No, not at all.
But it is relevant in the contex of a non-WG'd-wilson tweeter, because a WG allows more low frequency output from a tweeter with low distorsion.
Indeed.. but I do remember a time when enlarging a tweeter enclosure trended. I did it to a few myself.
You are right.
It’s not really a fair Apple to apples comparison.
My guess is that closest would be stock D2904/7100
@HiFiCompass measured it, and even that suggests it could handle a XO point around 1.2KHz; of based on exponential sine sweeps. (At 1-2KHz 94-96dB, H3 less than 60dB down)
I like to crossover as high as practically possible, but my goals/constraints are different.
It’s not really a fair Apple to apples comparison.
My guess is that closest would be stock D2904/7100
@HiFiCompass measured it, and even that suggests it could handle a XO point around 1.2KHz; of based on exponential sine sweeps. (At 1-2KHz 94-96dB, H3 less than 60dB down)
I like to crossover as high as practically possible, but my goals/constraints are different.
Last edited:
Theory Audio design appears to be making that a goal, using small shallow waveguides with CD's.
Company seems to be about making a marriage of the best of home audio, with the best of prosound.
https://www.theoryprofessional.com/technology
Last edited:
I believe this thread is clearly exposing the different design philosophies working with 1" dome tweeters, in their designers own opinions of maximizing the design in two different specific ways.
1) How to design a small dome to cross over as low as technically possible, making it appear to work (on paper) as well as it's possibly designated. That means (to my understanding) - "look what we did with this 1" tweeter... it can handle a 1.2k@LR24 HP. That lets us use an extremely bandwidth limited cone driver for the lower mids, right in the middle of the most sensitive range of the human hearing, where it can easily detect and disect a severe phase, timing and dynamic balance alteration." Yes, some sarcasm being inserted here in this quote...
2.) How to use alternatives to the ubiquitous, flawed 1" dome in such a way to deliver the best aspects of its strengths in musically accurate audio delivery, mainly to round out the shortcomings of the cone driver that has been pushed to its bandwidth edge. This is to extend the critical midrange it strives to reproduce, without interrupt of the range where the human ear can easily perceive any wrong-doings, pushing the midbass/midrange past the critical range of 300 - 4k (ish).
So these are 2 completely different design constraints and goals made by different technical philosophies of execution. One is by overzealous tweeter design people and the other by overzealous midrange(cone) driver people.
Which one do you believe has gotten the closest to delivering the most accurate sounding, believable delivery of a recorded sonic event, both live or in studio, using the most advanced methods of "production" (or lack of it)? Here's where it devides the philosophies once again...
The people who IMO, with out question, were the most successful at extending the CD/WG approach the furthest were JBL with their M2 design. They put all their eggs in one basket and by far achieved the widest bandwidth in the midrange and above with only one transducer. They did this very well in regards to a delicate balance of musicality and technical detail.
The people who were the most successful at delivering the most bandwidth from a single cone midbass/midrange or planar type driver were (with great reserve) several designers, each with their own set of accepted compromises. All these guys however have one important goal in common, which is not breaking up the critical 300 - 4k(ish) range of audio. Some of these guys IMO are Harbeth, Jordan, Quad, Heil, Accustat, Soundlab, Vandersteen, Spendor, Thiel (early), KEF (partial), LeCleech... to name a few.
The people who were successful at pushing the dome tweeter itself farthest as a component in conjunction with a bandwidth limited cone driver cut off between 300 hz - 1k...1.5k-ish range whiledelivering the same subjective and objective level of musical performance, accuracy and believability of the previously mentioned wide range cone/midbass/midrange were... crickets... (insert blank here). The Wilson's would be a great example of not achieving this goal, along with some other overpriced designs.
You can't physically get the leve of performance from a small dome tweeter crossed as humanly low as it will go with steep filters and special chambers, etc. You're chasing the wrong design goal for a wide banded, high output/dynamic range capable system without significant bandwidth limitations in the top end.
I agree the ribbon based HF system is the most accurate sounding depending on crossover points and ribbon bandwidth. Of all types of transducer designs, the ribbon is the most capable in its available bandwidth. With a WG, this type of driver could deliver the bandwidth needed for a large cone drive 2 way.
1) How to design a small dome to cross over as low as technically possible, making it appear to work (on paper) as well as it's possibly designated. That means (to my understanding) - "look what we did with this 1" tweeter... it can handle a 1.2k@LR24 HP. That lets us use an extremely bandwidth limited cone driver for the lower mids, right in the middle of the most sensitive range of the human hearing, where it can easily detect and disect a severe phase, timing and dynamic balance alteration." Yes, some sarcasm being inserted here in this quote...
2.) How to use alternatives to the ubiquitous, flawed 1" dome in such a way to deliver the best aspects of its strengths in musically accurate audio delivery, mainly to round out the shortcomings of the cone driver that has been pushed to its bandwidth edge. This is to extend the critical midrange it strives to reproduce, without interrupt of the range where the human ear can easily perceive any wrong-doings, pushing the midbass/midrange past the critical range of 300 - 4k (ish).
So these are 2 completely different design constraints and goals made by different technical philosophies of execution. One is by overzealous tweeter design people and the other by overzealous midrange(cone) driver people.
Which one do you believe has gotten the closest to delivering the most accurate sounding, believable delivery of a recorded sonic event, both live or in studio, using the most advanced methods of "production" (or lack of it)? Here's where it devides the philosophies once again...
The people who IMO, with out question, were the most successful at extending the CD/WG approach the furthest were JBL with their M2 design. They put all their eggs in one basket and by far achieved the widest bandwidth in the midrange and above with only one transducer. They did this very well in regards to a delicate balance of musicality and technical detail.
The people who were the most successful at delivering the most bandwidth from a single cone midbass/midrange or planar type driver were (with great reserve) several designers, each with their own set of accepted compromises. All these guys however have one important goal in common, which is not breaking up the critical 300 - 4k(ish) range of audio. Some of these guys IMO are Harbeth, Jordan, Quad, Heil, Accustat, Soundlab, Vandersteen, Spendor, Thiel (early), KEF (partial), LeCleech... to name a few.
The people who were successful at pushing the dome tweeter itself farthest as a component in conjunction with a bandwidth limited cone driver cut off between 300 hz - 1k...1.5k-ish range whiledelivering the same subjective and objective level of musical performance, accuracy and believability of the previously mentioned wide range cone/midbass/midrange were... crickets... (insert blank here). The Wilson's would be a great example of not achieving this goal, along with some other overpriced designs.
You can't physically get the leve of performance from a small dome tweeter crossed as humanly low as it will go with steep filters and special chambers, etc. You're chasing the wrong design goal for a wide banded, high output/dynamic range capable system without significant bandwidth limitations in the top end.
I agree the ribbon based HF system is the most accurate sounding depending on crossover points and ribbon bandwidth. Of all types of transducer designs, the ribbon is the most capable in its available bandwidth. With a WG, this type of driver could deliver the bandwidth needed for a large cone drive 2 way.
@tktran303 I do partially agree this being an apples to oranges comparison.
@5th element That 3rd order peak is definitely a product of the tweeter. It responds the same way a driver of its type would. The WG is doing its job, but not where the HD should be reduced, being its design limits as a direct radiating tweeter. If the cone driver would be as fault for this peak, it would be a sad indication from a technical design aspect. How hard would it be to notch out a 2.5k peak from a cone driver? Its not hard, including showing the company's design priorities.
Edit - The Revel226 (erroneously referred to by me as the 223) has a 5.25" cone mid. This size of driver wouldn't have any radial or axial breakup modes at 2.5k, unless it was of poor cone geometry design (doubtful). Reviewing the tweeter in this speaker, it looks more obvious to be at fault of the 2.5k peak IMO.
@5th element That 3rd order peak is definitely a product of the tweeter. It responds the same way a driver of its type would. The WG is doing its job, but not where the HD should be reduced, being its design limits as a direct radiating tweeter. If the cone driver would be as fault for this peak, it would be a sad indication from a technical design aspect. How hard would it be to notch out a 2.5k peak from a cone driver? Its not hard, including showing the company's design priorities.
Edit - The Revel226 (erroneously referred to by me as the 223) has a 5.25" cone mid. This size of driver wouldn't have any radial or axial breakup modes at 2.5k, unless it was of poor cone geometry design (doubtful). Reviewing the tweeter in this speaker, it looks more obvious to be at fault of the 2.5k peak IMO.
Last edited:
It seems funny how I always end up clashing with the guys who believe pushing tweeters to their breaking points is the best solution to an age old problem of simple physics.
There's only so far you can push that little 6 - 9 cm2 bit of diaphragm surface crossing below 2.5k. That's where almost all average design HF domes begin to struggle at higher levels. That limit can be pushed to about 2.2k with a stout WG design that doesn't hurt the effective top end limit. I haven't heard an exception to this guideline yet.
The T34B can pull it off just below the 2k mark with a WG, but the analog filter would be too steep for this to blend well with a typical cone woofer or midbass. I feel thats where we're at currently as a pinnacle in HF dome transducer design.
There's only so far you can push that little 6 - 9 cm2 bit of diaphragm surface crossing below 2.5k. That's where almost all average design HF domes begin to struggle at higher levels. That limit can be pushed to about 2.2k with a stout WG design that doesn't hurt the effective top end limit. I haven't heard an exception to this guideline yet.
The T34B can pull it off just below the 2k mark with a WG, but the analog filter would be too steep for this to blend well with a typical cone woofer or midbass. I feel thats where we're at currently as a pinnacle in HF dome transducer design.
For a softdome under $100 the main winner here is the Morel CAT308. It can do many things very well, easy to integrate and very durable. I've heard some designs with this tweeter sound very smooth crossed at 2.5k 2nd order with a decent 6" woofer. It was the only tweeter which didn't complain as it emitted its last note before the VC let go. This was in a 150W rated 2 way system playing at around 103dB with typically balanced spectrum of music (tweeter limited to 15W). This is the power level that almost all 1" class of domes let go. It didn't matter how low the tweeter goes if it can't keep up thermally. That takes lots of $ and would be better invested in extending cone driver performance instead.
Why try to solve this issue by designing super robust tweeters that cost $500, instead of using a 2-4" midrange driver....
That's the route Wilson went with WAMM and XVX anyway (with the 12MU)
That's the route Wilson went with WAMM and XVX anyway (with the 12MU)
So Wilson liked to cross his tweeters as low as possible, meaning in the upper midrange. Others prefer to not cross in the midrange and have one driver covering, say, 400-4000Hz. The midrange has far more content and energy so what is the advantage of stretching the tweeter over as large a frequency range as possible?
Don’t feel offended 😉 I referred to the post above mine. People don’t seem to understand a tweeter is just an electrodynamic driver which can be used around it’s fs. The only important thing is keeping the excursion within acceptable limits. Just like cone drivers.
Understanding these issues allows a good cross, and where this is achievable it wouldn't be fair to imply there'll be shortcomings. Waveguided speakers that meet the woofer in the narrowing upper region have to navigate these issues, and can be made seamless in the right hands.
This is an example. Your suggestion here makes good advice, but a successful cross can be made at any frequency. While it isn't easy to do, there's no need to fear that there will always be something wrong with it.
@AllenB I agree, this is a big problem with most systems and the easiest way to deal with this (on paper) is to meet the woofer in its wider radiating lower midrange section. This however necessitates crossing in the lower mids - not as much of an issue "musically" speaking if its well under 1k with a medium sized driver and a soft slope filter. Over that, it becomes a problem, starting to audibly degrade transient response using steeper filters. On my JBL 4430s they were right on the edhe.of this feit using a huge biradial and big cone. These were very good in terms of a big 2 way for accuracy in timing.
On the topic of large JBL 2 ways, they have always somehow looked at the problems of the large format 2 way monitor in a "least destructive" way. Those big horns can really reproduce a long, flat relative phase curve. Even with the issue of diffraction in a mass produced item being in the picture, they still manage to get it somewhat right and correct.
Attachments
Yes, people who own those usually say good things about them. With a modern waveguide they would be even better. Is there a reason they shouldn't?
I'm not sure I share you confidence in being able to properly hear flat phase.. but I do feel I can hear smooth response and power (consistent in room). Of course, we have to work with phase to make that happen, but I think it's helpful to identify the issues that matter.
I'm not sure I share you confidence in being able to properly hear flat phase.. but I do feel I can hear smooth response and power (consistent in room). Of course, we have to work with phase to make that happen, but I think it's helpful to identify the issues that matter.
I think one advantage to a waveguide is to be able raise a tweeter's crossover, not to lower it. Here we ignore the loading advantage, and just focus on the fact that the waveguide matches directivity to the midrange. Granted you need a fairly well-behaved mid, and a small or elliptical waveguide for ctc reasons, but this can allow a crossover to jump from 2khz to 3khz.
I don't know that I agree that no one has successfully done the waveguide and lower tweeter crossover thing. Revels have been formally and informally reviewed by many people at GTGs and strained tweeters are not something I have ever heard as a complaint. I've also never heard that complaint for Dutch & Dutch, March Audio, or Buchardt.
I don't know that I agree that no one has successfully done the waveguide and lower tweeter crossover thing. Revels have been formally and informally reviewed by many people at GTGs and strained tweeters are not something I have ever heard as a complaint. I've also never heard that complaint for Dutch & Dutch, March Audio, or Buchardt.
Last edited: