How many subs?

[thexder]

Hi all!

I was just researching how to time align multiple subs (more on that later) and did some measurements with two and three subs.
The difference between two and three is subtle but somehow three make the bass more fuller and deeper.
All measurements were done with room correction software. No manual settings were used.

Here is the file. I deliberately deleted almost all the information so there's no data whether there were 2 or 3 subs included in the measurement.
Legend: ABS = some absorption, XABS = more absorption

As an amateur I see very little difference on the graphs so I'm counting on you guys to share your thoughts.

 

[Matthew J Poes]

I’ll take a look when i get to my computer later.

What do you mean by no manual control? How did you align them?

As for absorption? What is the absorption and what did you do to add or take away absorption?

As for auto-room correction, was it rerun each time you changed things?

 

[thexder]

Hi,

thank you, Matthew.
Room correction did the time alignment. It supports two subs. Two of them were always equidistant.
It was rerun each time I changed positions.
The absorption: seven 8" thick absorber panels (24"x40", 10.000 Ns/m4). ABS means 4 panels at the front, XABS means 4 at the front and 3 behind the seats.

 

[Matthew J Poes]

Hey @thexder I tried to see what is going on but I too can't see any major differences here. Not even from the absorption. The issue may simply be that you are getting a very flat/smooth response with the correction.

Can you tell me what the measurements are? What is AUD? What about the 6 or 7? If I knew what this was, then it would help.

If you just want to prove to yourself the effectiveness of multiple subs, it's best to do this without EQ and to take measurements at different locations. You want to measure across all seating locations (or if there is just one, around the seating area). Then you measure deviance from a flat line that passes through the center. best way to do that is to normalize them to 0 and then calculate the deviance from 0. This has to be done in something like excel, REW doesn't do this. As you add in subs you will see the deviance go down. Without multiple measurements and EQ it can be hard to see the results (though the deviance measurement should be the same).

As for acoustic treatment, this was a point I made in my AXPONA presentation as well. 8" is thick and you would think that would absorb a lot of bass. The problem is that we are dealing with just stuff below 200hz where the wavelengths are long. As a result, the bass traps you built aren't doing a lot. That is common. You would need something a bit more efficient at absorbing low frequencies to show a difference, and realistically, probably even more (believe it or not).

You could try straddling a corner with the panels, that will make a good bit of difference. Spacing them from the wall would as well.

What are the panels filled with? That also makes a difference at this thickness.

If you do full bandwidth sweeps you will likely see the results of the panels. Likely above 200hz they are absorbing a good bit and your decay times probably went down. Remember that below that point we are dealing with big powerful reflections. If you look at your EDT and RT60 numbers you will see they are jumping around, that is to be expected in this range.

Overall that is a pretty good bass response so I wouldn't fret too much. If you really want to see the results graphically, we can try to come up with some experiments that show the effect. I will just need to see pictures of the room, sub placement, room dimensions, panel construction, and then we can try to come up with a plan. Let me know if you want to go that route.

 

[thexder]

Hi Matthew.

Thank you again for your reply.

AUD means audyssey. It was always activated and providing level, delay and EQ. Measurements 1-7 mean different sub positions. It turned out that some positions are better than the others which is logical. I know multiple subs are better as we have many facts to prove that (Geddes, Toole and even REW simulator). I'm just curious whether a good response can be achieved just with two since I have only one seat. I'm afraid due to space limitations most of us is having when considering acoustic treatment the more subs can be more than welcome.

I'm aware 8" isn't much but it turned out stacking those panels is useful when placing then between L/R. That means 24" thick absorber (6,5x1,6ft). That kind of thickness can be efficient but it would probably be needed more for maximum efficiency.
Straddling corners didn't do much so I dropped the idea at the moment.
The panels are, like I already stated, a rockwool wrapped in a cloth. Some of them are wrapped in a plastic foil so they don't absorb all the frequencies.

I know this isn't a bad response but if we look at the waterfall there is some heavy ringing present which needs to be addressed. Plus, looking at the ETC there are some reflections too.

 

[Matthew J Poes]

Hi Matthew.

Thank you again for your reply.

AUD means audyssey. It was always activated and providing level, delay and EQ. Measurements 1-7 mean different sub positions. It turned out that some positions are better than the others which is logical. I know multiple subs are better as we have many facts to prove that (Geddes, Toole and even REW simulator). I'm just curious whether a good response can be achieved just with two since I have only one seat. I'm afraid due to space limitations most of us is having when considering acoustic treatment the more subs can be more than welcome.

I'm aware 8" isn't much but it turned out stacking those panels is useful when placing then between L/R. That means 24" thick absorber (6,5x1,6ft). That kind of thickness can be efficient but it would probably be needed more for maximum efficiency.
Straddling corners didn't do much so I dropped the idea at the moment.
The panels are, like I already stated, a rockwool wrapped in a cloth. Some of them are wrapped in a plastic foil so they don't absorb all the frequencies.

I know this isn't a bad response but if we look at the waterfall there is some heavy ringing present which needs to be addressed. Plus, looking at the ETC there are some reflections too.

I’ll take a look again but I don’t recall seeing any heavy ringing. Really you shouldn’t see ringing if you have a flat response. This is minimum phase here so you can’t have ringing without a subsequent peak in the bass, which you don’t really have. Often when you see apparent ringing in a flat response it’s actually just noise.

Other possibilities could be maybe ports ringing at a slower rate than the woofers. Other similar things to that. If it’s that, room treatment won’t fix that.

When I built acoustic panels for a demonstration I was doing, to show a change in bass decay I actually had to build a huge bass trap (24”x24”x~32”) with 8lb density mineral wool in a special design to show anything. I had to use it in a small room (10’x12’), in a corner, and even then the effect was small. I could only show the change by comparing 1 octave filtered impulses.

When I talked to Geddes about this he mentioned how critical it was to get a lot of absorption over a wide area and how difficult that was to do after the fact. He said that was why he developed his lossy wall construction method. It’s what he simulated in his own PhD dissertation. That’s why I built my theater soundproof, the construction method is largely the same. After finding I still had plenty of issues and talking to him, I found out he filled the front of his room, behind the screen, with that 30ppi reticulated foam. Many feet thick.

Again, things looked quite good to me.

As for your question about the number of subs, smoothness improves with more LF sources as does evenness across measurement position. 3 is a good number but nothing wrong with 2. It will just be slightly less smooth/even. This is also somewhat room dependent so certainly some rooms might be just as smooth with 2 subs.

I may have missed this. So I know you measured across different sub positions. What about measurement positions? Was the mic always in the same place?

 

[thexder]

Trust me, it's there.

I know something isn't right because most of stereo tracks do not sound good.

Some time ago I achieved even flatter response with no EQ at all but there was severe ringing at 2nd mode (47 hz). Frequency response is just one detail and having this one flat does not mean others are near perfect too.

Traps, no matter what size they are, can be a waste of space if placed wrong. And corners aren't the best place to put them. Every book covering basic acoustic will confirm that.

I agree, things are alright. Not bad at all. But while movies sound fine I'd like to improve 2-channel reproduction. And there's for sure a room for that which can be seen on the plots.

Yeah, based on my experience two subs are a must. Most likely 3 are better, 4 are optimal. Especially if the room isn't a perfect rectangle. What me off is that there is a slightly better result with 3. And that little improvement has quite a pricetag!

Yes, when measuring aud1-aud7 mic was always at the same place.

Have you ever experienced with other types of absorbers? Membrane or with a mass spring system, etc?