Monoprice THX-365IW Review (In-wall Speaker)

[MrPeabody]

Interesting. At least two effects would potentially muck up the works. The leakage will permit front-rear cancellation, which will affect mostly the bass, although it should be a gradual effect starting at a wavelength about the same as the baffle width, which I would guesstimate to be maybe around 700 Hz, becoming stronger in effect as frequency decreases. I don't see any evidence of this at all; bass doesn't start to decline until ~150 Hz and is very abrupt at that point. As such, the leakage is only likely to do what you thought it might, i.e., make the enclosure lossier, which will mitigate the increase in Q that would otherwise occur with a sealed enclosure that isn't as large as it should be.

The other thing is the baffle step effect, but I don't see anything that looks like that either. The sharp dip at 950 Hz looks at first glance like it might a diffraction effect, because the off-axis responses swing the other way at that exact same frequency, which is a telltale feature of edge diffraction ripple. But the frequency doesn't seem to jive with the baffle dimensions. The prominent feature of baffle diffraction ripple is usually the first peak, which exaggerates the steepness of the baffle step and puts a big hump at the crest. This occurs at frequency corresponding to wavelength equal to the baffle width (assuming the drivers are located on the vertical midline, equidistant from both side edges), because the hump occurs with constructive interference, which means that the reflection must arrive to the listener in phase with the direct, and since the reflection is shifted 180 degrees due to the soft reflection, this will occur where one-half wavelength matches the distance from the center of the driver to the furthest edge. But here we have a 2" midrange that owns this part of the frequency spectrum, and it appears to be nearly the same distance from all four edges of the baffle you built, to a close approximation. As for the dip, it will occur an octave higher in frequency. If we use 1.5' for the approximate width and height (average) for the baffle you built, the prominent peak should be located at roughly 750 Hz. This means that the first dip should be found at around 1.5 kHz, which is not at all close to where that sharp dip is located.

But instead of rambling on this way, I really should have gone to bed a couple of hours ago.

 

[USER]

What that has to do with listening and a recommendation?
I assume no one forced you (band your hand behind your back) to test an in wall speaker.
Some fields, like HP require some fixtures. In wall speakers might need one too. If you wish to test in wall speakers, you might need some preparations. Or you may skip in wall. It is up to you.
I'm just saying, it is missing on your report. I thinks it is important.

Can you not see the many interesting discussions already going on in the thread? Is it all above your head? This is a work in progress, hence the questions in the evaluation. Maybe if you contribute something not snotty but actually intelligent and constructive, you can begin to understand that.

 

[B4ICU]

Can you not see the many interesting discussions already going on in the thread? Is it all above your head? This is a work in progress, hence the questions in the evaluation. Maybe if you contribute something not snotty but actually intelligent and constructive, you can begin to understand that.

"work in progress"

In Hebrew there is say, that you show a half way done job only to a donkey.
What's the rush?
Complete the prep. and then publish it done and dusted.

"Can you not see..."

Well, I can see some cheerleaders, applause everything that is uploaded to this site, but some do not meant to be applause.
I can see your good manners, lot's of acceptance, and a positive attitude. I'm with you, but just not yet...
Some others may not yest applause. Like me.
Everyone's free to take a side. It is obvious we are not taking the same side.

 

[phoenixsong]

Really interesting to see built-in speakers being measured now! I am wondering if in-ceiling speakers share the same target frequency response?

 

[GWolfman]

Thanks!

 

[amirm]

In Hebrew there is say, that you show a half way done job only to a donkey.

To a pisher, it might seem that way. For the rest of the people, there is plenty of proper information here.

My temporary teacher in elementary school said, "every classroom needs a clown; don't you try to be the one!" So think hard before you throw out an insult with zero content.

 

[milosz]

>>I am open to suggestions of how to vary or improve the setup for future testing. In-wall speakers are a huge category and would be good to have a standardized method of testing them<<

Build an entire room for each speaker tested, and rotate the Kippel instead of rotating the speaker.....

 

[Newman]

… I don't know the size "baffle" you created but I'm looking at it and estimating something around 562mm wide by 657mm tall (guesstimate). That likely explains the rolloff ~200Hz we are seeing here. I threw it in to VCAD and came up with the below for the estimation of baffle diffraction...

I snipped your image:

to compare with the measured response:

…and I think that’s a very informative correlation. Thanks

 

[Bear123]

On AVS, someone mentioned quite confidently that these speakers would have no trouble achieving reference level playback at 12 ft from 80Hz to 20,000 Hz "because they are THX certified". When I look at the distortion levels at 96 dB at 1m, I don't see any way these speakers are *remotely* close to reference capable at 12 feet, which would likely require another 15 dB of capability, not including eq which will make things even worse. As almost all low sensitivity speakers are also not capable. I'm aware this only matters if someone wants to listen at reference, but even 10 below reference seems like we might be asking a lot, since this would easily be 5 dB higher than the 96 dB sweep shown due to distance.

 

[hardisj]

It is not "worth" a shot at 13,000 euros plus shipping! It is also rather small since it is made for driver testing, not tall in-wall speakers. I am just not seeing the value of it.

no, no, no... I mean it might be worth setting up the run with the NFS baffle template as shown below to see if the results pan out (i.e., just do the front-half scan).

Looking at it now, I forgot there is a different calibration method for the baffle but - again - it might be worth looking at this as an option. Just providing a suggestion.

 

[Helicopter]

Thanks Amir. Great to see a foray into another category. I love in-wall speakers. I have a couple pairs of Polk RC85i speakers in my main listening room that work well for dancing with the kids or parties. Unfortunately, the contractor interpreted 'ear-height' for the tweeters as 6ft, which is about right when I am dancing or hosting a party, but that's about it. Mine are in MDF ship lap over OSB on insulated 2x6 studs... at least I got that right. I have wondered what it would be like to do a multi-channel setup with something better. I am pretty sure JBL, Revel, and Focal would be the only considerations. It is a little tricky figuring out what is best for a matching setup in the Revel lines. Focal has a line with beryllium tweeters that, though $$$$, are a lot cheaper than similar speakers in a fine hand finished cabinet.

If were testing these with an NFS I already had, I might find a wall in the garage and cut a 15.5x35" or so hole and then make 15.5x11.5 and 15.5x23.5 blanking covers. Then you could make 3 2-foot inserts out of a 2x4 piece of wall material, and 7 3-footers from a 4x8 sheet. I think I'd just use drill, jigsaw, sawzall, rotozip, or drywall knife for cutouts... that's what people use to install them in rooms.

Probably in the next 10 years, I will build a house with a setup using in-wall speakers in surround configuration. Also, my in-laws need to update their system and have some sort of in-wall speakers. I need to look. No center channel, so that will need to be addressed. To paraphrase Toole, multi-channel is better if you want to impress people.

 

[Grotti]

"work in progress"

In Hebrew there is say, that you show a half way done job only to a donkey.
What's the rush?
Complete the prep. and then publish it done and dusted.

"Can you not see..."

Well, I can see some cheerleaders, applause everything that is uploaded to this site, but some do not meant to be applause.
I can see your good manners, lot's of acceptance, and a positive attitude. I'm with you, but just not yet...
Some others may not yest applause. Like me.
Everyone's free to take a side. It is obvious we are not taking the same side.

Cheerleader here: what about stop complaining and add something of value to the discussion?

Even though Amirs approach of measuring in wall speakers is not near as perfect as his measurements of regular speakers are, he at least is trying to provide some meaningful data. Stop whining, take measurements as they are, try to learn something from the contributions here and take part in the discussion, if you have something relevant to state.
For me, in wall speakers are irrelevant since i will never invest in other than stereo, but it is much to learn here about loudspeaker physics, measurement problems and the interpretation of data.... It is called AudioScience

 

[Timbo2]

>>I am open to suggestions of how to vary or improve the setup for future testing. In-wall speakers are a huge category and would be good to have a standardized method of testing them<<

Build an entire room for each speaker tested, and rotate the Kippel instead of rotating the speaker.....

As long as that doesn't slow down Amir's release schedule I'm OK with that. Otherwise I'm going to have to complain and make a snarky post.

 

[MediumRare]

It wouldn't though. Bass will cause resonances in the drywall so will transmit back.

I am wondering if a practical test setup may include rolling off the bass with 80 Hz filter. This is how they will be universally used anyway so I don't see much benefit in quantifying their performance otherwise. This would relax the requirements for bass frequency measurements, back box, etc. What say you?

The issue of drywall is a real one, especially if the speaker mounts directly to the drywall and not to a stud. The resonance of the drywall and the reflectance of the baffle are key characteristics in the use case. I would absolutely built the front baffle out of drywall. Also, it's critical to seal the speaker box if the baffle is not "infinite". Not so much for bass but to reduce leakage of mid/hf sound as the Kilppel makes it way around the speaker. IMO that is invalidating the predicted bounces and therefore the in-room FR. Finally, Kilippel should not be measuring beyond 90+/- degrees at all.

Ps: no filter! Those intending to use a filter will be ignoring the non-sub score and bass, anyway. Many will use this for whole house audio and the full FR is essential to know.

 

[Remlab]

I have a feeling that the unsupported edges of the plywood could possibly be vibrating like crazy along with the test signals, and the gap between them causing potential severe cavity resonances

 

[Consilient]

Great new category to measure, Amir! One thought regarding testing consistency: many in-walls have backbox options from the manufacturer, so perhaps it makes sense to only test units with the provided backbox. It would at least minimize the internal wall effects.

 

[hex168]

The issue of drywall is a real one, especially if the speaker mounts directly to the drywall and not to a stud. The resonance of the drywall and the reflectance of the baffle are key characteristics in the use case. I would absolutely built the front baffle out of drywall. Also, it's critical to seal the speaker box if the baffle is not "infinite". Not so much for bass but to reduce leakage of mid/hf sound as the Kilppel makes it way around the speaker. IMO that is invalidating the predicted bounces and therefore the in-room FR. Finally, Kilippel should not be measuring beyond 90+/- degrees at all.

Ps: no filter! Those intending to use a filter will be ignoring the non-sub score and bass, anyway. Many will use this for whole house audio and the full FR is essential to know.

I agree that sealing the box makes sense for the above reasons and also for consistency. The volume should be sufficient not to raise the in-box Q and woofer resonant frequency much compared to open sides. Adding some fiberglass insulation should help with that. Box sides would also prevent the drywall edges from flopping around as much. Do the installation instructions from Monoprice say anything regarding fiberglass insulation behind the speaker? I suspect there may be significant effects caused by the parallel wall spacing with no damping.

 

[Francis Vaughan]

I think the main problem with in-wall speakers is that any simple mounting into dry-wall is inherently compromised. When measuring any given speaker we don't want to be measuring the limitations of the test rig every time a new speaker is measured. Which would suggest making the test rig as neutral as possible. Resonance problems or leakage problems in the wall are not the fault of the speaker manufacturer, and should not be present in a test. However the test rig needs to be representative of the important physcial conditions, so a depth of a standard wall stud is clearly correct, as is a volume no bigger than normally seen.
Adding fibreglass is interesting. The instructions for the Monoprice speaker are mute on the idea. Indeed they are mute on just about everything, being little more than a cutout template and a few diagrams. The CEA2024 spec cited earlier says no more than 40 litres volume, but nothing else.

There are lots of difficulties with trying to match real world uses. Does a manufacturer that provides installation instructions requiring sealer around the studs and noggins, plus insulation in the cavity, justify better review results than one that does not, even though it is the extra effort in installation that made the improvement in the result?

IMHO the speaker needs to be able to show itself at its best, and any poorer results due to the vagaries of installation assigned where they belong - in the installation. It may be that some speakers are more or less tolerant to installation issues, and those with a back-box might be one set. But unless a further complicating set of tests are done to characterise the lack of tolerance, it seems unreasonable to assume the worst.

 

[BrentW]

Thank you Amir for continuing to push the boundaries of speaker measurement and data sharing in the public domain. Based on my limited experience in the industry I would guess there are many OEMs that don't use nearly the same level of rigor / sophistication as Amir applies to measuring products in their own design and testing, so this is a real gift to have access to this kind of information on the internet. There have been gems like this before (Zaphaudio, etc.) that have come and gone - let's not forget what a great public service Amir is providing as we provide constructive criticism.

 

[hardisj]

I think the main problem with in-wall speakers is that any simple mounting into dry-wall is inherently compromised. When measuring any given speaker we don't want to be measuring the limitations of the test rig every time a new speaker is measured. Which would suggest making the test rig as neutral as possible. Resonance problems or leakage problems in the wall are not the fault of the speaker manufacturer, and should not be present in a test. However the test rig needs to be representative of the important physcial conditions, so a depth of a standard wall stud is clearly correct, as is a volume no bigger than normally seen.

Well, the 2034 spec explicitly states how to test an in-wall speaker, which I mentioned earlier but will link again below. It seems they also talk about in-wall speakers that contain enclosures but I'm not 100%. Point being, I don't see the point in Amir reinventing the wheel. The issue is the size of the baffle. But maybe making a scaled down version is OK, and using the NFS template for "baffle" (as I illustrated in an above post) to do the processing.