• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. 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!

Philharmonic BMR Speaker Review

Dennis Murphy

Major Contributor
Technical Expert
Joined
Mar 17, 2020
Messages
1,071
Likes
4,535
Pretty sure it's the RAAL 64-10, it's why it needs to cross above 3k.
First off, thanks be to Erin for his detailed analysis. That was quite an investment in time and effort, and I, as a speaker designer, particularly appreciate it. There's a lot to digest here, and I'll have a more detailed response a little later. For now I just want to discuss the distortion plots a little, since I think that's the only real negative that jumps out at you. The RAAL is the culprit, and it's pretty much inherent in the design approach. All ribbons will show an increase in distortion at the bottom of their operating range. That's the flip side of the extremely low mass of the ribbon element. Even with the very strong magnetic field present in these designs, the ribbon will flex more than a typical 1" dome as the wave lengths increase.

The RAAL is probably a little worse in this respect at very high output levels because Aleksandar deliberately avoided any significant horn loading in the face plate in order to maximize horizontal dispersion. That lowers sensitivity and means the ribbon has to work a little harder to produce a given SPL. That puts us solidly back into the world of trade-offs. When I was working with the RAAL in the early BMR's, I did an alternative design using a very high quality Aurum Cantus AMT. (It's used in the premium level RBH monitor). Even though distortion was lower, the AMT version just didn't sound as open, and I couldn't detect any improvement in clarity traceable to the reduced THD. I was still concerned enough to send a BMR off to the Canadian NRC for a test in their anechoic chamber. They test THD at 90 dB at two meters. I've attached the results, which show distortion ranging from 1 - 2.7% in the problem area. It would be less if the measurement were taken at the standard 1 meter, 90 dB, and much less at normal levels. I consider this acceptable for the BMR's intended use, which is mainly for full-range playback on music recorded with natural ambiance in a concert venue at moderately high to fairly high SPL's. I know there are those who think distortion above .5% is audible, but I'm not sufficiently convinced of that to sacrifice other attributes to meet that threshold at ear bleeding SPL's.
 

Attachments

  • 1595120532229.png
    1595120532229.png
    325 KB · Views: 279
Last edited:

Dennis Murphy

Major Contributor
Technical Expert
Joined
Mar 17, 2020
Messages
1,071
Likes
4,535
Salk is using the original design. The change to aligned drivers was because Dennis had to change cabinets.
Not quite. Jim is using a picture of the old design. What's coming out of his shop is the newer in-line version. I switched to in-line drivers because (1) People didn't like the looks of the offset drivers, (2) For the most recent Chinese BMR's, I didn't trust the factory to ship matched mirror-image pairs, and (3) The in-line version measured and sounded at least as good. BTW, the BMR tested by NRC had offset drivers and the off-axis response is a little different.
 

napilopez

Major Contributor
Forum Donor
Joined
Oct 17, 2018
Messages
2,111
Likes
8,446
Location
NYC
Man, I literally just updated my paragraph on this. I was driving my kid and dog to the creek and it kind of *hit* me that I didn't think to mention this aspect. I definitely agree that the PIR may not cover edge cases such as this. But I don't know for sure. I do know that this speaker did not sound remotely as bright as the PIR curve indicates. I agree there is more than meets the eye here. And I also believe the target curve in comparison with a speaker like this may be a bit misleading.

Yep, there are two relevant quotes from the Olive preference paper part II.

First, there's this one, which we've talked about a couple of times around here:

The degree of tilt varies among curves for Test One and the larger sample. Test One includes mostly 2-way designs whereas the larger sample includes several 3-way and 4-way designs that tend to have wider dispersion (hence smaller negative target slopes) at mid and high frequencies. This suggests that the ideal target slope may depend on the loudspeaker’s directivity.

Emphasis mine. And then there's this other choice quote that I do not think has been discussed here as much:

The speakers with the flattest sound power had rising frequency responses on-axis and/or reduced low frequency extension. Both are necessary compromises to achieve flat sound power for speakers that have a rising directivity at higher frequencies. Such speakers represent the vast majority of all speakers sold. A speaker with constant, flat directivity could theoretically satisfy the flat sound power criterion and still achieve high preference ratings, so long as it had a smooth on-axis response well-maintained off-axis. However, such speakers are not widely available."

Here Olive goes a bit further, indirectly but clearly saying that the target slope of the PIR is possibly/likely contingent upon the directivity. In other words, constant directivity designs like the D&D 8C and very wide direcitivty designs, which essentially approach constant horizontal directivity, likely do not need the -1dB per octave slope.

I do think wide directivity may contribute a little to brightness, especially in a small reflective room, but most of the flat slope is heard as an expanded soundstage/larger apparent sources rather than brightness. I wonder if it would be the opposite of a speaker had very wide vertical directivity though. I'm betting it would.

This doesn't apply to you, but it's why I always worry when some take 9-10dB down slope of the PIR far too literally, and another reason to be cautious with the preference score's if a speaker features unusually wide of narrow directivity. I personally don't think we should be trying to correct the slope of the PIR when we EQ unless the on-axis isn't flat. I prefer to EQ to the speaker's own directivity slope, though thankfully most speakers aren't too far off from the norm.

I also tend to believe that the wider a speaker is, the more acceptable flaws in the PIR are, but unfortunately there's only really own study I know that discusses this, and it's very old.
 
Last edited:
OP
hardisj

hardisj

Major Contributor
Reviewer
Joined
Jul 18, 2019
Messages
2,907
Likes
13,908
Location
North Alabama
This doesn't apply to you, but it's why I always worry when some take the ideal slope of the PIR too literally, and another reason to be cautious with the preference score's if a speaker features unusually wide of narrow directivity. I personally don't think we should be trying to correct the slope of the PIR when we EQ unless the on-axis isn't flat. I prefer to EQ to the speaker's natural directivity slope, though thankfully most speakers aren't too far off from the norm.

I didn't take it that way but I think it's perfectly fair to note this even toward me. This is not my dayjob so I take caution to assume too much and welcome learning not necessarily for my own edification but also to make sure I don't say misleading things. I appreciate you bringing this topic up. I may ping Mr. Olive and see if he has any insight on this test result wrt what you've quoted above.

I also agree that this might be a case where someone would potentially try to EQ the response to match the target and I believe that would be a bad thing. This speaker doesn't exhibit any audible brightness to my ear; in fact it sounded relatively tonally balanced. I don't know how much the direct vs reflected aspect weighs in a speaker with such wide horizontal directivity (nearly 90° at some points). It's an interesting specimen in the aspect of how the science is correlated to things like the PIR and what it ultimately means in the grand scheme of things.
 

whazzup

Addicted to Fun and Learning
Joined
Feb 19, 2020
Messages
575
Likes
486
@hardisj
Just curious why the woofer looks like it had participated in the Normandy beach landings.

Edit: My bad! That's how it's designed..... :D
 

MZKM

Major Contributor
Forum Donor
Joined
Dec 1, 2018
Messages
4,244
Likes
11,486
Location
Land O’ Lakes, FL
This doesn't apply to you, but it's why I always worry when some take 9-10dB down slope of the PIR far too literally, and another reason to be cautious with the preference score's if a speaker features unusually wide of narrow directivity. I personally don't think we should be trying to correct the slope of the PIR when we EQ unless the on-axis isn't flat. I prefer to EQ to the speaker's own directivity slope, though thankfully most speakers aren't too far off from the norm.

Yep, if a speaker sounds neutral to you, don’t EQ the slope to meet any “standard”, just correct irregularities.

As for the preference rating, that is why 1 reason I believe Olive’s formula based on 13 bookshelves did not translate well to tower speakers. The other is there is more bass with towers.

Thats to a certain someone :), we have the data on the 13 speakers, and I going to try to make an adjusted formula that does not care about directivity, while roughly keeping with Olive’s breakdown between direct sound, sound power, and bass.
 

AnalogSteph

Major Contributor
Joined
Nov 6, 2018
Messages
3,338
Likes
3,278
Location
.de
All ribbons will show an increase in distortion at the bottom of their operating range. That's the flip side of the extremely low mass of the ribbon element. Even with the very strong magnetic field present in these designs, the ribbon will flex more than a typical 1" dome as the wave lengths increase.
Part of that is due to an inherent air leak, addressed by sealed ribbon tech. Right now it's patented but could be interesting long-term.
 

Dennis Murphy

Major Contributor
Technical Expert
Joined
Mar 17, 2020
Messages
1,071
Likes
4,535
Part of that is due to an inherent air leak, addressed by sealed ribbon tech. Right now it's patented but could be interesting long-term.

Thanks for the link. Maybe RAAL will license the patent? I'm not really tempted to replace the RAAL because the off-axis response doesn't look any better than a typical dome. But I am surprised it took this long for someone to figure out you have to seal the air gap.
 

bigjacko

Addicted to Fun and Learning
Joined
Sep 18, 2019
Messages
721
Likes
359
I have noticed that for near field driver measurement, both woofer and tweeter goes up around 1k to 2k Hz. The volumes were pretty high, only -10db compare to intended output. Is it a bad thing or it does not matter much?

The crossover for woofer and BMR looks like to be 550 Hz. Why choose a crossover point this low? Because of directivity or phase alignment? I think the 400 to 500 Hz distortion spike comes from the BMR. The distortion is actually not bad because at normal listening level of 88db, it is less than 1%, but I am still curious why choose crossover at 550 Hz.
 

Dennis Murphy

Major Contributor
Technical Expert
Joined
Mar 17, 2020
Messages
1,071
Likes
4,535
I have noticed that for near field driver measurement, both woofer and tweeter goes up around 1k to 2k Hz. The volumes were pretty high, only -10db compare to intended output. Is it a bad thing or it does not matter much?

The crossover for woofer and BMR looks like to be 550 Hz. Why choose a crossover point this low? Because of directivity or phase alignment? I think the 400 to 500 Hz distortion spike comes from the BMR. The distortion is actually not bad because at normal listening level of 88db, it is less than 1%, but I am still curious why choose crossover at 550 Hz.

I can't help you out with the near field driver measurements, because I don't know what they mean or why they are relevant. What matters is how the individual drivers measure at the intended listening distance and at one meter. For the BMR, either at 3 meters (design distance) or one meter (usual measuring distance), the individual driver slopes are monotonically declining and follow smooth 4th order roll offs over most of their range(See attached).

As for the woofer-midrange crossover frequency, the actual crossover point at the listening position is 610 Hz. That doesn't strike me as low, given that this model BMR mirange is usually crossed a good bit lower than that (e.g. Cambridge Audio Aero 2), and the factory specs for operating range are 250 Hz on up. Plus, the acoustic slope is very steep, so the mid shouldn't be unduly stressed. However, the crossover point was chosen mainly to use only the smoothest portion of the Scan 8545 woofer. I wanted the first portion of the woofer roll off to be as close to target as possible to optimize phase and system frequency response. That said, the next generation of BMR's (available in October) will have a new woofer that is much smoother in the 800 Hz - 2 kHz range, and I've not only bumped the crossover point up 200 Hz, but I'm using 2nd order acoustic slopes to improve sensitivity due to greater bandpass gain.
 

Attachments

  • BMR DRIVER SLOPES.png
    BMR DRIVER SLOPES.png
    12.6 KB · Views: 233

3ll3d00d

Active Member
Joined
Aug 31, 2019
Messages
207
Likes
173
Zoomed and annotated for the arrival time of each drive-unit’s impulse. You can see the drivers are not time-aligned and arrive distinctly separately. The difference in time between the woofer (2.5ms) and the midrange (1.33ms) is approximately 1.2ms. To put this in perspective, 1.2ms is approximately 16 inches.

Philharmonic%20BMR%20Step%20Response_zoom.png
nice review. I don't think this is true though. Google suggests this uses an LR4 at 600 & 3500Hz so you'd expect to see these 3 separate peaks in the step response.

To illustrate, a quick sim in vituixcad of that same XO and a port tune of about 30Hz gives which is pretty much the same as the measured step response.

1595149448723.png
 

edechamps

Addicted to Fun and Learning
Forum Donor
Joined
Nov 21, 2018
Messages
910
Likes
3,620
Location
London, United Kingdom
I definitely agree that the PIR may not cover edge cases such as this. But I don't know for sure. I do know that this speaker did not sound remotely as bright as the PIR curve indicates. I agree there is more than meets the eye here. And I also believe the target curve in comparison with a speaker like this may be a bit misleading.

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.
 

bigjacko

Addicted to Fun and Learning
Joined
Sep 18, 2019
Messages
721
Likes
359
I can't help you out with the near field driver measurements, because I don't know what they mean or why they are relevant. What matters is how the individual drivers measure at the intended listening distance and at one meter. For the BMR, either at 3 meters (design distance) or one meter (usual measuring distance), the individual driver slopes are monotonically declining and follow smooth 4th order roll offs over most of their range(See attached).

As for the woofer-midrange crossover frequency, the actual crossover point at the listening position is 610 Hz. That doesn't strike me as low, given that this model BMR mirange is usually crossed a good bit lower than that (e.g. Cambridge Audio Aero 2), and the factory specs for operating range are 250 Hz on up. Plus, the acoustic slope is very steep, so the mid shouldn't be unduly stressed. However, the crossover point was chosen mainly to use only the smoothest portion of the Scan 8545 woofer. I wanted the first portion of the woofer roll off to be as close to target as possible to optimize phase and system frequency response. That said, the next generation of BMR's (available in October) will have a new woofer that is much smoother in the 800 Hz - 2 kHz range, and I've not only bumped the crossover point up 200 Hz, but I'm using 2nd order acoustic slopes to improve sensitivity due to greater bandpass gain.
Thank you for reply. I saw the review of BMR on hificompass, the distrotion is already more than 1% at 500 Hz, so that's what my concern was. This review showed better distortion than I expected, the crossover is perfect that the lobing is not visible even with 40 degree off vertically. I have seen the facebook post regarding new driver, they said the mid voicing is a bit different for the new one, what is your opinion on the new one?
 

pierre

Addicted to Fun and Learning
Forum Donor
Joined
Jul 1, 2017
Messages
963
Likes
3,048
Location
Switzerland

GelbeMusik

Senior Member
Joined
Apr 19, 2020
Messages
445
Likes
290
(The link to the flat-pack leads back to Your home page.)

The ribbon, sealed or not dictates a higher x/o frequency, but still it distorts considerably. We have uneven directivity in the vertical, the possible mechanical vulnerability, the high price. For not more than a tiny bit of wider dispersion in the tops of the hights.

The BMR again dictates a higher x/o due to otherwise too high of distortion figures. Even bigger problems in respect to vertical directivity.

The woofer is expensive, especially when compared to more contemporary offers. With a more competent mid the x/o would be more relaxed, the woofer could have been way cheeper. Alltogether with less intermodulation, where the woofer is always the main contributor.

So, it seems funny, but the measurements don't look that bad. Two things appear less fortunate to my eye.

- The curved sidewalls are inacceptable, design-wise.

- The distortion, whether perceptible or not, must not be. One might argue, that these aren't that bad, because of and so forth, o/k. But from the standpoint of technical evaluation they are at least a hiccup.

One might take a specific distribution of maximum level versus frequency into account, as it is demanded by a typical music program. To some accuracy it can be discribed as this:

https://www.hifi-selbstbau.de/images/stories/grundlagen/waveana/Track260MaxL.png

So, the demand in the deepest bass region is considerably less than mostly anticipated. Demand in lower treble is way less than in the mids. But even with the lowest levels the ribbon distorts in the 2..3% region.

A second irregularity is the interference that comes from the baffle dimensions. The discontinuity in the dispersion comes up around 3..4kHz.

And third the compression has a hot spot exactly there.

I think the ribbon is clearly the archilles heel for the overall design. I assume it is the most expensive single part. Does it make sense, really?
 

richard12511

Major Contributor
Forum Donor
Joined
Jan 23, 2020
Messages
4,335
Likes
6,700
(The link to the flat-pack leads back to Your home page.)

The ribbon, sealed or not dictates a higher x/o frequency, but still it distorts considerably. We have uneven directivity in the vertical, the possible mechanical vulnerability, the high price. For not more than a tiny bit of wider dispersion in the tops of the hights.

The BMR again dictates a higher x/o due to otherwise too high of distortion figures. Even bigger problems in respect to vertical directivity.

The woofer is expensive, especially when compared to more contemporary offers. With a more competent mid the x/o would be more relaxed, the woofer could have been way cheeper. Alltogether with less intermodulation, where the woofer is always the main contributor.

So, it seems funny, but the measurements don't look that bad. Two things appear less fortunate to my eye.

- The curved sidewalls are inacceptable, design-wise.

- The distortion, whether perceptible or not, must not be. One might argue, that these aren't that bad, because of and so forth, o/k. But from the standpoint of technical evaluation they are at least a hiccup.

One might take a specific distribution of maximum level versus frequency into account, as it is demanded by a typical music program. To some accuracy it can be discribed as this:

https://www.hifi-selbstbau.de/images/stories/grundlagen/waveana/Track260MaxL.png

So, the demand in the deepest bass region is considerably less than mostly anticipated. Demand in lower treble is way less than in the mids. But even with the lowest levels the ribbon distorts in the 2..3% region.

A second irregularity is the interference that comes from the baffle dimensions. The discontinuity in the dispersion comes up around 3..4kHz.

And third the compression has a hot spot exactly there.

I think the ribbon is clearly the archilles heel for the overall design. I assume it is the most expensive single part. Does it make sense, really?

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

overg

Member
Joined
Jul 19, 2020
Messages
8
Likes
19
I think the ribbon is clearly the archilles heel for the overall design. I assume it is the most expensive single part. Does it make sense, really?

The suggestion that the Raal is the achilles heel in the BMR is absolutely baffling to me, and I can only conclude this is one of those cases where the things being measured (distortion, vertical dispersion) have absolutely no correlation to the audible sound you hear, or at least the average listener's preference for that sound. I'm fairly sure I've read every review or forum post about the BMR that can reasonably be found via a google search. And aarons915 (who posted earlier in this thread) is literally the only person, ever, that I've seen complain about the Raal or about this speaker sounding too bright.

I first discovered the BMR a couple of years ago, when a Reddit poster described how they trounced the B&W's he was replacing. Multiple other posters chimed in singing the BMR's praises, so I put them on my "intrigued" list. Over the next few months I kept running across random praises thrown the speaker's way, so I did a deep dive, read through the entire gigantic Philharmonic thread on AVS, and then read all of the other random threads, blogs, and reviews google could find. The praise was pretty much universal (especially given the price, which was an absurd $1,350 at the time). And the Raal is without a doubt the most lauded feature of the BMR. It was almost universally praised for both it's clarity and complete absence of any sort of listening fatigue.

Then the Audioholics review came out, and cemented the BMR as my target purchase. Unfortunately, that is right when Dennis had his health issues and had to shut down Philharmonic, leading me to believe it just wasn't meant to be. But when Dennis recovered and revealed he had about a dozen pairs left to sell for $1,500 (including shipping), I jumped all over them. I've now owned my pair for a little over a year, and listened to hundreds of albums on it. I can't think of a complaint I've had about the Raals, and that's not even my personal favorite part of the speaker (I'm in awe of just how clear the midrange is). I've certainly never heard any audible distortion or lack of clarity from the highs, and they have never once triggered any ringing in my ears or even fatigue.

As another data point, Dave Fabrikant of Ascend Accoustics has mentioned more than once that they've had numerous customers compare their dome versus Raal versions of speakers, and the preference for the Raals is overwhelming. I believe Ascend may be the highest volume dealer of a Raal based speaker, and they also offer dome versions for most of those. And if I recall correctly, Dave has revealed that he has no profit-margin motive to sell Raal over anything else . . . he just does so due to listener preference.

And as a final note, if you know anything at all about Dennis Murphy, it should be that he has always aimed for incredible "bang for the buck" speaker designs, and if he feels something he designs is only as good as the competition, he generally doesn't keep selling it. If you read the AA review, Amir discusses how ridiculously low cost that speaker was. And if you check out the kit prices Erin cites in his review here, you'll note that the cost of the component parts is just about exactly what Dennis was selling these for fully finished and assembled.

Point being, the guy was trying to sell speakers for a great bargain, and was essentially giving away his labor to do it. I find it pretty much impossible to believe that if there were a more cost effective alternative to the Raal, that Dennis wouldn't have used it.
 
Last edited:
Top Bottom