The Subwoofer Integration Thread (1 Viewer)

[JohnRice]

This is something I've been seriously planning recently, so I figured I'd just go ahead and start it.

This thread is for information about how to truly and effectively integrate subwoofers into a room. Along with others here, like @John Dirk and with months of encouragement from @Dave Upton , I've been amazed at the improvements in sub integration that can be accomplished with minimal investment of money, but a fair amount of investment in time.

A lot of us have quite a bit invested in our systems, and analyzing integration of subs and the room using Room EQ Wizard (REW) is something more of us should be taking advantage of. So I thought I'd start a thread dedicated to the topic.

I'll be back with some of my own experiences with the changes and improvements I've accomplished, and hopefully others will contribute as well.

Please post curves showing the changes and the resulting improvements whenever possible.

 

[JohnRice]

So, where to start. The fact is, digging into REW is intimidating. The software isn't designed with a consumer style easy interface and you have to do research into how to use it with your system. I'm far from entirely understanding the best way to use it.

Like a lot of people, I'm using a Calibrated MiniDSP UMIK-1, with a calibration file that is available for $110 plus shipping from the link. Honestly, it's one of the greatest bargains in all of audio, considering the improvement it can make in your system. You load the calibration file into REW when you first start it to tell the software the true frequency response of your specific microphone. I'm not 100% certain which microphone orientation is best, so I used the 90 degree calibration with the microphone aimed straight up. Not that it matters all that much, but the computer is a base model M1 MacBook Air.

First I'm just going to show a before and after from my living room system. In between these two curves were a lot of just generally learning how to work with everything. There is a smoothing option on the curves, and it's not at the same setting for all graphs, but that's just one of the learning things. I started with it at 1/12 and ended at 1/3. I don't think there's an ideal setting for all situations. It depends on how precise you want the results to be, which I suspect varies with what you are dealing with.

So, here's where I started in my Living Room system, which has ELAC Debut v2 speakers and a single SVS SB-2000 subwoofer. The three readings are for the left and right ear, and the center. I later decided not to measure everything this way, and just do the center until I had made changes. This was also after I'd moved the sub to the front right corner of the room. It had been sitting to the left of the front right speaker. You can see there are peaks a around 10Hz and just below 40Hz. As big as they look, this is still a huge improvement from where I started.

I then used the built-in EQ of the receiver to manually smooth out the response of the system outside the sub's range and achieved this. This was done simply by looking at the curve and making adjustments that seemed reasonable, running another sweep, making further adjustments and so on. I did this individually for all three front speakers. If it sounds like it's time consuming... it is.

You can see how much I was able to smooth out the response in the range above the subwoofer. Again, this is after I improved the subwoofer's range by trying different locations, but I guess I didn't save those curves, so I can't post them.

One interesting thing is that I then checked the phase of the subwoofer by adjusting the distance setting in the receiver. The physical distance from the main viewing spot to the sub is 12', but I noticed there was a drop around 75Hz, so I decided to try sweeps and change the distance setting to see if I could smooth it a bit. I tried settings ranging from 12' to 8'. You can see that each time I reduced the distance, the low spot was reduced, but a new one started forming at about 130Hz. This test showed that the low at 75Hz was mostly smoothed without producing the second one at 10'. So, the best setting was to "incorrectly" set the sub at 10'. That one adjustment smoothed that frequency by about 5dB. I seriously doubt any automatic room correction could accomplish this.

You'll notice there are still some moderate highs and lows in the sub range. The EQ of the receiver isn't capable of smoothing them, but they are moderate enough they could be solved with an MiniDSP... Or the internal PEQ of a sub like SVS offers. Now I wish my SB-2000 was a SB-2000 Pro, which has that feature.

The end result, at no cost aside from the mic purchase isa subwoofer that's incorporated so much better into the room, it's like an entirely new system.

 

[John Dirk]

Nice work, John! I hope @Sam Posten 's ban doesn't extend to this thread because my real-world results coincide with some comments he made during he and @Brian Dobbs podcast with @Dave Upton back in February. When I reviewed the SVS SB4000's I concerned myself only with audio and trying to get results I felt would fairly approximate their capabilities for most users. The problem is, to get these results, I had to make sacrifices in other areas I was not willing to live with on a permanent basis. Namely, I had to position my front speakers in a way that blocked a significant portion of my screen and move my seats further back than I generally prefer.

Tonight I did some more extensive placement experimentation. First off I moved one of the SB4000's to the right side wall while leaving the other to the side of my left front speaker. Next I moved the SB4000 to the rear of my room, while the other remained next to the left front speaker. These are the only other valid options in my room but both produced essentially identical results with pronounced nulls at about 30 and 80 Hz. In my case I was able to make slight improvements with the SB4000's PEQ's but nowhere near flat and the nulls remained, albeit less pronounced. Based on Dave's comments in the podcast, I believe this is where you enter the realm of bass traps. I can't say my ears are unhappy but you don't know what you're missing until you experience it, so... If only there were free in-home trials for bass traps.

I'll try to post some curves tomorrow. Lugging those 100 pound subs around [even with a hand truck] is serious work. Feeling kind of defeated right now.

 

[JohnRice]

Yeah, feeling defeated might be an unavoidable aspect. It's complicated. My HT room has presented far fewer obstacles.

Reagrding your 80Hz null, what I have to say is 14'. That's the length of an 80Hz soundwave. How does that help? Not entirely certain, but I suspect it means something. Maybe 10.5', which is 75% of 14'. I believe a wall at 10.5' will wreak havoc at the frequency.

 

[Dave Upton]

Nice work, John! I hope @Sam Posten 's ban doesn't extend to this thread because my real-world results coincide with some comments he made during he and @Brian Dobbs podcast with @Dave Upton back in February. When I reviewed the SVS SB4000's I concerned myself only with audio and trying to get results I felt would fairly approximate their capabilities for most users. The problem is, to get these results, I had to make sacrifices in other areas I was not willing to live with on a permanent basis. Namely, I had to position my front speakers in a way that blocked a significant portion of my screen and move my seats further back than I generally prefer.

Tonight I did some more extensive placement experimentation. First off I moved one of the SB4000's to the right side wall while leaving the other to the side of my left front speaker. Next I moved the SB4000 to the rear of my room, while the other remained next to the left front speaker. These are the only other valid options in my room but both produced essentially identical results with pronounced nulls at about 30 and 80 Hz. In my case I was able to make slight improvements with the SB4000's PEQ's but nowhere near flat and the nulls remained, albeit less pronounced. Based on Dave's comments in the podcast, I believe this is where you enter the realm of bass traps. I can't say my ears are unhappy but you don't know what you're missing until you experience it, so... If only there were free in-home trials for bass traps.

I'll try to post some curves tomorrow. Lugging those 100 pound subs around [even with a hand truck] is serious work. Feeling kind of defeated right now.

This sounds like you're dealing with the dominant axial nulls in your room. One way to save yourself a little time, is to use a tool like amroc to determine where the nulls are worst in your space. You can then try to put your subs in "good" spots:

amroc - THE Room Mode Calculator

HTML5 room mode calculator. Read and hear the tune of axial, tangential and oblique modes. Scientific sources included.

amcoustics.com

You mouse over a given frequency to see where the mode is most excited and try to move your sub away from it. With two subs, you can locate them such that each sub excites a different mode, and use EQ (MiniDSP) to have each sub make up for the other's shortcomings. This lets you get away with much more flexible placement.

An alternative is to use an RTA app on your phone, and use it to determine where you have the worst issues by moving your phone around the room with the sub in the listening position:

Once you find two spots where your Sub 1 and Sub 2 measure with different nulls/peaks, you are ready to combine the two using a miniDSP. I strongly recommend you watch these 3 videos in order and follow them as your guide. Your room will never be the same:

 

[JohnRice]

Thanks Dave. I've been working my way through those videos and a few others you've posted.

I hope others will also pitch in, with advice or questions. I'm still very new to this.

One question I have for @Dave Upton is in those curves for my living room, there's a peak centered around 40Hz and a low centered around 17Hz. Is that the type of situation that a MiniDSP is suited for? I've placed the sub so that there appear to be no nulls. Of course, the low at 17Hz might be a limitation of the sub, which is a moderate model (SB-2000) in a decent sized room. My SVS PB12-Plus/2 will be looking for a new home shortly, and it could actually go in the opposite front corner of the room.

 

[Dave Upton]

Thanks Dave. I've been working my way through those videos and a few others you've posted.

I hope others will also pitch in, with advice or questions. I'm still very new to this.

One question I have for @Dave Upton is in those curves for my living room, there's a peak centered around 40Hz and a low centered around 17Hz. Is that the type of situation that a MiniDSP is suited for? I've placed the sub so that there appear to be no nulls. Of course, the low at 17Hz might be a limitation of the sub, which is a moderate model (SB-2000) in a decent sized room. My SVS PB12-Plus/2 will be looking for a new home shortly, and it could actually go in the opposite front corner of the room.

Absolutely. You'll position your other sub in a position without those peaks/nulls and use the miniDSP to balance them out

 

[John Dirk]

I'm working through the videos as well. Almost done with the first one, which has already been extremely helpful. Unfortunately, placement options are all but non-existent in my room but I have been able to achieve marked improvement by just playing with placement [to the degree possible] distance and phase settings.

EDIT:

How's that?

 

[JohnRice]

@John Dirk I'm guessing the green curve was an early one? I can't really read the graphs. You might save them larger and as PNG instead of JPG so there aren't any compression artifacts. I've been saving them at 1,000 to 1,500 pixels wide. They're still very small files, at about 90-100 KB each.

 

[John Dirk]

@John Dirk I'm guessing the green curve was an early one? I can't really read the graphs. You might save them larger and as PNG instead of JPG so there aren't any compression artifacts. I've been saving them at 1,000 to 1,500 pixels wide. They're still very small files, at about 90-100 KB each.

My first response to this post seems to have disappeared so forgive me if this is eventually a double post.

You're correct. The green curve is the one measured after the first round of corrections were applied. I'll work on saving the files in a larger size but you can also zoom your browser to make them readable. Here is what things looked like beforehand, so a pretty decent improvement.

 

[JohnRice]

I'm working through the videos as well. Almost done with the first one, which has already been extremely helpful. Unfortunately, placement options are all but non-existent in my room but I have been able to achieve marked improvement by just playing with placement [to the degree possible] distance and phase settings.

View attachment 96166

EDIT:

How's that?

View attachment 96176

You got a 9dB improvement of that major low spot. Amazing how much of a difference placement can make, even when your options are limited, isn't it?

 

[Dave Upton]

My first response to this post seems to have disappeared so forgive me if this is eventually a double post.

You're correct. The green curve is the one measured after the first round of corrections were applied. I'll work on saving the files in a larger size but you can also zoom your browser to make them readable. Here is what things looked like beforehand, so a pretty decent improvement.

View attachment 96173

Have you considered getting a MiniDSP?

 

[John Dirk]

I'm wondering about the MiniDSP 2x4HD as well. I know you use one but I'm not understanding at the moment how this would be integrated so I've held off. Namely, how do make it play well with your Processor? How do you have yours connected?

 

[Dave Upton]

I have my Anthem feeding a 2x4HD with a single output, and then use the 2x4HD to sum, time align and EQ my 3 subs.

The MiniDSP is far, far more advanced than any basic on-sub EQ, supporting 1024 FIR taps. This means you can auto-generate a correction curve in REW and export to the MiniDSP, saving a ton of time.

The primary other advantage being you can also adjust delay for each sub output on the miniDSP, which is important for time alignment

 

[John Dirk]

I have my Anthem feeding a 2x4HD with a single output, and then use the 2x4HD to sum, time align and EQ my 3 subs.

With this setup how do you prevent the Anthem from processing the signal before it is output to the MiniDSP?

 

[JohnRice]

It only does what you tell it to. Maybe I’m oversimplifying it, but it doesn’t do anything on its own.

 

[Dave Upton]

It only does what you tell it to. Maybe I’m oversimplifying it, but it doesn’t do anything on its own.

You calibrate the subs with room correction defeated, so you have ARC/Audyssey/Dirac off, calibrate with only crossovers set, and then once you have the miniDSP calibration done, you do your actual room correction measurements/calibration.

 

[John Dirk]

You calibrate the subs with room correction defeated, so you have ARC/Audyssey/Dirac off, calibrate with only crossovers set, and then once you have the miniDSP calibration done, you do your actual room correction measurements/calibration.

But once you do run room correction how do you ensure it won't muck up the subs?

 

[Dave Upton]

But once you do run room correction how do you ensure it won't muck up the subs?

That's now how DSP works. DSP uses available computing power and signal (available voltage swing) manipulation to fix issues in the frequency, phase and time domains. You can daisy chain multiple DSP devices and as long as you account for delay, there is no downside. The automated systems all have some limits (your AVR trim can only go up/down so many dB). When you correct in front of the AVR/Processor, you're simply fixing some of these issues upstream, letting the DSP on the unit work less hard and get a better result.

If you use Audyssey (and this is one of the reasons why it's so dated now and basically garbage compared to the other solutions), you can't make adjustments but it is still attempting to correct a response that is less ugly than before and will sound better.

If you use ARC/Dirac etc, you will see any corrections proposed in the audio band below the crossover in the software. You can then choose to accept any proposed filters (which you generally do), or set a high-pass filter for correction, or manually adjust any of the proposed parameters.

The takeaway here is that you want your processor to be correcting the flattest/most optimal response possible so the DSP can achieve a flat response.

 

[JohnRice]

Something I hadn't considered is that the MiniDSP can make phase adjustments that are frequency dependent, rather than just making sweeping changes. Is that what you're saying? It completely makes sense that would solve a lot more problems than the internal controls can.

BTW, no doubt I'll be delving into the MiniDSP world. I have this really annoying habit of wanting to fully understand stuff, rather than just using it. So, the phase (again, pun intended) I'm in right now is experimenting and trying to understand the dynamics ( ) of it all. I spent a couple more hours analyzing with REW with my living room system last night. Experimenting more with placement and so on. I expect that system is the first one I'll use the MiniDSP with, because there is a pronounced peak and low that positively cannot be fixed. The HT room seems to be a lot less challenging, but we'll see how dual subs perform on their own.

I think Dave will agree that it's still best to eliminate and reduce as many room problems first. Then go to electronic solutions.

Thanks for linking the videos. I've been working my way through them. My complaint is that so often they tell "what" but don't necessarily explain "why".