help me interpret this graph

Hi,

I'm a complete newb and don't understand nothing about the measurement I did, so could someone tell me how bad is my graph?
I used a digital splmeter (ar824) using C weighing, fast and level at 60, xo on sub at 80. Unplugged both speakers leaving only the sub (ACI Titan).
Any help of where my problems are will be greatly appreciated. Thanks in advance

[brucek]

You would benefit from equalization. The room is providing some nice boost at 20Hz, but you would want to reduce that peak down to the level of the 30Hz area. The same can be said for the 40Hz peak. It would be especially noticable. Pretty good response though for a sub with a spec of 20-250Hz ± 3db.

The signal rise at about 18Hz is the signal going into the noise. It can be ignored below that dip. It's not signal.

The area from 45Hz-55Hz may respond to a bit of gain, but sometimes gain filters can be more trouble than they're worth.

Next you need to measure with the mains turned on and check the area around the crossover. The phase control on the sub (and sometimes the distance control on the receiver) can be used to clear trouble there.

My comments assume you have an equalizer... i.e. BFD FBQ.....

fast and level at 60

No, use slow and set your targets and listening position levels to 75dBSPL.

brucek

Thank you for the reply.
About your comments bruce:

My comments assume you have an equalizer... i.e. BFD FBQ.....

Nope, I don't have, just treated my room with 6 absorbers and 17 bass traps and tested the room to see the corrections I have to do (preferably without using an eq)
It's a dedicated room for listening only stereo music.


Back to the graphs
I made a mistake on my earlier post, the test was performed with the splmeter set to slow, and aprox at 75 db adjusting the sub vol.
Now I plugged the speakers (Sonus Faber Eletra Amatro II which are monitor with back ported woofers) and made new tests
First graph: green=only sub phase 180,yellow=sub phase180+spkers
Second graph:green=only sub phase 180,blue=sub phase 0+spkers
Third graph=first+second graph
Fourth graph=sub phase 0+spkers

Please tell me what do you think and problem areas. Thanks again

[Wayne A. Pflughaupt]

Please tell me what do you think and problem areas. Thanks again.

Basically the same thing that brucek told you before: 20 Hz, 40 Hz, and the area between 45-55 Hz.

Regards,
Wayne

[brucek]

Please tell me what do you think and problem areas

OK, a few comments.

Normally it's best to use a vertical axis scale from 45dB to 105dB and a horizontal scale from 15Hz to 200Hz on all your graphs. This provides a consistency for interpretation for everyone. You can set the axis with the Graph Axis Limits button.

The phase control only has an effect when a sub and mains are played together. It has no effect when playing the sub alone.

Looks like the sub has the smoothest transition when its phase is at 180. The sharp dip won't be audible.

Unfortunately room treatments have very little effect on peaks and valleys at subwoofer frequencies <80Hz. For the low sub frequencies equalization is required.

brucek

[Wayne A. Pflughaupt]

Nope, I don't have, just treated my room with 6 absorbers and 17 bass traps and tested the room to see the corrections I have to do (preferably without using an eq)

Again, what brucek said. What bass traps primary do is decrease low frequency decay times. The audible effect is a tightening up of the bass, as the “overhang” is reduced or minimized.

One of our resident bass traps gurus, Ethan Winer, has comparative graphs showing that traps can reduce peaks and dips to some extent in the mid-to-upper bass region, but not drastically. Fully eliminating those will require equalization, especially below 60-80 Hz or so where traps have little if any effect.

If you are dead set against equalizing, fine, but I assure you that 40 Hz peak especially is audibly degrading sound quality. All things considered, though, I’d say your without-EQ curve rates as one of the better ones we’ve seen.

Regards,
Wayne

Thank you all for your helpful responses.
Just some questions,

1-If I buy an eq, how much difference will I notice in sound quality?
2-Why almost everybody is concerned about the low range and not mid/high range? i only see graph for low freq.
3-An eq will help to tame mid/high freq too?
4-Could you recommend a not so expensive eq for treating my problem areas?
Thanks again

[brucek]

1. That's hard for anyone to know. You would have to try it and find out. A peak at low frequencies tends to overwhelm the sound at that frequency. You may be hearing 40Hz and be missing information at 30Hz.

2. The low frequencies from a subwoofer are equalizable. The frequencies above that come from the mains are best dealt with using other methods such as room treatment.

3. No, you don't use an equalizer on the mains. It affects the subwoofer only.

4. BFD 1124 or FBQ 2496. They're available at music stores, or on-line, or e-bay. See the BFD Guide on using the BFD and REW software for software to set it up.

brucek

[Wayne A. Pflughaupt]

A different perspective from someone who uses full-range equalization:

1. Judging from your response graph, I’d say you’d see a noticeable improvement. Your 40 Hz peak is pretty wide, and a full 16 dB in severity. There’s no way eliminating that thing wouldn’t make a big difference. I’ll bet that 8-dB peak at 20 Hz is muddying things up at least some. And the depression between ~45 – 60 Hz is more than 1/3-octave wide. Typically wide depressions like that are going to be much more audible than ultra-narrow ones. Overall, I think you’d get audible improvements with a sub EQ.


2 / 3. I don’t think I’d say no one is concerned about the mid and high range. It’s just that it’s easier to insert a corrective equalizer into the sub signal chain than it is the mains. Also, with the upper frequencies, much of the raggedness you see in a graph looks a lot worse than it sounds.

The function of room treatments for the upper frequencies (i.e., above the subwoofer) is the same with bass traps – i.e., reducing signal decay times (with the upper frequencies, long decay times are typically manifested as reverberation or echo). Above about 500 – 1000 Hz, the point where the room becomes less of a factor in response (as far as modal issues are concerned), room treatments are not going to eliminate response problems your speakers might have. Now, if you have a really live and reverberant room, treatments can appear to “tame” the high frequencies by stopping them from endless bouncing around the room. But if your room is already reasonably “dead,” as it would be if you had wall-to-wall carpet and regular room furnishings, additional treatments are not going to do anything for say, a broad response peak at 4 kHz your speakers might have.

For instance, my speakers are a bit “soft” above 8 kHz, yet my room is a bit “live,” given the wood floors. If anything, additional treatment would make a problem like that even worse. So in at least some cases, equalization for the mains makes sense.

But the problem with outboard equalization for the main channels is that it’s a pricey endeavor. Not only do you have the cost of the equalizer for each channel you want to EQ, you will probably need outboard amplification for those channels as well, as fewer and fewer receivers these days seem to be offering amp-in jacks in addition to the usual pre-out jacks.

Fortunately, many recent-model receivers have fairly flexible equalization built in that can be very useful. In most cases, that should be all you need. I’d say if a few tweaks from the receiver’s EQ can’t take care of what you perceive to be a problem with your speakers, then you should probably upgrade them.


4. As recommended, the Behringer 1124 is a good bang-for-the-buck parametric EQ. If you’re still on the border, you might pick one up on eBay for typically $50-75, and flip it in the unlikely event that you’re not happy with the results.

Regards,
Wayne

So to clear things up cause I'm a little confused, you say measuring mid/high frequencies is of no use?
I now know I have to equalize the low range so I will buy the 1124, the problem now will be the rest of the range.
As I said, I already treated the room and the difference in sound quality is awesome, but I want to know if I can improve the room even more.
My idea was to measure the full range so I knew whether I needed more treatment because I already treated the first reflection points and also placed absorbers to suppress flutters effects and a poly in the rear wall.
Finally, I don't have a receiver, but an integrated amp without proccesor so I can't eq the mains.
Sorry, but still very confused.

[Wayne A. Pflughaupt]

So to clear things up cause I'm a little confused, you say measuring mid/high frequencies is of no use?

Hmm... If that’s what you got out of my last post, I need to enroll in a “effectively getting your point across” class.

Certainly, if you are going to try to EQ your mains, or see if perhaps they could use it, or see any effects you might be getting from treatments, you need to measure the mid/hi frequencies. If you haven’t already done so, you should move up to the calibrated Behringer mic, rather than the SPL meter, which is effectively “done” at about 7000 Hz.

Regards,
Wayne A. Pflughaupt

Just one quick question concerning the first graph, what causes normally the peaks that I see at 20 and 40 hz? Also, I generated a waterfall graph of sub+spkrs but it doesn't look right.

[Wayne A. Pflughaupt]

what causes normally the peaks that I see at 20 and 40 hz?

I’m probably not the best candidate to be answering this one, but bass response is greatly affected by the room, modal issues that have to do with specific frequencies reacting to the room’s physical dimensions. Much of is mathematics, having to do with a combination of the frequency in question and the distance between the sub and various boundaries – the floor, ceiling, left-, right-, and across-the-room walls. Move the sub to a different location, where the distances from the boundaries is different, and you’re going to get a reading that doesn’t resemble this one much at all. Same thing if you take the sub to a different room in your house, especially one that’s drastically different.

For instance, take a look at these two graphs of four subs, taken in two vastly different rooms:

As you can see, in each room all subs exhibit peaks and troughs at the same frequencies, but notice that each room has its own peak and trough frequencies.

I generated a waterfall graph of sub+spkrs but it doesn't look right.

What’s wrong with it? It looks fine to me. The bass traps are giving you extremely low decay times all the way down to 45 Hz or so. Even from 45 Hz to ~25 Hz decay times are excellent. It’s only below 25 Hz that decay times are long, but that’s very typical, as I understand the effects of bass traps.

Regards,
Wayne

What’s wrong with it? It looks fine to me. The bass traps are giving you extremely low decay times all the way down to 45 Hz or so

I think not, considering the 1000 ms time range. Now look when it's set at 400(both cases window ms set at 500):

[JohnM]

The 20 and 40Hz peaks are most likely both modal resonances of your room - is your room about 28 feet long? If you try measuring with the microphone in a few different places you can see how the amplitudes of those resonances varies with location, moving your seating position may give you a much more even response. You can also try measuring with the microphone placed in some alternative locations for your sub, the locations at which the 20Hz resonance is smallest could be good places to move the sub to.

My room is 26,3' x 16' x 9,5'h. I tried moving the sub to several places but there was no change at 20 hz.
I then tried moving the spl meter and after a lot of tries I found a spot where the 20 and 40 hz peaks become flat. It's like 3' farther from the speakers. I seated there (almost middle of the room) and noticed an improvement in bass which was annoying me but then I noticed a loss in soundstage, a little echo and was outside the free reflection zone. Oh well, compromises, I returned to my original position, I will try correcting the peaks with eq.

[JohnM]

If you place the sub halfway along the longest wall you should see a big reduction in the 20Hz peak. The peak is the first length mode of the room, the mid-length position is a null for that mode hence the improvement when you sat near the middle, placing the sub halfway along the length means it will not be able to excite that mode (or any of its multiples).

[clubfoot]

berlasso wrote:

My room is 26,3' x 16' x 9,5'h. I tried moving the sub to several places but there was no change at 20 hz.
I then tried moving the spl meter and after a lot of tries I found a spot where the 20 and 40 hz peaks become flat. It's like 3' farther from the speakers. I seated there (almost middle of the room) and noticed an improvement in bass which was annoying me but then I noticed a loss in soundstage, a little echo and was outside the free reflection zone. Oh well, compromises, I returned to my original position, I will try correcting the peaks with eq.

Once you've found the seating location that gives the smoothest bass response that follows the target, then add the mains and move them to a location that gives the smoothest integration with the sub. Don't forget to adjust distances in your processor, if you can. Then use polarity and phase for the final tweaking.

I moved the sub to where JohnM said and got a better curve, but obviously when I added the mains, the peaks returned to the original position. Silly me< I didnt have that in count before losing all day trying sub's positions.

Clubfoot:the seating postion would be at the center of the room and I've read that acoustically it's not recommended to sit there. I placed my seat and speakers following the 38% rule, and everything sounds wonderfull, except some low freq. notes that I supposed are the ones at 20 and 40 hz.

[JohnM]

If you are feeding your mains through an AV processor try them set as "small" (bass limited) rather than "large" (full range) so that their output below the crossover frequency is reduced, they will then not contribute so much to the 20Hz mode. You could also try placing them 25% along the length and see if that helps the 40Hz peak.