Shallow speakers for home theater

[babinsky]

Hi all,
For my first post, I would like to first thank for all the contributors to this forum. All the info in here is gold, thanks!
Also, sorry in advance for the length of the post.

Now what brings me here. I would like to setup a home theater in the coming year, and would love to have shallow speakers, that can play nicely above 80Hz (I will use subwoofers below that).
My requirements for the speakers:
1. shallow speakers, ideally 10cm or less. I don't care about height/width though.
2. wide enough dispersion for all listeners to comfortably enjoy the movies.
3. SPL wise, I am targeting around 80-75 dB average for each speaker at each seat, and some headroom of ~15 dB SPL.
4. All speakers will be eq-ed (if that is a word) using dirac/audyssey. So I am not necessarily looking for a specific frequency response profile out of the box, just enough distortion-free output and eq-ability.
5. Easy enough to build for a first set of speakers. I am a decent woodworker, but have no experience with speaker building (e.g. no knowledge of crossovers).
6. I am looking at building 11 speakers and 4 subwoofers. So cost wise, would be nice to keep it under 150 per speaker.
7. All speakers will be on-wall.

I looked at commercially available options, and it seems they are all out of reach pricewise (e.g. Grimani systems Rixos-S and Ascendo 6).

Looking around the forum, it seems there are quite a few positive review of Markaudio drivers, and they seem to fit the bill, however I did not find reference builds there with a "shallow enough" design. I hence tried the following design under winISD:

• CHR-90 or CHP-90 for front speakers. Those are the farthest from listeners and would require the most output and smallest angle to cover. These drivers seem efficient enough and should provide the SPL levels / dispersion desired.
• CHR-70 or CHP-70 for other speakers. As those are closer, so they will need less output but need to cover wider angles. IIUC, smaller fullrange drivers will start beaming at higher frequencies right? Don't know if those are enough dispersion wise though.
• sealed cabinets: it seems those drivers are happy in a sealed enclosure when no bass extension is needed.
  For the CHR-90, a 7L box has an f3 at 93Hz and an f6 at 70Hz. With 30Watt, it seems to produce enough SPL with 95 db at 3m (and still some headroom left for power and cone excursion). I am thinking this should be enough for my needs, assuming I will also eq them (so some headroom is required I guess), and putting them on-wall should give some gain.
  For the CHR-70, same principe. It seems a 5L box produces the desired level of SPLs (about same f3/f6 as the above).
• Dimensions: I tried to use irrational ratios (square roots of primes) to get nice shapes. For example, for a 5.5L box I get 26.4cm width, 5.5cm depth and 37.9cm height internal. This is very shallow and I wonder if there could be issues with back panel too close to the driver? Also, what is the advice for such a box wrt bracing?
• Driver placement: I have not read much about driver placement. Would it make sense to just put it a 1/3rd height, and at center wrt width?

None of the above is fixed (driver choice, dimension, ...) and I am open to suggestions.

Here are extra info on the room, speaker/listeners placement, and some computations I made, if that helps :
- the Room is 5.3m x 3.7m (black spots are seats, blue box is couch, red boxes are speakers, red cones cover 40 degrees angles -- ignore the surround glitch, will fix my script)

- Here is a table of speaker/listeners distances (in meters), and the associated attenuation if that helps

Sorry for the long post... and thanks in advance!

 

[Mooly]

Test. Ignore.

 

[hm]

for such a room size IMHO its overkill. 5.2 would be enough, i would take five FLUTE Sats an two subs, http://www.hm-moreart.de/12.htm

 

[mattstat]

I think your output estimates are optimistic for the front speakers. A 3.5 inch driver is small for typical home theater fronts unless your listening levels are modest. Displacement limited output at 80 Hz seems low to me, but I'm not sure what all your assumptions are. Given your stated output goals, I would go for the biggest drivers (volume displacement) that will fit within the depth and price requirements. If you don't have experience with the Mark Audio drivers, I would also start off by building one test speaker and seeing if it does enough for front use or for your other positions.

Have you looked at the Parts Express kits? For lower priced speakers, they tend to be liked. The documents give all the components and schematics required, so you could skip the full kit, buy the parts, and build your own boxes. I'm not sure what the prices/availability are like in France though.

https://www.parts-express.com/speaker-components/speaker-subwoofer-kits

As long as the driver depth doesn't interfere with your 10 cm max target, the enclosure shape can be pretty much whatever you want as long as the volume is the same (for a normal ported or sealed enclosure). Doing that and putting them on the wall will both affect frequency response, but if you are going to EQ them anyway, that's not a big concern in my opinion.

 

[babinsky]

Thanks for the replies

I think your output estimates are optimistic for the front speakers.

Could you help with a more realistic approach? For the attenuation formula I used -20 * log(distance). This also seems to align with winisd simulation (see below).

A 3.5 inch driver is small for typical home theater fronts unless your listening levels are modest. Displacement limited output at 80 Hz seems low to me, but I'm not sure what all your assumptions are. Given your stated output goals, I would go for the biggest drivers (volume displacement) that will fit within the depth and price requirements.

That was exactly my plan. I did a few simulations using winsid, here is what I got for CHR-90 (7L), CHR-120 (14L) and CHN-110 (9L). The signal was set to 3m, and wattage was max possible without exceeding xmax, and documented nominal power. This is what drew me to CHR-90 (even though CHN-110 is still tempting).

If you don't have experience with the Mark Audio drivers, I would also start off by building one test speaker and seeing if it does enough for front use or for your other positions.

That was exactly my plan, but wanted some feedbacks before starting the build (especially on driver selection).

Have you looked at the Parts Express kits? For lower priced speakers, they tend to be liked. The documents give all the components and schematics required, so you could skip the full kit, buy the parts, and build your own boxes. I'm not sure what the prices/availability are like in France though.

Yep looked at that thanks. It seems those are mostly for US market and shipping to France is costly. Even the parts used are more expensive here. Also I did not find much about high enough SPL, shallow speakers (mostly small desktop boxes or larger boxes/towers).

Regarding already existing builds, there are a bunch on Markaudio website and here, but non completely filled the depth requirement. The best match I found was these but WinISD simluation shows it would not be enough output for front speakers.

 

[babinsky]

for such a room size IMHO its overkill. 5.2 would be enough

Agree, this is mostly the end-game plan 🙂. For start I will likely do 5.2 and upgrade progressively from there. For subs I have plans based on this but this is for another thread I guess 🙂 .

i would take five FLUTE Sats an two subs, http://www.hm-moreart.de/12.htm

Do I understand correctly this is 45cm deep?

 

[mattstat]

Could you help with a more realistic approach

The distance calculation/attenuation table seemed OK. It's the maximum output for the volume displacement that seems optimistic. Rough calculation I ran earlier was for 94 dB at 80 Hz and 1 meter for 85 cm Sd and 7 mm one way travel. In a more involved modeler, I'm getting 100 dB at 1 meter, 30 watts (though I think that may be a full space vs. half space difference between the two). Subtracting 12 dB for your distance brings that down to 88 dB at the listening position. That's a bit short of your 90-95 dB peak level at the listening position. I don't use WinISD, so I'm not sure where the discrepancy between my estimates and yours are coming from.

Part of my opinion is also based on personal experience with similar sized drivers and how loud they can play without strain. I don't have direct experience with the Markaudio ones, so I'm not sure if they're out of gas at Xmax or if they can play cleanly beyond that. Some drivers do, some don't. I'm sure someone with more direct experience with those will join the discussion.

 

[babinsky]

I see thanks. So I should take WinISD simulation with a bigger grain of salt then 🙂. In any case, Do you still think this is not enough since 80Hz will be the XO freq with subs? If so, would you think CHN-110/CHR-120 would be enough (they have bigger Vd) ?

 

[mattstat]

It really depends on how hard you are going to push things. Small drivers can do remarkably well for what they are, but for dynamic source material and higher playback levels, larger drivers often provide benefits.

Smaller drivers will typically have better off-axis behavior and high frequency extension/smoothness though.

The 40° coverage cones for the fronts seem adequate in your layout, so the larger CHR-120 may be fine for what you are doing. The directivity shown here seems good to 15 kHz at +/- 20°.
https://www.diyaudio.com/community/threads/markaudio-chr-120-test-and-review.393476/

Just going by volume displacement derived from Xmax, changing from the CHR-90 to CHN-110 should provide 3 dB more output capability. The CHR-120 should be capable of another 4 dB over the CHN-110. From an output standpoint, the CHR-120 seems like the better fit for what you want to do on the front speakers. Until you build one and give it a listen, it's all theoretical though.

 

[babinsky]

It really depends on how hard you are going to push things.

I actually made the test with my current sound bar by pushing level to about where I liked it / was comfortable with (ignoring distortion). It was hitting peaks of 92 db and was between 75-80 db on average. This is why I mentioned those values above. I assumed that if we don't go over xmax, there should be no excessive distortion, hence why I am trying to hit those marks without exceeding xmax (let meknow if that assumption is wrong).
TO me it seemed overshooting already as winISD curves are "anechoic", and reflection in the room should give higher SPL reading in real room right ?

Small drivers can do remarkably well for what they are, but for dynamic source material and higher playback levels, larger drivers often provide benefits.

This is a good point thanks. Can you share what characteristic I should look for in a speaker driver to have good dynamics? From common measurements, it seems FR, polar map and IMD are measured, but I can't see anything that would tell me how good a speaker is with dynamic material.

The 40° coverage cones for the fronts seem adequate in your layout, so the larger CHR-120 may be fine for what you are doing. The directivity shown here seems good to 15 kHz at +/- 20°. https://www.diyaudio.com/community/threads/markaudio-chr-120-test-and-review.393476/

This is exactly why I picked 40 degrees 🙂

Just going by volume displacement derived from Xmax, changing from the CHR-90 to CHN-110 should provide 3 dB more output capability. The CHR-120 should be capable of another 4 dB over the CHN-110. From an output standpoint, the CHR-120 seems like the better fit for what you want to do on the front speakers. Until you build one and give it a listen, it's all theoretical though.

Thanks for the confirmation. The choice of driver is not limited to markaudio, I just picked those based on their popularity on this forum and because there was some measurements available. I will wait a little if other members have some feedback on driver selection, and will start building test speakers.

 

[perceval]

Other companies that offer drivers available in Europe are FaitalPro and SBAcoustics, the latter even has a French distributor.

Both of them offer a wide range of drivers, and are premium products.
Faitals are more ”pro” oriented, with driver capable of playing loudly and handling power.
SBA is more like a spin-off of ScanSpeak in Indonesia, with very well balanced drivers that are reasonably priced.

I will also second that 3” Drivers for HT is probably not the best approach. Sure, on paper, they look the part, but in real life, they lack the “meat” required. Such a small cone cannot push that much air.

 

[babinsky]

Thanks indeed SBAcoustics is available. For some reason, those drivers are not labeled as fullrange so I missed them. Would you have any recommendation for my use case?

 

[perceval]

They do now have a "full range" one, a coax, a a few wide range as well.

The 2½″ SB65WBAC25-4 is a great performer that extends high, near a tweeter could do. It's really good with percussion, woodwinds and string instruments.

The 3″ SB10PGC21-4 is a very good even performer, even with a peak of distortion at 4kHz, which may or may not bother you.

Some even consider the 4″ SB12MNRX2-25-4 as close enough to a full range. It's a bit rugged on top, but if your ears are not too young, it might not bother you.

Now, being small, they would need woofer help for a bit more punch. SBA have plenty of good ones, 6.5" might be ok for your small space. Especially if adding a subwoofer (or two not placed in symmetry to take care of room nodes).

I did mention a bigger full range, the new 8” SB20FRPC30-8. It works sealed, vented and in a transmission line. Quite the feat! Being 8", it will have more punch than the small ones you are looking at. It will work in a 20-liter sealed enclosure with a Fs of about 80Hz. You can then ask the subs to fill lower. Very reasonable price on these drivers as well.

On the other hand of the spectrum, they also have a new coax driver that looks mighty impressive, the 7.5" Satori MT19CP-8. For about the same size box as the SB20, in a vented design, you will get about 50Hz. I have the TangBand version of this, and I expect the SBA to perform as good if not better. That will be a little pricier to have LCR with the MT19CP !

TangBand is another company will good drivers, but they all need a tweak or two to sound great. I have bought a few, and none was to my liking out of the box. But with some notches here and there (or EQ), they sound quite good.

Would 20-liter enclosures feel too big for you? If not, I'd get the SB20. Put them on stands and around the room. Very good performers and they won't break the bank. Myself, I can't wait to get my hands on a MT19 pair... but that will have to wait a bit. 🙂

Bonne lecture!

 

[babinsky]

Thanks a lot for the recommendations, I will try to model those.

I actually had looked at SB20FRPC30-8 during my initial research phase (in particular this thread), but when I model it in a sealed enclosure it seems to perform similar to the Markaudio CHN-110. It seems to be due to similar Vd for both as SB20FRPC30-8 has twice Sd but half xmax. However, the reason I did not look more into SB20FRPC30-8 is:

• it requires almost 2x cabinet volume for similar output
• I did not find much data for off-axis, whereas for Markaudio drivers I found more data on that (in particular this and this).

 

[Harry72]

Have you looked at the kits at Soundimports?

https://www.soundimports.eu/en/diy-kits/diy-component-packs/

 

[perceval]

• it requires almost 2x cabinet volume for similar output
• I did not find much data for off-axis, whereas for Markaudio drivers I found more data on that (in particular this and this).

Of course the SB20 will need a bigger enclosure.
Edit: Haven't had time to check for the CHN110, but Madisound suggest a 18 to 25 liters enclosure, while I mentioned the SB20 will need about 20 liters. Don't know where this "double the cabinet size" came from....

But the MA has a sensitivity of 88dB, as the SB20 has 92dB.
And maybe you could tell me what you liked about the data from those two sites (which are not the CHN110, by the way).
The two have shown that the data from MA and the measured ones from testers were quite different. Alright, different conditions apply, but the break ups don't show as much on MA's data.
And both exhibit close to 10dBs break up mode, yet, the tester called it smooth?
Like I said, there's no perfect drivers, and my set of ears will be different from yours, as is with expectations.

If you desperately try to stay with a 3 to 4" driver, may I suggest another route?
The TABAQ is a proven design, that offers a very satisfying reach. It can be placed on the floor, right against the wall, it can be hung on a wall, etc...
Paired with a good driver, it might be ok for you. Plenty of drivers have been tested and reviewed on the TABAQ thread.

A mini HT with TABAQs spread around could be nice for limited space.

 

[mattstat]

(ignoring distortion)

That's kind of hard to do in practice. Things sound louder when they are distorted, so you may turn them down because of that. A more capable speaker may inspire you to turn things up more, so having some extra output capability may be beneficial. This is me judging things from a distance and using generalizations though.

trying to hit those marks without exceeding xmax (let meknow if that assumption is wrong).

Generally this is a safe approach. Most drivers will play cleanly to Xmax, as long as the manufacturer is being truthful about it. Some play well beyond it, but it's not always predictable which ones will, and data like this is often lacking.

winISD curves are "anechoic", and reflection in the room should give higher SPL reading in real room right ?

From what I'm seeing online, WinISD is modeling half space, so there may not be as much room gain as you were thinking.

Can you share what characteristic I should look for in a speaker driver to have good dynamics

Part of it is not running a speaker up against its limits. Having headroom in the system is useful all the way around. Sometimes size constraints have to take priority, but they come at a cost. Same thing with amplifier power. So overall, I'd say try to get as much sensitivity, cone area, volume displacement, and headroom across the board as you reasonably can. Some of those things tend to fight each other (like sensitivity and volume displacement, especially in smaller drivers), so it's a balance.

 

[babinsky]

Thanks for all the feedback !

Haven't had time to check for the CHN110, but Madisound suggest a 18 to 25 liters enclosure, while I mentioned the SB20 will need about 20 liters. Don't know where this "double the cabinet size" came from....

This comes from winISD simulation. The higher volume requirement seems to stem from better bass extension. Given that I don't care about bass below 80Hz, I tried to get the smallest possible enclosure while keeping power requirement below documented nominal power and xmax for each driver at 80Hz. In other words, I am trading efficiency for box volume (if that makes any sense).

Here is what I get for CHN-110 (9L green), SB20FRPC30-8 (20L red) and CHR-120 (14L blue). The two Markaudio boxes are limited by power, the SB20FRPC30-8 is limited by xmax.

I could push the SB20FRPC30-8 to be power limited with a 9L box, and I get the following picture (with a steeper roll-off above 100Hz).

So indeed I could get more output above 100Hz from the SB20FRPC30-8 but it seems to not extend to 80Hz as well as CHR-120 in a sealed box (maybe the solution would be a vented box of higher volume?)

But the MA has a sensitivity of 88dB, as the SB20 has 92dB.

And maybe you could tell me what you liked about the data from those two sites (which are not the CHN110, by the way).
The two have shown that the data from MA and the measured ones from testers were quite different. Alright, different conditions apply, but the break ups don't show as much on MA's data.

Sure. I agree the Markaudio official chart is ridiculous, but it seems to be pretty similar to the ones measured by third parties (just scale is different). Regarding cone breakup, it seems the SB Acoustic driver exhibit the same behavior from the official charts (so 10-15db drops/bumps starting at 3KHz).
As I mentioned in my requirements, I would like:

• to have wide enough dispersion, so that all seats can enjoy the movie (e.g. no more than 10dB difference between seats)
• to have eq-able speakers. IIUC, directivity index is key here. In those polar maps, the "lines" are monotonic, which would indicate that there is no directivity error. Let me know if I am misinterpreting those polar charts and I am wrongly linking those to eq-ability. In the SB acoustics charts, it seems the 30degree off-axis does not match the on-axis well above 3KHz, and indicates the polar map would not look as smooth. Let me know again if I am misinterpreting something.

If you desperately try to stay with a 3 to 4" driver

I have no particular passion for 3"/4". Mostly trying to understand the trade-offs on speaker design 🙂 (output levels, directivity, ...)

A mini HT with TABAQs spread around could be nice for limited space.

Do I get it right it is an MLTL design? Other than lower extension, would it help above 80Hz ?

 

[babinsky]

Things sound louder when they are distorted, so you may turn them down because of that. A more capable speaker may inspire you to turn things up more, so having some extra output capability may be beneficial. This is me judging things from a distance and using generalizations though.

Fair enough. Distortion was audible when I did that experiment, but it was not painfully audible 🙂. I will redo the experiment closer to the speaker to avoid going in distortion territory.

From what I'm seeing online, WinISD is modeling half space, so there may not be as much room gain as you were thinking.

Thank you very much for that information ! That will save me from hoping more than what the simulation predicts.

Part of it is not running a speaker up against its limits. Having headroom in the system is useful all the way around.

Ok, I guess this matches what I am trying to do: 95db actually accounts for the headroom I am looking for (average listening should be 75-80db, so it accounts for 15-20 db headroom). Do you think this would be enough headroom or am I am being too optimistic?

 

[babinsky]

Have you looked at the kits at Soundimports?

https://www.soundimports.eu/en/diy-kits/diy-component-packs/

Thanks for the link. I looked at a few of those and it seems most of those are bookshelf speaker with limited SPL output. Some have higher output capabilities but are much bigger boxes (tower speakers). Is there a particular box I should look for in that list?