• Welcome to ASR. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Dedicated HT build: 4 sealed 18" subs + MSO, HTPC-based – multichannel DAC and object-based audio questions

pablitho

Member
Joined
Sep 16, 2023
Messages
50
Likes
4
Hi everyone,
I'm planning a dedicated home theater in an existing room and I'd like to get input from the ASR community before buying anything.

ROOM AND CONSTRAINTS- Dimensions: 3.0 m W x 5.8 m L x 2.7 m
Walls: drywall layer mounted on masonry, polyester fiber in the gap; suspended drywall ceiling (10-20 cm void above)- Single row, 2-3 seats- 110" 16:9 acoustically transparent screen covering the full front wall; a baffle wall is planned.

Budget for the whole audio system: roughly 15-30k EUR.

WHAT IS ALREADY DECIDED- 4 sealed subwoofers, BMS 18N862 drivers in 55 cm cubic enclosures.
Two at the front (under the screen), two at the rear (~1 m behind the seats), subs must stay on the floor.-
Pro amplification: Powersoft for the subs, fronts behind the screen, still undecided (considering JTR Noesis 212RT, or some cinema pro speakers.
Source: Windows HTPC with JRiver Media Center + madVR- Planned layout: [e.g. 7.4.4 / 5.4.4 – to be defined]

GOALS: My main goal for the low end is a smooth response across the 2-3 seats and above all, better control of modal decay (sealed subs chosen for transient response). My plan is to use MSO for the 4 subs.
I understand MSO mainly reduces seat-to-seat variation and modal peaks rather than "removing"room modes, so I'd like to hear what's realistic in a room this size.

QUESTIONS:
1
. MSO with 4 subs - Is the front/rear placement sensible for this room, or would you change it?
Sub-only or subs+mains configuration? Would you allow all-pass filters? How many measurement positions for a 2-3 seat single row?

2. Multichannel DAC for the HTPC - I'd like a USB DAC/interface with at least 16 analog outputs, connected directly via USB (no bridges).
Currently looking at the MOTU 16A (2025). Which one would you choose today, and are there independent measurements I should look at?

3. Object-based formats with an HTPC. As far as I understand, JRiver decodes TrueHD Atmos only as the 7.1 bed, so no height objects. How would you handle Atmos/DTS:X in an HTPC-centric system? Is there a software solution, or is bitstreaming to a processor (e.g. miniDSP Tide16) the only realistic route?

4. HTPC + DAC vs dedicated processor. I'm weighing three architectures: HTPC + 16ch DAC (full DSP freedom, FIR, MSO), miniDSP Tide16 (Atmos + Dirac ART).
With 4 subs and decay control as the priority, which would you pick? Does MSO make sense alongside Dirac Bass Control/ART, or is it one or the other?

5. Gain staging. Any advice on matching DAC output level to Powersoft input sensitivity to minimize hiss with high-sensitivity cinema speakers?
 
The room is tiny for this amount of low frequency output. Also, it looks likme you will need earplugs.
 
Thanks for the feedback. To clarify the reasoning: the goal of 4 subs is not maximum SPL but distribution. More sources in different positions give MSOmore degrees of freedom to reduce seat-to-seat variation and modal peaks.Playback level is set by calibration (reference at most, often below),so the system won't be louder than with one sub, but each driver will workat a fraction of its excursion, with lower distortion and plenty of headroom.

That said, I agree the room is small (~47 m3, first axial modes around30/57/64 Hz), so pressure gain below ~30 Hz will help. The walls are drywallover an air gap on masonry, so I expect it to be somewhat leaky at very lowfrequencies rather than a rigid box.
 
I would eather buy Minidsp Tide 16. Too many problems with multichannel htpc past 8 channels
 
Thanks, that's helpful. The Tide16 is high on my list for exactly that reason:as far as I understand, JRiver only decodes the 7.1 bed of TrueHD Atmos, soanything beyond 8 channels on an HTPC is upmixed rather than native objects.When you mention problems past 8 channels, do you mean driver/ASIO issues,Windows configuration, or the lack of object decoding?

Have you used aTide16 yourself?
My current idea would be a hybrid: keep the HTPC (JRiver + madVR) as thesource and bitstream to the Tide16 for decoding, Dirac and DAC.

One more question for anyone with hands-on experience: with 4 subs, canDirac Bass Control match the precision of MSO in smoothing modal peaksacross seats? And since my main goal is modal decay, does ART actuallyreduce decay times in a measurable way (e.g. waterfall or RT60 below 100 Hz)compared with an MSO-optimized setup? Has anyone compared the two on thesame system?
 
Haven't tried MSO myself, but I have learned a thing or two by calibrating my system and watching home theater designers on Youtube. MSO does a a great job optimizing frequency response, but does so at the cost of time alignment. Result is loss of the impact, everything sounds "muddy", even if the graph if very pretty. Learning to use wavelet spectrogram in REW was a bit an eye opener. There you can see, how your subs and mains combine through the whole bass range. I think this is one of the first things you need to grasp, if you're after tight and fats bass. Ben Goff and Ryan Charpentier talk regularly about this.
 
Thanks, that's a very useful point.
I'll definitely use the wavelet spectrogram and group delay in REW to verify any optimization, not just the magnitude response.
My understanding is that MSO works with phase-aware measurements but optimizes the summed magnitude, without explicitly constraining group delay.
So I'd expect the "muddy" results to come mostly from all-pass filters or large delays.

My plan is to constrain MSO (no all-pass, limited delays) and check the result in the time domain before accepting it.For those who have used MSO: does that match your experience? And do youhave specific Ben Goff or Ryan Charpentier videos on sub/main time alignment you'd recommend?
 
Thanks, that's helpful. The Tide16 is high on my list for exactly that reason:as far as I understand, JRiver only decodes the 7.1 bed of TrueHD Atmos, soanything beyond 8 channels on an HTPC is upmixed rather than native objects.When you mention problems past 8 channels, do you mean driver/ASIO issues,Windows configuration, or the lack of object decoding?

Have you used aTide16 yourself?
My current idea would be a hybrid: keep the HTPC (JRiver + madVR) as thesource and bitstream to the Tide16 for decoding, Dirac and DAC.

One more question for anyone with hands-on experience: with 4 subs, canDirac Bass Control match the precision of MSO in smoothing modal peaksacross seats? And since my main goal is modal decay, does ART actuallyreduce decay times in a measurable way (e.g. waterfall or RT60 below 100 Hz)compared with an MSO-optimized setup? Has anyone compared the two on thesame system?
Both. Lack of atmos decoding and extra channels. Although I have 6 subs due to 8 channels in my Okto and Windows in general I am restricted to 4 at the moment. The other 4 channels used are front left/right and surround left/right. I am looking forward to buying Tide 16 for that reason.
 
You could get similar results using good dual opposed 12" like the SB Acoustics SB34PWL76 in 40cm cubes and save some room. I Modelled both your suggestion and the 12"s in the 00 sim and the results were surprising.
 
Both. Lack of atmos decoding and extra channels. Although I have 6 subs due to 8 channels in my Okto and Windows in general I am restricted to 4 at the moment. The other 4 channels used are front left/right and surround left/right. I am looking forward to buying Tide 16 for that reason.
Thanks, that clarifies it. So in your case the channel limit comes from the8-channel Okto plus Windows' 7.1 speaker configuration. With a 16-channelASIO interface the channel count wouldn't be the issue, but the lack ofAtmos object decoding still is, which is the real deal-breaker for me too.

Good luck with the Tide16. If you get it, I'd be very interested in yourexperience with Dirac Bass Control and ART on 6 subs.
 
You mention decay control, in that case my understanding is you need Dirac ART, other stuff only controls decay in the sense of lowering the level of peaks to get them below the noise floor sooner, not actually canceling reflections. If it were me with this budget I'd go for the Tide for are. I haven't tested any of this myself but if you look at the various before and after screenshots people post of ART it's definitely a level up.

Also as an alternative to the Noesis you might want to look at the Ascilab c8c, maybe with bass stand, they have really excellent measurements and are close in price. It's not a horn so it won't go as loud, but I feel like you probably don't need 4x 12" + 4x 18" in that size of room.
 
Last edited:
You mention decay control, in that case my understanding is you need Dirac ART, other stuff only controls decay in the sense of lowering the level of peaks to get them below the noise floor sooner, not actually canceling reflections. If it were me with this budget I'd go for the Tide for are. I haven't tested any of this myself but if you look at the various before and after screenshots people post of ART it's definitely a level up.

Also as an alternative to the Noesis you might want to look at the Ascilab c8c, maybe with bass stand, they have really excellent measurements and are close in price. It's not a horn so it won't go as loud, but I feel like you probably don't need 4x 12" + 4x 18" in that size of room.
Thanks, very useful. Agreed on ART being the only active approach to decayhere. That's a big reason the Tide16 is now my front-runner. One nuance:at low frequencies modes are often minimum phase at a given seat, so EQthere also shortens decay, not just the level. The problem is that it onlyworks well at one position, which is where ART (or 4 subs + MSO) should help.
 
It seems like you are heading in the right direction, but I'll be blunt in hopes of removing any doubt :-)

My understanding is that MSO and DLBC do roughly the same thing, you don't use them together, and they don't do much to control LF decay.

Controlling decay below 150 Hz with absorption takes VERY thick panels. So use ART for controlling decay below 150 Hz, and absorption for first reflections and decay above 150Hz. Personally, I'm incredibly happy with the upgrade from DLBC to ART. When you first hear the room with much of the LF ringing removed, it sounds like you accidentally loaded a target curve with reduced LF. Then you increase the LF in the target curve from +6dB to +9dB, then to +12dB, then to... and you bask in the glory that is four big subs and ART.

You have the first length mode at 30Hz, which also means you have a second length mode at 60Hz, which stacks up with the modes at 57 and 64Hz. I think your room dimensions strengthen your case for four large subs and Dirac ART.
 
I'm planning a dedicated home theater room my self, I will go with four 12" subs (each in a closed shallow box, scan speak driver) in an SBA configuration behind an AT screen with an absorber in the back.
 
It seems like you are heading in the right direction, but I'll be blunt in hopes of removing any doubt :-)

My understanding is that MSO and DLBC do roughly the same thing, you don't use them together, and they don't do much to control LF decay.

Controlling decay below 150 Hz with absorption takes VERY thick panels. So use ART for controlling decay below 150 Hz, and absorption for first reflections and decay above 150Hz. Personally, I'm incredibly happy with the upgrade from DLBC to ART. When you first hear the room with much of the LF ringing removed, it sounds like you accidentally loaded a target curve with reduced LF. Then you increase the LF in the target curve from +6dB to +9dB, then to +12dB, then to... and you bask in the glory that is four big subs and ART.

You have the first length mode at 30Hz, which also means you have a second length mode at 60Hz, which stacks up with the modes at 57 and 64Hz. I think your room dimensions strengthen your case for four large subs and Dirac ART.
Thanks a lot, bluntness appreciated! Your point about the 2nd length mode at ~60 Hz stacking with the width and height modes matches what I found in a simple modal simulation of the room: the 55-70 Hz region was the one I couldn't fix.

I'm at a crossroads between two approaches:
- Acourate (linear-phase FIR on the HTPC) with the 4 subs in a DBA, which a friend experienced with Acourate recommends
- the Tide16 with Dirac Live + ART, which already includes the ART licence

Since you've made the jump from DLBC to ART, a few questions if you don't mind:

1. What is your setup: processor, number of subs and where they are placed, room size?
2. Do you have before/after measurements (waterfall or decay) you could share?
3. Did you ever compare ART with a DBA, or with FIR correction like Acourate or Audiolense?
4. Does ART care about sub placement? Is 4 subs at 1/4 and 3/4 of the width, two front and two back, a good layout for ART?
5. How much headroom did ART cost you, and do your mains take part in ART or are they crossed over to the subs?

Thanks again!
 
Haven't tried MSO myself, but I have learned a thing or two by calibrating my system and watching home theater designers on Youtube.

Okay, so you haven't used it yourself, so any subjective observations you might make about it are someone else's.

MSO does a a great job optimizing frequency response, but does so at the cost of time alignment.

Did you get this information from Home Theater Designers YouTube channel? If so, can you provide a link to the specific video in which this is claimed?

Version 2 of MSO, which came out in May of 2024, introduced SPL maximization, which aims to match the acoustic phase of the subs at each listening position. When used without all-pass filters, this is equivalent to what's often called "time alignment": that is, what you might obtain manually by using the REW alignment tool to maximize the output by adjusting sub delays while observing the summed response. An entire page of the user's guide is devoted to explaining this and how it came about. Also, in the "Tips and Tricks" section of the manual, there is a dedicated topic titled Getting Optimum Delays With Sub-Only Configurations, which expands on the idea further.
 
Last edited:
Okay, so you haven't used it yourself, so any subjective observations you might make about it are someone else's.
Correct.
Did you get this information from Home Theater Designers YouTube channel? If so, can you provide a link to the specific video in which this is claimed?
Also correct. They covered this topic multiple times, one I can remember is in this video at around 45min mark:

Also, they are not the only ones making similar observations, that's why I took notice.

To be clear - as I've stated, I have no personal experience with MSO and much of this stuff is above my head. But I have some experience calibrating and then evaluating.
I have achieved very pretty looking frequency graphs, subs flat down to 20-25hz, but in the end I just couldn't enjoy them - everything sounded "off". And I'd have spent hours and hours adjusting delays, etc.

Then, while trying something else, I hooked up one inferior sub, no EQ at all, just the delay was adjusted and it just sounded more "right". Had no idea what was going on - frequency response was not bad, but clearly worse than in the system before.

Then I've watched this video:
(around 35min mark)
Also found this one:
(from ~25min)

And then went back to check my measurement result, and sure enough, the simple subs was way better incorporated in time domain, while multiple subs with heavier EQ were scattered.

I'm not say "do this" or "do that", and MSO might have the tools to make everything right, but, at least for me, this whole time alignment thing was very revealing. So I just share my experience of what I've ancountered.
 
Thanks a lot, bluntness appreciated! Your point about the 2nd length mode at ~60 Hz stacking with the width and height modes matches what I found in a simple modal simulation of the room: the 55-70 Hz region was the one I couldn't fix.

I'm at a crossroads between two approaches:
- Acourate (linear-phase FIR on the HTPC) with the 4 subs in a DBA, which a friend experienced with Acourate recommends
- the Tide16 with Dirac Live + ART, which already includes the ART licence

Since you've made the jump from DLBC to ART, a few questions if you don't mind:

1. What is your setup: processor, number of subs and where they are placed, room size?
2. Do you have before/after measurements (waterfall or decay) you could share?
3. Did you ever compare ART with a DBA, or with FIR correction like Acourate or Audiolense?
4. Does ART care about sub placement? Is 4 subs at 1/4 and 3/4 of the width, two front and two back, a good layout for ART?
5. How much headroom did ART cost you, and do your mains take part in ART or are they crossed over to the subs?

Thanks again!

Sure, happy to give any input! With that said, my experience is admittedly rather limited, my objectives and preferences are different, and my room is very different. I've done my best in the past several months to learn about small room acoustics, speakers and subwoofers, treatment, and DSP. I'm primarily interested in music, usually with a very healthy LF boost.

My room not ideal to say the least. It is an odd shape and is full of openings. It's also my living room, so aesthetics and speaker/sub placement are very much a consideration.

1792509213867.png


-Tide16
-4x Starke Sound SW15 subs
-Buchardt A200 mains running the cardioid tune (cannot treat front wall due to TV and space)
-2ft wide by 6ft tall absorption panels effective down to ~125Hz on left and right wall first reflections

Less than I year ago I ran a simple 2.1 system with Dirac Live RC (no BC or ART). Then I started learning and experimenting. My current speakers and subs are sufficient for the levels I previously listened at, but I now realize that I never liked to listen loud because the sound was terrible. With multiple subs, ART, absorption panels, and the A200 mains limited to 150Hz, I now listen much louder and love it! I'm currently building new subs (4x DIY sealed subs, each with dual opposed Dayton SS15-22 drivers and 2000W) and planning new speakers to allow much more output and lower distortion.

I have measurements, but nothing worth sharing. Part of my learning process has been learning how to measure in a consistent manner and not change too many variables at a time. I can only give you my subjective experience. I set up the fours subs and the Tide16, did the Dirac measurements, loaded a BC filter with a +6dB LF lift. I listened for a couple hours and it was a stunning improvement over anything I previously had. Then I loaded an ART filter also with a +6dB LF lift, and it sounded like there was LESS LF output. Then I changed to a +9dB LF lift and turned the volume up... I was in love. ART removed enough LF ringing to make a very obvious difference to me, and I very much prefer it.

I have no experience Acourate or Audiolense. I have no experience with single or dual bass arrays, but I think I have a correct high-level understanding of how each of them works. I didn't pick up that you were considering a DBA when you mentioned the HTPC option. Yes, my understanding is that ART, Trinnov Waveforming, a DBA, or an SBA with rear absorption are all valid LF decay control methods. I can't speak to which is better generally, or better in your specific situation. Given my room, I think ART is the only option I have.

If I had a dedicated room that was rectangular and not full of openings, I would give serious considerations to the ART alternatives. Do you ever visit the AVS DIY subforum (https://www.avsforum.com/forums/diy-speakers-and-subs.155)? Plenty of guys on that forum that have experience with large DIY subs, DBAs, SBAs, and ART.

I like the simplicity of the Tide16 and ART. The Tide16 also looks nice in my living room. If I had a dedicated room, I'd me more inclined to deal with the extra complexity of the HTPC option. ART is now also available for PC and MacOS. Maybe you build a HTPC and try out ART along with other DSP/DBA options you are considering? Since you have a dedicated room, can you accommodate an SBA (all subs up front with specific SBA placement, and a false rear wall filled with insulation/absorption)?

My understanding and limited experience is that ART mostly cuts, and does not try to fill in deep nulls. So yes, sub placement is still a big factor. That said, I get the impression from my reading about ART that it allows for more flexible sub placement than Waveforming, DBA, or SBA. In my less than ideal room with my sub placement I still have a null in the upper 40's, but it's not very deep. I need to test if placing a sub in one of the ceiling corners would take care of the null. Or maybe the additional output from the DIY subs will help ART to deal with it.

Yes, I would think 4 subs at 1/4 and 3/4 of the width, two front and two back, would be a good layout for ART. I think those positions will properly excite two of the room modes and prevent deep nulls that ART cannot fill.

I don't think I can quantify a headroom reduction with ART. There is the obvious output reduction due to the subs/speakers also outputting out-of-phase sound to reduce LF ringing. Maybe a larger concern is ART having to do lots of cutting to bring down peaks, so as to not have to boost nulls too much. I will say that I'm taking the more is better approach with my DIY subs (8x 15" drivers with lots of excursion and power).

My ART settings are atypical, as I currently have tiny main speakers. To reduce distortion in the mains I set the ART F Support Low value to 149Hz for the mains and the ART F Support High value to 150Hz for the subs. ART does not have a mains to sub crossover like Dirac BC, but I've essentially set ART to do such a crossover. This works as my front subs are co-located with the mains. With larger speakers you would typically you would allow subs to run up to 150Hz and your speakers to run down to a frequency that is appropriate for them, 50Hz for example. In that case ART is using both the subs and mains in that 50-150Hz region to optimize the sound. I don't have experience with multiple larger speakers contributing below 150Hz, but I've read folks reporting excellent results doing so.

If you are interested in infrasonic playback, make sure you read up on the challenges of infrasonic with ART. This honestly might be a deal breaker for ART for you.

There are lots of folks complaining about various things about the Tide16. You can't please everyone, but it might be worth reading the various Tide16 threads to make sure none of those are sticking points for you.

Sorry, that was lots of words, not many clear answers, and no measurement ;-) Hopefully some of it was somewhat helpful.
 
...at least for me, this whole time alignment thing was very revealing.

Floyd Toole has discussed subwoofer "time alignment" elsewhere. Here is a post of his in which he brings that up. I'll quote the relevant portion below.
Floyd Toole said:
Multiple subwoofers are now accepted as being useful, but not all schemes are equally effective.
Some “room EQ” algorithms, and some DIY persons, time align the first arriving, direct sound, from each of multiple subwoofers. The goal is to ensure delivery of a synchronized direct sound. But:
1. In small rooms the direct sound is swamped by reflected sound – it is not the dominant factor in what we hear.
2. Time domain distortion of low-frequency transient sounds is dominated by the build-up and decay of room resonances, so the first objective should be to attenuate room resonances, which also exhibit frequency response peaks – what we hear is dominated by the spectral peak, not the ringing (Section 8.3 in the 3rd edition of my book). What matters is how the subwoofers couple with the room modes when all are operating simultaneously. This requires understanding the superposition of the transfer functions (amplitude and phase) at the listening location.

That post was written before the 4th edition of his book came out. The relevant section corresponding to 8.3 of the 3rd edition is 14.5.11 in the 4th edition.

The purpose of SPL maximization in MSO is to do what is often referred to as "time alignment", not to synchronize early-arrival sounds, but to ensure the subs combine as efficiently as possible. The second purpose is to ensure that each sub is driven equally hard relative to its capability, so the SPL capability is not limited by the weakest sub. Getting that right requires extra effort when using dissimilar subs. SPL maximization is not something that's done after the fact as a band-aid. It is an integral part of multi-stage optimization, which is the recommended approach for the current version, and has been since 2024.
 
Floyd Toole has discussed subwoofer "time alignment" elsewhere. Here is a post of his in which he brings that up. I'll quote the relevant portion below.


That post was written before the 4th edition of his book came out. The relevant section corresponding to 8.3 of the 3rd edition is 14.5.11 in the 4th edition.

The purpose of SPL maximization in MSO is to do what is often referred to as "time alignment", not to synchronize early-arrival sounds, but to ensure the subs combine as efficiently as possible. The second purpose is to ensure that each sub is driven equally hard relative to its capability, so the SPL capability is not limited by the weakest sub. Getting that right requires extra effort when using dissimilar subs. SPL maximization is not something that's done after the fact as a band-aid. It is an integral part of multi-stage optimization, which is the recommended approach for the current version, and has been since 2024.
I will not dispute any of that.

My understanding, simplified:

Subwoofer response manipulation through various heavy handed EQ methods increases group delay (I hope it's the right term in this case) - energy peak in time becomes scattered through different frequencies. I've seen that shown on Youtube, I've seen that in my own measurements, I've seen that in other post of the members on this forum. My understanding is limited, and maybe there are other ways to see/tackle the same issue (I haven't seen one yet), but wavelet spectrogram in REW is very easy way to visualize the issue and see, what EQ does to bass/time combo.

If there is a way to take correct steps in MSO to ensure proper time alignment through the whole range - great. I'd like to see that.

Right now I'm in the process of redoing my whole setup and will go DBA or ART route, so I will only invest time in MSO if those two fail miserably somehow.
 
Back
Top Bottom