Thank you!
This is good to know. I will continue to explore adding two sealed subs.
One way you might be able to implement this without more channels could be with a miniDSP 2x4HD and Multi Sub Optimizer (MSO). This is what I'm exploring. The R&L sub channels are sent into miniDSP and processed with filters generated with MSO to output 4 channels (one for each sub) that minimize variations in total response (like Toole's book Sound Reproduction). I'll use my two subs to take measurements in 4 locations and simulate in MSO the total response and decide if worth adding two more subs - most likely it will be worth it, per so many references of users with 3 or 4 subs.
Your explorations into OB midrange/treble, their xo and dispersion patterns are very interesting. I'm doing MTM with a large AMT tweeter and 8" mids and have a fair amount of testing to do. Need to re-read your explorations and findings.
The extra subs give the option to improve mean spatial variation, so the sound around the room will be more even. If you want to have good bass in multiple positions this is a good choice. If you always sit in the same seat or a small area then improving spatial variance doesn't matter so much and adding more subs might not work any better, as in Welti's research two subs had the highest output, going to 4 reduced the maximum output when optimizing for improved spatial variation.
Thank you. I was aware of the spatial variation piece, but was intrigued by the reduced output with 4 vs 2 subs. Is this because 4 subs optimized for minimum variations have an average output smaller than 2 unoptimized subs, but the response from the two subs is likely to be uneven so while perceived as louder they are also likely to have some peaks and nulls in their response?
Todd Welti's words
"Obviously, wall midpoint locations result in optimum room response, based on the std anyway. Not surprisingly, symmetrical configurations seem to work better than non symmetrical ones. Four subwoofers results in the most symmetrical configuration and the best results, but with significantly less (normalized) low frequency output than two subs. Time did not permit full optimization using more than four subwoofers."
https://www.harman.com/documents/multsubs_0.pdf
"Obviously, wall midpoint locations result in optimum room response, based on the std anyway. Not surprisingly, symmetrical configurations seem to work better than non symmetrical ones. Four subwoofers results in the most symmetrical configuration and the best results, but with significantly less (normalized) low frequency output than two subs. Time did not permit full optimization using more than four subwoofers."
https://www.harman.com/documents/multsubs_0.pdf
I should draw attention to the word normalized in the quote above. You might get more output with 4 subwoofers than 2, but it will be significantly less than expected from doubling the number. The graphs and slides around the quote on page 18 might help explain it better.
Thanks for sharing. I've gone through the paper and Welti doesn't seem to introduce the use of delaying signals to different subs. Floyd Toole describes this in Sound Reproduction and explores different subs configurations (1, 2, 4 placed in corners, midwall, etc) without and with electronic sound field management, and basically concludes it absolutely works and with 4 subs their precise location is not as critical.
Indeed, what is minimized is the seat-to-seat variation, but from running the MSO tutorial I see the optimized response also makes the variation at each seat very significantly smaller than before the correction. Can't tell how representative is this simulation of what will happen in the real world, though. Worth exploring though.
The information I posted came from a presentation because it was easier to find the specific reference to 2 vs 4 subs and output.
The actual paper is here in case you haven't seen it
https://audioroundtable.com/misc/Welti_Multisub.pdf
Everything Floyd Toole describes comes from the above and related experiences. Sound field Management is covered in the paper and MSO is using a very similar process.
The point I was trying to make is whether reduced spatial variation is important to you in your situation. Do you listen in one chair only or all around the room or with multiple listeners?
In a video interview Welti covers some of that by saying that if you listen in the one spot you can easily put a single high output subwoofer in a room corner to excite the maximum modal response and then equalize it flat at the listening position.
If your listening position is in a difficult spot modally within the room then placing the subwoofers to avoid creating nulls in that region would be worthwhile even for a single position. Use the information and apply it to your situation to see what approach would work best.
The actual paper is here in case you haven't seen it
https://audioroundtable.com/misc/Welti_Multisub.pdf
Everything Floyd Toole describes comes from the above and related experiences. Sound field Management is covered in the paper and MSO is using a very similar process.
The point I was trying to make is whether reduced spatial variation is important to you in your situation. Do you listen in one chair only or all around the room or with multiple listeners?
In a video interview Welti covers some of that by saying that if you listen in the one spot you can easily put a single high output subwoofer in a room corner to excite the maximum modal response and then equalize it flat at the listening position.
If your listening position is in a difficult spot modally within the room then placing the subwoofers to avoid creating nulls in that region would be worthwhile even for a single position. Use the information and apply it to your situation to see what approach would work best.
Welti & Devantier paper and study ignores vertical modes... I have found it a very difficult issue, because a sitting person has ears at near half of the (small) room height, so a strong null is unavoidable. I even had one sub on top of the bookshelf in my HT, without help. One at half height near listener would help...
Here my HT in stereo (L/R measured separately) with a single sealed 2x15" sub, minidsp 2x4HD LR2 xo at 100Hz. Upper lines at ear height when sitting and lower curves at butt height on the sofa...
Here my HT in stereo (L/R measured separately) with a single sealed 2x15" sub, minidsp 2x4HD LR2 xo at 100Hz. Upper lines at ear height when sitting and lower curves at butt height on the sofa...