The making of: The Two Towers (a 25 driver Full Range line array)

Beolab 90 measurements: Bang & Olufsen BeoLab 90 loudspeaker Measurements | Stereophile.com

Not the STEP response I was hoping to see:
117Beo90fig5.jpg


Reading trough the text on that page, it does use mid/side EQ, interesting and not surprising at all. I knew there would be similarities with my own DSP path.

I get the impulse/STEP thing... these speakers are meant to play without any delay (however short) to back up their TV etc. Still it is a compromise. Now I get why Geoff, B&O's 'tonmeister' thought the Eric Clapton live recording was disturbed by it's unfiltered low frequency material. I did not think that at all, in fact, it enhances that recording for me. Bringing me closer to that performance.
 
During these Christmas days we've listened to a lot of (online) radio and we watched a movie which was fun. We viewed "a Christmas Carol (2009)".

At one point I checked the RadioShack SPL meter, it had peaks over 100 dB at our listening couch. Especially during the chase sequence with the ghost of Christmas future following Ebenezer through the streets of London.
attachment.php


All in all not to many challenges in this movie in the low end, the arrays handled it well. The scare factor worked though, both my girlfriend as well as my boy freaked out regularly. Many realistic 'doors slamming shut' throughout this movie and similar effects.

The door knob scene was fun...
EJM90F.gif


Internet radio worked reasonably, the quality of it did disappoint me though. We were listening to a countdown from the "top 2000 of all time" songs and I did get to pick out some songs we liked that worked remarkably well.
 

Attachments

  • maxresdefault.jpg
    maxresdefault.jpg
    56.4 KB · Views: 426
I make it a point to watch as many Xmas movies as I can in December... I have to try to get into the mood as there's no snow, and my house is the only one with a tree around here!

I also give a viewing to our students of a Xmas movie, and funny enough, this year was the same movie as yours!

Happy holidays!

(as for me, I'm already back to work... Xmas is not a holiday here... I'll have a winter holiday with the Chinese New Year which is at the end of January this year)
 
Beolab 90 measurements: Bang & Olufsen BeoLab 90 loudspeaker Measurements | Stereophile.com

Not the STEP response I was hoping to see:
117Beo90fig5.jpg


Reading trough the text on that page, it does use mid/side EQ, interesting and not surprising at all. I knew there would be similarities with my own DSP path.

I get the impulse/STEP thing... these speakers are meant to play without any delay (however short) to back up their TV etc. Still it is a compromise. Now I get why Geoff, B&O's 'tonmeister' thought the Eric Clapton live recording was disturbed by it's unfiltered low frequency material. I did not think that at all, in fact, it enhances that recording for me. Bringing me closer to that performance.
MartinLogan Masterpiece Renaissance ESL 15A loudspeaker Measurements | Stereophile.com

not much better either. yet I know they are incredible pair of speakers. you seem to think your able to evaluate a speaker with measurements?
 
No, you get the wrong idea, I know I can :).

As I already mentioned, the STEP shown from the Beolab 90 is not trying to compensate for the timing error introduced by the crossover topology. Perfectly understandable for a speaker that has to play without any initial delay.

Where exactly did I say the Beolab 90 is a bad speaker? Or have a bad sound? Do not try to put any meaning in my words that aren't there, please. I applaud the effort that was put into that marvellous piece of engineering.

Why would you come to that conclusion is beyond me. Yes, I prefer a time coherent result. That doesn't make me the boogie man denying any other speaker will have or achieve excellent results without it.

Maybe you should read what John Dunlavy had to say about it, his word will probably have way more meaning, at least that's what I'm hoping for:
John Dunlavy: Oh, no. Listening comes later. Because if you stop to think about it, no loudspeaker can sound more accurate than it measures. It may sound worse, or it may sound sweeter, prettier, but if we're talking about absolute accuracy—the ability of the speaker to reproduce as perfectly as possible whatever's fed to it—such a system can never sound more accurate than it first measures. So we try to get the greatest accuracy we can achieve from measurements. Then we begin doing what some people might call "voicing," because the best set of measurements are still open to interpretation.

You could have a speaker that makes ±1dB, for example. And you say, "Gosh, how can you better that?" But what if it's +1dB over an octave and a half, say from 1-2.5kHz? And then it suddenly jumps down to -1dB, for a total change of 2dB, for the next octave? You're going to hear a spectral imbalance. The specs might look great—golly, here's a speaker that measures ±1dB—but it's not going to sound nearly as good as a speaker that is up and down 1dB every third of an octave. In the real world we're used to hearing that—reflections from the walls of the room cause similar variations—and we tune that out.

It may come as a surprise—this is giving away a trade secret—but when I design a loudspeaker, I first design it by looking at the step response. I find that by playing around with the crossover network while observing the step response in real time, any change I make is immediately available. When I get the step response right, everything else goes along. It's implicit. It goes along for the ride.

So, to those who like to call us "technocrats" or whatever, we would suggest that those who design by "voicing" loudspeakers are working with an enormous number of perturbations. Did they have their ears examined the morning they designed the speaker? What kind of a room did they voice it in? What program material did they listen to to voice it? You go on and on and on. The permutations and combinations are mind-boggling, to say the least.

When I was a little brat and color photography was just coming in—this was back in the late '30s—grass tended to look greener than green and all reds were fire-engine red, and the sky was super-blue. And everyone was saying, "Oh, aren't these photos wonderful!" It was almost better than real! It was surrealistic instead of realistic. And people initially liked it.

But what we're finding more and more is that you get tired of that. Nature is pretty clever; we've evolved naturally under the influence of our environment, and nature has tended to provide us with a color spectrum that, over a long period, is more satisfying than surrealistic colors. The same is true with sound. We find that a good percentage—not everyone, certainly—of young people today like "boom-boom boxes" on the bottom end, they like maybe a high end that rises a bit, and that's fine. That's like the surrealistic color on old postcards. But in the long run, you grow tired of it.

Recently I read in one of the hi-fi magazines that accuracy doesn't mean anything: "As all violins sound different, there cannot be a concept of what an accurate violin actually sounds like." To me that's illogical. If I buy the latest violin recording made with a very rare 17th-century violin or whatever, I want to hear what that particular violin sounds like. I don't want to reproduce it so it sounds like a generic violin.

Read more at Loudspeaker designer John Dunlavy: By the Numbers... | Stereophile.com

Now chew on that and learn why he said that. I agree with his opinion. Read the full interview, if you want more after that I can provide it. It makes perfect sense. Even though many will deny the importance.

Looking at the Martin Logan speakers and it's STEP:
117MLRenfig5.jpg

That IS much better (than the Beolab 90) from a timing point of view. Did you miss that? Chances are they measured it up close and did not record the STEP I'd like to have seen.

If you start to lookup the STEP responses from Dunlavy's speakers things might become more clear:
99DAL4fig4.jpg


Though a lot more is needed to examine the speaker results as a whole to form any opinion on it. Time coherency is just one of many attributes I was after. Tonal balance is a way bigger pet peeve for me. But if you can have both...

The Beolab 90 project focussed on room integration, and I've followed that project as it went as that's an area I've played with too.
I was curious to see their STEP response, I've even asked for it on Geoff's blog. You can learn so much from measurements, if you take the time to really dig in.
I'm not on a soap box, you either agree or disagree with me. Both are fine by me. Even if it's neither :).
I do not condemn anyone who thinks otherwise, I just happen to think John Dunlavy makes some indisputable claims on the topic of musical reproduction. It isn't the easiest road to travel, it's not "just" a phase edit to get what he talks about. In fact how he got the results was even more of a challenge, I would have loved to see what the project mentioned on page 5 of that article would have been capable of.
 
Last edited:
No, you get the wrong idea, I know I can :).

Let's put some more word in on this subject. Basically this trip or thread shows my learning curves trough the world of measuring speakers as well as documenting it's build. I cannot put to words the many things I've learned along the way. I didn't do this to present myself on a pedestal or make myself more important than I am... I just tried to put into words what I have learned along the way, while measuring myself to better sound.
I'm not done learning, doubt if that will ever happen :). But I do know I managed to measure, learn and achieve better sound that way out in my room. No more, no less.

I am convinced measurements tell us what's going on. It's not a voodoo art. The learning parts are knowing what to look for. The real magic is in the music. Not in the reproduction. Though every detail counts.
 
It probably came out wrong, was more a question. sorry.

on the positive side, it extracted from you interesting information.
thanks
No, you get the wrong idea, I know I can :).

As I already mentioned, the STEP shown from the Beolab 90 is not trying to compensate for the timing error introduced by the crossover topology. Perfectly understandable for a speaker that has to play without any initial delay.

Where exactly did I say the Beolab 90 is a bad speaker? Or have a bad sound? Do not try to put any meaning in my words that aren't there, please. I applaud the effort that was put into that marvellous piece of engineering.

Why would you come to that conclusion is beyond me. Yes, I prefer a time coherent result. That doesn't make me the boogie man denying any other speaker will have or achieve excellent results without it.

Maybe you should read what John Dunlavy had to say about it, his word will probably have way more meaning, at least that's what I'm hoping for:


Now chew on that and learn why he said that. I agree with his opinion. Read the full interview, if you want more after that I can provide it. It makes perfect sense. Even though many will deny the importance.

Looking at the Martin Logan speakers and it's STEP:
117MLRenfig5.jpg

That IS much better (than the Beolab 90) from a timing point of view. Did you miss that? Chances are they measured it up close and did not record the STEP I'd like to have seen.

If you start to lookup the STEP responses from Dunlavy's speakers things might become more clear:
99DAL4fig4.jpg


Though a lot more is needed to examine the speaker results as a whole to form any opinion on it. Time coherency is just one of many attributes I was after. Tonal balance is a way bigger pet peeve for me. But if you can have both...

The Beolab 90 project focussed on room integration, and I've followed that project as it went as that's an area I've played with too.
I was curious to see their STEP response, I've even asked for it on Geoff's blog. You can learn so much from measurements, if you take the time to really dig in.
I'm not on a soap box, you either agree or disagree with me. Both are fine by me. Even if it's neither :).
I do not condemn anyone who thinks otherwise, I just happen to think John Dunlavy makes some indisputable claims on the topic of musical reproduction. It isn't the easiest road to travel, it's not "just" a phase edit to get what he talks about. In fact how he got the results was even more of a challenge, I would have loved to see what the project mentioned on page 5 of that article would have been capable of.
 
It probably came out wrong, was more a question. sorry.

on the positive side, it extracted from you interesting information.
thanks

Hahaha, well you got me going :D

I'm just as sensitive and fanatical as all the rest, no doubt. It comes from passion (at least I like to think so ;)).
If you do want more Dunlavy material, I kept a series of images found on a forum, I tried to find them again. Browsing Google you'll be able to find a DOC file containing all of his old posts on rec.audio.* newsgroups:

http://www.google.nl/url?sa=t&rct=j&q=&esrc=s&source=web&cd=5&ved=0ahUKEwj_zMn6kJXRAhUIMFAKHVQAC-cQFgg_MAQ&url=http%3A%2F%2Fwww.avsforum.com%2Fforum%2Fattachment.php%3Fattachmentid%3D9901&usg=AFQjCNHR8kwKjpvJIDcYyhoLl845QGS3Jw&sig2=i7fHgp7OnCldvBRNNxQJRg

Meanwhile I'll try and hunt down the images, I don't feel right about uploading them as I did not scan those myself.

I have to add: I equally enjoy reading Toole, our own forum visitors Geddes, Smith and Danley, as well as Keele and many more. There's so much to learn from all of them.
 
Last edited:
Due to the holidays my speakers have been working overtime. Best part: my girl constantly asks me if we can play music over the speakers. Still busy with the countdown of the top 2000 songs of all time.

In between a movie here and there. Reminds me of 2 years ago when I first had sound from these two speakers. Things have improved though. Measurements can guide the way.

Meanwhile I'm having a hard time not to think of what adding two 12" woofers could do :cool:. If only I could make up my mind as to what is needed. I would buy the Scan Speak 30W in a heartbeat, but can't help seeing a more wobbly low end than the 32W in the charts. Is this the vented pole piece vs the dust cap vent on the cheaper Disco sub? The 32W is out of my league. I simply can't afford that. Are there any valid competitors? Without adding more mms like most of the Dayton subs?
As you guys know by now, measurements can convince me.

Looking for 12" woofers, low Fs peak, about 15 mm x-max, lots more x-mech, preferably with a nice flat BL curve and mms around 130 gram (or less).

I did eye Car subs, like Alpine's 'Type R' but the lack of info on that is saddening. Nice BL curve though, would love to see the rest of the Klippel measurements.

I'm not looking for bigger than 12". Just the best mix between quality and budget. So far the 30W is still my prime candidate. The SB acoustic sub didn't beat it from what I've seen measurement wise. The less I have to fix in DSP, the better the results can be. So a clean impulse is the first thing I'll look for.
 
Peerless XLS need not apply, but the Peerless XXLS is basically the same or similar enough to Scan Speak's 30W. Funny enough, the Scan Speak is more affordable.
I will have more than 2, I already have 50 real tiny ones :) that will still continue to contribute. I'm only considering adding 2 subwoofers to gain (even) more control over the whole thing.
 
Last edited:
Due to the holidays my speakers have been working overtime. Best part: my girl constantly asks me if we can play music over the speakers. Still busy with the countdown of the top 2000 songs of all time.....

.....:up:

.....Meanwhile I'm having a hard time not to think of what adding two 12" woofers could do :cool:. If only I could make up my mind as to what is needed. I would buy the Scan Speak 30W in a heartbeat, but can't help seeing a more wobbly low end than the 32W in the charts. Is this the vented pole piece vs the dust cap vent on the cheaper Disco sub? The 32W is out of my league. I simply can't afford that. Are there any valid competitors? Without adding more mms like most of the Dayton subs?
As you guys know by now, measurements can convince me.......

For me it looks like all SS charts are wobbly below 100Hz be it Discovery, Revelator, or Illuminator except that lonely 32W/4878T00, so maybe it has been measured more carefully in more bass friendly environment because its their top model. For example look the wobbly signs in SS datasheets for 10F/8824 plus 26W/4558T00, and then see it isn't replicated in Klang & Ton or Voice Coil measurements here http://www.scan-speak.dk/datasheet/reviews/10F-8424G00-KlangTon2010-2.pdf, http://www.scan-speak.dk/datasheet/reviews/D2604-833000_26W-4558T00-VoiceCoil_2011-2.pdf.

ooh and Happy new year : )