planet10 said:
I don't see the why you use the word "discontinuity" there, cause phase is not discountinous in the xover region. It doesn't suddenly jump from one value to another. Just check out some xover sum phase curves.
drivers no matter how perfect are only an element of a loudspeaker system which
contribute to the quality of the system it doesnt mean the system is good just
by using perfect drivers it is how the system will perform according to its electrical
and accoustic parameters as a whole in other words its synthesis of its result will
determine its quality
contribute to the quality of the system it doesnt mean the system is good just
by using perfect drivers it is how the system will perform according to its electrical
and accoustic parameters as a whole in other words its synthesis of its result will
determine its quality
Discontinuities like the phase shift in a properly designed multiway? I'd have to disagree with you about them being so obvious.planet10 said:
Feyz said:
If you study the measured group delay and the phase response of the driver, study the same for XOs, combine the two, then study the phase and group delay of the combination, then it probably would shed some light.
B4 said:
Hans L said:
In case I didn't mention it, I definitely agree with Hans L on this (DW is still in business after ? years). I'm just not sure (from very limited experience) that they do 'sound good'. But that could simply be that I expected more of speakers that are the price of a small car.
Feyz said:
I think he's just using the term to express a more rapid change or phase inversions.😉
Just as a deviation from the Wilson bashing, I wonder if there is anything we can learn from the past.
In 1976 "HiFi For Pleasure" ran an extensive test of thirty (well-known at the time) loudspeakers from America, AR,JBL,etc,(and poor old Bose), Britain, Germany and Japan, Yamaha, Sansui etc. They ranged in size from the dimutive LS3/5 to the Wharfedale Airedale at over 4 cub.ft.
The "blind" listening tests were in a small hall, before a panel of seven experienced listeners.
The first set of tests was "live" versus "mono recorded" of voice and single instruments, including drums, acoustic guitar, cymbels etc. The listeners were asked to judge for realism, nothing else.
Top was the JR149, the LS3 was 6th, the KEF 103 8th, Tannoy 12" dual-concentric 17th, the Quad ESL 19th, the Airdale 25th
The second set of tests was recorded stereo music, with comments and points given for overall subjective frequency response, colouration by frequency bands, clarity imaging etc. etc.
Here the order changed very substantially, possibly indicating realism has nothing to do with listening pleasure. The simple two-ways generally did better than the bigger three ways, despite less bass. The Quad still did poorly, but the text said that the panels comments varied much more widely for these than for any other speaker, depending where they were seated . The LS3 was still rated 4th, but the Tannoy came up to 3rd.
Axial and 45 degree freq response were given, and many of these were at least as good as modern speakers.
Some attempt was done at statistical analysis, but frankly no really definite conclusions were drawn, except to say that the mid range was most critical, that resonances (colouration) of any type were bad, and that in general flatter on axis frequency response gave best results. Broad deviations could be tolerated, as could dips, but sharp peaks could not.
To me the response curve bore no relation to the listener results.
The only conclusion I can draw from this test is that the five speakers that are still highly prized, did not do very well, and that 25 have vanished without trace.
In 1976 "HiFi For Pleasure" ran an extensive test of thirty (well-known at the time) loudspeakers from America, AR,JBL,etc,(and poor old Bose), Britain, Germany and Japan, Yamaha, Sansui etc. They ranged in size from the dimutive LS3/5 to the Wharfedale Airedale at over 4 cub.ft.
The "blind" listening tests were in a small hall, before a panel of seven experienced listeners.
The first set of tests was "live" versus "mono recorded" of voice and single instruments, including drums, acoustic guitar, cymbels etc. The listeners were asked to judge for realism, nothing else.
Top was the JR149, the LS3 was 6th, the KEF 103 8th, Tannoy 12" dual-concentric 17th, the Quad ESL 19th, the Airdale 25th
The second set of tests was recorded stereo music, with comments and points given for overall subjective frequency response, colouration by frequency bands, clarity imaging etc. etc.
Here the order changed very substantially, possibly indicating realism has nothing to do with listening pleasure. The simple two-ways generally did better than the bigger three ways, despite less bass. The Quad still did poorly, but the text said that the panels comments varied much more widely for these than for any other speaker, depending where they were seated . The LS3 was still rated 4th, but the Tannoy came up to 3rd.
Axial and 45 degree freq response were given, and many of these were at least as good as modern speakers.
Some attempt was done at statistical analysis, but frankly no really definite conclusions were drawn, except to say that the mid range was most critical, that resonances (colouration) of any type were bad, and that in general flatter on axis frequency response gave best results. Broad deviations could be tolerated, as could dips, but sharp peaks could not.
To me the response curve bore no relation to the listener results.
The only conclusion I can draw from this test is that the five speakers that are still highly prized, did not do very well, and that 25 have vanished without trace.
soongsc said:
Thanks for the suggestion, will do that. Oh but wait, I had done that many times already 😉
This has turned into one of those meaningless internet discussion threads. I am off of this thead.
There is no such thing as an ideal lodspeaker and there will never be one. There are always tradeoffs to be made. Unfortunately all parameters like response linearity (on- and off- axis ), amplitude response, transient response and nonlinear distortion have to be traded against each other in some way.
Which loudspeaker sounds best is still a matter of taste. One can't actually tell how a speaker sounds only how it is perceived and this one varies greatly between individuals.
As I already mentioned I once had the opportunity to listen to the Wilson Audio Sophia. IMO it "sounded" nice. It sounded much more smooth in the midrange than my Mangers for instance and I am sure that there are many people out there who might like that sound. But still my Mangers "sound more real" to me.
I don't think that many Wilson owners buy their speakers without listening to them first. Since they are not a tool (Wilsons used as monitors is a rumour IMO) they don't have to sound accurate and neutral but pleasurable to their owners.
Regards
Charles
Which loudspeaker sounds best is still a matter of taste. One can't actually tell how a speaker sounds only how it is perceived and this one varies greatly between individuals.
As I already mentioned I once had the opportunity to listen to the Wilson Audio Sophia. IMO it "sounded" nice. It sounded much more smooth in the midrange than my Mangers for instance and I am sure that there are many people out there who might like that sound. But still my Mangers "sound more real" to me.
I don't think that many Wilson owners buy their speakers without listening to them first. Since they are not a tool (Wilsons used as monitors is a rumour IMO) they don't have to sound accurate and neutral but pleasurable to their owners.
Regards
Charles
Sophia
(sorry for my english)
I have listened too the Sophia just last week in a "controlled" situation. A friend tested the Sophia in his room for 15 days and I know his electronics and room very well.
My opinion is simply this speaker do not "sound". In the same contest
a pair of Ascendant have sounded much better ( and I do not love the avalon particulary) .
Just my opinion but Pass x1-x250-teac-wadia are good components
to judge a pair of loudspeakers. My friend haven't purchased
these Wilson (8.500 euro- used).
Cheers,
(sorry for my english)
I have listened too the Sophia just last week in a "controlled" situation. A friend tested the Sophia in his room for 15 days and I know his electronics and room very well.
My opinion is simply this speaker do not "sound". In the same contest
a pair of Ascendant have sounded much better ( and I do not love the avalon particulary) .
Just my opinion but Pass x1-x250-teac-wadia are good components
to judge a pair of loudspeakers. My friend haven't purchased
these Wilson (8.500 euro- used).
Cheers,
Very much my point.
Thirty years ago a good frequency response (by itself) did not create a good speaker. We know today this is so. Those judging Wilson on that basis alone are wasting their time.
But on the other hand, speakers that were considered "the best" by listeners thirty years ago have vanished off the face of this earth, while others considered less good are prized. Why?
We are conditioned by what we hear reproduced, and by fashion.
We accept we cannot reproduce live music faultlessly, despite modern advances in technology and materials, so choose on the basis of what we prefer.
Who is game to say the purchasers of Wilson are wrong. Just their choice is not ours.
Yes, modern speakers are (usually) better. But which of today's speakers will be around in thirty years.
Thirty years ago a good frequency response (by itself) did not create a good speaker. We know today this is so. Those judging Wilson on that basis alone are wasting their time.
But on the other hand, speakers that were considered "the best" by listeners thirty years ago have vanished off the face of this earth, while others considered less good are prized. Why?
We are conditioned by what we hear reproduced, and by fashion.
We accept we cannot reproduce live music faultlessly, despite modern advances in technology and materials, so choose on the basis of what we prefer.
Who is game to say the purchasers of Wilson are wrong. Just their choice is not ours.
Yes, modern speakers are (usually) better. But which of today's speakers will be around in thirty years.
Freq response
About frequency response of a loudspeaker:
Measuring a single speaker compared to two speakers playing the same sound as in stereo is completely different. Have you ever thought about how big anomalies there is in the frequency response of the interference field of the two sound waves coming from the two speakers? Add to that complex head diffraction effects. The sound that is supposed to go into your ears as it would in the original space where the performance took place is badly distorted. Yet, this is what the stereo relies on: the interference field of the two speaker signals.
Do a test: place a single loudspeaker in front of you and place a microphone close to your ear and measure the freq response. Then place two speakers in stereo configuration playing the same signal to form a phantom image between the speakers and measure the freq respopnse with the same microphone (and head) position. You see that the frequency response is not the same. This is one indication why stereo does not sound the same as live. There are more reasons too, of course. Still we are using stereo to listen and still we are using freq response measurements of a single speaker as a figure of merit.
There is a limit up to which it is beneficial to improve the performance of a single loudspeaker (and listening room) when the total system performance is considered. More is to be gained by changing the stereo system itself. The biggest limitation of two speaker stereo is the lack of producing the ambience of the original recording place, I think, but that is another story.
- Elias
About frequency response of a loudspeaker:
Measuring a single speaker compared to two speakers playing the same sound as in stereo is completely different. Have you ever thought about how big anomalies there is in the frequency response of the interference field of the two sound waves coming from the two speakers? Add to that complex head diffraction effects. The sound that is supposed to go into your ears as it would in the original space where the performance took place is badly distorted. Yet, this is what the stereo relies on: the interference field of the two speaker signals.
Do a test: place a single loudspeaker in front of you and place a microphone close to your ear and measure the freq response. Then place two speakers in stereo configuration playing the same signal to form a phantom image between the speakers and measure the freq respopnse with the same microphone (and head) position. You see that the frequency response is not the same. This is one indication why stereo does not sound the same as live. There are more reasons too, of course. Still we are using stereo to listen and still we are using freq response measurements of a single speaker as a figure of merit.
There is a limit up to which it is beneficial to improve the performance of a single loudspeaker (and listening room) when the total system performance is considered. More is to be gained by changing the stereo system itself. The biggest limitation of two speaker stereo is the lack of producing the ambience of the original recording place, I think, but that is another story.
- Elias
Hi Elias
Your observations are not a big issue IMO as long as the off-axis response doesn't show too many irregularities.
The remaining measurement differences should then be related to the HRTF and are therefore not our enemy but our friend.
Intensity stereophony does of course not caputure all the ambience that there is. It is mainly a one dimensional (i.e. horizontal) capturing of an event with some clues in the vertical and depth.
But the end-result is quite good if you look at the relatively small added complexity of a second channel.
Regards
Charles
Your observations are not a big issue IMO as long as the off-axis response doesn't show too many irregularities.
The remaining measurement differences should then be related to the HRTF and are therefore not our enemy but our friend.
Intensity stereophony does of course not caputure all the ambience that there is. It is mainly a one dimensional (i.e. horizontal) capturing of an event with some clues in the vertical and depth.
But the end-result is quite good if you look at the relatively small added complexity of a second channel.
Regards
Charles
The off axis response has nothing to do with what I just said. I was talking about the difference when a natural plane wave reaches listener's ear canals compared to situation in stereo where false plane wave is created in the interference field. You know how stereo works?
- Elias
- Elias
Time spent to apply the knowledge is much more fun. Getting the speakers together for comparison would show the obvious.😀
Elias
Are you talking about the peaks and dips through interference when both speakers carry the same signal ?
If yes this has nothing to do with measuring speakers IMO (not even with their quality). Since
1.) It happens with good and bad speakers equally.
2.) It happens independantly of the speaker being minimum-phase or not.
3.) You don't usually feed your two speakers the same signals apart from centered phantom sources and bass signals. The latter being not very relevant because
4.) the problem exists only at short wavelengths if you are seated correctly. Which is then further
5.) minimised by the HRTF
Regards
Charles
Are you talking about the peaks and dips through interference when both speakers carry the same signal ?
If yes this has nothing to do with measuring speakers IMO (not even with their quality). Since
1.) It happens with good and bad speakers equally.
2.) It happens independantly of the speaker being minimum-phase or not.
3.) You don't usually feed your two speakers the same signals apart from centered phantom sources and bass signals. The latter being not very relevant because
4.) the problem exists only at short wavelengths if you are seated correctly. Which is then further
5.) minimised by the HRTF
Regards
Charles
My point is that the two speaker stereo does not sound the same as live even we have speakers with infinitesimal ripple in freq response. And I gave some reasons to it. That is why I think it is better to improve the stereo system itself rather than going after some ideal speakers and spend huge amount of effort.
Does it make sense to measure the system components and tweek and polish them endlessly if the system itself is not up to the task? For me the task the system should do is to create an illusion close to reality so that I would be fooled into thinking that I was in the original recording space when the original performance took place. Others may have other tasks for their stereo systems.
What comes to the HRTF, that is where all the evil is in case of stereo. Consider a natural sound source in front of you. You can locate the sound coming from this natural source because we have learned to do so during evolution. Then consider a phantom image created by two stereo speakers. Due the high freq part of the HRTF, namely pinna localisation cues, the ear+brain combination will locate the sound to the speakers rather than into the phantom location. Then you got the image splitted: at low freqs, where ITD cues rules, the image is quite good and located at phantom centre, but at high freqs, where pinna cues rule, the sound seems to be coming from two distinct locations neither beeing the intented location. Now, how this can produce any realism like satisfactory performance. Well, obviously it can not so the stereo sound is not going to sound like live. Never.
- Elias
Does it make sense to measure the system components and tweek and polish them endlessly if the system itself is not up to the task? For me the task the system should do is to create an illusion close to reality so that I would be fooled into thinking that I was in the original recording space when the original performance took place. Others may have other tasks for their stereo systems.
What comes to the HRTF, that is where all the evil is in case of stereo. Consider a natural sound source in front of you. You can locate the sound coming from this natural source because we have learned to do so during evolution. Then consider a phantom image created by two stereo speakers. Due the high freq part of the HRTF, namely pinna localisation cues, the ear+brain combination will locate the sound to the speakers rather than into the phantom location. Then you got the image splitted: at low freqs, where ITD cues rules, the image is quite good and located at phantom centre, but at high freqs, where pinna cues rule, the sound seems to be coming from two distinct locations neither beeing the intented location. Now, how this can produce any realism like satisfactory performance. Well, obviously it can not so the stereo sound is not going to sound like live. Never.
- Elias
Elias said:
Speakers have the widest diversion of performance. The closer they get to being phase flat, resonant free, linear, reduced lose, compression etc. then changes in other parts of the system can be more clearly identified.
That's a good point. As soon as we have the perfect transducer we can start to develop the "ultimate multichannel reproduction chain". My assumption is that none will give the same illusion/cost ratio as good old stereo.
I once attended a 2+2+2 demo and that one was way cool but remember that it means three times the amount of amps speakers etc.
Regards
Charles
- Status
- Not open for further replies.