[konut]

Great thread. Time to goose it!

 Some anecdotal observations. I've been into audio awhile. Prior to the introduction of the B&W DM6, in 1976, there was not much attention paid to t&pa. Of course there were the Altec and JBL coaxials, as well as the Magnaplanar Tympani IDs, which I owned, the Quads, and Lowthers, which have been around forever. But they weren't touted as being linear phase. Then in 1973 the Dahlquist DQ10 appeared and it seems more and more designs since then are more cognizant of these considerations.
      I'll never forget the first time I heard the DM6s. Until then I had no conception of sound stage. The DM6s changed all that. The DQ10s had that attribute as well. That is, one had a more realistic sense of the acoustic space the recording was made in. The Magnaplanars, to some extent, had that quality, but being dipoles, they had to be in an almost perfect acoustical setting before that sense of "acoustic disbelief' was evident. More often than not the room they were in influenced the  sound to the extent that it became a 'wall of sound'. Not an altogether unpleasant experience mind you, just not a realistic sound stage.
     The essence of the t&pa benefit is the realistic sense of sound staging.  But, as its been stated in previous posts, its the implementation  that determines the success of the design in this regard.  In 2 or 3 way designs the interaction of the drivers and crossovers can trip up even the most careful designers. It's also been my observation that in all but concentric, or coincident designs where the tweeter resides either in the center of the cone or at the rear of the bass driver, the sweet spot is very narrow requiring 'head in a vice' seating. The reason for this has been explained very well in previous posts. The latest innovation to mitigate,to some extent, this effect is to use an open baffle, or dipole, in the midrange driver.
        In single driver designs the compromise made is that dispersion suffers as frequency rises leading to a lessening of the sweet spot as well. This is not all bad as this lessens side and floor reflections of higher frequencies (its not a flaw, its a feature!). The other penalties of extended range drivers is linear frequency response, and fall off at the frequency extremes.
       So what to do? In my case the journey has been influenced my exposure , as a younger man, to a couple of full range JBL drivers. As a sophomore in college a friend, who I worked with on the coffee house sound crew, purchased a couple of 4 1/2" full range JBL drivers intended for car use. He cut some holes in large cardboard boxes and filled them with dirty laundry. We were both amazed at the quality of sound that this dorm room setup made. Later, when I was in the sound reinforcement business, I heard a PA consisting of columns  in which 5 10" fullrange JBL resided per side. These also impressed my with a coherency that was rare at the time.
          A couple of years ago, when I started researching a new speaker purchase, I came across Audiocircles, as well as numerous other web sites devoted to music reproduction. It became evident to me that in order to successfully implement a multi-driver system with anything other than first order cross-overs, one needed to use outboard electronic xovers. The added expense and complexity of such a system all but ruled out this alternative for this frugal audiophile. The growing interest and technological innovation in extended range drivers brought back memories of those JBLs I'd had experience with. Thanks to DMason's insatiable curiosity, and seemingly unlimited budget, my 'leg work' was all but done for me as he explored the best of the best when it came to coaxial and full range drivers. As a result of his explorations, I settled on the Omega Aperiodic 8, to which I've added a Rocket UFW-12 subwoofer, as my main system.
     To be sure, this is not a perfect solution. No design currently devised is. I can't back this up with scientific data, but experientially speaking, having listened to hundreds of speakers, it is very rare to have a conventional multi-driver system, with the drivers mounted vertically on a baffle, convey a stable, convincing, wide sound stage without spending over, at least,$3000 . To those of us who have made the leap to extended range drivers, with supplementation, we are unlikely to go back to multi-driver designs where a cross-over is used where the ear is most sensitive, ie 80-5000hz.
      I realize I have made some gross generalizations regarding the use of time alignment and phase alignment interchangeably. They are indeed 2 different things. Also, there are unique examples of speakers that do not conform to the general broad statements I have made. Please refrain from rejecting the general thrust of my observations out of hand due to those exceptions.      aa

[SET Man]

      ... As a sophomore in college a friend, who I worked with on the coffee house sound crew, purchased a couple of 4 1/2" full range JBL drivers intended for car use. He cut some holes in large cardboard boxes and filled them with dirty laundry. We were both amazed at the quality of sound that this dorm room setup made...

Hey!

       Man! now that is what I call a good no-money college student ingenuity Make me wonder if the sound and smell would be improve with a load of clean laundry

         ...To those of us who have made the leap to extended range drivers, with supplementation, we are unlikely to go back to multi-driver designs where a cross-over is used where the ear is most sensitive, ie 80-5000hz...
     

   Yup, I agreed with you on this. I'm also using a Single driver, homebrew Fostex in my case with helper supertweeter

    In my small room these work well for me It ain't perfect but within the limit ( e.g. SPL) of those drivers the sound it great!

    One memorable coaxial driver system I've heard is the big old fashion looking Tannoy Turnbery  at one of the member, Hogg. If I had a much bigger room I might have to check them out

Take care,
Buddy

[PaulHilgeman]

The essence of the t&pa benefit is the realistic sense of sound staging.  But, as its been stated in previous posts, its the implementation  that determines the success of the design in this regard.  In 2 or 3 way designs the interaction of the drivers and crossovers can trip up even the most careful designers. It's also been my observation that in all but concentric, or coincident designs where the tweeter resides either in the center of the cone or at the rear of the bass driver, the sweet spot is very narrow requiring 'head in a vice' seating. The reason for this has been explained very well in previous posts. The latest innovation to mitigate,to some extent, this effect is to use an open baffle, or dipole, in the midrange driver.

This is exactly what I am experiencing with the Ronin 1.0d, controled by the DEQX's linear phase crossovers.  The t&pa properties of this speaker are outstanding, and the effect on realistic placement and size within the soundstage is very very good.

Better terms that we should maybe start using are "location specific transient perfect" with the distiction of "location free transient perfect".  The former would apply to non-coincident designs, the latter to coincident and single driver systems.  Any of the multi-driver systems in either of these categories should employ transient perfect crossover design.  Impulse perfect works too.  This implies not only do the signals arrive at the same time and polarity, but also with no group delay delaying certain frequencies around the crossover point, and a flat, extended frequency response both high and low.  This will result in a very narrow and tall impulse response.  The overall concept of Transient Perfect may or may not be John Kreskovski's idea, but it is at least a philosophy widely spread by his research.

Regards,
Paul Hilgeman
Nomad Audio

[DeeCee]

As a sophomore in college a friend, who I worked with on the coffee house sound crew, purchased a couple of 4 1/2" full range JBL drivers intended for car use. He cut some holes in large cardboard boxes and filled them with dirty laundry. We were both amazed at the quality of sound that this dorm room setup made.

Wow, talk about your flashback... I thought me & a friend were the only oddballs to try this...

Back in 8th grade a friend & I had a couple of no name speaker parts and for fun decided to stuff them in a cardboard box using a blanket and a sweater (or something) with a cap as a "crossover" (not even soldered) for the tweeter...

Oddly, they weren't bad at all... Sigh! The good ol' days! (Pure magic with his home built 1W per channel stereo amp!)

Heh,
DeeCee

[PaulFolbrecht]

Really enjoyed the tech-talk in this thread!  This is very cool stuff to learn about. 

Of course, I knew that T&PA can't be close to the be-all, end-all as most of the most expensive and highly regarded speakers in the world don't bother with it. 

Perhaps the speakers I liked so much at RMAF sounded so great in part because of their adherence to T&PA, or perhaps that is entirely ancillary.  No way to know, of course.

I was wondering if the experts who have posted in this thread might consider one other question: In T&PA speakers, what is preferable: a coincidentally-mounted midrange driver/tweeter (ala Thiel) or more conventional, separate drivers.  I see theoretical advantages for both... another case of trade-offs.  Would be interested in exploring this further here.

Paul

[JLM]

Sex (bigger, flashier, more complicated, newer) sells, not fundamentally sound/basic design that has been around for decades.  I learned that in college from reading through old stereo magazines and seeing the same "fads" reintroduced over and over in slightly different variations.  Huge profit margins pay for advertising campaigns and get more press, regardless of how they sound.  After 40 years, the hype still stinks the same. 

Over those past 40 years the only "discovery" in speaker design I'm aware of has been mathematical modeling of transmission line cabinets (ala Martin King and TBI).  Computer advances have also allowed consumers to manipulate signals for a variety of reasons.  And advances have been made in materials.  That's about sums up advancements in loudspeakers.

Ultra expensive anything turns me off for all of the above reasons.

All things being equal, I prefer a coaxial design, as long as the tweeter doesn't block woofer.  A point source is the ideal imaging and soundstaging, although frequencies under say 80 Hz are normally more ideally located differently (at room boundries) from the higher frequencies.

[PaulHilgeman]

Over those past 40 years the only "discovery" in speaker design I'm aware of has been mathematical modeling of transmission line cabinets (ala Martin King and TBI).  Computer advances have also allowed consumers to manipulate signals for a variety of reasons.  And advances have been made in materials.  That's about sums up advancements in loudspeakers.

More recently we have had the 'discoveries' of T-Shaped pole pieces, Copper/aluminum above and below the VC Gap, many strides have been taken in the design of the spiders themselves, the surround edge resonance problems of many older drivers have gone away due to more scientific modeling of drivers as they are designed.  Motor design has come a HUGE way for low distortion, long excursion woofers, look at the Adire drivers, or Ascendant Audio's drivers, using XBL^2 and the like using dual-gap ideas.  TC Sounds is now doing things with non-uniform VC windings for incredibly flat BL curves.  Cone materials have come a long way in diversity, not necessarily improvements.  One of the biggest things in the last 40 years has been the ability to design crossovers that were more than just a cap, an inductor and a resistor tuned by ear.  While these designs can be good, a majority of the designs around here wouldn't exist without CAD on the crossover. 

Have a good one guys,
Paul Hilgeman

[PaulFolbrecht]

I thought I would post Roy of Green Mountain's response to my question about coincident vs. non-coincident mid/tweet arrangements:

"Regarding co-incident drivers:  Any 1" dome tweeter, below 5kHz, wants to radiate in an omni-directional pattern. This means that much of the sound you would hear from a co-axial mid/tweeter driver in the lower treble comes to you secondarily by way of bouncing off of that mid's cone. Even a short horn around the tweeter cannot prevent this (because it is short). A high-order crossover can help reduce that secondary 'splash' off the mid's cone, by chopping off the low-end of the tweeter even more rapidly."

So, it's about dispersion.  That makes a lot of sense to me.  (Talking to Roy, it becomes apparant that there is not the tiniest detail of speaker design he hasn't spent massive amounts of time studying.  Of course, this is no tiny detail.)

[PaulHilgeman]

So, it's about dispersion.

Lets fuel this fire... 

I think you (He, Roy) mean diffraction.

I feel that the absence of off-axis-nulls and lobing at and around the crossover region (thus proper power response) trumps diffraction by a long shot, but again, thats my opinion based on my massive amounts of research.

I'm sure (not sarcastic) both principles are great, I just believe in mine more .

'night

Paul Hilgeman