Michael,
I would take issue with that statement in that it is true only if the amplitude response is also constant over that room angle. There in lies the problem, (and Earl's objection). The room sound field is 3 dimensional. It changes with position, vertically, horizontally, and distance. Given that you do obtain a crossover that results in a speaker with minimum phase band pass response, then changing vertical position must change the alignment of the drivers relative to the listening point (since the propagation distance changes) and there must be some resultant change in the way the drivers sum. The magnitude of this change may be insignificant if the spacing between the drivers is acoustically small (small fraction of a wave length), but typically it is not. Thus even if the response remains MP different EQ would be required to account for the potentially different amplitude over that room angle. One size doesn't fit all. You truly need a minimum phase, constant directivity system.
I'm sure you realize this but by not stating it you leave your comments open ended.
Another point is that this will only apply to the free field response since once room effects enter the picture (both in the modal region and the reverberant region) MP behavior may be lost.
When you look at the in room response it can be though of a the sum of output if an infinite series of low pass (or band pass if you include the roll off of the system) filters with different corner frequencies and amplitude. This sum may not yield a MP response. At very low frequency, just around the room fundamental, the response tends toward MP since only on mode is contributing to the response. As the frequency rises the sum depends on several modes and is typically not MP and will depend additionally on the position of the observation point. When you enter the reverberant range of the response it becomes a matter of summing the reflections to the direct sound. In this range the result is highly dependent on the magnitude of the reflections compared to the direct sound. It may seem strange that a delayed reflection can be summed to the direct sound and, in some cases, still result in a MP response at the observation point, but that is the way it comes out.
Not to mention that, assuming the net response is MP throughout the room, room effects will alter the amplitude spatially even if the speaker is truly constant directivity. Thus the best you could ever hope for using room eq it to obtain the desired room response curve at a single position in the room, or perhaps over a narrow window. The perfect speaker is always at the mercy of the room. That is not to say that room eq, appropriately applied, can not improve gross room problems, but it is not capable of correcting a three dimensional varying sound field. Hopefully we all recognize that.
I like this concept
.But for me it's nothing more than the concept of straight wire with gain (=swwg
).(In France we have a famous specialist in 3 letters audio concepts, not good...)
I'm your fan - seriousely - I really like how exceptionally clear you put things !
I like this concept
But for me it's nothing more than the concept of straight wire with gain (=swwg
(In France we have a famous specialist in 3 letters audio concepts, not good...)
Hej Erling,
unfortunately my command of Swedish is limited to simplest of common phrases and a bit of reading (technical mostly). I wish I could spend more time in Sweden.
I expect that crossing a dome tweeter that low would exhibit some stress; maybe levels you used aren't that demanding ? Or perhaps the material ? Simpler recordings (eg solo piano, small ensemble etc.) and white/pink noise are much better tools in my experience. Broadband noise shows beaming, breakup and polar problems exceedingly well (perhaps too well!).
Do you have distortion measurements similar to what you have done for MJAO ? It would be interesting to see how Vifa+Eminence fare close to crossover region. I am guessing polars are not going to be that hot though, there is simply too much difference in radiation pattern of a 15" unit and 1" dome at 1kHz.
Can you be more specific please ? I'm not sure who might that be ...
Cheers,
Bratislav
unfortunately my command of Swedish is limited to simplest of common phrases and a bit of reading (technical mostly). I wish I could spend more time in Sweden.
I expect that crossing a dome tweeter that low would exhibit some stress; maybe levels you used aren't that demanding ? Or perhaps the material ? Simpler recordings (eg solo piano, small ensemble etc.) and white/pink noise are much better tools in my experience. Broadband noise shows beaming, breakup and polar problems exceedingly well (perhaps too well!).
Do you have distortion measurements similar to what you have done for MJAO ? It would be interesting to see how Vifa+Eminence fare close to crossover region. I am guessing polars are not going to be that hot though, there is simply too much difference in radiation pattern of a 15" unit and 1" dome at 1kHz.
Can you be more specific please ? I'm not sure who might that be ...
Cheers,
Bratislav
I don't claim wide experience or (even slightly) expert experience. And I've put my foot in it more than once.
But, after testing numerous hi-fi midranges I have settled on the pro 18sound 6ND430.
I'm using an 18sound XT1086 waveguide which has a lower pattern maintenance down to 1.6 kHz (although, even though this is quoted by 18sound, they also somehow claim 1.2 kHz!).
I'm running a NSD1095n compression driver, which I used to cross at 1.6 kHz.
(I REALLY tried to like the DE250, but now I have four of them lying around).
First break-up of the 6ND430 seems, by freq response and impedance plot, to be about 3.5 kHz, with a major break-up at 5 kHz.
So, since it is so easy to make adjustments, and comparisons, quickly with DEQX, I've tried numerous crossovers. (By the way, there have been many obviously true comments about DSP only equalising for one spot. However with CD, I can get virtually identical freq responses over an arc of about 50 degrees - with gradual minor parallel roll-off above 10 kHz).
Anyway, I've now settled, with the new midrange, on a 2.5 kHz crossover i.e. not low at all (a change since I last posted, Erling) (and while 1.6 kHz sounded best, and gave the best impulse response, at 48 dB acoustic symmetrical per octave, 2.5 kHz sounds best, and gives the best impulse response, at 120 dB per octave??!- minor tweeter pre-ringing)
There is a very minimal blip in polar continuity at 2.5 kHz compare to 1.6 kHz.
Contrary to my expectations the higher xover is cleaner and more detailed (something not quite right with the XT1086 down that low?).
(I mainly inhabit the Audiocircle open-baffle forum- the OB in my signature).
David
But, after testing numerous hi-fi midranges I have settled on the pro 18sound 6ND430.
I'm using an 18sound XT1086 waveguide which has a lower pattern maintenance down to 1.6 kHz (although, even though this is quoted by 18sound, they also somehow claim 1.2 kHz!).
I'm running a NSD1095n compression driver, which I used to cross at 1.6 kHz.
(I REALLY tried to like the DE250, but now I have four of them lying around).
First break-up of the 6ND430 seems, by freq response and impedance plot, to be about 3.5 kHz, with a major break-up at 5 kHz.
So, since it is so easy to make adjustments, and comparisons, quickly with DEQX, I've tried numerous crossovers. (By the way, there have been many obviously true comments about DSP only equalising for one spot. However with CD, I can get virtually identical freq responses over an arc of about 50 degrees - with gradual minor parallel roll-off above 10 kHz).
Anyway, I've now settled, with the new midrange, on a 2.5 kHz crossover i.e. not low at all (a change since I last posted, Erling) (and while 1.6 kHz sounded best, and gave the best impulse response, at 48 dB acoustic symmetrical per octave, 2.5 kHz sounds best, and gives the best impulse response, at 120 dB per octave??!- minor tweeter pre-ringing)
There is a very minimal blip in polar continuity at 2.5 kHz compare to 1.6 kHz.
Contrary to my expectations the higher xover is cleaner and more detailed (something not quite right with the XT1086 down that low?).
(I mainly inhabit the Audiocircle open-baffle forum- the OB in my signature).
David
I'm not surprised by this result. I've never liked the sound of a tweeter pushed too low; the less the diaphragm moves, the better they sound - at least to me. The tradeoff, of course, is a higher crossover, and polar patterns that aren't as pretty. I prefer the sound of a higher crossover - provided it doesn't creep into the range of the mid or LF driver breaking up.
The kind of program material that exposes excessive tweeter excursion is well-recorded massed chorus; this has a very dense spectrum that is strongly affected by trace amounts of high-order IM distortion.
Anyone that listens to live choral music knows how rare it is for a hifi system to sound like a live (unamplified) event - most hifi systems change the sound from scores of individual singers to an undifferentiated roar of distortion. This isn't one of those close-your-eyes-and-think-about-it things; the sound is either right, or horribly wrong. And minimizing distortion in the critical 1~5 kHz range seems to be the most effective action to take.
If all you listen to is movie soundtracks or minimalist-recorded trios and quartets, this type of distortion is less of a concern. You may never notice or object to it. Music with a sparse spectrum is not going to stress the tweeter nearly as much. I suspect much of the reason audiophiles prefer different types of speakers - with different tradeoffs in the crossover and tweeter - is that they listen to different types of music, which in turn creates different demands on the system.
Part of the reason I'm exploring the more difficult path of working with large-format compression drivers is the 800 Hz ~ 2 kHz region sounds cleaner (to me) than small-format drivers - with 2.7 to 4 times the diaphragm area of a small-format driver, the bigger one is going to sound different in the lower part of the working range. The tradeoff, of course, is losing the top octave.
Last edited:
The other day I had a desire to listen to some classical music - I hardly ever do. I don;t have many recordings, but I found all of them horribly recorded - unlistenable. SO I went and bought some Duetch Grammophone ones and they were horrible too. What is a "well-recorded massed chorus" so that I can listen to what you are talking about. Thus far my conclusion about classical music is that its all so badly recorded that its impossible to judge anything from it.
This is a classic point of view that I object to. If the system is good then it can't matter what kind of source material to listen to. Its only when there are problems that different sources might delineate different errors. So what you are saying can only be true for poorer systems, not really good ones. You would claim that there aren't any really good ones, I would disagree.
I would recommend listening to the Duetch Grammophone recordings with polarity of both speakers inverted at the same time. If you don't notice any improvement.
Try Reference Recordings with normal polarity.
Last edited:
I was going to suggest Reference Recordings too.
When I came to visit your place, Dr. Geddes, back a few years now, I played a disc from RefRec that you asked about. You said something about not having any good classical and asked to see the disc.
I had brought, EXOTIC DANCES FROM THE OPERA....bottom of this page
http://www.referencerecordings.com/Minnesota.asp
There are a number of great pieces on that disc and it is well recorded. Its not Choral music though. For choral, I only know Arvo Part, but my disks or Part are not well recorded.
When I came to visit your place, Dr. Geddes, back a few years now, I played a disc from RefRec that you asked about. You said something about not having any good classical and asked to see the disc.
I had brought, EXOTIC DANCES FROM THE OPERA....bottom of this page
http://www.referencerecordings.com/Minnesota.asp
There are a number of great pieces on that disc and it is well recorded. Its not Choral music though. For choral, I only know Arvo Part, but my disks or Part are not well recorded.
That is true in my experience. To me choral music is the best to test a system. I don't cross my tweeters low so I don't have the over excursion problem. However, the distortion above 1kHz is harder to get rid of. Most HiFi equipment have high PSRR/CMR at lower frequencies but above about 1kHz the PSRR/CMR degrade. I think lower frequency problems are often caused by the room and the speakers, while higher frequency problems are often caused by the amplifiers and the source.
The distortions above 1kHz is best revealed by very well recorded choral music, far more revealing than even a piano or a violin.
Below are two very well recorded SACD of choral music I use to test my system:
Immortal Nystedt by 2L : The music is not the easiest to listen to. I listen to Track 11 and 12 only, which only add up to about 3 minutes. The best live music I have ever attended to was at the Canterbury Cathedral in East England. The choral music sent my soul above heaven! I have been dreaming about listening to that type of sound again from a HiFi! I found no CD recordings provide that kind of musical experience. This 2L disc is not there, but may be the closest I can find. The church in which the music was recorded is very reverberant, so only the most outstanding systems can make the CD sound great, and for a lesser system, it is like what Lynn said - change the sound from scores of individual singers to an undifferentiated roar of distortion.
Grieg: Choral Music - Grex Vocalis by 2L: Less reverberant but outstanding acoustic and clarity. The music is far easier to listen to. However, this SACD should sound good in almost any reasonably good systems.
Regards,
Bill
Last edited:
Well yeah, sure. But how many systems are good enough to do big orchestral or choral music justice? Very few. Very, very, few. That's not a slight to many good systems - getting the the orchestral mass right is extremely difficult. It is my goal - but by no means easy to achieve. It can be done.
If you find most orchestral music poorly recorded that says two things to me:
1) You have ears and brain good enough to reorganize it's wrong.
and
2) Your system is not up to the task. (This includes the room)
Point 1 is the curse, point 2 is the goal. With the right system you will enjoy all the recordings because of the music. And you will really love the well recorded ones.
FWIW, I am not there yet - far from it. Several years of work ahead. Pop, jazz, folk - no problem. Big orchestras and choruses? Ha! Carmina Burana can humble almost any system....
The closest approximation to real, live, full-scale choral music was a demo I heard at the BBC Research Labs of a first-generation, discrete quadraphonic mastertape made on a Studer 1/2" machine with no Dolby processing. Playback was through a quartet of internally bi-amplified (using Quad amplifiers) Spendor studio monitors in a small, heavily damped room.
I've heard Beethoven's 9th many times in concert halls (with no amplification, thank you very much), and this was the first time I'd ever heard this piece - at full symphonic volume - with no audible distortion, and clear differentiation of the individual singers. The audience got excited (this was a live recording of Last Night at the Proms), and started clapping along, and every handclap was discretely and clearly audible for at least 20 feet around, in all directions.
The Shadow Vector decoder that Audionics was pitching to the BBC was patched into the tape feed, so we could do a direct four-channel discrete vs BBC matrix vs SQ matrix comparison - and yes, as you might expect, the matrix presentation blurred the spatial resolution, although neither the BBC nor the SV decoder flattened out the spatial presentation (the main failing of contemporary gain-riding decoders, like CBS full-logic systems).
I've heard a few - very few - prototype systems that were in the same league, but almost no commercial systems, much less the offerings at hifi shows. My attempts to play Carmina Burana at hifi shows usually result in prompt ejection from the room, especially when I turn it up to concert volume. I've come to conclusion that Diana Krall, Cowboy Junkies, Jazz at the Pawnshop, et al, are used at hifi shows because they don't demand much from the system, and audiophiles tolerate this stuff, for some weird reason.
Having heard what I heard at the BBC back in 1975, there's no reason it can't be done now, some 34 years later. Just because most folks think Dolby Digital and iPod MP3's are hifi doesn't mean it should set the quality standard for the industry. The 176.4/24 PCM mastertapes made by Keith Johnson (former MIT professor of Physics) are the best I've heard so far, and comparable in quality to the BBC demo.
John Atwood attended the Computer Audiophile Symposium, and said it was the highest-quality presentation he's heard to date - I'm looking forward to their demos at the this year's RMAF.
I've heard Beethoven's 9th many times in concert halls (with no amplification, thank you very much), and this was the first time I'd ever heard this piece - at full symphonic volume - with no audible distortion, and clear differentiation of the individual singers. The audience got excited (this was a live recording of Last Night at the Proms), and started clapping along, and every handclap was discretely and clearly audible for at least 20 feet around, in all directions.
The Shadow Vector decoder that Audionics was pitching to the BBC was patched into the tape feed, so we could do a direct four-channel discrete vs BBC matrix vs SQ matrix comparison - and yes, as you might expect, the matrix presentation blurred the spatial resolution, although neither the BBC nor the SV decoder flattened out the spatial presentation (the main failing of contemporary gain-riding decoders, like CBS full-logic systems).
I've heard a few - very few - prototype systems that were in the same league, but almost no commercial systems, much less the offerings at hifi shows. My attempts to play Carmina Burana at hifi shows usually result in prompt ejection from the room, especially when I turn it up to concert volume. I've come to conclusion that Diana Krall, Cowboy Junkies, Jazz at the Pawnshop, et al, are used at hifi shows because they don't demand much from the system, and audiophiles tolerate this stuff, for some weird reason.
Having heard what I heard at the BBC back in 1975, there's no reason it can't be done now, some 34 years later. Just because most folks think Dolby Digital and iPod MP3's are hifi doesn't mean it should set the quality standard for the industry. The 176.4/24 PCM mastertapes made by Keith Johnson (former MIT professor of Physics) are the best I've heard so far, and comparable in quality to the BBC demo.
John Atwood attended the Computer Audiophile Symposium, and said it was the highest-quality presentation he's heard to date - I'm looking forward to their demos at the this year's RMAF.
Last edited:
May I ask a stupid question?
How can I judge the playback is good (enough) if I don't have an experience of the real thing?
OK, I (sort of) remember the sounds of a high school chorus (good or bad performances). I've been among the singers and also at the position of conductor, and in the audience of course. But I haven't been to any live concert of, say, Carmina Burana, this kind of scale and performance level. How can I imagine or judge the playback is "right" or not? Any particular phenomenons I can observe?
BTW, this is good on DVD, I think the performance is excellent:
http://www.amazon.com/Orff-Carmina-...=sr_1_5?ie=UTF8&s=music&qid=1253852557&sr=8-5
(excluding the soprano in Beethovon No.9, badly pickuped or mixed )
How can I judge the playback is good (enough) if I don't have an experience of the real thing?
OK, I (sort of) remember the sounds of a high school chorus (good or bad performances). I've been among the singers and also at the position of conductor, and in the audience of course. But I haven't been to any live concert of, say, Carmina Burana, this kind of scale and performance level. How can I imagine or judge the playback is "right" or not? Any particular phenomenons I can observe?
BTW, this is good on DVD, I think the performance is excellent:
http://www.amazon.com/Orff-Carmina-...=sr_1_5?ie=UTF8&s=music&qid=1253852557&sr=8-5
(excluding the soprano in Beethovon No.9, badly pickuped or mixed
)
It doesn't take an audiophile - if it sounds like a woodshop has just started up in the next room, it's a safe assumption that something, somewhere is terribly distorted. I can assure you that radial arm saws are not part of the score of Carmina Burana or Beethoven's 9th, but an addition courtesy of so-called "high fidelity".
Now, the origin of the degradation is another matter. Yes, there are many atrocious and extremely distorted recordings. But in mastertape form, at least, there are recordings of full-scale symphonic and choral music that are subjectively free of distortion. They are not common, but they do exist.
The real shocker was hearing Stravinsky in the concert hall for the first time. On (commercial) recordings, much of 20th-Century classical music sounds appalling - dry, harsh, and atonal. But in the hall, it was intricate, very dense at times, but also remarkably beautiful sounding. I was brought to the point of tears several times during the performance - I'm kind of a softie that way, but I didn't expect it at a performance of Stravinsky's Rite of Spring.
I had never heard it sounding that way on any recording, on any system, and came to the sad conclusion that some types of music are unrecordable at the current state of the art. The denser and more atonal the original score, the less chance it can be recorded with fidelity to the original performance.
Now, the origin of the degradation is another matter. Yes, there are many atrocious and extremely distorted recordings. But in mastertape form, at least, there are recordings of full-scale symphonic and choral music that are subjectively free of distortion. They are not common, but they do exist.
The real shocker was hearing Stravinsky in the concert hall for the first time. On (commercial) recordings, much of 20th-Century classical music sounds appalling - dry, harsh, and atonal. But in the hall, it was intricate, very dense at times, but also remarkably beautiful sounding. I was brought to the point of tears several times during the performance - I'm kind of a softie that way, but I didn't expect it at a performance of Stravinsky's Rite of Spring.
I had never heard it sounding that way on any recording, on any system, and came to the sad conclusion that some types of music are unrecordable at the current state of the art. The denser and more atonal the original score, the less chance it can be recorded with fidelity to the original performance.
Last edited:
I also find massed choral music to be a severe test. Most hi-fi systems just cannot get that right. Lynn, I think you are on the right track here.
Radial arm saws !! Whew! OK, I never heard that in my system. It's a good sign
I can not tolerate shouting upper-mid / lower-high very much. I found it's almost a common problem in vocals in many POP recordings. They are drilling my head.
Lynn Olson said:
But I haven't heard THAT kind of clarity, either
Seems a long way to go....
Reproduced choral music sounding bad
If you have a bunch of hot singers belting away at the heavy bits of Beethoven or Orff think about what their singer's formants are doing and where they're landing: 2 - 5 kHz. The live sound can actually be unpleasant, sometimes, if you are close enough. The unpleasantness is distortion created by our hearing. These self distortion products can mask lower level detail. I've heard this effect from a single tenor singing fortissimo and I was sitting in the balcony of a large theatre! That BBC dip was there for a reason.
The Cambridge type choral sound can have the same effect because of the combination of singers having very similar formants and the reverberant space emphazing the effect.
Defects in speakers and electronics can create similar effects but if equipment quality is decent then perhaps it's time to think about a bit of EQ.
If you have a bunch of hot singers belting away at the heavy bits of Beethoven or Orff think about what their singer's formants are doing and where they're landing: 2 - 5 kHz. The live sound can actually be unpleasant, sometimes, if you are close enough. The unpleasantness is distortion created by our hearing. These self distortion products can mask lower level detail. I've heard this effect from a single tenor singing fortissimo and I was sitting in the balcony of a large theatre! That BBC dip was there for a reason.
The Cambridge type choral sound can have the same effect because of the combination of singers having very similar formants and the reverberant space emphazing the effect.
Defects in speakers and electronics can create similar effects but if equipment quality is decent then perhaps it's time to think about a bit of EQ.