KBK said:
So please ... don't leave us in darkness ... Shine the light and tell us what is listenable to you ! I've heard quite a few speakers in my life, but by no means I've heard them all.
What are those magic cookies ?
in my country I use to that fact that everybody are experts of building speakers. 🙂 I like to listen to this expert made speakers. they are nice speakers 🙂
the more complex filter the more "right" sound is. right is not nessesary "natural"
past two years I like 1st order series pasive filters. if woofer sloped acousticaly due enclosure axis , ther is nothing to prevent to use 1st order series filters with well behaved drivers.
last week I did some experiment with no cap in tweeter series xover , and find it very three dimensional and not harsh sounding xover. no wonder why sonus faber use it. regards wilsons.
they are flat spekers. thought they are designed to use at home or studios, not anechoic chamber.😉 try to measure wilson spekers mooved in room centerfrom 1 meter ON AXIS and you will get +-7db .
the more complex filter the more "right" sound is. right is not nessesary "natural"
past two years I like 1st order series pasive filters. if woofer sloped acousticaly due enclosure axis , ther is nothing to prevent to use 1st order series filters with well behaved drivers.
last week I did some experiment with no cap in tweeter series xover , and find it very three dimensional and not harsh sounding xover. no wonder why sonus faber use it. regards wilsons.
they are flat spekers. thought they are designed to use at home or studios, not anechoic chamber.😉 try to measure wilson spekers mooved in room centerfrom 1 meter ON AXIS and you will get +-7db .
elviukai said:
In my country there is a saying that opinions are like light shining devices - everybody has one 😀
Seriously, I'm old enough to see the same old adage of "revolutionary solutions" that magically improve sound reproduction to previously unheard levels. Like every other science, true breakthroughs in loudspeaker design are few and far between. Yes, the overal direction is forward, but steps are rather Small (with apologies to Thiele 😉 )
Bratislav
There are so many different designs out there that it's hard to find one with drivers that are live enough to be alive, yet dead enough to not have sound of their own. I think probably finding drivers that can perform as such is very difficult. For people that know how to analyze driver diaprams and motors know what ideally can solve the problem while the solution might not be financially beneficial because most people don't care so much, and companies cannot survive.
So the quality of candidates reflect the quality of the mass customers. Same as any democratic process.😀
So the quality of candidates reflect the quality of the mass customers. Same as any democratic process.😀
RobWells said:
I still think i'd be doing that actively (as in a couple approriately sized coupling caps in the amp)
dave
Re: Re: Re: Re: Re: Re: Re: Q
It was a harsh reply to a harsh generalisation by KBK.
There was no mention of active/passive, only that loudspeakers with 4th order filtering are all bad.
planet10 said:
It was a harsh reply to a harsh generalisation by KBK.
There was no mention of active/passive, only that loudspeakers with 4th order filtering are all bad.
Re: Re: Re: Re: Re: Re: Re: Re: Q
Many people have different experiences and different preferences. It would be nice to understand whether such preferences were just based one a few tries or some technical analysis. Very high order XO seems interesting, yet, due to the massive components, there might be more signal loss along the line making it sound a bit dead. Probably the higer order the more benefit there is to use a digital XO like the 2496.
ShinOBIWAN said:
Many people have different experiences and different preferences. It would be nice to understand whether such preferences were just based one a few tries or some technical analysis. Very high order XO seems interesting, yet, due to the massive components, there might be more signal loss along the line making it sound a bit dead. Probably the higer order the more benefit there is to use a digital XO like the 2496.
I don't see why everyone has to beat up on Wilsons. Every time I hear one, young or old, I like the sound. They're not the best speaker in the world and they're certainly not worth the money new, but they are a very well designed speaker.
The price is part of its marketing. If you make a superior product you're better off pricing it high and ******* some people off than trying to compete on price or market a product's value. Being expensive is more of an endorsement than engineering prowess. Most customers don't know anything about cabinet resonances, crossover slopes, etc. But they see a heavy price tag, a reputation and high build quality and they're sold. I would guess most owners are happy with them.
Lots of the enjoyment of a speaker or any audio product is suggestive. The best brains in the industry recognize this. When something looks good we spend more time on the better aspects of how it sounds. I love Sonus Fabers. I think they're heavily compromised designs. I find many problems with their sound, but they're so beautiful I forgive them. I would love to own some.
I think many people get hung up on the finer aspects of design that they ignore the big picture. Many engineering choices - which take heroic feats of thought and imagination - make changes that are so subtle you have to ABX them to be sure, if then. (Take the debate over a speaker's impulse response!) When I got into design - after many years of being involved with music - I had to train myself to hear what speakers sound like. Most people never do. Most people just listen to music. (Yes, there are 'audiophiles', who are critical like engineers without a real understanding of what they're hearing.)
Anyway, my point is that I think many engineering-oriented people miss the fact that ultimately you're designing a very highly specialized piece of furniture - not a lab instrument. You have to compromise to the fact that people like these beautiful, expensive things that make music. If you're designing gear for normal people most of your design choices will be lost on them. I don't mean they don't understand them, I mean they won't hear them.
People talk about Wilson like he's a criminal. He's selling overpriced, yet well engineered gear to people who can afford it. Many of the high end manufacturers are doing more of a service to the world than the companies like Yamaha who pump out 500$ home theater setups that sound like dookie. Everytime I see one of those systems my heart hurts. The margins are so thin they had to fire the designer first! Why not charge a little more and design something 50% better? Makes me sick. (Or even respected companies like Linn - their new home theater setup rolls off at 80 to an 8-inch sub!!! The only thing an 8-inch sub produces is harmonic distortion.)
The price is part of its marketing. If you make a superior product you're better off pricing it high and ******* some people off than trying to compete on price or market a product's value. Being expensive is more of an endorsement than engineering prowess. Most customers don't know anything about cabinet resonances, crossover slopes, etc. But they see a heavy price tag, a reputation and high build quality and they're sold. I would guess most owners are happy with them.
Lots of the enjoyment of a speaker or any audio product is suggestive. The best brains in the industry recognize this. When something looks good we spend more time on the better aspects of how it sounds. I love Sonus Fabers. I think they're heavily compromised designs. I find many problems with their sound, but they're so beautiful I forgive them. I would love to own some.
I think many people get hung up on the finer aspects of design that they ignore the big picture. Many engineering choices - which take heroic feats of thought and imagination - make changes that are so subtle you have to ABX them to be sure, if then. (Take the debate over a speaker's impulse response!) When I got into design - after many years of being involved with music - I had to train myself to hear what speakers sound like. Most people never do. Most people just listen to music. (Yes, there are 'audiophiles', who are critical like engineers without a real understanding of what they're hearing.)
Anyway, my point is that I think many engineering-oriented people miss the fact that ultimately you're designing a very highly specialized piece of furniture - not a lab instrument. You have to compromise to the fact that people like these beautiful, expensive things that make music. If you're designing gear for normal people most of your design choices will be lost on them. I don't mean they don't understand them, I mean they won't hear them.
People talk about Wilson like he's a criminal. He's selling overpriced, yet well engineered gear to people who can afford it. Many of the high end manufacturers are doing more of a service to the world than the companies like Yamaha who pump out 500$ home theater setups that sound like dookie. Everytime I see one of those systems my heart hurts. The margins are so thin they had to fire the designer first! Why not charge a little more and design something 50% better? Makes me sick. (Or even respected companies like Linn - their new home theater setup rolls off at 80 to an 8-inch sub!!! The only thing an 8-inch sub produces is harmonic distortion.)
Re: Re: Re: Re: Re: Re: Re: Re: Re: Q
The sharp slopes tend to 'trap' resonant noise in the fiters themselves, and do nothing to tame noise created in the drivers, at both harmonic and subharmonics of their passband..ie, outside of the crossover-amplifier's capacity to damp. The temporal 'smear' is heard by the ear and locates itself to the driver, by the ear, as a noise that is not musical, or is a distinct and separate part-that is not part of the original signal. The drivers then, of course, are briging attention to themselves as distinct and separate devices, that are not connected to the music signals.
The very large part of 24db (passive) crossovers have some low grade parts in them, due to their sheer mass of components. This, of course, dictated by the economics. Not that 'doing it right ($$$!!) would provide any better results. So, internal reflections within the crossover components as a group..temporal noise..and the out of band noise is left hanging on it's own. No damping. Not that lower slope crossovers are much better. It's all the art of compormise, and the high slopes have advantage in polar etc response patterns, but fall flat elsewhere. Then, the hammering that the amplifiers endure, as Planet10 speaks of. More grist for the mill.
The solution COULD be digital crossovers, except for the fact that they perform worse, outside of their one trick pony aspect.
Doing the rounds at the last Canadian Audio show at the Sheraton in downtown Montreal this past March, and speaking directly with the desigers, we all, almost to a man(!), agree..that digitial may be a nice 'one trick pony', but it ain't ready for prime time. And no, Shin, we're not Neanderthals or idiots. We'd all gladly hop onto a device, method or system of design or implementation method that brings us better quality and better sound, advantage in the market, etc, etc....but digital -stated flatly-....does notyet do so. This, with regards to the 'highest end' 🙂 Sorry. 🙂
In my humble experience, you start at 6db per oct..and work your way up to not much more than a compensated 12, which is an effective 18. I also like compensated 12's and 18's..where the compensation 'leg' is quite far up the slope. I try to do this on one driver alone, it tends to make for better phase intergration. Maybe play with some resistive/LCR loading of a way effective/efficent tweeter, and get some phase alignement/agreeance that way. Sonus Faber's method works well, but requires lots of headroom in a tweeter, ie EFFICENCY. You need a big crowbar (capacity to insert LCR combinations to crowbar phase intergration) and it can only be enacted with lots of resitance (in this case, not futile) and requires that you start with high efficency. However, too much resistance can create a situation where the driver cannot be effectively damped. A serious problem. Using two amps for level matching can solve thse problems/issues to a great degree, as less resistance can be used and simple phase flips on amp connections can help..but of course, crossovers need to be designed around specific set-ups. All inclusive. and thus, complex for end users. Not an issue for DIY/Personal use.
And keep tweeking it, over a very long term of analysis. A year or so, per crossover, in my experience. Sometimes you get lucky, and can stop after about 2 months or so. Depends on the driver combinations. Then you've possibly got something.
Active/opamp has most of the same phase issues as passive so it ain't all that much of an imporvment. It is, yes, but results will differ, of course. IMHO.
soongsc said:
The sharp slopes tend to 'trap' resonant noise in the fiters themselves, and do nothing to tame noise created in the drivers, at both harmonic and subharmonics of their passband..ie, outside of the crossover-amplifier's capacity to damp. The temporal 'smear' is heard by the ear and locates itself to the driver, by the ear, as a noise that is not musical, or is a distinct and separate part-that is not part of the original signal. The drivers then, of course, are briging attention to themselves as distinct and separate devices, that are not connected to the music signals.
The very large part of 24db (passive) crossovers have some low grade parts in them, due to their sheer mass of components. This, of course, dictated by the economics. Not that 'doing it right ($$$!!) would provide any better results. So, internal reflections within the crossover components as a group..temporal noise..and the out of band noise is left hanging on it's own. No damping. Not that lower slope crossovers are much better. It's all the art of compormise, and the high slopes have advantage in polar etc response patterns, but fall flat elsewhere. Then, the hammering that the amplifiers endure, as Planet10 speaks of. More grist for the mill.
The solution COULD be digital crossovers, except for the fact that they perform worse, outside of their one trick pony aspect.
Doing the rounds at the last Canadian Audio show at the Sheraton in downtown Montreal this past March, and speaking directly with the desigers, we all, almost to a man(!), agree..that digitial may be a nice 'one trick pony', but it ain't ready for prime time. And no, Shin, we're not Neanderthals or idiots. We'd all gladly hop onto a device, method or system of design or implementation method that brings us better quality and better sound, advantage in the market, etc, etc....but digital -stated flatly-....does notyet do so. This, with regards to the 'highest end' 🙂 Sorry. 🙂
In my humble experience, you start at 6db per oct..and work your way up to not much more than a compensated 12, which is an effective 18. I also like compensated 12's and 18's..where the compensation 'leg' is quite far up the slope. I try to do this on one driver alone, it tends to make for better phase intergration. Maybe play with some resistive/LCR loading of a way effective/efficent tweeter, and get some phase alignement/agreeance that way. Sonus Faber's method works well, but requires lots of headroom in a tweeter, ie EFFICENCY. You need a big crowbar (capacity to insert LCR combinations to crowbar phase intergration) and it can only be enacted with lots of resitance (in this case, not futile) and requires that you start with high efficency. However, too much resistance can create a situation where the driver cannot be effectively damped. A serious problem. Using two amps for level matching can solve thse problems/issues to a great degree, as less resistance can be used and simple phase flips on amp connections can help..but of course, crossovers need to be designed around specific set-ups. All inclusive. and thus, complex for end users. Not an issue for DIY/Personal use.
And keep tweeking it, over a very long term of analysis. A year or so, per crossover, in my experience. Sometimes you get lucky, and can stop after about 2 months or so. Depends on the driver combinations. Then you've possibly got something.
Active/opamp has most of the same phase issues as passive so it ain't all that much of an imporvment. It is, yes, but results will differ, of course. IMHO.
KBK said:
There's so much that is false or exaggerated in that post that I don't care to waste time offering a more grounded perspective.
KBK said:
We've done this little dance before, did it change anything? No.
We're both spouting the same stuff as back then hence my complete disinterest in wasting further time and energy, I'm guessing you feel the same way.
Wilson Audio know exactly what they are doing. Making products that sell to a particular niche market for lots of money. Are you guys jealous or something? 🙂
Re: Q
Can we have some facts, please ? You know what they say about opinions ... You have measured this "noise" of course ? Can we see some results ?
Try as hard as I could, ATC100s sound wonderfully coherent, despite a passive 24dB filter in them. Can't really say I can localise separate drivers, but hey, I've been called deaf by some golden ears before. So let's measure the sucker ... nope. IMP doesn't show any "noise" above expected harmonic residuals. What should I be looking for ?
I see ... you don't really understand much of the filter theory, do you ? Look up "Linkwitz-Riley" and read up the part that talks about phase and group delay preservation.
I just love to read about unrecognised geniuses that with one hand wipe off all of the knowledge built over centuries. I guess Butterworth, Chebyshev, Cauer, Bessel, Gauss, Messrs L-R have wasted their time ?
I suggest try your theories with IEEE, submit the article and see how they respond.
KBK said:
Can we have some facts, please ? You know what they say about opinions ... You have measured this "noise" of course ? Can we see some results ?
Try as hard as I could, ATC100s sound wonderfully coherent, despite a passive 24dB filter in them. Can't really say I can localise separate drivers, but hey, I've been called deaf by some golden ears before. So let's measure the sucker ... nope. IMP doesn't show any "noise" above expected harmonic residuals. What should I be looking for ?
I see ... you don't really understand much of the filter theory, do you ? Look up "Linkwitz-Riley" and read up the part that talks about phase and group delay preservation.
I just love to read about unrecognised geniuses that with one hand wipe off all of the knowledge built over centuries. I guess Butterworth, Chebyshev, Cauer, Bessel, Gauss, Messrs L-R have wasted their time ?
I suggest try your theories with IEEE, submit the article and see how they respond.
Re: Re: Q
Messrs L-R? That's funny how you snuck them in there too! In addition to their well known contributions in stereo modulation of FM they also wrote the defacto bible on turn signals.😀
Bratislav said:
Messrs L-R? That's funny how you snuck them in there too! In addition to their well known contributions in stereo modulation of FM they also wrote the defacto bible on turn signals.😀
This is my opinion as well. Maybe it is their customers who don't know what Wilson is doing. 😉
As I already mentioned the Sophia is the only Wilson speaker that I heard personally so far. IMO it sounded nice with the music that was played at the given location together with the components it was demoed with. It didn't sound as REAL as I am used to and I wouldn't buy it personally. But I can imagine that there must be quite some people around that will like its sound.
I don't need any third-party expert who helps me to build this opinion, I am old enough by myself. If this speaker's sound is torturing someone else - so be it ......
Regarding fourth order X-overs:
They have their merits regarding IMD and polar response but I don't use them personally either. And that they allow less control of the driver's nonideal behaviour sounds logical as well if they are passive.
Active is the only way to go for me. But by saying that active filters were haunted by the same phase problems as their passive counterparts is dismissing the fact that there are other active filter topologies around than those boring oldfashioned strings of Sallen-Key sections.
It IS possible to use active filters that take the drivers' behaviour into account and that also sum flat in amplitude and phase.
Developing these takes mathematical skills, the use of measuring equipment and listening tests.
But with listening tests alone it is impossible to develop a decent speaker. Be it an active or a passive one of either transient-perfect or -imperfect behaviour.
Regards
Charles
- Status
- Not open for further replies.