2. speakers
  3. 154828-20253

What the 802 and 3 Decades can tell us about trends

Hi Everyone,

For the longest time I’ve been saying that high end speakers can be very trendy. What taste-makers claim to be superlative sounds one decade may no longer be in another, and the trouble recording engineers have as a result. The ideal loudspeaker is not a lab instrument in a vacuum. While we can in fact measure what neutral is, we cannot also say you should buy X because of it, buy what you like.

The same is true for recordings. Played on different era gear, they may sound better or worse. How we each resolve this issue is not for me here, but I think we have enough data for the B&W 802 to talk about how it has evolved and we have a modern speaker designer talking about how differently he would voice the same speaker today. The S3 version is, I believe, from the early 1990s.

If you get anxious about dissecting your brand of speaker, and that happens to be B&W, this post is not for you. Just stop right here. On the other hand, if you are curious about contrasting takes on speaker design, in terms of calendars and designers, this may be a fun thread for you.

First, we have to start with the Troels upgrade of the 802 S3 because he includes measurements for the original as well as his mods:

http://www.troelsgravesen.dk/B&W-802s3_Upgrade.htm

Under "Measurements," look on the left. Troels knows what he’s doing, so we’ll take his work at face value. That is the measurement for the 802 S3 as he received it.

Let’s compare that to the modern 802 measurements from Stereophile, (2016, figure 3) here:

https://www.stereophile.com/content/bowers-wilkins-802-d3-diamond-loudspeaker-measurements

The 1990 era speaker is flat from about 100 Hz to 1 kHz and then shelves down about 4-5 dB. Strong bass and presence, with subdued mid to tweeter response. Not exactly a V shape, and look in particular around 2 kHz, the trend there is upwards. This range is important as it’s been used to accentuate imaging after this.

Second, look at the 2016 measurements. A dip around 2-3 Kz, and two peaks, one at 4 kHz and one at 10 kHz. This is rather the trend for modern B&W for the past few years. That peaky response IS the B&W sound, and yes, it is very very important in giving them the character they have. As experts have written, the first thing buyers notice is the frequency response of the speaker, then smooth off axis response.

Lastly, lets look at how Troels takes the original speaker and re-defines it. He follows the classic B&K curve here. As smooth as possible, and downward sloping. There’s significantly less variation anywhere in the tweeter range, and this is no longer either a shelved or V shaped curve. Under his page, look on the right under Measurements to see his solution.

Objectively, of the three, Troels has the closest to a neutral sounding B&W 802. The factory 802 S3 and D3 models are tuned very differently from each other, in large part because the tweeter has significantly different sonic signature.

So what are we left with? Trends. This posting is all about trends and changes and fads and how we are all involved in this process. Think of this as a general law of audio relativity. What seems to sound neutral to us today, certainly would not have 30 years ago, and vice versa. Reviewers and buyers and recording artists are making this all up as we go along, so buy what you like because there will never be one right, perfect answer.
erik_squires
ct0517's avatar
ct0517
2,041 posts
 
Erik
Your article would be IMO much more fun and interesting, if your author Troels had set up the 802 s3 as designed from the factory optimized with the BAF.  Then one could see what his mods accomplish versus a properly set up 802M s3.  

https://www.kenrockwell.com/audio/b-w/800-series-bass-alignment-filter.htm

taken from ^^

Bowers and Wilkins' Matrix 800 series of speakers from the late 1980s through the mid-1990s were designed directly to John Bower's vision as part of a system incorporating an active, dedicated low-level filter in the preamplifier signal path to optimize the bass performance, not only in frequency, but also in the time and phase domains.
This dedicated bass equalizer, also called a variable high-pass alignment filter, was often included with the speakers since its frequency response is an integral part of the systems' sixth-order design.
Without this equalizer, the naked B&W Matrix speakers are a vented fourth-order design, specifically in a Bessel alignment. "Fourth-order" is an engineering term that refers to all vented and passive-radiator speakers; sealed boxes are "second order."

Troels comments at the start of the article.
  
Before starting any measurements I set up the speakers in my living room and replacing my ATS4, I experienced a somewhat smaller speaker with less visceral impact in the bass area and a somewhat less weighty lower midrange presentation, giving less fullness to vocals.


The sonic symptoms he describes can be attributed to running without the bass alignment filter. I say this based on personal experience having owned at one time all of the matrix line. Sill own matrix 800's.

Now the thing that is even more curious to me. He does link the owners manual in his article. The owners manual clearly states the need to use the BAF to achieve the frequency response that is documented. 

Cheers

mgrif104
371 posts
 
It’s an interesting discussion.  I had the opportunity to work as an assistant audio engineer in a well regarded performing arts center back in the mid 80s, as I put myself through grad school.  I gained valuable experience from recording symphonies to doing sound reinforcement for all sorts of groups (including Alison Krauss!).  What I learned is that there is no such thing as neutral.  The room has a “sound” which can and should be tuned.  Commercial audio speakers are unbelievably colored - and tuned in the room to remain that way by any audio engineer worth their salt or it sounds unnatural.  Microphones are chosen for their sound. The mixing board has a sound.  I had to be careful with mic cable length and path.  Both analog and digital recorders have a sound.  

This all goes to Erik’s point - by what you like as there is no such thing as neutral as even our live music benchmark is not neutral.  Measurements in an anechoic chamber can only tell a designer so much. 

Of course, This isn’t to suggest there is  little value in moving up or to new toward a better component. I recently replaced my speakers. I hear a good deal more information through the new ones, despite them being far less tipped up in the treble.  And, the music is presented more pleasingly.  

My suspicion (can’t really be tested) is that if I were able to compare what I just purchased to a similar level 70s/80s/90s model from the same manufacturer (KEF in this case), I believe they would sound quite different, even if the measured frequency response suggested similarity.  There are many reasons to suspect this, but I am likely to also be accused of drafting a too long post - so another time.

Summary: buy what you like, but there is merit in the pursuit of “better”.  And, the companies making our gear are doing the same.
my $.02.

mike_in_nc
1,163 posts
 
Interesting analysis.

Yes, speakers are getting brighter. And yes, many have a bump up around 10 kHz to add immediacy.

I think the less irritating response of old speakers is one reason for the boom in "vintage" audio.
erik_squires OP
16,404 posts
 
The sonic symptoms he describes can be attributed to running without the bass alignment filter. I say this based on personal experience having owned at one time all of the matrix line. Sill own matrix 800’s.

@ct0517

You are right, this would have been good for completeness, but also, look at what they are doing, they give a boost at 26 Hz or so, but is about +0.5 dB by 100 Hz. Troels charts only start at 50 Hz, and his upgrade does not actually change the bass very much, except to add a notch filter. Most of what he does is fix the midrange, and point the tweeter down. Shame we can’t listen to all 3. :) In other words, the bass to mid/treble shelf would have remained, unless I missed something else about the bass alignment filter.

Based on my own in-room measurements, I’d be curious if I’d even need it. There’s a lot of bass in small rooms. :) The bass alignment filter may have been nothing more than specmanship. Look, look, we can do 20 Hz! When in fact the speaker plus room did not require it.
ct0517's avatar
ct0517
2,041 posts
 

Erik

the real genius behind the BAF.

please see the attached graph that B&W England sent me many years ago.

https://photos.app.goo.gl/4kkpoZsV6jJERspy7

Now this is for the 801 matrix, but it applies to the whole matrix line.

See how the db levels drop quickly on the 801 below 100hz. Trust me the graph for the 802 with the much smaller box is a far steeper slope.

**************

The BAF provides multiple benefits.

One it ensure a better slope to the bottom most octave and keeps the ohm requirements higher so it is EASIER on the amp.

Two - it provides for full range reproduction of the music if the notes are there - and in rooms which would not normally support this.

The only B&W matrix speaker that is full range without the BAF is the 800 matrix. But if it is in a room that doesn’t support the bottom two octaves ? That’s where the BAF bump fixes things.


**************

something else to note.

The B&W John Bowers Matrix Era coincided with the Krell / D’Agostino era. B&W and Krell were partners. Like B&W and Classe.today. fwiw B&W bought Classe but I think they - B&W - sold the Classe unit.

Now..... Dan D’Agostino had Matrix 800’s in his living room !

I know, because I have discussed this with him.

The BAF was so Integral to the execution of this Matrix line that Krell came out with their own version of the B&W BAF .

https://photos.app.goo.gl/r6jQSw1Cqu7NwtRM6

Years ago I tried 4 of the different branded B&W BAF’s out there and the Krell is the best one - I felt & the one I kept.

Cheers Chris
More to discover