Wanted: Proof of multiple subs and sub EQ

[Ethan Winer]

This is mostly for Amir because I know he's up on this topic, but I'm glad to hear from anyone else who has graphs or (better) an REW data file and is willing to share. I'm writing an article about bass problems and solutions in home-size rooms, and I want to be as fair and accurate as possible. So toward that end:

1) I'm looking for proof that using multiple subwoofers can reduce modal ringing. I've heard proponents claims you can place a sub "where it won't excite room modes," but that makes no sense to me. I've asked several multi-sub proponents for Before/After waterfalls, but they never had any actual proof of this claim.

2) Likewise, I'm looking for proof that subwoofer EQ can reduce modal ringing more than a tiny amount.

My only requirement for graphs is that the Before and After measurements must have been made with identical microphone placements. If the mic moves even an inch or two between sweeps, that alone will create differences large enough to invalidate the data. As an example of what I'm looking for, the attached graphs show the response in a largish room with and with the Dirac room EQ system engaged. The mic was not moved between measurements.

Thanks!

 

[Purité Audio]

Surely extra subs can't help with a fundamental room mode , I can see how placing one in a null caused by second or third order etc might help?
Keith.

 

[DonH56]

To me modes relate to boundary conditions and standing waves in the room; ringing relates to time response and transient behavior. So I am not sure exactly what you are seeking. I would expect additional subs to help with frequency response but not with ringing.

Frequency response in my room was much improved by adding subwoofers placed near key points to drive nulls. Not sure I measured the ringing, though I could, but my room is pretty atypical in that it is very heavily dampened with a plethora of absorbers. I still have some fine-tuning to do, just haven't had time and energy (working too much!)

 

[Fitzcaraldo215]

The few measurements in this were a little hard to read, and the measurement protocols are not fully explained. So, this might not directly address Ethan's questions. But, here is what appears to be a recent commentary that is relevant to the subject generally by Floyd Toole:

http://www.audioholics.com/room-acoustics/history-of-multi-sub-sfm

 

[h.g.]

1) I'm looking for proof that using multiple subwoofers can reduce modal ringing. I've heard proponents claims you can place a sub "where it won't excite room modes," but that makes no sense to me.

What is modal ringing?

To prove to yourself what multiple subwoofers can do go into the room simulation part of your REW program and move a couple subs around and look at what changes. Then change the phase between the two subs and look at what changes. My guess is that you will then see your "modal ringing" of some of the modes change for the better and some for the worse.

What is being done is using the output from one sub to reduce the output of another over part of the frequency range and in part of the room. It is not efficient in that you need to pump more power into the room but if done well the result after cancellation should be a substantially smoother transfer function with lower fatter peaks (less "modal ringing"?) than you would get with a single subwoofer.

 

[amirm]

Hi Ethan. I have some of that already created using REW in this thread: http://www.audiosciencereview.com/f...urements-understanding-time-and-frequency.25/

As you see there, I show in time domain how ringing is reduced as filters are applied to peaks. This is one example:

The graph shows simultaneously that the amplitude of the peak is reduced around 53 Hz, and how the ringing dies sooner (light brown).

After applying four filters, this is what we get:

The filtered response in blue is both flatter and lasts less in time domain.

The key to seeing that is using the right parameters in REW. That is the purpose of that article. Without it, such observations can be lost.

 

[amirm]

On multiple subwoofers, there is also this article: http://www.audiosciencereview.com/forum/index.php?threads/subwoofer-low-frequency-optimization.15/

And these two graphs showing one sub in the corner versus four subs in each corner:

What is important here is that we made the response seating location independent. By doing so we are therefore erasing the room impact. Once there, the four subs can be EQ'ed collectively without having different effects in different seating locations.

 

[Ethan Winer]

Thanks for all the replies.

Don and Ken, I agree that multiple subwoofers can improve the response, but I don't see how they can reduce ringing decay times. Don, do you use REW? If you could measure at the listening position with both subs, then turn off one sub and measure again, that'd be perfect. REW has a display mode that shows LF decay time separately from the response.

Fitzcaraldo, Floyd Toole's data is not complete enough or explained well enough to be adequate proof for me. His waterfall "Subs 2+3" shows a lower peak, but that could make it seem like less ringing. This is why the other REW display mode I just mentioned is so useful. Also, not the time and frequency resolution of his waterfalls seem too coarse to be useful.

h.g., thanks that's a great idea and I'll try it. However, for the most compelling proof I also need actual measurements rather than a simulation.

Amir, that's awesome. But it's difficult for me to see how much the ringing is reduced because the peak level was also reduced. Would you be willing to send me your file? Then I can scale the data to see more of the decay, and use REW's spectrogram view to isolate only the decay times. As for the multi-sub data, I understand that multiple subs can improve the response. What I question is whether that can also reduce ringing.

Thanks very much guys!

 

[h.g.]

h.g., thanks that's a great idea and I'll try it. However, for the most compelling proof I also need actual measurements rather than a simulation.

I would suggest that the most compelling proof comes from understanding what is going on rather than some measurements off the internet.

You haven't told me what "modal ringing" is. I ask because I think you may be mixing up two separate things. Even if you achieve a perfectly flat transfer function in a reverberation chamber (no visible "modes" on your decay plot) it will still take forever to decay and so will not be suitable for high quality music reproduction.

 

[NorthSky]

Part 1: http://www.audioholics.com/subwoofer-setup/multiple-subwoofer-setup-calibration-1
Part 2: http://www.audioholics.com/subwoofe...-setup-calibration-1/multiple-subwoofer-setup
Part 3: http://www.audioholics.com/subwoofe...tiple-subwoofer-setup-listening-and-adjusting

__________

♦ http://www.audioholics.com/home-theater-calibration/bass-optimization-for-home-theater
♦ http://www.acousticfrontiers.com/20...ow-to-use-parametric-eq-to-flatten-your-html/

 

[Ethan Winer]

I would suggest that the most compelling proof comes from understanding what is going on rather than some measurements off the internet.

The best proof is measurements. I understand fully what's going on, but empirical evidence trumps theory every time.

You haven't told me what "modal ringing" is. I ask because I think you may be mixing up two separate things. Even if you achieve a perfectly flat transfer function in a reverberation chamber (no visible "modes" on your decay plot) it will still take forever to decay and so will not be suitable for high quality music reproduction.

I assure you I'm not mixing up anything. Home-size rooms don't behave like the reverb rooms in a professional acoustics lab. (Though even those rooms have modes that ring individually at very low frequencies.) Do you know what the Schroeder frequency is? I'm asking for data about two specific things. Do you have anything I can use?

--Ethan

 

[Ethan Winer]

Part 1: http://www.audioholics.com/subwoofer-setup/multiple-subwoofer-setup-calibration-1
Part 2: http://www.audioholics.com/subwoofe...-setup-calibration-1/multiple-subwoofer-setup
Part 3: http://www.audioholics.com/subwoofe...tiple-subwoofer-setup-listening-and-adjusting/

I've seen Gene's work and it doesn't address what I'm asking. I know he thinks it does but it really doesn't.

http://www.acousticfrontiers.com/20...ow-to-use-parametric-eq-to-flatten-your-html/

Yes, Nyal's work is more compelling because he includes decay information. I emailed him yesterday asking for clarification on a few things, and hopefully he'll reply. In the mean time I figured I'd ask here since I knew Amir has a lot of good data.

--Ethan

 

[AJ Soundfield]

The best proof is measurements.

http://www.genelec.com/documents/publications/aes111th_2.pdf
Yeah, forget audibility.
http://www.aes.org/e-lib/browse.cfm?elib=7868
http://www.akustinenseura.fi/wp-content/uploads/2013/08/goldberg.pdf
And heaven forbid injecting 6db less power at LF....with the exact same onset response. No, let's create energy, then turn it into heat. Yay.
Ethan, you use one sub with "bass traps" right? Are you familiar with lateralisation? Or the work of this guy:

Just curious. Hope all is well with you.

 

[h.g.]

The best proof is measurements. I understand fully what's going on, but empirical evidence trumps theory every time.

Not if you are an engineer and want to get something done which is why science takes the form it does.

I assure you I'm not mixing up anything.

But are you in the best position to judge that?

Do you know what the Schroeder frequency is?

Yes. (I also know what is inside reverberation chambers.)

I'm asking for data about two specific things. Do you have anything I can use?

Probably not given I am not confident what you mean by "modal ringing" and your being interested in pictures rather than how things work. I would only have asked a question about the work done by the sound next anyway.

 

[iridium]

Amir,
I read your Subwoofer / Low Frequency Optimization
By Amir Majidimehr
I tried to post a reply to the article; would not work [my text was drawn through]. So here it is:

Great Amir,
Now you should take it to the next step & go beyond Keith Yates' CFD.
#1 Develop a small robot that travels all spaces that are directly connected to the room when music is playing. The robot would use lasers to accurately 3D map every surface. The robot would have a periscope to 3D map soffit lighting, etc.
#2 The 3D map would then be shared by a small army of robots that emit sound waves of all frequencies. This little army would bit by bit use different frequencies separately & in combination while taking different positions in the room to develop the Amir SUPER CFD analysis result.
#3 Do not forget to KISS the software GUI.
#4 Retire & live-off the royalties.

iridium.

 

[amirm]

Actually, you are not too far from an idea I have on that front!

 

[DonH56]

Hey Ethan,

I agree, not sure how multiple subs can change ringing, except perhaps by flattening the FR they reduce the apparent magnitude at the listening position, or by simple wave interference..

I have REW, a fairly recent acquisition, will see what I can do. May take a while, my day started at 6 am in the lab and I'm still dealing with some emails and test results now at 10 pm. Been a long year already... OTOH there's a blizzard warning here tomorrow so I may have some extra time.

Good to "see" you again, BTW. - Don

 

[amirm]

Blizzard this late in the season???

 

[NorthSky]

I don't see Mike in this thread.

 

[amirm]

I don't see Mike in this thread.

Sometimes Blizzard means the real thing. You know, a big snow storm .