Speaker Shootout - two of the most accurate and well reviewed speakers ever made - Page 44 - AVS Forum | Home Theater Discussions And Reviews
Forum Jump: 
 1624Likes
Reply
 
Thread Tools
post #1291 of 1494 Old 08-31-2017, 10:27 PM
AVS Forum Addicted Member
 
noah katz's Avatar
 
Join Date: Apr 1999
Location: Mountain View, CA USA
Posts: 21,563
Mentioned: 3 Post(s)
Tagged: 0 Thread(s)
Quoted: 1095 Post(s)
Liked: 380
Quote:
Originally Posted by andyc56 View Post
In this post, Noah seemed to suggest that SFM is no different from conventional EQ.
Water under the bridge, but that's not what I was suggesting at all.

I was asking only about SFM setup complexity.
noah katz is offline  
Sponsored Links
Advertisement
 
post #1292 of 1494 Old 08-31-2017, 10:39 PM
AVS Forum Addicted Member
 
noah katz's Avatar
 
Join Date: Apr 1999
Location: Mountain View, CA USA
Posts: 21,563
Mentioned: 3 Post(s)
Tagged: 0 Thread(s)
Quoted: 1095 Post(s)
Liked: 380
Quote:
Originally Posted by Floyd Toole View Post
...Because of the efficiency gains in these schemes, some or all of the subs can be smaller...

Something just occurred to me re SFM.

I have two huge built in subs behind the front and rear midwalls.

If I add a third in the one available corner, is there any reason to think that one of the existing ones wouldn't become one requiring much less drive?
noah katz is offline  
post #1293 of 1494 Old 08-31-2017, 10:43 PM
Member
 
Join Date: Aug 2017
Posts: 68
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 46 Post(s)
Liked: 36
Quote:
Originally Posted by noah katz View Post
Water under the bridge, but that's not what I was suggesting at all.

I was asking only about SFM setup complexity.
I queried NK about that but we only talked past each other for a while. My question remains whether averaged proper multiple precision level-only measurements and precision level-only EQ per same-polarity sub would or could result in the same outcome at the listeners' ears as SFM.

(An acoustician friend suggested that perhaps it additionally addressed what is called frequency dispersion, though noted that audibility of that was unlikely in air, just conceivable.)

Last edited by davidrmoran; 08-31-2017 at 10:44 PM. Reason: fix error
davidrmoran is online now  
 
post #1294 of 1494 Old 09-01-2017, 03:52 AM
AVS Forum Special Member
 
Join Date: Dec 2014
Posts: 2,985
Mentioned: 36 Post(s)
Tagged: 0 Thread(s)
Quoted: 1227 Post(s)
Liked: 701
The simple answer is with EQ only, no SFM would not be achieved. You need to be able to adjust delay and volume independently for each sub as well. Decent placement of the subs will give the consistent seat to seat response, if one has subs stacked in the same corner SFM will not be as effective. Once subs are placed in the room you can adjust the delay between each sub to smooth out frequency response. You can EQ per sub to get more precision but even without EQ you can usually get a good result with just using the delay and of course volume going to each sub. Once this is set then the LFE channel can be EQ'd with the combined response of the subs, because in the end we need to EQ the LFE channel. And lastly the phase needs to be set between the speakers and sub to have good integration at the crossover frequency. SFM is just applying techniques that have been done by calibrators for a long time and automating the process. Since this can be a tedious process it is a nice thing to automate

Like many auto EQ systems they have some advantages and disadvantages. For instance one can use multiple filters to wave shape a correction because not all things that need to be fixed are nice and neat sine waves that an inverse correction with a single filter can easily be made. This is one area where machines can be make a squiggly line flat (or to a target curve) much better than humans. The problems with many auto EQs is they often don't know if something should be corrected or not since they often don't know the root of the anomaly. I haven't used SFM but I imagine by using the delays between subs to smooth response it can get to a better starting point for EQ than without using individual sub delays.
andyc56 likes this.
Ellebob is online now  
post #1295 of 1494 Old 09-01-2017, 07:49 AM
AVS Forum Addicted Member
 
noah katz's Avatar
 
Join Date: Apr 1999
Location: Mountain View, CA USA
Posts: 21,563
Mentioned: 3 Post(s)
Tagged: 0 Thread(s)
Quoted: 1095 Post(s)
Liked: 380
Quote:
Originally Posted by Ellebob View Post
...if one has subs stacked in the same corner SFM will not be as effective...
Not sure about that; why wouldn't the same SFM principles apply?

Seems like PEQ combined w/delay could (though it might not) remedy a null/peak.

Last edited by noah katz; 09-01-2017 at 07:53 AM.
noah katz is offline  
post #1296 of 1494 Old 09-01-2017, 09:04 AM
Advanced Member
 
bigguyca's Avatar
 
Join Date: Oct 2007
Posts: 701
Mentioned: 9 Post(s)
Tagged: 0 Thread(s)
Quoted: 393 Post(s)
Liked: 270
Quote:
Originally Posted by R Harkness View Post
BTW,

Just to add one more thing from my recent speaker auditioning....

Though I adore the sound of my current Thiel speakers, [/B][/U]I'd been seriously looking for a possible replacement for the last 6 months or so, simply because the speakers have to be placed in a way that somewhat impedes entrance into my living room. The smaller, or at least less deep, the speaker is, the better (as the Thiels are tall, and their depth is the biggest problem). Inches make a difference. It happens that thus far, I can't find something I like enough.

That's just to state why I've been going to many audio shows, and audio dealers to audition speakers. My recent encounter with the Revel speakers was at a (very nice) high end shop and I was able to listen to speakers between around $3,000 up to $30,000. As is often the case, the speakers were hooked up to extremely expensive accessories. All speaker for instance hooked up via mind-numbingly expensive Nordost speaker cabling and other expensive interconnects, crazy priced AC cables, and in the the case of some, car-priced amplification as well.

Yet, as has usually been the case, I could hear nothing "better" than I do at home. No more detail revealed, nothing more believable, nothing actually as fully satisfying.
This to me is just another repudiation of the claims about the importance of super high end cables and assorted audio jewelry. At home, my speakers are connected to a long run of basic beldon speaker cable, with basic, competently constructed interconnects. Yet it does all the detail and realism - more to my ears - than I hear from high end speakers using ultra-expensive cables. It really drives home the central role of the choice of speakers (and room interaction) vs getting caught up in spending nutty money on audiophile cables. (Fortunately, I actually cured myself of spending money on most audiophile cables/AC cables many years ago, using my own blind testing of cables).
If your Thiel's speakers are a Jim Thiel design then they have 1st order crossovers. Due to the characteristics of these crossovers and the resultant affect on frequency response, and perhaps other parts of the design, Thiel's speakers generally measure in line with these measurements from Stereophile of the Thiel CS3.7 loudspeaker:

https://www.stereophile.com/content/...r-measurements

or the Thiel CS2.4:

https://www.stereophile.com/content/...r-measurements

Specifically the frequency response of Thiel speakers on and off axis is very uneven. Contrast this contrasts with the measured results from say the Revel Salon2 or other results presented on this thread or available from the Stereophile website. If you have become accustomed to a typical Thiel design then most of Harmon's designs will sound very different from your reference. If you like the Thiel speakers then it wouldn't seem to require too much hand waving to predict that you may not like Harmon designs.
bigguyca is online now  
post #1297 of 1494 Old 09-01-2017, 09:13 AM
Advanced Member
 
bigguyca's Avatar
 
Join Date: Oct 2007
Posts: 701
Mentioned: 9 Post(s)
Tagged: 0 Thread(s)
Quoted: 393 Post(s)
Liked: 270
Quote:
Originally Posted by torii View Post
well i had.....but i only narrowed it down to 5k and under....sry for copy paste but it is what it was.... why did i go where i did? thats alot of competition and i listened to 5+ pair that cost more than i bought. granted many i didnt listen to cause of form factor aesthetics...but still no smart guy has an....

  1. Bowers & Wilkins (12)
  2. Definitive Technology (5)
  3. Elac (3)
  4. Focal (10)
  5. GoldenEar (6)
  6. JBL (2)
  7. Klipsch (5)
  8. Magnepan (3)
  9. McIntosh (3)
  10. Polk Audio (13)
  11. PSB (5)
  12. Revel (5)
  13. Sonus faber (9)
  14. SVS (2)
  15. Tannoy (7)
Does the list represent your ordering of the brands based on some metric(s)?

Is there a significance to the numbers in parenthesis?

Are you a fan of E.E. Cummings? If so, you may wish to remove one item from the list.
bigguyca is online now  
post #1298 of 1494 Old 09-01-2017, 09:17 AM
AVS Forum Addicted Member
 
sdurani's Avatar
 
Join Date: Oct 2001
Location: Monterey Park, CA
Posts: 24,978
Mentioned: 108 Post(s)
Tagged: 0 Thread(s)
Quoted: 5074 Post(s)
Liked: 3547
Quote:
Originally Posted by Ellebob View Post
You need to be able to adjust delay and volume independently for each sub as well.
Yup, SFM is supposed to use independent level, delay and single-band PEQ for each sub to improve seat-to-seat consistency. However, last time I checked (albeit a few years ago), the level adjustment parameter in SFM had been turned off for ARCOS calibrations. Back then the SFM algorithm was willing to do things like sacrifice 9dB of level to get half a dB of greater consistency. You tell a piece of software to maximize consistency and that's exactly what it will do, irrespective of everything else. So, in the field, the level adjustment was off and SFM was using only delays and single-band PEQ. Haven't kept up on whether this has changed over the last few years.
andyc56 likes this.

Sanjay
sdurani is online now  
post #1299 of 1494 Old 09-01-2017, 09:40 AM
AVS Forum Addicted Member
 
noah katz's Avatar
 
Join Date: Apr 1999
Location: Mountain View, CA USA
Posts: 21,563
Mentioned: 3 Post(s)
Tagged: 0 Thread(s)
Quoted: 1095 Post(s)
Liked: 380
Quote:
Originally Posted by sdurani View Post
So, in the field, the level adjustment was off and SFM was using only delays and single-band PEQ. Haven't kept up on whether this has changed over the last few years.

Looks like it has:

Quote:
Originally Posted by Floyd Toole View Post
In my room the power distribution among the subs is 100%, 25%, 6.3%, 6.3%...
sdurani and andyc56 like this.
noah katz is offline  
post #1300 of 1494 Old 09-01-2017, 09:44 AM
AVS Forum Special Member
 
Join Date: Dec 2014
Posts: 2,985
Mentioned: 36 Post(s)
Tagged: 0 Thread(s)
Quoted: 1227 Post(s)
Liked: 701
Quote:
Originally Posted by noah katz View Post
Not sure about that; why wouldn't the same SFM principles apply?
SFM could apply but not for multiple seats. PEQ applies now for a single sub but again not for multiple seats. Placement is still the issue. If we had only on sub ideal placement would be somewhere in the middle of the room and would lose a lot of output. So we use the virtual sub techniques by using two or more subs placed at opposite sides of a given dimension. To give us output and.better seat to seat consistency. What adjusting delays gives us is the ability to virtually move the subs instead of physically moving them to the best location. But to virtually move them we still have to start ok n opposite side of that dimension.

Harman's research of finding good locations for 2 or 4 subs for seat to seat consistency still requires EQ for good response. But when you do use EQ it works for all seats then. Using delays of SFM allows those locations to not be as precise but the subs still need to start at opposite side or close.

SFM I think will be a great tool because when you get more than 2 subs the process gets tedious. SFM also allows us to not have that perfect placement of corners, 1/4, midpoints and will work for more 'real' rooms. But as with many things in acoustics it starts with location, location, location
Gooddoc likes this.
Ellebob is online now  
post #1301 of 1494 Old 09-01-2017, 10:15 AM
AVS Forum Addicted Member
 
noah katz's Avatar
 
Join Date: Apr 1999
Location: Mountain View, CA USA
Posts: 21,563
Mentioned: 3 Post(s)
Tagged: 0 Thread(s)
Quoted: 1095 Post(s)
Liked: 380
Quote:
Originally Posted by Ellebob View Post
SFM could apply but not for multiple seats...

I believe that contradicts itself, as well as much of what Dr. Toole has said previously.

Also, I wasn't restricting it to just two sub locations; the co-located sub could work in conjunction with a 4th sub at a 3rd location.

But even with two locations, at a minimum a co-located sub adds the degree of freedom of another PEQ freq, and yet another if SFM cannot adjust phase of the PEQ freq independently from the rest of the passband.

Last edited by noah katz; 09-01-2017 at 10:47 AM.
noah katz is offline  
post #1302 of 1494 Old 09-01-2017, 10:21 AM
Member
 
Join Date: Aug 2017
Posts: 68
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 46 Post(s)
Liked: 36
The Thiel horizontal radpats I have seen are not good (have not measured any myself) but I am not persuaded this is inherently because of its first-order xovers. Way back when, those were deemed to be a very good goal by the best designers, and I have seen smooth off-axis responses from some such. It depends on how well-behaved and pistonic the drivers are. Otoh, with the off-axis stitching smoothness of something like the Infinity P362, whose FR have by angle I measured extensively, I have a new appreciation for the steeper slopes. So it's all in the engineering and taking the right measurements during design.
davidrmoran is online now  
post #1303 of 1494 Old 09-01-2017, 10:56 AM
AVS Forum Special Member
 
Join Date: Dec 2014
Posts: 2,985
Mentioned: 36 Post(s)
Tagged: 0 Thread(s)
Quoted: 1227 Post(s)
Liked: 701
If all the subs are co-located within a 1/4 wavelength of the crossover frequency (3.5 feet for 80z crossover) SFM would not be that effective for seat to seat consistency. Basically when you place subs close together their response is treated as one. You can adjust the delay between the two to alter the frequency response but not for seat to consistency since acoustically the room still sees them as one sub. Maybe Dr. Toole could explain this if I am incorrect. It would certainly work if you have a couple subs located together and a 3rd or 4th sub elsewhere in the room.

Last edited by Ellebob; 09-01-2017 at 11:22 AM.
Ellebob is online now  
post #1304 of 1494 Old 09-01-2017, 11:06 AM
AVS Forum Addicted Member
 
R Harkness's Avatar
 
Join Date: May 2001
Location: Toronto, Ontario, Canada
Posts: 14,358
Mentioned: 10 Post(s)
Tagged: 0 Thread(s)
Quoted: 2110 Post(s)
Liked: 1820
Quote:
Originally Posted by bigguyca View Post
If your Thiel's speakers are a Jim Thiel design then they have 1st order crossovers. Due to the characteristics of these crossovers and the resultant affect on frequency response, and perhaps other parts of the design, Thiel's speakers generally measure in line with these measurements from Stereophile of the Thiel CS3.7 loudspeaker:

https://www.stereophile.com/content/...r-measurements

or the Thiel CS2.4:

https://www.stereophile.com/content/...r-measurements

Specifically the frequency response of Thiel speakers on and off axis is very uneven. Contrast this contrasts with the measured results from say the Revel Salon2 or other results presented on this thread or available from the Stereophile website. If you have become accustomed to a typical Thiel design then most of Harmon's designs will sound very different from your reference. If you like the Thiel speakers then it wouldn't seem to require too much hand waving to predict that you may not like Harmon designs.
Thanks for the reply. Your take on the Thiel 3.7s seems different than most others as just about every review found them to be, in practice, extremely neutral sounding. Secretes Of Home Theater Fidelity measured a very smooth in room response iirc. And to me, again, I do not hear obvious bulges and frequency dips in the sound (I sit between 6 and 7 feet from mine, in a room designed with an acoustician, FWIW). Also, as I mentioned, instead of the Revel speakers sounding different from what I'm used to, I'd mentioned it was the opposite: I was struck by how similar they sounded. I've never heard any speaker sound closer to the Thiels (giving me the same impression of evenness top to bottom).

Your reasoning seems to assume different speaker measurements would predict what I would and wouldn't like (to some degree). But then I'm not sure how that would explain why I like so many other speakers that measure differently from the Thiels, e.g. these Hales:

https://www.stereophile.com/content/...r-measurements

Or any number of speakers I've owned or auditioned.

I've also loved speakers that have been designed specifically for neutrality and wide even dispersion, even room power response, e.g. these Waveform speakers (designed with Paul Barton using NRC facilities, he's a proponent of double-blind tests as well iirc):

http://www.audio-ideas.com/reviews/l...mach_solo.html

I still own a pair of Waveform speakers and love them. But they don't sound precisely like the the Revels, timbrally, to my ear.

I also much prefer the sound of my MBL omni-directional monitors. They are clearly designed for similar on/off axis sound, and apparently measure that way.

https://www.stereophile.com/content/...r-measurements


"The 101E Mk.II is remarkably flat throughout the midrange and treble,..."


I have the monitor version which uses the same midrange/treble omni-drivers, so I'm likely getting similar sound - even midrange/treble with wide, even dispersion. That's certainly how they sound - no speaker I know sounds so similar no matter how I move around the speaker.

Or I could talk about the Quad ESL 57 electrostatics. I'm sure their measurements will deviate from the Revel measurements, and I'd guess they would not be chosen over the revels in blind tests (?). And yet I can sit transfixed for hours in front of the Quads in a way that the Revels just don't seem to do for me.

So, again, I'm not yet getting to what degree I should predicate what I like, or predict what I like, on measurements. Again, this is FAR from saying measurements aren't useful as they obviously are. It would be simple to produce a set of measurements so "bad" that it would predict just about everyone here would deem the sound terrible. But withing the range of "not terrible" or "pretty good to good" things seem to get trickier.

I don't see any reason why I would be different from most people, if I took part in the HK blind testing. I'd bet I would prefer what most people prefer, and probably pick, blind, Revel speakers over most or all contenders behind the screen. So I have little doubt the HK research is predictive in that respect. It's also, it seems obvious, predictive to a degree outside the blind tests: they really have identified many areas of what makes "good" speaker design, in terms of what people want to hear. And in practice, even in non-blind conditions, I could tell the Revels were indeed very well designed speakers from their excellent sound.

It's just that last tiny bit down the stretch....that part where we say "this is a speaker whose sound makes me want to own it" were, at least outside the lab, I'm not seeing the predictions working out. And since that last "in the field" experience of how people actually buy speakers ends up being the crucial step, it seems I can't predicate my purchase decision on the fine research done by HK or Dr. Toole. And that's the puzzling part I'm trying to get at. Why not?
HK speakers are designed via the best scientific research we have in terms of predicting what people prefer. And yet outside the lab, in the world of speaker shopping, especially among picky audiophiles, many people who hear them still prefer the sound of other speakers. Why?

Is there something not quite described in the measurements, or not yet interpreted in the measurements, that would explain why, for instance, I find the sounds of instruments more timbrally truthful or believable on certain speakers vs others, even against some speakers that measure optimally for "listener preference?" I dunno.
Scotth3886 likes this.

Last edited by R Harkness; 09-01-2017 at 11:09 AM.
R Harkness is online now  
post #1305 of 1494 Old 09-01-2017, 11:08 AM
AVS Forum Addicted Member
 
sdurani's Avatar
 
Join Date: Oct 2001
Location: Monterey Park, CA
Posts: 24,978
Mentioned: 108 Post(s)
Tagged: 0 Thread(s)
Quoted: 5074 Post(s)
Liked: 3547
Quote:
Originally Posted by noah katz View Post
Looks like it has:
I'm guessing they still keep an eye on it, in case it ends up trading lots of headroom for a little more consistency. BTW, I found it refreshing that the creators of SFM weren't absolutist (max consistency at the expense of everything else) about their own algorithm but instead were practical about it. Same with subwoofer placement. Going by the graphs in Welti's paper, 4 subs at the midpoints of 4 walls have slightly better consistency than 4 subs in 4 corners. Again, practicality wins out over absolutism. Harman usually recommends corner placement because it is worth giving up the last little bit of consistency in exchange for the large gains from corner loading.
Quote:
Originally Posted by noah katz View Post
I believe that contradicts itself, as well as much of what Dr. Toole has said previously.
It's not mentioned often but SFM can be set to solve for: greatest consistency from seat to seat or max output from subwoofer interaction or flattest response at the main listening position. I doubt the latter two have ever been used as part of an ARCOS calibration, but the SFM algorithm itself is not limited to solving for multiple seats.

Sanjay
sdurani is online now  
post #1306 of 1494 Old 09-01-2017, 11:26 AM
Member
 
Join Date: Aug 2017
Posts: 68
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 46 Post(s)
Liked: 36
Quote:
Originally Posted by R Harkness View Post
Thanks for the reply. Your take on the Thiel 3.7s seems different than most others as just about every review found them to be, in practice, extremely neutral sounding. Secretes Of Home Theater Fidelity measured a very smooth in room response iirc. And to me, again, I do not hear obvious bulges and frequency dips in the sound (I sit between 6 and 7 feet from mine, in a room designed with an acoustician, FWIW). Also, as I mentioned, instead of the Revel speakers sounding different from what I'm used to, I'd mentioned it was the opposite: I was struck by how similar they sounded. I've never heard any speaker sound closer to the Thiels (giving me the same impression of evenness top to bottom).

Your reasoning seems to assume different speaker measurements would predict what I would and wouldn't like (to some degree). But then I'm not sure how that would explain why I like so many other speakers that measure differently from the Thiels, e.g. these Hales:

https://www.stereophile.com/content/...r-measurements

Or any number of speakers I've owned or auditioned.

I've also loved speakers that have been designed specifically for neutrality and wide even dispersion, even room power response, e.g. these Waveform speakers (designed with Paul Barton using NRC facilities, he's a proponent of double-blind tests as well iirc):

http://www.audio-ideas.com/reviews/l...mach_solo.html

I still own a pair of Waveform speakers and love them. But they don't sound precisely like the the Revels, timbrally, to my ear.

I also much prefer the sound of my MBL omni-directional monitors. They are clearly designed for similar on/off axis sound, and apparently measure that way.

https://www.stereophile.com/content/...r-measurements


"The 101E Mk.II is remarkably flat throughout the midrange and treble,..."


I have the monitor version which uses the same midrange/treble omni-drivers, so I'm likely getting similar sound - even midrange/treble with wide, even dispersion. That's certainly how they sound - no speaker I know sounds so similar no matter how I move around the speaker.

Or I could talk about the Quad ESL 57 electrostatics. I'm sure their measurements will deviate from the Revel measurements, and I'd guess they would not be chosen over the revels in blind tests (?). And yet I can sit transfixed for hours in front of the Quads in a way that the Revels just don't seem to do for me.

So, again, I'm not yet getting to what degree I should predicate what I like, or predict what I like, on measurements. Again, this is FAR from saying measurements aren't useful as they obviously are. It would be simple to produce a set of measurements so "bad" that it would predict just about everyone here would deem the sound terrible. But withing the range of "not terrible" or "pretty good to good" things seem to get trickier.

I don't see any reason why I would be different from most people, if I took part in the HK blind testing. I'd bet I would prefer what most people prefer, and probably pick, blind, Revel speakers over most or all contenders behind the screen. So I have little doubt the HK research is predictive in that respect. It's also, it seems obvious, predictive to a degree outside the blind tests: they really have identified many areas of what makes "good" speaker design, in terms of what people want to hear. And in practice, even in non-blind conditions, I could tell the Revels were indeed very well designed speakers from their excellent sound.

It's just that last tiny bit down the stretch....that part where we say "this is a speaker whose sound makes me want to own it" were, at least outside the lab, I'm not seeing the predictions working out. And since that last "in the field" experience of how people actually buy speakers ends up being the crucial step, it seems I can't predicate my purchase decision on the fine research done by HK or Dr. Toole. And that's the puzzling part I'm trying to get at. Why not?
HK speakers are designed via the best scientific research we have in terms of predicting what people prefer. And yet outside the lab, in the world of speaker shopping, especially among picky audiophiles, many people who hear them still prefer the sound of other speakers. Why?

Is there something not quite described in the measurements, or not yet interpreted in the measurements, that would explain why, for instance, I find the sounds of instruments more timbrally truthful or believable on certain speakers vs others, even against some speakers that measure optimally for "listener preference?" I dunno.
When a serious and experienced listener like yourself is used to (and enjoys) the horizontal radpat of MBL and Waveform, I don't think it's at all a foregone conclusion that you would prefer Revel in a blind test.

With Quads my experience is similar to yours but this is sitting pretty close. Farther away not at all. One thing for sure, they're nothing like a point source in the standard definition, however much the company has always waved its hands about that.
davidrmoran is online now  
post #1307 of 1494 Old 09-01-2017, 12:26 PM
Member
 
Floyd Toole's Avatar
 
Join Date: Jun 2017
Location: California
Posts: 189
Mentioned: 12 Post(s)
Tagged: 0 Thread(s)
Quoted: 197 Post(s)
Liked: 771
[QUOTE=R Harkness;54717954]Agree. Ultimately speaking.

"They've established what type of sound most people will prefer.

What exactly does that say about the usefulness of that research to me as a consumer."

This indeed is where the rubber hits the road. In a situation where all factors were equal - namely if you were a participant in those double-blind tests at Harman (not HK, by the way; that is a vanishingly small brand now) - then we could carry on an intelligent discussion because your opinion would either have fallen in with the majority, or not, in which case we could ask: why? Historically, it has been shown that prior biases, such as your long-term adaptation to the Theils, has not been a factor, as many persons who have voted down their "favorite" products can attest. The most likely factor is your personal hearing performance.

I have gone on about this earlier in one or other of these forums, and I get very serious about hearing performance in the new book, including an entire chapter. It is not trivial, and it is widespread - including me - so I know whereof I speak. I used to be a superb listener - consistent, observant, a useful "measuring instrument". But, in our computer controlled tests we track the performances both of the loudspeakers and the listeners. In my 60's my judgments showed increased variations, although my audiometric tests looked good. I was also finding it more difficult to make up my mind about small differences - things that in my prime would have been simple. In spite of the hit to my ego, I retired from the listening roster. I still had opinions, still do, but they are mine and mine alone, not to be used as guidance by those with normal hearing.

But, according to may friendly audiologist, I have "normal" hearing. But audiometric tests and criteria are based on speech intelligibility - they have nothing to do with perceiving the subtleties of music, or hearing small distortions. And, for those dedicated to "imaging" it turns out that even people exhibiting normal audiograms can experience something called "hidden hearing loss", in which binaural hearing is affected. We lose some of our ability to separate sounds in space and to adapt to complex sound fields. The degradations can be huge, and it has been found among young listeners.

So, enjoy what you enjoy, and don't worry about it.

For me, I would place much more value in a spinorama to reveal the potential neutrality of a loudspeaker than my own judgment of it. Living with Salon2s, Voice2 and Gem2s in a 7.1 system is a genuine treat, but not all recordings are . . . That is the "other" variable, the one virtually ignored by subjectivists.
Floyd Toole is online now  
post #1308 of 1494 Old 09-01-2017, 12:42 PM
Member
 
Floyd Toole's Avatar
 
Join Date: Jun 2017
Location: California
Posts: 189
Mentioned: 12 Post(s)
Tagged: 0 Thread(s)
Quoted: 197 Post(s)
Liked: 771
Quote:
Originally Posted by noah katz View Post
Not sure about that; why wouldn't the same SFM principles apply?

Seems like PEQ combined w/delay could (though it might not) remedy a null/peak.
If you have good measurements, a flexible DSP box and a lot of patience, a DIY version of SFM is possible:

DellaSala, G. (2016). “Bass Optimization for Home Theater with Multi-Sub + mDSP”, https://www.audioholics.com/home-the...r-home-theater, March 03, 2016.
Floyd Toole is online now  
post #1309 of 1494 Old 09-01-2017, 12:47 PM
Member
 
Floyd Toole's Avatar
 
Join Date: Jun 2017
Location: California
Posts: 189
Mentioned: 12 Post(s)
Tagged: 0 Thread(s)
Quoted: 197 Post(s)
Liked: 771
Quote:
Originally Posted by davidrmoran View Post
The Thiel horizontal radpats I have seen are not good (have not measured any myself) but I am not persuaded this is inherently because of its first-order xovers. Way back when, those were deemed to be a very good goal by the best designers, and I have seen smooth off-axis responses from some such. It depends on how well-behaved and pistonic the drivers are. Otoh, with the off-axis stitching smoothness of something like the Infinity P362, whose FR have by angle I measured extensively, I have a new appreciation for the steeper slopes. So it's all in the engineering and taking the right measurements during design.
Yes, when two drivers overlap output unpleasant things can happen; different at different angles. That is why low-slope crossovers are not universal - indeed they are rare, used as a means of differentiation. The other reason is that transducers, especially fragile tweeters, are exposed to power input at frequencies where they are not contributing significantly. Blown drivers are a risk.
Floyd Toole is online now  
post #1310 of 1494 Old 09-01-2017, 12:52 PM
AVS Forum Special Member
 
Join Date: Jun 2015
Posts: 1,556
Mentioned: 28 Post(s)
Tagged: 0 Thread(s)
Quoted: 986 Post(s)
Liked: 329
sorry guys, I haven't kept up with this thread.


Did any of the guys who preferred the M2 ever post their summary/review?


Did the original poster ever post measurements? I know he was swamped after the test.

THANK YOU!!! for your selfless contribution to the audio world
Anders Torger for Brutefir
Thomas Drugeon for Rephase
John Mulcahy for room eq wizard
Denis Sbragion for DRC room correction
Juha Hartikainen for Winisd
1201 is online now  
post #1311 of 1494 Old 09-01-2017, 12:55 PM
AVS Forum Addicted Member
 
beastaudio's Avatar
 
Join Date: Dec 2010
Location: Western NC
Posts: 12,647
Mentioned: 291 Post(s)
Tagged: 1 Thread(s)
Quoted: 4117 Post(s)
Liked: 3435
Quote:
Originally Posted by 1201 View Post
sorry guys, I haven't kept up with this thread.
That's because it was basically impossible to there for a minute...

European Models do not accept banana plugs
Belly of the Beast: Bass Bunker Theater
beastaudio is online now  
post #1312 of 1494 Old 09-01-2017, 01:06 PM
AVS Forum Addicted Member
 
noah katz's Avatar
 
Join Date: Apr 1999
Location: Mountain View, CA USA
Posts: 21,563
Mentioned: 3 Post(s)
Tagged: 0 Thread(s)
Quoted: 1095 Post(s)
Liked: 380
Quote:
Originally Posted by Floyd Toole View Post
If you have good measurements, a flexible DSP box and a lot of patience, a DIY version of SFM is possible:

DellaSala, G. (2016). “Bass Optimization for Home Theater with Multi-Sub + mDSP”, https://www.audioholics.com/home-the...r-home-theater, March 03, 2016.
Thanks very much, I'll look at that.

Also, I was hoping you'd weigh in on whether a co-located sub with independent signal processing counts as SFM.
noah katz is offline  
post #1313 of 1494 Old 09-01-2017, 01:10 PM
AVS Forum Addicted Member
 
DonH50's Avatar
 
Join Date: Feb 2010
Location: Monument CO
Posts: 10,332
Mentioned: 30 Post(s)
Tagged: 0 Thread(s)
Quoted: 2170 Post(s)
Liked: 1751
There were one or two pro-M2 listeners who responded earlier in this (massive) thread; probably a number of pages back now.

I am not sure John posted measurements and he was banned for soliciting (seemed like sour grapes or some such nonsense to me but I do not have all sides of the story; hard to avoid such a perception in a thread like this) so won't be back until 9/12/2017. You could try a PM or emailing him directly or through his website.

On an "open" SFM: @andyc56 has a multi-sub optimizer (MSO) program that is excellent by all accounts and works with a miniDSP unit to optimize multiple subs in a room. Probably about the best thing going for those of us for whom SFM is unobtanium for one reason or another. Andy has posted in this thread and has his own MSO thread in the subwoofer forum.

"After silence, that which best expresses the inexpressible, is music" - Aldous Huxley
DonH50 is online now  
post #1314 of 1494 Old 09-01-2017, 01:30 PM
Member
 
Floyd Toole's Avatar
 
Join Date: Jun 2017
Location: California
Posts: 189
Mentioned: 12 Post(s)
Tagged: 0 Thread(s)
Quoted: 197 Post(s)
Liked: 771
Quote:
Originally Posted by noah katz View Post
Thanks very much, I'll look at that.

Also, I was hoping you'd weigh in on whether a co-located sub with independent signal processing counts as SFM.
SFM can handle subs in any location but because the sound fields from each end up being superimposed, co-locaton simply increases the power output potential at that location. It is better to spread them around so that the bothersome modes can be more effectively addressed. It is correctly stated that the goal of SFM is to reduce seat-to-seat variations, after which global EQ can work for more people. Stated differently, it is a method of manipulating room modes - the root cause of seat-to-seat variations. Reducing these variations automatically reduces the amplitudes of room modes. This can be seen in the final "after" room curves - in the best solutions there is no evidence of consequential room resonances - bass is "tight"; for everybody. A recent investigation (not ours) of room mode control schemes showed that active methods like SFM reduce modes to well below the threshold of audibility. None have the positional flexibility for subs or listeners that SFM has, and many require perfectly rectangular rooms: SFM does not.
andyc56 likes this.
Floyd Toole is online now  
post #1315 of 1494 Old 09-01-2017, 02:40 PM
Member
 
Join Date: Aug 2017
Posts: 68
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 46 Post(s)
Liked: 36
>> "They've established what type of sound most people will prefer."

Established it for the conventional (beaming w frequency) hor radpat, correct? (Answer: really good seams.) Is that fair to say? While there was an occasional planar involved, iirc, the Harman work has not methodically blind-compared different radpats nor tested to see what radpat type people prefer in the first place. There is no reason to prefer the most perfect implementation of the conventional (which arose simply out of necessity, natural cone/dome behavior, and became a virtue). I think investigation remains to be done in this area, meaning if you enjoy MBL, Ohm, BeoLab5, Allison, dbx SF, Keele CBT, BA E100, or bi planars wide or narrow or Linkwitzian, you might well prefer it blind to the Revel.
davidrmoran is online now  
post #1316 of 1494 Old 09-01-2017, 03:15 PM
Advanced Member
 
bigguyca's Avatar
 
Join Date: Oct 2007
Posts: 701
Mentioned: 9 Post(s)
Tagged: 0 Thread(s)
Quoted: 393 Post(s)
Liked: 270
Quote:
Originally Posted by davidrmoran View Post
The Thiel horizontal radpats I have seen are not good (have not measured any myself) but I am not persuaded this is inherently because of its first-order xovers. Way back when, those were deemed to be a very good goal by the best designers, and I have seen smooth off-axis responses from some such. It depends on how well-behaved and pistonic the drivers are. Otoh, with the off-axis stitching smoothness of something like the Infinity P362, whose FR have by angle I measured extensively, I have a new appreciation for the steeper slopes. So it's all in the engineering and taking the right measurements during design.
Quote:
Originally Posted by Floyd Toole View Post
Yes, when two drivers overlap output unpleasant things can happen; different at different angles. That is why low-slope crossovers are not universal - indeed they are rare, used as a means of differentiation. The other reason is that transducers, especially fragile tweeters, are exposed to power input at frequencies where they are not contributing significantly. Blown drivers are a risk.
radpats = radiation patterns or are you a New England fan or ? Texting isn't a my specialty.

To expand on the horizontal radiation patterns and 1st order crossovers:

The effect on the horizontal radiation pattern could be called a secondary effect. The typical primary effect is that the crossover from the midrange to the tweeter is moved up in frequency due to the effect Dr. Toole notes above, that is, to protect the tweeter. The crossover frequency in a system using a 4th order crossover might be at 2 KHz. In a 1st order system the crossover might be at 3 kHz, but more likely 4 kHz. The directivity of a driver is related to its diameter. The larger the driver diameter, the narrower the directivity, the more the speaker "beams", or is directional. (Whatever your favorite term.)

When a typical midrange driver with narrow directivity at 4 kHz crosses over to a tweeter with wide directivity at 4 kHz the response off axis doesn't match well between the units. There will typically be a broad off axis rise somewhere above the crossover frequency.

One way to "fix" this problem is to use a smaller diameter midrange. This however causes a similar problem at the crossover between the woofer and midrange in a three-way system, since these drivers off axis response it now dissimilar since the crossover frequency likely had to be raised. Another way to address the problem is with a waveguide on the tweeter that allows it to better match the directivity of the mid-range. The thoughts here don't address waveguides or speakers like the M2 with a compression driver.

Most any horizontal off axis measurement in Stereophile for a speaker will show some effect of this off axis response mismatch between the tweeter and midrange drivers, even for speakers for higher order crossovers. The Salon2 avoids this effect because it is a well designed four-way speaker that has drivers of a size that match well at each crossover. Such a design is more challenging and costs more to implement of course.

An added advantage of the four-way design with steep crossovers is that each driver has moderate loading at the low end of its band pass which means its distortion is reduced due to less excursion. Ultimately this reduces multi-tone intermodulation distortion a higher outputs. This is another reason the Salon2 is well liked in my opinion. My opinion is that one of the reasons symphonies are rarely used for speaker demonstrations is that high SPL levels bring out this distortion.

Here a Klippel piece on multi-tone intermodulation distortion (IM). Note how the multi-tone IM is high that the harmonic distortion.

https://www.klippel.de/know-how/meas...istortion.html

This site has some really nice images from Neumann that show the advantages of four-way speakers (likely with high-order crossovers) over three-way and two-way speakers in the area of multi-tone IM.

http://www.neumann-kh-line.com/neuma...?Open&term=TIM
genesplitter likes this.
bigguyca is online now  
post #1317 of 1494 Old 09-01-2017, 04:30 PM
Okv
Senior Member
 
Okv's Avatar
 
Join Date: Jul 2010
Location: Aalesund, Norway
Posts: 312
Mentioned: 23 Post(s)
Tagged: 0 Thread(s)
Quoted: 115 Post(s)
Liked: 102
Quote:
Originally Posted by davidrmoran View Post
>> "They've established what type of sound most people will prefer."

Established it for the conventional (beaming w frequency) hor radpat, correct? (Answer: really good seams.) Is that fair to say? While there was an occasional planar involved, iirc, the Harman work has not methodically blind-compared different radpats nor tested to see what radpat type people prefer in the first place. There is no reason to prefer the most perfect implementation of the conventional (which arose simply out of necessity, natural cone/dome behavior, and became a virtue). I think investigation remains to be done in this area, meaning if you enjoy MBL, Ohm, BeoLab5, Allison, dbx SF, Keele CBT, BA E100, or bi planars wide or narrow or Linkwitzian, you might well prefer it blind to the Revel.
And this is where it starts to get really interesting. But you must start somewhere, and the Harman studies - which they promote and publish - I find is helpful to establish a sort of basic foundation to build upon. They also include at least one planar dipole in their tests, if I understood the presented information correct.

If you look around at what is available, most of the speakers, even the ones considered as measuring well, has a similar radiation - it narrows at high frequencies, and they are omni at low frequencies. And the ones that are different, often get good reviews and owners like them.

With larger horns it is possible to control radiation much better, you can actually choose a pattern. The narrowing top is possible to fix, at least improve a lot. Further down in frequency requires larger horns, so large they are difficult to accept by most people. But controlled and narrow radiation, even if the pattern collapses at lower frequencies, can sound very different from the typical hifi-speaker of today. You get more insight into the recording, more space and room, more clarity and separation between sound objects.
BufordTJustice likes this.

I am Øyvind Kvålsvoll, owner and founder of Kvålsvoll Design.
Okv is offline  
post #1318 of 1494 Old 09-01-2017, 05:21 PM
Member
 
Join Date: Aug 2017
Posts: 68
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 46 Post(s)
Liked: 36
Totally concur in the importance of the Harman work.

Toole's new book, I am told, does get into designs of much broader horizontal radiation (which of course chiefly means upper midrange and treble) --- part-omni, equi-omni, shaped-omni (beamformed), and so on. So advances in this area are to come, although as is regularly pointed out with multichannel the investigation becomes perhaps less necessary and the subject less interesting.

It would be good to be able to include modern horns in any hor radpat discussion, as some horns have sounded extraordinary (to my declining but semi-educated ear), and I like the phrase 'insight into the recording'.
davidrmoran is online now  
post #1319 of 1494 Old 09-01-2017, 06:00 PM
AVS Forum Addicted Member
 
R Harkness's Avatar
 
Join Date: May 2001
Location: Toronto, Ontario, Canada
Posts: 14,358
Mentioned: 10 Post(s)
Tagged: 0 Thread(s)
Quoted: 2110 Post(s)
Liked: 1820
Thanks Prof. Toole!

[quote=Floyd Toole;54729532]
Quote:
Originally Posted by R Harkness View Post
Agree. Ultimately speaking.

"They've established what type of sound most people will prefer.

What exactly does that say about the usefulness of that research to me as a consumer."

This indeed is where the rubber hits the road. In a situation where all factors were equal - namely if you were a participant in those double-blind tests at Harman (not HK, by the way; that is a vanishingly small brand now) - then we could carry on an intelligent discussion because your opinion would either have fallen in with the majority, or not, in which case we could ask: why? Historically, it has been shown that prior biases, such as your long-term adaptation to the Theils, has not been a factor, as many persons who have voted down their "favorite" products can attest. The most likely factor is your personal hearing performance.
Uh-oh, perhaps I've started an "unintelligent discussion."

Sorry about the "HK." It's reflex. I still have an excellent HK reciever (had it since the 90's) and the word "Harman" seems permanently attached to "Kardon" in my head.

It makes sense that prior biases aren't a big factor; if they were, it seems to me the research wouldn't have shown the trends you have outlined here (because I presume most people in the tests own a sound system they are used to, yet the numbers show people converge anyway). That's why I didn't presume my owning Thiels at the moment wouldn't necessarily mean I'd choose them in a blind test (I switch between the Thiels and 3 other speaker brands, different designs that I own. I like mixing it up).


Quote:
Originally Posted by Floyd Toole View Post
So, enjoy what you enjoy, and don't worry about it.
All I got's my own ears. But the folks who inhabit this place rarely just sit back and enjoy; we tend to be interested in the "why" of things as well. (Though admittedly, I do a lot less of that these days, more enjoying).

Quote:
Originally Posted by Floyd Toole View Post
For me, I would place much more value in a spinorama to reveal the potential neutrality of a loudspeaker than my own judgment of it. Living with Salon2s, Voice2 and Gem2s in a 7.1 system is a genuine treat, but not all recordings are . . . That is the "other" variable, the one virtually ignored by subjectivists.
Agreed about the measurements, of course.

Though I'm not sure how "subjectivists" ignore the fact of varability in recordings. That's a common topic in the audiophile community, including even the most execrable subjectivists :-)

Speaking of the subjectivist camp: as I keep railing about, on one hand audiophiles have just about the worst protocol one could have for coming to reliable conclusions about the sound of gear, especially sonic differences (and if they exist or not).

On the other hand, I do find the language developed in the audiophile community pretty useful. I mean, if all of us consumers had enough experience in speaker design, measuring, correlating that with what we hear etc, we could more likely go by specs, look at measurements and have a good idea of how a speaker will sound. (Though, then again we have the problem of actually finding good consistent measurements for many speakers. We only have magazines, like Stereophile and a few others doing it, and I always see those measurement techniques criticized, so even then we can't be sure).

In lieu of this, if we are talking about audio gear that does sound different - e.g. speakers - I have found some subjectivist reviewing both entertaining and useful. Certain reviewers seem pretty consistent in describing the sonic characteristics of a speaker in a way that just nails it, when I've had experience with the same speaker. But I can see how someone like yourself will have little time for that if you have so much experience you can look at speaker measurements and tell if you'd like them and what they'd sound like.

Thanks again and I look forward to purchasing your new book.
Floyd Toole likes this.

Last edited by R Harkness; 09-01-2017 at 09:43 PM.
R Harkness is online now  
post #1320 of 1494 Old 09-01-2017, 06:14 PM
AVS Forum Addicted Member
 
R Harkness's Avatar
 
Join Date: May 2001
Location: Toronto, Ontario, Canada
Posts: 14,358
Mentioned: 10 Post(s)
Tagged: 0 Thread(s)
Quoted: 2110 Post(s)
Liked: 1820
This thread reminds me that a lot of audiophiles (including writers) often raise a sort of simple heuristic when it comes rejecting the influence of bias in their findings. It goes like this:

"I wasn't biased to hear X,

"In fact if anything I was biased to not hear X, "

"But it turned out I heard X, "

"Therefore my hearing X wasn't due to bias."


This misunderstands the nature of Bias effects, thinking that it only arises if one has a preconceived bias. But we also have other biases, such as the fact that if we are merely listening to hear a difference, we will often perceive a difference, even if one isn't actually audible. And once you think you hear a difference, you may also think you like it better. I remember being somewhat fooled by this in the 90's when I first started checking out things like audiophile AC cables. I tried a selection of ever more expensive cables, given to me to try for free. I was a skeptic about AC cables making any difference and none did...until one cable seemed to. It seemed to make an obvious difference to my ear. And I thought "Well, I sure didn't expect it, so it can't be expectation bias, maybe it's real." But then I did a blind shoot out against regular stock AC cables and the difference vanished. I think blind testing (especially double blind of course if it can be managed) is one of the most educational things anyone can do if you are interested in audio, or even just the scientific method.

(And if I've made any mistakes in the above, I'm very happy to be corrected).
R Harkness is online now  
Sponsored Links
Advertisement
 
Reply Speakers

Thread Tools
Show Printable Version Show Printable Version
Email this Page Email this Page


Forum Jump: 

Posting Rules  
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off