What kind of shootout was this?
EDIT: OK, if you're talking about those three speakers in the picture, then the individual drivers aren't being compared. You should know better than that.
Well, how would you compare them then ? On their own ?
In this case, all mids are operating in similar ranges (except of course ribbons which are not low passed). And it is midrange-reach source (voice, cello) that is used for testing.
I could simply measure them but that is not the point. I believe in measurements (and companies who build those speakers do too, so there was no point measuring freq. linearity etc), but some things simply cannot be measured.
In this case, all mids are operating in similar ranges (except of course ribbons which are not low passed). And it is midrange-reach source (voice, cello) that is used for testing.
I could simply measure them but that is not the point. I believe in measurements (and companies who build those speakers do too, so there was no point measuring freq. linearity etc), but some things simply cannot be measured.
My opinion is that raw measurements tell most of the story, which is enough to determine the quality of a driver or its suitability for various applications.
The facts that are missing from today's measuring practices are only missing because we don't know what to measure, or how to interpret the data we get. I'm pretty sure at least one law of physics would be violated if something audible was happening that was unmeasurable.
Also, in case you get the wrong idea, I do think the ATC is the best mid of that bunch... but back to the topic at hand - the baffles aren't even the same, and how are you certain of the crossover points and types? Furthermore, if isolating a driver was as easy as picking the right type of music, then you should've just disconnected the tweeter and the woofer(s), if those speakers used parallel crossovers. Plus, the dispersion is going to be different. Perhaps the shootout should've been done in an anechoic chamber.
The facts that are missing from today's measuring practices are only missing because we don't know what to measure, or how to interpret the data we get. I'm pretty sure at least one law of physics would be violated if something audible was happening that was unmeasurable.
Also, in case you get the wrong idea, I do think the ATC is the best mid of that bunch... but back to the topic at hand - the baffles aren't even the same, and how are you certain of the crossover points and types? Furthermore, if isolating a driver was as easy as picking the right type of music, then you should've just disconnected the tweeter and the woofer(s), if those speakers used parallel crossovers. Plus, the dispersion is going to be different. Perhaps the shootout should've been done in an anechoic chamber.
Last edited:
The point is, people who build those speakers are seasoned pro's. Both VAFs and Ambiences are held in highest regard in Australia and there is no reason to doubt that drivers we talk about are implemented in their most optimal fashion.
They all cross to the mid in ~400Hz region. ATCs and ribbons are cut sharply (they have to be, there is simply no driver area to go any lower), Excels much more gently (nominally 6dB/oct filters, but with some additional spices added to keep resonances and phase in check).
BTW, I can disconnect other drivers in VAFs and Ambiences but not in ATC. Besides, that would give ribbon an unfair advantage, as it would go to 20kHz, while other two are capped at ~3kHz or so.
They all cross to the mid in ~400Hz region. ATCs and ribbons are cut sharply (they have to be, there is simply no driver area to go any lower), Excels much more gently (nominally 6dB/oct filters, but with some additional spices added to keep resonances and phase in check).
BTW, I can disconnect other drivers in VAFs and Ambiences but not in ATC. Besides, that would give ribbon an unfair advantage, as it would go to 20kHz, while other two are capped at ~3kHz or so.
Last edited:
ATC SM75-150SI've heard all Dynaudio mids (52A/F, 54A/F, m560 and 76) and ATC SM75/150S is simply better. Not sure why (double spider, insane magnet, efficiency, tacky coating - who knows?), but it simply is THE best dome I've ever heard. Loud OR quiet, it just sounds right. Detailed but never irritating, uber dynamic but never shouting, smooth but never dull. Just natural. D52 would become shouty under stress/large SPL (dome rocking?). ATC just changes gears and waits for listener's ears to give up
Note Dynaudio chose not to use dome mids in Evidence variants ...
Thanks for linking to that.
If that ATC driver is considered to be a good example of a dome midrange, then consider me unimpressed.
Zaph's comment is interesting: "Very nice, smooth frequency response with only a minor narrow band dip at 4.5KHz.", indeed looking at the frequency response it doesn't actually look too bad, (although Zaph's measurements always look rather smoothed to me) however frequency response by itself can be very misleading.
That "minor narrow band dip" at 4.5Khz is not quite as innocuous as it appears when you look at the Cumulative Spectral Decay - its obvious that its an anti-phase resonance of the centre of the dome, and a very severe high Q one at that.
This is the first major breakup mode of a soft dome driver, where the centre of the dome oscillates out of phase with the perimeter, causing a notch in the steady state response. When the signal is removed the outside of the dome stops relatively quickly, but the centre continues to oscillate by itself for a long time. (Conventional 1" soft dome tweeters usually have this same resonance - but an octave or so higher due to their smaller size)
It's difficult to judge due to the scale of the axes on the CSD but it looks like that resonance has only decayed about 10dB after 3ms. This is a very audible resonance despite the innocuous looking dip on the frequency response, and at that frequency it will give an edginess and fatiguing sound to vocals if not dealt with. (Some people might say it had "bite" or was "revealing")
Ideally you want resonances in that critical frequency range to decay at least 20dB in well under 1ms to be unobtrusive.
The high Q nature of the resonance is such that you wouldn't be able to flatten it out and eliminate the time domain ringing, so your only option would be to cross it over at 2Khz or below with a steep filter (24dB/oct) and notch it out with a deep notch filter - much like you would with a metal cone mid.
If you have a 3" driver that you're forced to cross over at 2Khz or below with a complicated filter, what's the point ? There are plenty of 3" and larger cone drivers which don't have severe resonances so low in frequency...
Last edited:
I have used few mid domes in my builds already and always liked the result. Recently I build simple and cheap 3-way for colleague at work. I had 2" dome from radioshack on the shelf for a long time, so I got small Dayton dome tweet from PE to accompany it. Mounted both closely on baffle kinda like HiVi duo is. I got cheap Jamo 6" to do the rest. I turned out that Jamo was rolling off early and perfectly, no breakups, where the dome mid took over. The whole speaker was heavy, I built square box for woofer from ceramic tiles, yet quite compact, long story short, guy is immensly impressed.
But I have two much more serious projects done with domes, both dayton domes, aluminium and the soft dome, both projects turned out wonderfully.
I like dome mids. I will use them in the future. I have not done anything with ATC dome yet, but I might. What you see as tiny dip and resonance does not discourage me. I see amazing polar response and low distortion.
Many cone drivers have much worse waterfall with more breakups and resonances. I have been at RMAF and heard some amazing systems utilizing dome mids.
But I have two much more serious projects done with domes, both dayton domes, aluminium and the soft dome, both projects turned out wonderfully.
I like dome mids. I will use them in the future. I have not done anything with ATC dome yet, but I might. What you see as tiny dip and resonance does not discourage me. I see amazing polar response and low distortion.
Many cone drivers have much worse waterfall with more breakups and resonances. I have been at RMAF and heard some amazing systems utilizing dome mids.
I think I forgot the point, I could build simple 2-way for the collegue, he would not care, yet the transition from 6" mid/woofer to the 1" tweeter is not ideal. Dome fit the gap, covered the voices with low distortion, have great polar response, and the result is one league better speaker. I am pretty certain about that. In this case. There might be cases, like when dedicated 4" cone mid is used, where dome mid would not improve the things, I agree.
What would your proposed crossover frequencies and slopes be for inserting a dome midrange between a 6" midwoofer and a 1" dome tweeter ?
Don't you think addition of a second (arguably unnecessary) crossover point in the most critical part of the spectrum would cause more problems than it solves ?
Don't you think addition of a second (arguably unnecessary) crossover point in the most critical part of the spectrum would cause more problems than it solves ?
approximately 800/3000
I loosely followed zaph's 3.5...but used holm impulse program to adjust for different drivers used
Yes and no, this is complicated question and hard to unswer.
I like how Troles explains the need to insert dome mid here:
Design Criteria
I believe there is too much disparity between 6-8" mid/woofer and 1" tweeter. Even the tweeter goes low enough and can power handle it, the character of the polar response it too different. 6-8" woofer runs into problems at the upper region, beaming, breakups....thats where dome mid comes into the rescue.
Yes, it breaks the mids into another band, but its the sound in the end what matters. And my personal experience and listening, assuming right execution, is that dome mid is beneficial.
I talked with Bill Dudleston at RMAF about their speakers, (I loosely copied his Whisper) about similar thing, as you know they have four mids on open baffle with upper mid and tweet (this time both planars). I asked why they decided to use two planars...could cone mids go high enough for just a tweeter to take over? His response was mostly power handling, and polar response were the factors.
Anyway, I hope you know what I want to say. Look, often theory is one thing and the practice is another. Ideally we should just have one speaker for all the frequencies, right? In reality sound of some of those fullrange executions speaks for itselfs...they suck!
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.