,
Wayne A. Pflughaupt

 

[PaulDA]

My next plan is to re-run the measurements with the mains + sub and see what combined curve results. I wanted a bassline measurement of the sub alone to see what I needed to work with. Am I right, though, in thinking that the lower frequency curve is reasonably flat (to about 50-56 hz or so) and the major problem is the dip? Also, a dip (or null) is a lot harder to fix (if at all) via EQ, as opposed to peaks? Lastly, in theory (and I will run measurements when I get a spare moment), if I set my xover to 60 (below the dip) and let my mains handle more (they're spec'd at F3 51 hz), could that possibly alleviate the dip? Just curious. I'm not opposed to adding EQ or traps or both (though they'll have to be phased in--not made of money ) but I'm also interested in interim solutions as well as understanding how things work.

I must say that despite the current null (perhaps because it is rather narrow--uncorrected readings around it are: 62 hz = 77dB, 63 hz = 67.5dB, 64 hz = 71dB, 65 hz = 79.5dB, 66 hz = 85dB) I don't have any serious subjective issues. I don't find my bass bloated or boomy, nor do I find too thin (though there are some bass guitar passages and occasional kick drums that a bit more laid back than they should be). Mostly I find the blend between the sub and the mains is not as smooth as it could be.

 

[oliverLim]

Wayne,

I checked both Sub only and Sub with Mains using TrueRTA and there is no phase problems due to the signal before and after the crossover at both the sub and mains signal. It drops very smoothly. The only thing I have not tried is to lift my Sub about 1 to 1.5 feet up to see if there is improvement. There is a few locations for my sub placement in my room that does not have this 60hz dip but I don't believe I can convince my wife to accept those locations

So the only way would be to take the dip as a base line and drop all other frequency to that level. Do I lose anything such as headroom by doing that?

Thanks
Oliver

 

[PaulDA]

I re-ran the frequencies (started at 20hz this time) with mains on, at 60, 80, 100 and 120hz xovers. (-20dB from reference to avoid running the sub hot again. Gain settings are still the same as last time). Still haven't figured out how to post graphs. Sorry about the messy numbers layout, it looks alright when I'm typing the message but not after it's posted.

FREQRAW 80hzSPL (comp)RAW SPL 100hzSPL (comp)
16011.5011.5
1808.008.0
206976.568.576.0
227480.57480.5
257984.07984.0
2880.584.58185.0
31.581.584.58285.0
3682.585.082.585.0
408385.58385.5
4578.580.57880.0
507071.571.573.0
5673.575.07374.5
6364.566.06667.5
7173.575.073.575.0
807576.57677.5
897475.575.577.0
10077.579.57880.0
1118081.079.580.5
1257272.57474.5
142.57575.57575.5
1607574.574.574.0



RAW SPL 120hzSPL (comp)
011.5
08.0
6976.5
7480.5
79.584.5
8185.0
8285.0
82.585.0
8385.5
7880.0
7273.5
73.575.0
6667.5
7374.5
7677.5
7677.5
7880.0
7980.0
7474.5
7474.5
7473.5


RAW SPL 60hzSPL (comp)
011.5
08.0
68.576.0
7379.5
7984.0
8084.0
81.584.5
8284.5
8385.5
7981.0
7071.5
70.572.0
5960.5
72.574.0
7475.5
7273.5
7880.0
80.581.5
7272.5
7676.5
74.574.0

 

[Wayne A. Pflughaupt]

Oliver,
I’m not clear on this. Are you using an equalizer? That’s the only way I know of that you could easily “drop all other frequencies.”

Regards,
Wayne A. Pflughaupt

 

[oliverLim]

Wayne,

I have the BFD connected currently. Basically what I am doing now is boosting the 60hz dip by about 4dbs. Problem is my crossover sounds best at 80hz, so from 80hz to 60hz , there is a dip of about 10dbs, and then to 40hz, there is a peak of about 10dbs. From 40hz to 20hz, there is a slight house curve of about 8dbs. But due to the 60hz valley, bass impact such as arrows hitting or punches are just not there. I know it is not a black hole as it responses well to boost. Meaning each 1db boost does increase SPL by 1db.

What I was hoping to do was to have 60hz level to be my new base line and EQ all others down to that level. Thats still about a 15db cut on almost all FR. Would this be good or advisable and do I lose anything. Ultimately it would be the sound that will confirm everything. But I would like to know if I should go ahead to do it.

Oliver

 

[Wayne A. Pflughaupt]

Thanks Oliver, I get the picture now. To answer your question:
The rise between 40 and 20Hz may be aggravating the problem. I’ve found that a house curve sounds best when you level it off at about 32Hz – i.e., keep it flat below that point. You might try it and see.

Regarding the house curve above 32Hz, things sound most natural to me when you get it to where test tones at about 100Hz and 32Hz sound like they’re at the same level. Naturally, when that’s accomplished the SPL meter will show 32Hz measures louder than 100Hz. The goal at that point is to get response to smoothly “ramp up” between the two figures.

Regards,
Wayne A. Pflughaupt

 

[MingL]

Wayne, I helped oliver measure and dial in his sub+bfd a couple of weeks back. (Measured with a ECM8000/mic-pre/truerta, using sweep and pink noise.)

IMO, I measured his listening spot and found out that there was a suck-out of about 60hz-75hz; 15dB down from average. But this suck out largely disappeared at the adjacent seats and 3ft to the front and back of the listening spot.

Because of this phenonmenon, I decided not to correct for this wideband 15dB dip using EQ as it will screw up the response for the other seats, as well as prevent the sub from overworking.


Concluding that it was a placement problem, we placed his PCU at his listening position and measured the response at few potential sites for placing the sub around the room.I think this is a good method for objectively finding a good spot for sub placement. And just as expected, the dip was measured and the response was the near identical.

I felt that this case was purely an issue of sub location. It was just unfortunate that his particular choice of sub position produced a dip. I advised oliver that the sub needs to be repositioned drastically elsewhere in the room, in those places that measured & sounded good while the sub was in the listening position.

What are your opinions?

 

[Wayne A. Pflughaupt]

Hmm. This sounds like one of those “shoebox” rooms with all-parallel walls...

If the problem is appearing at only one seat, it makes more sense to me to move the seat rather than the sub. (I’m going to hazard a guess that that seat is currently dead-center in the room – right?) You noted the problem went away 3 feet in front of and behind the seat. Any chance the seat could just be moved there?

Here’s what concerns me: If you move the sub to where it measures (and sounds) right at that one seat, there’s a pretty good chance in this case that it will get poor results everywhere else – i.e., just the opposite of what you have now.

Regards,
Wayne A. Pflughaupt

 

[MingL]

Most 95% of houses in singapore have concrete all-parallel walls in the rooms. Concrete, parallel and rectangular rarely make a good combo where bass is concerned. Since concrete walls do not absorb bass energy and largely reflects them back, the interaction of the sound waves will construtively and destructively interact.

And this was very evident in that there was a less serious 25-30hz dip in tandem with the 60hz dip.

In oliver's case, I also suggested him to move his seat 1m backwards or 1m to the sides. But that meant the plasma screen was now 1m too much to the right or the walkway behind the sofa was blocked.

So, its either move the sofa or move the sub.

 

[oliverLim]

Yupe. MingL has said what I needed to say! Just to add that I am not sitting near the center of the room. My room is 15' x 17' x 8'. Almost square all right. I seat about 5 feet from the right wall and about 4 feet from the back wall. Actually the near the back left of the room, it opens up slightly to the left with a square of about 4 feet by 6 feet. So a odd shape room at the back.

I just got my USB sound card with a mic preamp and the ECM8000. so I will be able to measure my room response again. When MingL was helping me with the measuresments, we were not able to save all the measurements after playing around with placements for nearly 7 hours! The system just decided to crash when he tried saving it.

I basicaly have 3 choices. Move the sub to a location that seems to have this 60hz not affecting it. thats where the corner of the small odd shape area I was talking about. Or to lift the sub upwards to see if it helps. Or to run my REL Q100E at the same time as I believe it does not have a dip in 60hz. I would hope to have some charts out over this weekend but was hoping for further advise on what I should try out first. MingL has already kindly given me some further suggetions

Sounds like a back breaking weekend!

oliver

 

[Wayne A. Pflughaupt]

Sounds like you have a tough situation there, Oliver. Keep us posted on your progress!

Regards,
Wayne A. Pflughaupt