Which product will do the magic?

Hi folks,
I have a problem with my SVS PB12 Plus.
The problem comes form the sub placement. I know it for sure but I can not move the sub.
From the graph atached you can see I have big deeps and peaks that can not be cured with the onboard PEQ and Adussey MultEQ.
Which parametric equalizer should I have to flatten the sub curve?

I have read all the BFD informative section and I noticed that boosting is much worse then decreasing signal. Can I use a Beringher to decrease the peaks and flatten the curve and then increase the sub level on the sub itself?
As the deeps and peaks are very high in dB would a Beringher product be able to compensate such big differences?

Thanks for your help.

ciao

[brucek]

Can I use a Beringher to decrease the peaks and flatten the curve and then increase the sub level on the sub itself?

It appears to be mostly peaks.

The worst peak is at 33Hz and then another at 71hz. A BFD could certainly reduce those.. If you have the headroom left you can turn up the subwoofer amplifer to compensate.

I would stronly suggest the use of REW software to do your measuring instead of the manual method. It will give a much improved picture of the situation.

brucek

[Wayne A. Pflughaupt]

Hi Enzo,

A BFD could probably help with some of your problems, but with a 35 dB diffential between the worst peaks and depressions, don't expect miracles. Taking the edge off the two worst peaks is probably the best you can expect. Don't try to "flatten the curve" by pulling that 33 Hz peak all the way down to where 71 Hz is - that would be disasterous.

You don't necessarily avoid boosting by cutting only, especially if you re-adjust the sub level upwards after equalizing, as explained here.

As you note, your biggest problem comes from placement. That sub should be able to get down to 18 Hz in most rooms, with proper placement.

Regards,
Wayne

Hi Brucek,
I think I have headroom. As far as I understand headroom stands for spare power and, if so, I have my gain knob turned up about 30% from zero and the AVR set to -5dB.
I tryed to use REW having as mic the RS SPL meter. The result scared me. So I went for the manual measure.

Thanks for your answer.

ciao

Hi Wayne,
I do not know how this translates in english but "you are turning the knife in the cut". I know that a proper placement will make the magic, but....I can not!
On the other hand i understand that a parametric equalizer would have limits in solving my problem. I did not understand how much I can correct without destroying the sound quality. Form peak to deep, how much I can hope to get as difference in dB?

Thanks for you answer. I like when I do not feel pushed in purchasing something and instead get warned about possible unsatisfaction.

ciao

[brucek]

The result scared me. So I went for the manual measure

.
That result looks quite correct except for the scale you're using.

Change the scale to vertical = 45dB-105dB and horizontal = 15Hz-200Hz. The axis scale is set using the Graph Limits icon button in the top right corner.

Don't use the manual method. REW is quite a bit more accurate.

Repost your REW graph with those axis set.

brucek

Ok here is the same graph with the graph limits you required.
I also attach a new graph which I obtained setting the PEQ on the sub, tuning the sub to 12 Hz and putting the phase at 180°.
It looks much better now I think.
It remains to be seen if MultEQ can adjust it more.

ciao

[brucek]

It looks much better now I think.

No, with respect, you're getting a really bad response from a sub that has a stellar response curve, and should be flat to below 20Hz. You just can't EQ or MutiEQ those results. You need to do something about the positioning.

BTW, always reset your levels in REW to test at 75dB, each time you make any changes.

brucek

brucek wrote:

No, with respect, you're getting a really bad response from a sub that has a stellar response curve, and should be flat to below 20Hz. You just can't EQ or MutiEQ those results. You need to do something about the positioning.

BTW, always reset your levels in REW to test at 75dB, each time you make any changes.

brucek

Brucek, do you want me to cry?

I tryied to move the sub to a corner, but it did not change that much. It changed though, but it is far away from a flat curve. However, by now, I must live with it positioning. May be in the future we will redo the furnishing and I will move it to a corner.

The dilemma will be how to know in advance, once I will redo the furnishing, how it will sound in its designed new position.....

Beside that, I have to tell you that moving the RS SPL meter just a couple of foot, I got a totally different response curve. I didn't realize before that the response curve is so critical to sub positioning and mic positioning. It seems to like a lottery to find the right positioning.

I also wonder how much the RS SPL meter is reliable in measuring sub response. It seems to be consistent in the measures it takes, but...reliable?

However, thanks again for your continued support.

Appreciated.

ciao

enzo

[brucek]

I also wonder how much the RS SPL meter is reliable in measuring sub response.

It's completely fine for sub measurement when the correct calibration file is used.

how it will sound in its designed new position

Place the sub in your listening position and walk around and do measurements in different positions. It's a lot easier than moving the sub around.

Also you can use the RTA feature in REW with its internal signal generator set to Periodic Pink Noise and walk with the meter while watching the response dynamically on the screen.

brucek

Well Brucek,
I did not talk about reliability just for the sake of it.
I noticed that if I take two measures one after the other, without touching anything the graphs I get are never the same.
Small differences appears every time. I do not think this is normal but I may be wrong.

ciao

[thewire]

Is the MultiEQ setting your speakers to large and then you are setting them back to small after the setup? Does it help to set them to large and do you still see the peek on them as well? I'm not positive but I think sending less info to the subwoofer might help if you are running the level loud enough that it is troublesome or even hurting your hearing for that matter. We don't want that to happen. Maybe for at least until you improve the peek by other means? If your speakers can handle 30Hz - 40Hz you could probably listen to them at a reasonable volume set as large without risk to damaging them. This is the best suggestion I could think of for now.

Examples of some measurements I took from my XA2 analog output playing tones on my subwoofer with setting speakers to large versus small.


edit: first the mains to see if things are alright.

440Hz small


440Hz large


edit: then the subwoofer

Tones with speakers set large


Tones with speakers set to small


Sorry if this is only confusing and does not help. I have a tendency to make more complex than they really are has been said.

Steven,
thanks for your info.

MultEQ sets my Front and Center always as large. I always turn them to 80Hz.
I see you point in sending less info to the sub.
What I do not understand is your Graphs.

Play tones what does it means?
Sinus waves?

How can you read the graph with so many tones played in? (how can you have in the same graph more than one tone shown?)

Today I tried to use the RTA moving the RS SPL Meter around the room.

The result was really surprising: peaks changing to deeps and deeps changing to peaks moving the mic only few feet. This makes me wonder how much sens it have to measure in one single position and then chasing after peaks using the onboard PEQ of the sub.

I used the Pink Periodic Noise.

I do not know if this is normal and if it is I do not understand which is the sense of measuring in this way. May be the RS Meter should be substituted by a better mic, or may be multiple mics should be used, mixed, and the resulting response measured....I do not know...I am just wondering.

ciao

[brucek]

The result was really surprising: peaks changing to deeps and deeps changing to peaks moving the mic only few feet. This makes me wonder how much sens it have to measure in one single position and then chasing after peaks using the onboard PEQ of the sub.

In my experience, there are areas that 'suffer' drastic changes in response when you only move a few feet. There are also areas that 'enjoy' a fairly consistent response over a fairly wide area. It's a crapshoot.

If you find your seating / listening position area is quite erratic, you can see the problems associated with EQ'ing of a specific frequency for a single measurement spot. If a position a few feet away suffered the opposite results at that frequency, then the EQ that corrects the main listening spot would only be aggrevating the new spots response.

If you find you have a fairly decent listening area with some reasonable changes over the measurement area, it may be advisable to use the averaging feature of REW to create an EQ profile that makes the entire area 'OK'. This involves taking up to eight measurements, then taking an average of those responses (on the Averaging Tab) and then creating filters to correct that average response. If you had a BFD, you could have one profile for multiple seating and another for the single critical listening spot.

The RTA is quite useful, and the Radio Shack meter is certainly adequate, specifically because with RTA we're not concerned about 'absolute' accuracy of SPL level, but instead we want to see 'relative' change from one position to the next.

As discussed, use Periodic Pink Noise (Pink PN) from the generator, with ~1/24 octave, with a Rectangular window, perhaps a bit of averaging ~2-4, and a FFT of 65536, and a dBFS axis (~-60 or -70 to 0). Go to Settings / View Tab and turn off bars to get a smoother looking line. First set up your levels with a normal response measure of the main listening position, then start the RTA. If you need to touch up the input level from the Settings screen by using the Input RMS readout on RTA screen it's OK.. The static response measures are selectable to place on the RTA screen for comparison.

What I do not understand is your Graphs.

Yeah, not really the best use of RTA. Rather, set up as I discussed above.
Below is a quick test of my poor office sub. I first took a standard response measurement, and then switched to RTA and set up as shown on the graph without moving the microphone.
Now I'm ready to move that mic around and see what happens to the bottom RTA plot as I move around in comparison to the static upper plot. Very revealing.
Gold line = Standard frequency response measurement of sub only.
Black line = RTA readout of sub only.

REW RTA screen


brucek