MSO- Multi Sub Optimizer Beginners Thread

[alexf40]

Just a couple thoughts. I would say never gain match. Especially with subs that are different. Gain matching only works with identical subs. Now, the reason gain matching doesn’t work, imo, is that almost always, they end up in different part of the room, with different relationships to the MLP. So if you have one in a corner, and one on an open wall where it leads into a kitchen, the response will be completely different. Ime, level match works better.
Also, I’ve found asio to be nothin it a headache, and unnecessarily complicated.

Yes I'm still really confused over this. I get the difference in principle between gain and level matching. But in practice I'm a very unclear. As with gain matching, unless I've been doing it wrong, it ends up almost the same anyway, both subs are almost identical distance from the MLP, speaker layout is symmetrical.

I've been putting the sub in the center of the room on the coffee table with the mic about 6" from the sub, adjust the gain until it hits 75dB the same as the pink noise from the mains. Swap subs over and do the same. Then put them back and remeasure to see how much they've dropped below the 75dB from the MLP. Then put them back up on the coffee table doing the same thing increasing their volume. (I've done this so many times now in hindsight I should have worked out the actual dB reading they need to be to gain match before moving them back, would have been as simple as one extra measurement before I moved them).

In fairness, I haven't tried actually level matching from where they are with the mic in the MLP. I just can't work out how they'd be that much difference. And is the very small difference they'd be at reference level even worth the hassle as I never listen to anything that loud anyway so can't see why I'd need any minimal difference in headroom. I've only done it this way as I don't know what I don't know and everything I've read says I should gain match.

But surely even identical subs in non indentical positions wouldn't be any different to non-indentical subs in different positiions due to the room layout? Also what about the gain post EQ, surely this as well would mean they are no longer gain matched?

I'll start by saying that you should set range to optimize in MSO a little wider, perhaps 20Hz to 150hz. That will give you more room to decide where to set the final Crossover post AVR calibration. Remember that with a sub only optimisation you are looking to create the best combined subwoofer response so that when you perform your auto calibration with the AVR, it detects a good enough bass response so it doesn't have to tweak it. It will just find the best distance and crossover point to then integrate it with the other speakers.
I'm in the middle of a few things now, so I will review your file later and add any comments that are needed. Enjoy the toddlers.

I suppose I was just assuming that the cross over would still be 80Hz, as all my other speakers would easily play down to 80Hz. But I haven't tested it and is something I should do, but will try this as the second stage once I manage to be able to get the basics working first. Also as I was only able to sweep to 200Hz which included some cross over from the mains I thought it was better to have more room for the MSO optimisation. I wish there was some clearer documentation on the explanation of this part. Mind you it's free software, and a bloody amazing tool, and incredibly generous of Andy to give it away free, so I completly understand why it's not going to have the same documentation as a paid platform.

Thanks for your help in looking in to the files. Good news is that I finally managed to get ASIO4ALL working properly and can now choose output channels to all speakers. Once I have some time with everyone out the house I'll try and perform all the sweeps again and see if anything is considerably different.

If anyone else out there is having issues with ASIO4ALL I came across another alternative https://github.com/dechamps/FlexASIO

 

[William Lemmerhirt]

Yes I'm still really confused over this. I get the difference in principle between gain and level matching. But in practice I'm a very unclear. As with gain matching, unless I've been doing it wrong, it ends up almost the same anyway, both subs are almost identical distance from the MLP, speaker layout is symmetrical.

I've been putting the sub in the center of the room on the coffee table with the mic about 6" from the sub, adjust the gain until it hits 75dB the same as the pink noise from the mains. Swap subs over and do the same. Then put them back and remeasure to see how much they've dropped below the 75dB from the MLP. Then put them back up on the coffee table doing the same thing increasing their volume. (I've done this so many times now in hindsight I should have worked out the actual dB reading they need to be to gain match before moving them back, would have been as simple as one extra measurement before I moved them).

In fairness, I haven't tried actually level matching from where they are with the mic in the MLP. I just can't work out how they'd be that much difference. And is the very small difference they'd be at reference level even worth the hassle as I never listen to anything that loud anyway so can't see why I'd need any minimal difference in headroom. I've only done it this way as I don't know what I don't know and everything I've read says I should gain match.

But surely even identical subs in non indentical positions wouldn't be any different to non-indentical subs in different positiions due to the room layout? Also what about the gain post EQ, surely this as well would mean they are no longer gain matched?

Lots of good questions. I’ll go in no particular order.
Identical subs in non identical locations would not be the same as non identical subs in identical locations. This is because two different subs will have different phase responses. So even IF they’re actually in theoretically identical locations, their responses will be different with the room and different from each other. That means their response to each other is unpredictable and you will have issues with cancellations.
The reason I like level matching, is because it absolutely takes the room and placement into account. That also means audyssey, REW, ypao, MSO or any kind of Eq and level matching is directly related to real use case.
Simple gain matching has to be done with identical subs since their responses, and capabilities also have to match each other. Otherwise it’s a crap shoot and back to unpredictability.
Not sure where you’ve read so much about gain matching being “the way” to do it, but Ime gain matching is less practiced compared to level matching.
The point I guess is that to me, the differences aren’t about headroom and output capability. It’s about how the subs work with/against the room and LP.

Is your room symmetrical and completely closed off? If not, then the symmetry of speakers doesn’t matter.
Hope some of that makes sense lol. Been a long day...

 

[Wellz]

Just bringing this over from another thread that delt with Gain Matching vs Level Matching, where Craig (a contributor) wrote:

"When you gain-match, you want to measure the raw output of the subwoofer, with as little impact from the room as possible. That is why it is recommended to move the sub to the middle of the room... it reduces the levels of the reflections from the room boundaries.

Once you've gain-matched them, and then replaced the subs into their in-room positions, they will still be driving the same amount of energy into the room, they will still have the same headroom, the same point where they go into compression and distortion, etc.

They may measure differently at the (arbitrary) primary listening position due to the room reinforcement from the boundary reflections, but they will be driving the same electro-mechanical energy into the space. That is the whole point of gain-matching."

Then there is also this post from Gene DellaSala where he uses the term Level Match instead of Gain Match, but what he is describing is what I (and others call Gain Matching): https://www.audioholics.com/subwoofer-setup/multiple-subwoofer-setup-calibration-1/multiple-subwoofer-setup

 

[alexf40]

I finally had some quiet time to have another go at this, and I've seen a superior improvement in using ASIO over Java. Although far from perfect, I have a far better response than prior, which is a massive improvement as using Java measurement sweeps in REW I generally had worse performance after using MSO.

I still have some teething issues it appears as I'm getting a sub trim out in MSO of +15dB and a gain on one sub of -24dB.

I'm assuming I'm doing something wrong when setting the gain on the subs. For pure laziness today, I used level matching rather than gain matching.

I used the signal generator to play pink noise (500-2000Hz) out of REW to FL and turned up the AVR volume until the REW's SPL Meter (C/Slow) in MLP was at 75dB.

I then used the signal generator to play pink noise (30-80Hz) out of REW to the subs independently until they were at 75dB. The Klipsch sub I had to turn down to a quarter of the way on the gain.

So should I do the calibration a different way to ensure that the AVR doesn't have to be at max settings?

Also is there any way to set the max trim AVR output? As my Sony only goes up to 10dB.

Thanks
Alex

Green - Both subs MLP before MSO
Blue - Both post MSO

This is the pre, post and predicted, the gap between predicted and actual I assume is due to the 5dB restriction in the AVR trim.

Bright Blue - prior to optimisation
Red - prediction
Light Blue - actual after optimisation (prior AVR calibration)

 

[William Lemmerhirt]

Well the blue definitely looks better than green. But you should change your vertical axis to show from 45-105.
And also. You don’t want to run the sub channel trim in the AVR in the plus, and definitely don’t max it out. This can add distortion, and clipped signals. Not good.
After calibration, the trim should be around -5 to -7. Then you can bump a little to taste. Usually 2-5db. I use three subs, and set each one to about 70db individually. Then combined they come out at about 78-79. I also have a minidsp for eq and house curve, but that’s another thing.
If you’re gonna level match, turn off one sub, and set the other to about 72. Then switch them, and set the other one to 72. Then turn on the second one and check the results. Set the trim so you get about 77-78. Se how you like it.
Also, does the AVR have internal test tones? I wonder what results you’d get from trying those.
And again. Do not max the sub trim.

 

[alexf40]

Well the blue definitely looks better than green. But you should change your vertical axis to show from 45-105.
And also. You don’t want to run the sub channel trim in the AVR in the plus, and definitely don’t max it out. This can add distortion, and clipped signals. Not good.
After calibration, the trim should be around -5 to -7. Then you can bump a little to taste. Usually 2-5db. I use three subs, and set each one to about 70db individually. Then combined they come out at about 78-79. I also have a minidsp for eq and house curve, but that’s another thing.
If you’re gonna level match, turn off one sub, and set the other to about 72. Then switch them, and set the other one to 72. Then turn on the second one and check the results. Set the trim so you get about 77-78. Se how you like it.
Also, does the AVR have internal test tones? I wonder what results you’d get from trying those.
And again. Do not max the sub trim.

Thanks William, well that's what I was trying. I just don't understand why the trim is so high. I deleted all levels, distances and EQ from the AVR. I set the subs to 75dB to match the 75dB mains, I'll give it a go with the AVR test tones.

I'm wondering if my issue with the trim is due to the different frequences used for the measurements?

Or could it be an issue with the AVR sub out of 2v and the miniDSP2x4 of 0.9v?

 

[William Lemmerhirt]

Thanks William, well that's what I was trying. I just don't understand why the trim is so high. I deleted all levels, distances and EQ from the AVR. I set the subs to 75dB to match the 75dB mains, I'll give it a go with the AVR test tones.

I'm wondering if my issue with the trim is due to the different frequences used for the measurements?

Or could it be an issue with the AVR sub out of 2v and the miniDSP2x4 of 0.9v?

I would definitely suspect the .09v. That’s one reason I always recommend the HD version. I don’t know off hand of a solution for that. Iirc, @lovinthehd has used the regular minidsp. Maybe he can help out.
Couple other side roads...
One, I’d be curious if you tried rew generated filters exported to the mini,(after level matching each) what the FR would look like.
Two, just for shts and giggles. Remove the mini, and just use a splitter, and set the levels, and phases manually. Then measure with rew and see how bad it is. I don’t remember if you have placement options or not. Might do better that way. Dozing off atm so, clarity is failing lol.

My familiarity with MSO isn’t very high. I looked at it for a few days a long time ago, and decided it wasn’t for me and my use case. I’m afraid I can’t offer you much there.

 

[lovinthehd]

The unbalanced 2x4 is limited to .9V output. The unbalanced can do 2.0V, believe HD is also capable of 2.0V. There are also different input level capabilities. I think even the .9V would work well with a lot of plate amps I'd think, tho....

I was just looking thru some of the latest postings on this a while ago, but not in depth.

 

[William Lemmerhirt]

miniDSP 2x4 HD

MiniDSP 2x4 HD, a USB DAC+DSP in a low cost package powered by a powerful Sharc 400MHz floating point DSP. The upgraded on-board 400 MHz Analog Devices SHARC processor also ena-bles substantial processing upgrades previously available only on more ex-pensive platforms, such as 96 kHz internal...

www.minidsp.com

At the bottom there’s a chart that compares the two. .9 for non HD only. You can increase the input sensitivity to 2v but I saw somewhere that will decrease its output capability, not sure why Can’t remember though. Bedtime lol.
For some reason I thought you had a non HD. Guess not!

 

[alexf40]

I would definitely suspect the .09v. That’s one reason I always recommend the HD version. I don’t know off hand of a solution for that. Iirc, @lovinthehd has used the regular minidsp. Maybe he can help out.
Couple other side roads...
One, I’d be curious if you tried rew generated filters exported to the mini,(after level matching each) what the FR would look like.
Two, just for shts and giggles. Remove the mini, and just use a splitter, and set the levels, and phases manually. Then measure with rew and see how bad it is. I don’t remember if you have placement options or not. Might do better that way. Dozing off atm so, clarity is failing lol.

My familiarity with MSO isn’t very high. I looked at it for a few days a long time ago, and decided it wasn’t for me and my use case. I’m afraid I can’t offer you much there.

Yes in hindsight I should have got the HD, I didn't know enough about them before I actually started to use one. Will try and see if anyone wants to buy it and upgrade.

As for your questions.

1. I've really struggled with REW trying to get the time delay etc to work to get a better response. I then choose MSO as it seem so much simpler for what I need in a sub only configuration.

2. Without the DSP there was a lot of cancellation between the subs frequency response.

EDIT: Sorry think I made a mistake earlier, should have labled them when I took the screen shot.
Think blue and green were independent subs,orange combined.

 

[lovinthehd]

miniDSP 2x4 HD

MiniDSP 2x4 HD, a USB DAC+DSP in a low cost package powered by a powerful Sharc 400MHz floating point DSP. The upgraded on-board 400 MHz Analog Devices SHARC processor also ena-bles substantial processing upgrades previously available only on more ex-pensive platforms, such as 96 kHz internal...

www.minidsp.com

At the bottom there’s a chart that compares the two. .9 for non HD only. You can increase the input sensitivity to 2v but I saw somewhere that will decrease its output capability, not sure why Can’t remember though. Bedtime lol.
For some reason I thought you had a non HD. Guess not!

I have one each of the balanced and unbalanced (this was before the HD existed....I'd have gotten the HD from the get go if it was available then).

 

[alexf40]

miniDSP 2x4 HD

MiniDSP 2x4 HD, a USB DAC+DSP in a low cost package powered by a powerful Sharc 400MHz floating point DSP. The upgraded on-board 400 MHz Analog Devices SHARC processor also ena-bles substantial processing upgrades previously available only on more ex-pensive platforms, such as 96 kHz internal...

www.minidsp.com

At the bottom there’s a chart that compares the two. .9 for non HD only. You can increase the input sensitivity to 2v but I saw somewhere that will decrease its output capability, not sure why Can’t remember though. Bedtime lol.
For some reason I thought you had a non HD. Guess not!

I read the Mini DSP manual earlier, there's a jumper to change 0.9v to 2.0v for input sensitivity, but it doesn't change the output at all. Output is always fixed to 0.9v on the standard 2x4.

 

[William Lemmerhirt]

I read the Mini DSP manual earlier, there's a jumper to change 0.9v to 2.0v for input sensitivity, but it doesn't change the output at all. Output is always fixed to 0.9v on the standard 2x4.

Correct.

 

[William Lemmerhirt]

I have one each of the balanced and unbalanced (this was before the HD existed....I'd have gotten the HD from the get go if it was available then).

Roger that.

 

[William Lemmerhirt]

Yes in hindsight I should have got the HD, I didn't know enough about them before I actually started to use one. Will try and see if anyone wants to buy it and upgrade.

As for your questions.

1. I've really struggled with REW trying to get the time delay etc to work to get a better response. I then choose MSO as it seem so much simpler for what I need in a sub only configuration.

2. Without the DSP there was a lot of cancellation between the subs frequency response.

View attachment 45288
Green and Orange independent subs, blue combined.

That looks odd... I’ll look again tomorrow lol.

 

[Wellz]

I finally had some quiet time to have another go at this, and I've seen a superior improvement in using ASIO over Java. Although far from perfect, I have a far better response than prior, which is a massive improvement as using Java measurement sweeps in REW I generally had worse performance after using MSO.

I still have some teething issues it appears as I'm getting a sub trim out in MSO of +15dB and a gain on one sub of -24dB.

I'm assuming I'm doing something wrong when setting the gain on the subs. For pure laziness today, I used level matching rather than gain matching.

I used the signal generator to play pink noise (500-2000Hz) out of REW to FL and turned up the AVR volume until the REW's SPL Meter (C/Slow) in MLP was at 75dB.

I then used the signal generator to play pink noise (30-80Hz) out of REW to the subs independently until they were at 75dB. The Klipsch sub I had to turn down to a quarter of the way on the gain.

So should I do the calibration a different way to ensure that the AVR doesn't have to be at max settings?

Also is there any way to set the max trim AVR output? As my Sony only goes up to 10dB.

Thanks
Alex

Hey Alex,
It seems like you are overlooking some of the key settings in MSO which allow you to set the limits/parameters for whatever hardware you plan on using.
Like I mentioned earlier, the MSO instructions are not necessarily written for folks who are just starting out with this hobby (ask me how I know).
It is for that reason that I took the time to write the summarized walkthrough for you in my earlier post.
Having said that, If you go back to the MSO manual you will find a section explaining all of that in great detail, but here is another quick summary:

In the MSO program menu, under TOOLS, you will find Application Options.
In this section you will find several folders with sub folders within them.
The first folder is FILTERS. In that one you can set the minimum and maximum GAINS limits(among other settings) that you want to allow MSO to use. These will also affect the range that your AVR and MiniDSP will be fed in the results.
Further down you will find the HARDWARE folder. In this one you will also find parameters that MSO need to know in order to create a proper filter report based on the hardware info that you input there. So there you will indicate such things as the measurement scale that your AVR uses to set Distances, etc.
*Both of those folders will allow you to set the limits of your AVR as well as your MiniDSP.

Now, with regards to the procedure that you posted above for level matching... whether you choose to Level match or Gain match your subs, JUST MAKE SURE that you are disconnecting the Main speaker (who's channel you are using to generate the sweep tone or Pink Noise) before you run the sweep, so that ONLY the Sub(s) is actually reproducing the sweep tone. And when you set the parameter in REW, make sure that you run a full signal sweep that covers your sub's frequency range.
When you run the sweep, (aside from the sub) the only Main speaker that should remain connected is the one being used for the Acoustic Timing Reference.

All of that is explained in my earlier post, but it is worth mentioning again.

Lastly, I'm not sure just how much you are expecting out of your setup, but the results you posted from MSO look pretty good (especially when compared to your raw measurements).

If I was in your shoes, I would stick to doing the Sub-only configuration with MSO, then input those results into your MiniDSP, and then finally run the auto-calibration from your AVR.

Be mindful that MSO will generate 2 different sets of suggestions in the Filter Report.
One will be the settings needed in order to reach the specific SPL TARGET level that you chose in Optimization Options under "REFERENCE LEVEL COMPUTATION".
The other will be just the results that you need to use in order get the subs to play nice with each other.
Those recommended settings (which are at the end of the FILTER REPORT ) look something like this, except that the recommended levels will be specific to your setup:

Final gain and delay/distance settings:

Complete gain settings for SPL target of 80 dB (optional):

Increase AVR sub out trim gain by 5.5 dB

Sub Channel 1 gain: -5.78 dB

Sub Channel 2 gain: -1.11 dB

Sub Channel 3 gain: -8.19 dB

Sub Channel 4 gain: -0.48 dB

Minimal gain settings, neglecting SPL target:

Sub Channel 1 gain: -5.30 dB

Sub Channel 2 gain: -0.62 dB

Sub Channel 3 gain: -7.71 dB

Sub Channel 4 gain: 0.00 dB

Delay settings:

Sub Channel 1 delay: 0.00 msec

Sub Channel 2 delay: 6.80 msec

Sub Channel 3 delay: 11.51 msec

Sub Channel 4 delay: 0.65 msec


Notice that in the above results, ONLY the "optional" results require that you adjust the AVR trim. This is based on the SPL of the original measurements that you took. But since all you want at this point is to get the subs playing well together, you should really just use the second set of results listed under "Minimal Gain Settings Neglecting SPL Target".

Use those setting along with the Biquad results that were generated, and then allow your AVR decide and adjust the Sub Trim level when you run the auto calibration.

Some final advice.
Now that you've made the financial investment into the equipment and audio gear, it is only to your benefit that you also invest some time reading and educating yourself on how to get the most out of it and why.
So even if you don't get time to re-run the actual calibrations anytime soon, I would suggest that you find some time to get more familiar with the process by just re-reading these post as well as the REW and MSO manuals.
All of these tools including have a learning curve, and it's best to take your time learning it.

P.S. In one of my earlier posts I also suggested that you watch Episode 7 of the HomeTheaterGurus channel on YOUTUBE. That episode in particular will give you a visual representation of what and how to do this subwoofer integration process, albeit, he does not use MSO for his results, but rather chooses to use REW to generate the "filters" aka EQ points and biquads needed to get a nice curve and great results. I strongly advice that you watch that episode because it also gives you some insight on WHY you want to do this correctly. In fact if you continue to have issues with MSO, I would recommend that you just set things up per the intructions on that Episode and enjoy the results!

Good luck, and happy listening!

 

[alexf40]

Hi @Wellz I did watch the video you linked to, well I had actually seen it before. I first saw this video and tried to follow it in REW before moving to MSO. I assume the problem I had with MSO and not using the ASIO driver was the same issue why I got poor results in REW.

I had set the max gain settings in the application settings and also in the filter settings to match that of the miniDSP +/- 15, and also the delay limits of +/- 7.2ms.

The hardware settings are all correct, but there is no setting within MSO to set the AVR sub trim limit.

However what I have realised I can do as a workaround is to reduce the dB limit of the filters to 10dB and that minimises the trim needed on the AVR.

 

[Wellz]

Hi @Wellz I did watch the video you linked to, well I had actually seen it before. I first saw this video and tried to follow it in REW before moving to MSO. I assume the problem I had with MSO and not using the ASIO driver was the same issue why I got poor results in REW.

I had set the max gain settings in the application settings and also in the filter settings to match that of the miniDSP +/- 15, and also the delay limits of +/- 7.2ms.

The hardware settings are all correct, but there is no setting within MSO to set the AVR sub trim limit.

However what I have realised I can do as a workaround is to reduce the dB limit of the filters to 10dB and that minimises the trim needed on the AVR.

Ok. that can work. However, just take a minute to think about why the SUB-ONLY Configuration could work best for you if you ignore the "Optional" MSO Filter Response suggestion for the AVR.
I'll explain;

If you use the Sub-Only configuration, then in essence what you will be creating is the best version of a "subwoofer" that your AVR can THEN integrate with your main speakers.
In order to do this properly, you FIRST need to have the subs combine (integrate/align) with EACH OTHER to create this "optimal subwoofer".

If you set this up properly( by first gain matching the subs and then combining the subs in MSO via the MiniDSP), then when it comes time for the AVR auto-calibration to integrate the "subwoofer" with the main speakers, the AVR will only "SEE" 1 subwoofer response that needs to be integrated in the system.

This new "optimal subwoofer" was FIRST optimized by MSO to suit the room and seating locations independent of any other gear or speakers that you have and will use. Which means that now the AVR auto-calibration will simply have to adjust the subwoofer channel's trim/gain as well as its distance setting in order to integrate it with the rest of the speakers, there should be no need for the AVR to correct the Subwoofer response.


Here is another way to attempt to get the combined output from your GAIN Matched subs to the 75db needed at the MLP:

1.Start with 1 sub placed at the location where it will remain once it is integrated with the rest of your system. Connect that sub directly to your AVR sub out channel, and then feed it the subwoofer channel test tone as if you were doing a manual channel gain setup. The AVR should now be sending a 75db signal to the sub.

2. Set up REW (using the REW SPL Meter function) and place the USB Mic at the MLP. Adjust the gain knob on the back of the sub until you get a 72dB reading at the MLP. now you have the gain of that sub set to give you a good SPL at the MLP.

3. You can now do a close mic SPL reading of that first sub (place the mic 4 inches away from the center of the cone) to see what the actual output is from the sub at close range (which minimizes the effect of the room on that reading). This should result in a much higher SPL reading then the 72dB that you measured at the MLP. Take note of what that SPL reading is, and then do the same close mic SPL reading on the rest of your subs. Adjust the gain knob of each sub so that the SPL output at close range matches the close range SPL output of the first sub. This will gain match the subs

4. Now that the subs are gain matched using the close mic, you can place them all at their final destination.

4. Now connect the subs directly to the MiniDSP and then connect the MiniDSP to the AVR using just 1 sub output from the AVR.
With the MiniDSP zeroed out per my original post, you can now send the same sub channel test tone (used above) to all of the subs combined via MiniDSP, and you should have an SPL reading at the MLP of at least the 72dB that the first sub put out. If the subs align well the combined output reading at the MLP should be higher than 75dB, if it isn't, then this means that they are not YET time aligned. This is what MSO will correct when it works its magic. So you don't need to try to manually adjust the phase on the subs. Simply leave them as they are, and allow MSO to calculate the delay requirements.

(***Note: As you combine gain-matched subwoofers that are time aligned properly, you will see a 2-3dB gain with each additional subwoofer. So in theory if you have 3 subwoofers, their combined output when measured at the MLP should end up close to 77dB).

Now that you have an "optimized subwoofer" you can run your AVR auto-calibration and it will adjust the gain and distances for the sub channel so that it integrates well with the Main speakers.

Good Luck.

 

[alexf40]

I've just recieved my 2x4 HD this morning, so will try and get around to doing it all again at the weekend. As far as I could tell I was doing anything wrong which would have caused such significant differences between the prediction and measured. My last attempt was much better but i wanted to give it a go with the HD to see if it's partly due to the limited voltage output.

I'm learning a lot as I go along, but really want to get this process ironed out before I finish my DIY subs. Which are taking far longer than planned.

 

[Beast722]

I know I still have a lot to learn and I'm sure I've done stupid things here. I have measured my subs, run MSO, and everything sounds great. The issue I'm having is occasional clipping. I am a bassaholic and love everything loud. What I don't know is if I am pushing my system harder than it can handle (amp clipping), or if I have done things in my chain to cause a clipped signal that I can correct.
My chain is:
Emotiva XMC-1 (running Dirac Live Full)
MiniDSP 2x4HD
2x PSA s3600i
1x SVS 20-39 PC+

My system is in a 16x16 room on a slab and the bass is amazing - i love it. What I don't love is when the most demanding low bass scenes happen - one of the s3600i will clip - loudly and very scarily. Other than for the briefest of moments, all is fine. Based upon what I know about the design of the s3600i, it shouldn't be possible for them to bottom out so it must be a clipped signal or the amp clipping. If it's the latter, I just need to turn it down. If it's the former, I'd like to fix it so I can turn the volume up without worry of blowing up my sub.

My current setup:
REW measurements with no EQ/DSP
MSO with a house curve:
20 10
30 8.6
50 7.1
60 5.7
70 4.3
80 2.9
90 1.4
100 0
Dirac with the house curve below:
10 -5
20 -5
25 -5.66
30 -6.23
35 -6.7
40 -7.15
44.9 -7.5
55 -8.15
63 -8.6
70 -9.0
75 -9.2
85 -9.6
100 -10.25
140 -12
200 -13.3

Any help/suggestions you can provide are greatly appreciated!