Philharmonic BMR Speaker Review

[R Swerdlow]

What kind of ports are included in the Optional Kit from meniscus audio? The flared ones or simple tubes?

The Meniscus Audio Bare Bones Kit does not include port tubes at all. The Full Kit adds a number of useful parts to those of the Bare Bones Kit, including two Precision Ports with wide flares on both inner and outer ends. In my opinion, these flared ports do improve the sound quality when compared to non-flared port tubes. They're well worth the added price of the Full Kit.

Why the baffle width is not negotiable? I'm asking because there is surely somewhere a picture with a much wider black BMR. I cannot find it right now. Also ,in the early specs of the BMR i found a 9" value for the width of the cabinet.

The proper balance between the woofer & mid-range driver for an 8" wide front baffle has been built into the crossover design. This is sometimes called baffle step compensation (BSC). If you change the front baffle width, it would require a redesigned crossover to provide the correct level of BSC. IMO, it would be best not to change that.

The active volume of the speaker. Correct me if I'm wrong. Cabinet volume minus small BMR driver box ,minus the volume dislocated by the other two drivers( the Scanspeak in particular) , minus the port tube volume(?) , minus the bracing wood piece , minus some dumping thickness(?)

As you estimate the internal volume of the bass chamber, don't include the volume of the damping material. It's porous and is still part of the internal volume that the bass driver "sees".

The Meniscus Audio Full Kit also includes the appropriate amounts & types of damping materials – another reason why I like the Full Kit. Dennis Murphy says the type & amount of damping material for the mid-range chamber is critical to the mid-range sound quality of this speaker.

 

[EEG]

Thanks for the feedback
Speaking of BSC ,I found :
https://www.salksound.com/blogtopic.php?id=2
Close related to:
https://www.salksound.com/blogtopic.php?id=7
Very simple explanations for newcomers.
So ,from the BSC point of view , is somehow better the offset version(the mid/high drivers) of the baffle ?

 

[Chromatischism]

Sounds like we need to come up with an offset value for PIR based on some kind of average dispersion width value.

Actually, the model should:

1. Incorporate room dimensions so that predictions can be made for outdoor speakers where side walls and ceiling = 0
2. Incorporate dispersion width so that indoor predictions can be more accurate

 

[R Swerdlow]

Thanks for the feedback
Speaking of BSC ,I found :
https://www.salksound.com/blogtopic.php?id=2
Close related to:
https://www.salksound.com/blogtopic.php?id=7
Very simple explanations for newcomers.
So ,from the BSC point of view , is somehow better the offset version(the mid/high drivers) of the baffle ?

Those two links are both about diffraction from the edges of the front baffle, but BSC and what Jeff Bagby calls Edge Diffraction cause different problems, with different solutions.

High frequency sound coming from tweeters (in 2-way speakers) or tweeters & mid-range drivers (in 3-ways) will interact with cabinet edges, causing small peaks and valleys in the frequency range of about 4 to 8 kHz. The exact frequencies are determined by the distances from driver to cabinet edge. So, these diffraction frequencies depend on front baffle width and driver layout.

If a mid-range or tweeter is centered in a tall narrow cabinet, there will be two driver-to-edge distances that are the same, creating diffraction peaks and valleys at the same frequencies. They add together, generating larger peaks and valleys.

If the tweeter or mid-range is slightly off center, there will be two slightly different driver-to-edge distances. They will generate peaks and valleys at differing frequencies. There will be more of them, but each will be smaller in amplitude.

The present version of the BMR speaker was designed with centered tweeter and mid-range, and the crossover reflects those locations. In older versions of this speaker, the tweeter and mid-range were located off-center, and the crossover took that into account. Dennis Murphy does pay close attention to details like these.

Placing the tweeter & mid-range off-center has a generally similar effect as rounded over front baffle edges have on a speaker. It's a nice touch, but makes a small difference in the measurable sound at higher frequencies. It can be measured, but it is not known just how audible a difference it makes.

In contrast, baffle step responses occur at lower frequencies, and are affected by external cabinet width. Rounded over edges make no difference at those lower frequencies.

 

[EEG]

How about mitered edges of the baffle at 45 degrees?

 

[R Swerdlow]

How about mitered edges of the baffle at 45 degrees?

Mitered edges address the same high-frequency edge diffraction problem as rounded-over edges do.

Mitered edges are generally easier to cut on a front baffle than rounded-over edges. Both are best done with the right wood working tools correctly set up. That can be a table saw tilted for 45° cuts, or a table-mounted fence-guided router with either a 45° chamfer bit or a ¾" radius round over bit. I would never try using those router bits with a hand-held router, only with a router-table. In my limited experience, I needed to practice these cuts on scrap boards several times before I tried the cut on the real piece.

I've never seen measurements that directly compare these two edge treatments, but I'd be surprised if they didn't perform similarly. Neither of them have a MAJOR effect on a speaker's performance. But they are nice touches, and they do have more effect than different types of binding posts .

 

[EEG]

Sorry to bother with other silly questions.
Are the drivers'(all 3) position lowered on the cabinet kit from the speakerhardware or the version used on Audioholics review comparing with the Salk design or other early designs? Considering the height is the same.
Or I'm just tired and it is an optical illusion?

 

[R Swerdlow]

Sorry to bother with other silly questions.
Are the drivers(all 3) lowered on the cabinet kit from the speakerhardware or the version used on Audioholics review comparing with the Salk design or other early designs? Considering the height is the same.
Or I'm just tired and it is an optical illusion?

The drivers do look as if they're lower in the first two photos. But the overall height of all those cabinets are not the same. I forget their heights, but there were some changes to the cabinet dimensions as well as the driver layout.

Here is a photo I have from 2018, of one of the BMR Road Trip speakers. That cabinet was sold pre-made by Parts Express, the sides are curved, and the drivers are vertically aligned, not off-set. Older versions had off-set mid-range and tweeter. If I recall, when PE stopped selling those cabinets, the cabinets Dennis next used were made in Taiwan, and may have been an inch or two shorter overall.

Why are you concerned about this?

 

[Haflermichi]

Putting a radius on baffle edges with a hand held router and round over bit is a piece of cake. No different than a bevel bit.
I just did that with my Hales Revelation Threes.
But of course, if you're not experienced, definitely practice on some scrap first.

 

[EEG]

AFAIK the tweeter shouldn't be at the same distance from the 3 edges of the baffle. Top, left side and right side That's obviously happens with all cabinet versions.
It's just a little professional distortion I guess ,I'm a mechanical engineer( although i do not practice that for ages) and i do observe lots of details in drawings and photos.

 

[R Swerdlow]

AFAIK the tweeter shouldn't be at the same distance from the 3 edges of the baffle.

I think that's general advice for round dome tweeters. But this ribbon tweeter is much narrower horizontally, and taller vertically.

 

[xarkkon]

Great work on the review! Found your website through the S400 review and found the BMR one there. I own a pair of the S400s and found my experience very similar to yours. Also keen on the BMR as my next DIY project (after my ongoing XLS encore). How would you compare the BMR and the S400 with each other?

 

[tuga]

Hi @hardisj may I suggest that complement the set with an actual (uncorrected) in-room response taken from your listening spot?
Also, would it be possible to draw a thin horizontal line at 1% in the THD plots?
Finally, which speakers do you use as reference?

P.S.: a CSD would be nice to have too.

 

[tuga]

I think the ribbon is the biggest reason why so many people love this speaker.

Love is irrational

 

[tuga]

This discussion about levels of distortion in the lower range of the BMR's tweeter ignores an important point. What percent distortion is audible?

I've re-posted the NRC measurement of the BMR speakers distortion (THD+Noise, expressed as SPL) where the speaker's frequency response, at roughly 90 dB is shown on the same graph.

At 2.5 kHz, the speaker produces about 90 dB, and the THD+N is about 58-60 dB (as measured by eyeball). The THD+N is 28-30 dB quieter. In my own experience when two sounds are played at the same time, and one is at least 25 dB louder than the other, the ear/brain tends to ignore the quieter sound.

At worst, at the slight dip at 3 kHz, the speaker's output is 86 dB, and the THD+N is about 58 dB. That difference, 28 dB, is still enough to easily ignore the distortion.View attachment 74144

28dB below full level is more than 3%. Depending on listener and genre/material it should be pretty obvious.
Some listeners may actually enjoy it because it makes the speaker sound "louder" at lower volumes, more "lively", PA-like gig loud.

For the discerning/demanding listener the distortion at 90dB/1m should and can be (well) below 1% (-40dB) from 100Hz upwards.

 

[tuga]

Just to clarify, I don't think the RAAL is bright, I just felt the BMR as a whole was a bit bright in my small room, I wish I had a larger room to test the theory that it's related to the close sidewall reflections but I didn't.

Wide-directivity speakers should perhaps come equiped with a HF tilt control for this very reason... The tonal balance is very room-dependent.

Can you prop some matresses/cushions up against the early reflection zones or are the speakers no longer in your possession?

 

[tuga]

I've always been sceptical that PIR is representative of the speaker's tonality or timbre. The research seems to indicate that we perceive these mostly in the direct sound (i.e. on-axis/listening window), and the off-axis data is mostly only relevant for spaciousness attributes (and consistency of reflected response). I know that some people on this forum swear by PIR and believe it subsumes everything else, but that's clearly not what @Floyd Toole thinks, it's not what the Olive Preference Rating model thinks, and it's not supported by the research, either.

If I get asked to judge a speaker's tonality or timbre from measurements, the Listening Window would be my best bet, not PIR.

I think that you're better off analysing on-axis, 15° and 30° curves than relying on Listening Window which is what, averaged over a 30°-wide angle? No head is that wide (well maybe an elephant's), and it's not wide enough to reflect wall interaction...

I find @John Atkinson lateral response family set of curves far more informative in regard to predicting tonal balance, although I wish he'd always produce the normalised and the non-normalised plots.
I also think it would be important to always produce a second CSD plot with raised response (as was done here) whenever the tweeter resonance peaks above 5dB; as currently it drowns the decay by as much as the tweeter peaks hiding potentially serious issues.

 

[R Swerdlow]

28dB below full level is more than 3%. Depending on listener and genre/material it should be pretty obvious.

Again, I ask what percent distortion is audible? This requires blind listening tests, done with appropriate negative and positive controls. What should be pretty obvious may very well be your opinion. So be it. But that doesn't pass for scientifically determined data.

 

[hardisj]

Great work on the review! Found your website through the S400 review and found the BMR one there. I own a pair of the S400s and found my experience very similar to yours. Also keen on the BMR as my next DIY project (after my ongoing XLS encore). How would you compare the BMR and the S400 with each other?

If I am being honest, I personally preferred the BMR. It had lower bass extension without mechanical failure and I liked the HF presentation more.

 

[hardisj]

Hi @hardisj may I suggest that complement the set with an actual (uncorrected) in-room response taken from your listening spot?
Also, would it be possible to draw a thin horizontal line at 1% in the THD plots?
Finally, which speakers do you use as reference?

P.S.: a CSD would be nice to have too.

Edit: I might go back and remeasure in the room but I can't promise I'll remember and/or find the time to do that. I have a stack of speakers and drive units that are taking precedent.

Yea, I can add the 1% THD line.

CSD plots are a waste of time, IMHO, mainly because you can play with the slices/window and floor and make the result look different from one run to the next. Also, you can see resonances in the FR data. It's just one more thing for me to have to keep up with on top of a laundry list of data I am already providing. That is why I do not provide them. I get people like them. I don't have any intention on spending time with CSD plots, however.