The Seas DXT tweeter does have some specific advantages over other WG "assisted" domes. The Morel CAT378 is a fabulous tweeter, but only crossed 1st order with 4 to 5uf cap along with an 8" midbass ie. Seas U22REX/P with a 1.5 mH air core coil + 3.1k LCR. Depending on baffle width, its a really accurate combo and delivers on almost every front. Horn solos sound pretty believable on a design like this.
Morel CAT378 https://hificompass.com/en/speakers/measurements/morel/morel-cat378
SEAS DXT https://heissmann-acoustics.de/en/test-seas-dxt-27tbcd-g/
But a wider wg can look like this https://www.somasonus.net/satori-tw29bn with Augerpro elliptic
Then of course the driver itself must have good motor and cone ...
or ribbon https://www.madisoundspeakerstore.com/ribbon-tweeters/fountek-neocd3.5h-horn-tweeter/?srsltid=AfmBOop1Dy3psHe3vww7irv-dSSzPD_ouMcpgjiLyep17kqamFKBD-AA
SEAS DXT https://heissmann-acoustics.de/en/test-seas-dxt-27tbcd-g/
But a wider wg can look like this https://www.somasonus.net/satori-tw29bn with Augerpro elliptic
Then of course the driver itself must have good motor and cone ...
or ribbon https://www.madisoundspeakerstore.com/ribbon-tweeters/fountek-neocd3.5h-horn-tweeter/?srsltid=AfmBOop1Dy3psHe3vww7irv-dSSzPD_ouMcpgjiLyep17kqamFKBD-AA
Again Wilson bashing... too long CTC in Mid-tweeter is typical
WAMM has huge CTC in MTM and that will most likely introduce vertical lobing issues. I haven't seen it measured but cheaper models seem to use LR4 acoustic
Revised speakers (post-David) seem to have better crossovers per measured performance https://www.stereophile.com/content/wilson-audio-specialties-sabrinax-loudspeaker-measurements
WAMM has huge CTC in MTM and that will most likely introduce vertical lobing issues. I haven't seen it measured but cheaper models seem to use LR4 acoustic
Revised speakers (post-David) seem to have better crossovers per measured performance https://www.stereophile.com/content/wilson-audio-specialties-sabrinax-loudspeaker-measurements
No this is not what's going on.
The distortion peak is from the midrange driver. It's a result of the metal cone ringing at about 8kHz. As I mentioned before look at the 4th and 5th order harmonics for confirmation of this. It seems like you might not be completely aware of how this phenomena presents with metal cones but this is the basic gist.
Cone rings at 8kHz.
8/3 = 2.67kHz.
When the driver produces a 2.67kHz tone it also generates a third harmonic distortion product at 8kHz. But because the cone also rings at this frequency the distortion product is amplified by the ringing and you get a peak in the distortion.
8/4 = 2kHz. The same thing happens with the fourth harmonic at 2kHz and you'd expect there to be a corresponding peak in HD4 at 2kHz. Look at the distortion plot and there's your peak.
8/5 = 1.6kHz. Same as above but for HD5 and it too has a corresponding peak in the distortion plot.
This is textbook metal cone behaviour and 8kHz is bang on for a typical frequency this would occur at in a driver of this size.
5th element, this is a good description of what is happening.
It brings the question of why measure distortion on-axis, since the peak may not exist and the response may be much lower off axis. More to the point, 0 degrees is far from representative of power.. However for all intents and purposes it may be enough to expect and account for the peak in the measurements.
It brings the question of why measure distortion on-axis, since the peak may not exist and the response may be much lower off axis. More to the point, 0 degrees is far from representative of power.. However for all intents and purposes it may be enough to expect and account for the peak in the measurements.
@5th element I see what you're saying, the down modulated 3rd order peak comes from based on the measurements I cited. If that is in fact the actual culprit, it's rather lame on Revel's account leaving that peak unattenuated. If a simple LCR would have solved it, that lack of detail would be unacceptable on a $7k set of speakers.
Looking at the Revel 226 specs shows the tweeter HP at 2.2k, which with a WG could be ok in terms of distortion. It isn't however so based on the crossover point, which should anyways be chosen so that it doesn't interfere with the delicate area of mid band.
If they can't even get a way with making a cone mid (of any material) behave at 2.5 -2.7k, it shows pretty poor engineering choices.
I don't use any aluminum cone or ceramic cone mid that is larger than 75mm. Any larger than this, it will leave some sort of distortion artifact in the outlaying breakup area, where it can potentially wreak havoc in unexpected ways (similar to those cited here).
For comparison sake, I have nothing against properly engineered metal domes like the M74A as long as they have a clearly defined mid peak at least double the intended LP point. It's a very safe bet to know this mid will behave without any down modulated issues. Its a proven driver with a good track record.
Again, I'm not 100% convinced the peak is in fact originating right where the logical math calculates it to be. We'd need to see the actual raw distortion plot of the mid and not just assume its there based on backwards math of an unknown driver. If it were the case, such oversight on high priced speakers is inexcusable.
Looking at the Revel 226 specs shows the tweeter HP at 2.2k, which with a WG could be ok in terms of distortion. It isn't however so based on the crossover point, which should anyways be chosen so that it doesn't interfere with the delicate area of mid band.
If they can't even get a way with making a cone mid (of any material) behave at 2.5 -2.7k, it shows pretty poor engineering choices.
I don't use any aluminum cone or ceramic cone mid that is larger than 75mm. Any larger than this, it will leave some sort of distortion artifact in the outlaying breakup area, where it can potentially wreak havoc in unexpected ways (similar to those cited here).
For comparison sake, I have nothing against properly engineered metal domes like the M74A as long as they have a clearly defined mid peak at least double the intended LP point. It's a very safe bet to know this mid will behave without any down modulated issues. Its a proven driver with a good track record.
Again, I'm not 100% convinced the peak is in fact originating right where the logical math calculates it to be. We'd need to see the actual raw distortion plot of the mid and not just assume its there based on backwards math of an unknown driver. If it were the case, such oversight on high priced speakers is inexcusable.
@5th element It may be productive for some to do destructive testing on cheap tweeters, but the data and knowledge derived from this would be rather inapplicable.
For starters, many lower priced HF drivers are designed to be durable in specific applications, which usually doesn't mean they'll sound good (at least not to my standards). If you want this sort of thing to be applicable to the real world of drivers typically chosen for building higher caliber systems, you'd have to test the actual drivers you're intending to use.
Most people wouldn't want to waste that sort of money, being the cost if the driver is high and they would likely choose the tweeter appropriately for their application (so power handling and max expected SPL wouldn't be a concern to start with).
If you want some real world data on safe area power handling, the old rules still apply. Most average sized dome tweeters won't reliably tolerate more than 10 - 15W continuously. Pushing a tweeter to destruction isn't a real world test, as it doesn't prove anything pertinent to the driver's performance capabilities. There are tolerances which come into play here, creating inconsistent results. Most people would want to know what the tweeter will safely survive, without failure.
For starters, many lower priced HF drivers are designed to be durable in specific applications, which usually doesn't mean they'll sound good (at least not to my standards). If you want this sort of thing to be applicable to the real world of drivers typically chosen for building higher caliber systems, you'd have to test the actual drivers you're intending to use.
Most people wouldn't want to waste that sort of money, being the cost if the driver is high and they would likely choose the tweeter appropriately for their application (so power handling and max expected SPL wouldn't be a concern to start with).
If you want some real world data on safe area power handling, the old rules still apply. Most average sized dome tweeters won't reliably tolerate more than 10 - 15W continuously. Pushing a tweeter to destruction isn't a real world test, as it doesn't prove anything pertinent to the driver's performance capabilities. There are tolerances which come into play here, creating inconsistent results. Most people would want to know what the tweeter will safely survive, without failure.
FWIW, here is the distortion plot of the Revel F328. It uses the same mid as the F226 but different tweeter and waveguide. Look at the third harmonic at ~2700Hz.
Here is the distortion measurement of the Revel F228. Same mid, tweeter, and WG as the F226. No distortion spike around 2700Hz.
Erin Klippel’d the Revel C426 (center channel) and it shows similar distortion products as the F226.
https://www.erinsaudiocorner.com/loudspeakers/revel_c426be/
Here’s a pic of me unpacking my Revel F228s and placed next to my F206s to show scale.
If the F228s had cone breakup issues with the mid, either I can’t hear them (according to the plot above I won’t) or I don’t know that’s what I’m hearing and instead assume it’s room related. They’re placed in a living room and not a well designed, dedicated listening room. It’s hard to resolve nonideal speaker behavior in a nonideal environment.
Here is the distortion measurement of the Revel F228. Same mid, tweeter, and WG as the F226. No distortion spike around 2700Hz.
Erin Klippel’d the Revel C426 (center channel) and it shows similar distortion products as the F226.
https://www.erinsaudiocorner.com/loudspeakers/revel_c426be/
Here’s a pic of me unpacking my Revel F228s and placed next to my F206s to show scale.
If the F228s had cone breakup issues with the mid, either I can’t hear them (according to the plot above I won’t) or I don’t know that’s what I’m hearing and instead assume it’s room related. They’re placed in a living room and not a well designed, dedicated listening room. It’s hard to resolve nonideal speaker behavior in a nonideal environment.
@looneybomber. If you're happy with your speakers, that's all that really matters. I'm not the type of guy who spends close to 5 figures on ready built speakers, so I'm automatically excluded as a potential customer. I build my own stuff to the level I prefer, mainly because I'm not satisfied with what's currently available.
Anyways, I'd be able to hear the issue(s) pointed out here and they would likely become fatiguing quickly, even at lower playback levels. The transient response alone is prohibitive for my likings, as well as the chosen filter points. Its just not a speaker I'd buy, period.
Anyways, I'd be able to hear the issue(s) pointed out here and they would likely become fatiguing quickly, even at lower playback levels. The transient response alone is prohibitive for my likings, as well as the chosen filter points. Its just not a speaker I'd buy, period.
Wonder what's going on? Just took a look at the F206 and it's clean no peak. You can see the peak in the midrange driver but it looks to be suppressed enough. I have a pair in my living room they sound good. These F22's are the new and improved?
Attachments
Hello
I don't think they are worried. The whole deal is you pick a listening position, work out the ear height and aim the boxes at the target point in space. So when you think about it the lobing is controlled by how the boxes are aimed. What's your take on this?
Rob 🙂
I'm just reading through this thread, but without Wilson Audio in mind I immediately felt the need to respond, a response that has not really something to do with Wilson Audio in particular:
It's true that listening is a very individual thing, depending on all the aspects you mentioned (and maybe even a lot more).
But in my eyes (or ears? ) it is pointless to optimise loudspeakers on this basis, firstly because the sense of hearing is too individual. It is better to use a linear design as the basis for EQ corrections according to taste, which, if the basis is linear, can be used in a more targeted way.
But even more fundamentally: a violin will not adapt to an individual hearing ability, it will always be a violin. So I want it to be reproduced as well as possible (sic!) by my playback equipment (electronics, room, ...), unless I have a hearing impairment that I want to compensate for.
You simply shouldn't confuse the two: using the violin as an example, this means that the violin and the person playing it are the art. Everything that is used for reproduction (electronics, space, ....) is engineering.
Such an oversight is very common and all things considered this is the only small fly in the, otherwise, superb performance of these speakers. They might be $7000 but they are still built to a price and adding an RLC, as per Purifi's datasheets for taming third order peaks in metal cone drivers, adds cost. Notching out the peak in the frequency response with a notch, that's in parallel to the driver terminals (and is usually how it's done), does nothing to address the distortion. It has to be one that's in series with the driver and this isn't widely known yet.
Then again if you've seen the horrific HD performance of a large number of expensive speakers you'd give the Revels a pass. I could argue that the gross linearity issues of the Wilson's in the very basic frequency response domain is inexcusable but here we are.
@Robh3606 The slight peaking in the mids around 4 - 6k that isnt so stable depending on listening angle would put me off, but that's my opinion. I'm not really in the business of evaluating speakers i haven't yet heard. Again, if you like them, that's all that really matters in the end.
No offense intended towards you in saying this, but Its interesting how some people will buy something, claiming to be happy with it and then question their preference based on other people's approval. I don't really like speakers with aluminum cones, unless they're treated in such a way to be well dampened, without any signs of mechanical driver breakup, yet not overdampened. That sort of speaker doesn't exist in the real world.
No offense intended towards you in saying this, but Its interesting how some people will buy something, claiming to be happy with it and then question their preference based on other people's approval. I don't really like speakers with aluminum cones, unless they're treated in such a way to be well dampened, without any signs of mechanical driver breakup, yet not overdampened. That sort of speaker doesn't exist in the real world.
The F206, of the previous generation, were built using SB Acoustics drivers. These, as you can see from the pictures, use SBs ribbed metal cones. This ribbing does an excellent job of helping to suppress the HD peaks usually associated with metal cone drivers.
In their great wisdom Revel decided the best approach would be to copy the SB drivers and make similar ones in their own factories. Whether or not the ribbed cones were omitted due to patent, or because they decided it simply wasn't needed as a way of cutting costs, who knows. But that's the reason why the F206 have a cleaner HD in the midrange.
@5th element I'd only consider a series LCR (paralleled components placed in series) among other methods of "fixing" the issue. This is how it would be done if I were designing the crossover. Having a strong 3rd order ringing peak only 28 - 33dB below the fundamental level is too much and shouldn't be so in that claimed caliber of speaker. Again, people claiming its "good enough"... sorry, not for me. I'd expect better performance from a $7k speaker and its not out of line to do so. Adding corrective measures to the xover wouldn't be that much cost for how much it will improve the speaker.
Again, just because you don't care or notice doesn't mean I feel the same way. This is always how salespeople will deal with customers, downplaying the issue to make the customer feel at fault for having higher standards or expectations. Unless the customer isn't sufficiently educated or enlightened to figure out for themselves whether its a deal breaker, the salesperson should have no need to push the customer into the decision. That's called manipulation and I'd immediately walk away. Something tells me you have a sales background.
Again, just because you don't care or notice doesn't mean I feel the same way. This is always how salespeople will deal with customers, downplaying the issue to make the customer feel at fault for having higher standards or expectations. Unless the customer isn't sufficiently educated or enlightened to figure out for themselves whether its a deal breaker, the salesperson should have no need to push the customer into the decision. That's called manipulation and I'd immediately walk away. Something tells me you have a sales background.
Then you're free to walk away but back in the real world the loudspeaker companies are in it to make money and cutting costs in the crossover is one way to do it. All manufacturers do this even up to the top of the line ranges. This is one of the reasons we DIY of course.
If the F206 was built with better spec drivers and the next series of speakers were designed/manufactured in house, that would indicate to me cost cutting and as the end result, a less refined product in favor of cutting costs. That shows the true motivation of the company. That's why I wouldn't ever be a customer. There's nothing wrong to expect better from something or someone. Its up to the individual consumer to make up their mind. In this day and age, where people are being convinced to like or want something rather than given the choice to make up their own mind, this is in line with politics. The customer is told its too confusing or complicated to have more choices of their own, but in reality its just manipulation.
@5th element Yes, 100 percent entitled to walk away. I dont know why this sorr of discussion needs to be a pi$$ing match.