In light of post #38 and since I've not received an answer to post #36, can I safely assume there was never a "later correction" to the supposed rise in the LS3/5a? I am asking since, "Its one thing understanding the development of speaker design,
quite another knowing about who, what, when and how these
things happened." And, I would certainly like to know about who, what, when and how these things happen so as not to misrepresent anything about the speaker I use. Any of you Luddites, I'm sorry, "British hifi nuts" have any objections?
quite another knowing about who, what, when and how these
things happened." And, I would certainly like to know about who, what, when and how these things happen so as not to misrepresent anything about the speaker I use. Any of you Luddites, I'm sorry, "British hifi nuts" have any objections?
Hi,
perhaps it's usefull: Believing the seller Rainer Krönke, this Audax Tweeter has been designed for the BBC. The rise at 13 Khz is remarkable.
Greets
http://www.audax-speaker.de/content.php?seite=shop/produkte.php&details=31
perhaps it's usefull: Believing the seller Rainer Krönke, this Audax Tweeter has been designed for the BBC. The rise at 13 Khz is remarkable.
Greets
http://www.audax-speaker.de/content.php?seite=shop/produkte.php&details=31
soufiej said:
According to Alan Shaw, the owner of Harbeth and the bloke who effectively designed the 11 ohm version at the BBC's request, the original 15 ohm model had a significant rising response over 10KHz, deliberately put there to assist in the mastering with analogue tapes. I'm still rooting through the loft trying to find my back issues of Hifi News from the late 1980s which have a comprehensive set of measurements of 15 ohm ands 11 ohm models, made in anechoic chambers, which show this data.
Regarding curves I really don't see a HF rise, maybe a rolloff in the earlier 15 ohm design. Some of the images of the LS3/5a I see, some with and without tweeter dome covers of perforated metal and heavier grill cloths, not sure if tweeter flanges have been rebated but surely all of this makes a difference here.
Scottmoose - Thanks for the response. I went up in the attic the other day but didn't find what I was looking for before it got too hot to stay up there. As I remember that article was fairly complete in its assessment of the two speakers as the revision was quite an event in HFNRR. However, I think it is one of the artciles which leads to much of the confusion and mythology of the 3/5a. (Of which there seems to be an abundance.) If the 08/98 article indicates a change on HF response, then it would somewhat contradict the article I sourced from the Unofficial website which was printed in HFNRR in 01/89. The article from 2004 in HNF from Shaw would seem to further mangle the facts. Whatever they might be. I also find it interesting that the two versions could be used for the "Shootout" without any mention of the dissimilar response. (Not to mention sub'd in use by the BBC staff.) The favorite speakers in that article bounce back and forth between the old and new versions with no preference shown for a specific frequency response from either unit. But, that was Ken Kessler and I suppose he could be just another Yank who some might accuse of re-writing the history of what the BBC accomplished.
sreten - I'm sorry, what am I supposed to be looking at in that last graph you posted? I don't see a rising high end response in either graph. But then again, I don't see any reference for when or where the graph originated or under what circumstances it was made. This seems to be a continuing problem with the graphs you bring us; doesn't it? Though the resolution of the graph is a bit tight, from what I see, the response of the "15 Ohm" unit at 20kHz is spot on the same level as the 1kHz response. Once again I will say that possibly you DIY'ers see things differently than I do, but the same level at 1k and 20k doesn't imply the same response as your "subjective" or "in-room" response graphs which you touted a while back as evidence of a substantial rising high end on the 15 Ohm version. (Up as much as 6dB; wasn't it?) Am I wrong here? And, I would suggest that a raw designation of "15 Ohm" doesn't indicate when the graph was made. It has been acknowledged the speaker went through a revision which ameliorated some of the response abnormalities which had crept into the design over the course of more than a decade. If you are showing me a 15 Ohm unit which is at the worst case of production, you've shown me nothing to indicate you aren't merely grasping at straws. Please, some more information for once. I know those of us not from a certain school or a particular area of London are boring to you (I've spent time on the British based Yahoo forum and I've seen how what someone says is judged by what school they attended), but I would like to know if I have misinterpreted what I have read about the BBC design. I would ask you to set aside your petulance and let's try to sort this out. Oh, and please give some consideration to my question in post #36.
So, let's sum up; shall we?
Was there a substantial rise in high frequency response in the original LS3/5a design?
Was it as much as 6dB up?
At what frequency does this supposed rise begin? Different graphs we've been shown show different results.
Are there conditions under which measurements could be taken which would indicate a variance in response from the same speaker?
Are purely subjective graphs of any value to this discussion?
If a major change of more than 6dB in frequency response was made, why didn't the BBC designate the speaker with a LS3/5b model?
Could two speakers with such drastically different frequency response go unnoticed as a pair as spec'd by the BBC policies?
Have I forgotten anything?
sreten - I'm sorry, what am I supposed to be looking at in that last graph you posted? I don't see a rising high end response in either graph. But then again, I don't see any reference for when or where the graph originated or under what circumstances it was made. This seems to be a continuing problem with the graphs you bring us; doesn't it? Though the resolution of the graph is a bit tight, from what I see, the response of the "15 Ohm" unit at 20kHz is spot on the same level as the 1kHz response. Once again I will say that possibly you DIY'ers see things differently than I do, but the same level at 1k and 20k doesn't imply the same response as your "subjective" or "in-room" response graphs which you touted a while back as evidence of a substantial rising high end on the 15 Ohm version. (Up as much as 6dB; wasn't it?) Am I wrong here? And, I would suggest that a raw designation of "15 Ohm" doesn't indicate when the graph was made. It has been acknowledged the speaker went through a revision which ameliorated some of the response abnormalities which had crept into the design over the course of more than a decade. If you are showing me a 15 Ohm unit which is at the worst case of production, you've shown me nothing to indicate you aren't merely grasping at straws. Please, some more information for once. I know those of us not from a certain school or a particular area of London are boring to you (I've spent time on the British based Yahoo forum and I've seen how what someone says is judged by what school they attended), but I would like to know if I have misinterpreted what I have read about the BBC design. I would ask you to set aside your petulance and let's try to sort this out. Oh, and please give some consideration to my question in post #36.
So, let's sum up; shall we?
Was there a substantial rise in high frequency response in the original LS3/5a design?
Was it as much as 6dB up?
At what frequency does this supposed rise begin? Different graphs we've been shown show different results.
Are there conditions under which measurements could be taken which would indicate a variance in response from the same speaker?
Are purely subjective graphs of any value to this discussion?
If a major change of more than 6dB in frequency response was made, why didn't the BBC designate the speaker with a LS3/5b model?
Could two speakers with such drastically different frequency response go unnoticed as a pair as spec'd by the BBC policies?
Have I forgotten anything?
Hi,
Its unbelievable your complaining about graphs that are all
from the original article that you originally posted the link to.
table of contents :
1977 BBC Version
1977 BBC Version part 2
1984 Spendor version
1989 Rogers version
1989 Rogers version part 2
1989 Review Context
1989 Measurements
1989 Follow-Up
1993 Harbeth version
1993 Review context > Harbeth Measurements
Harbeth Measurements part 2
BBC LS3/5a Specifications
The last graph I posted is on the very page that you posted :
http://www.stereophile.com/standloudspeakers/361/index10.html
edit : It clearly shows the old version is too hot in the treble > 3Khz.
(though not by a huge amount, the final octaves should
have a smooth roll-off due to tweeter dispersion, flat in
the top octaves in-room indicates a rising treble response.
It also illustrates the realignment of the top octaves in the Harbeth)
/sreten.
Its unbelievable your complaining about graphs that are all
from the original article that you originally posted the link to.
table of contents :
1977 BBC Version
1977 BBC Version part 2
1984 Spendor version
1989 Rogers version
1989 Rogers version part 2
1989 Review Context
1989 Measurements
1989 Follow-Up
1993 Harbeth version
1993 Review context > Harbeth Measurements
Harbeth Measurements part 2
BBC LS3/5a Specifications
The last graph I posted is on the very page that you posted :
http://www.stereophile.com/standloudspeakers/361/index10.html
edit : It clearly shows the old version is too hot in the treble > 3Khz.
(though not by a huge amount, the final octaves should
have a smooth roll-off due to tweeter dispersion, flat in
the top octaves in-room indicates a rising treble response.
It also illustrates the realignment of the top octaves in the Harbeth)
/sreten.sreten - No need to get testy again. It's bad for your health. It's after 11 AM here in Dallas and, with the time change, I suspect you will be toddling off to bed soon. I would hate to think you would have difficulty sleeping due to such a minor squabble over a now discontinued speaker. (Imagine one of those big smiley faces here, please.)
A simple note when you post a graph would be sufficient rather than expecting all of us involved in this truth seeking adventure to search through all the 3/5a files in existence. Could we agree that such notation is important, please?
Now on to the graph.
I repeat, what am I supposed to take away from this chart?
What you have shown, in case you missed my point, is a response curve which shows the speaker is flat at 20kHz when compared to a relative level at 1kHz. This is in contradiction to your previous posts which indicated a rising response which was significantly up at 20kHz when compared to a relative level at 1kHz. If I have misrepresented your case or misread the graph (you're so right, I am so baaaaad at interpreting charts), please instruct me where I've gone wrong without getting snippy about the whole affair. We are trying to find some truth here, not what is relative to anyone person's point at any one moment.
The 15 Ohm version did drift from the original spec by a significant amount over the course of the more than one decade during which the speaker was produced under BBC liscence. That is not in dispute.
Is it?
Does anyone here disagree about that issue?
That problem was the cause for the revision made in the late 1980's. But, this rising response in the 3kHz region you point to is not a rising response in the highest frequencies. That was your contention previously; was it not? +6dB at 20kHz. Correct? That is what you implied if not out right stated? That is why you posted the graphs showing a subjective and in-room response which mirrored your previous contention; is it not?
The rising response in the region just North of 1kHz was the result of the B100 midbass driver going out of spec. Is that not correct? The change was made to correct the individual variations from driver to driver and centered mostly on the surround material, did it not? (That the crossover changed is not, I think, relevant to the discussion at hand.) The midbass, though a very good driver, would seem to have little effect on the speaker system's response above, say, 10kHz. Correct? So, along with my previous questions, which you've avoided, please explain how a revision to the midbass driver would affect the repsonse at 20kHz. That still is the area we are debating; is it not? Or, has your argument changed to fit the present facts? By the way; what exactly is your argument now?
Sir, I assure you, I am not making this tough on you just to embarras you. I am truly in need of some education regarding the LS3/5a. Iam looking for information which I might be lacking. If you haven't noticed, I have great respect for the BBC engineers who designed this and many other simlarly voiced products. I would truly like to know the history of the product as best possible. I turn to those who seem to have significant knowledge of the who, what, when, why and how of the design. Can you help me sort out the contradictions I notice in your previous posts? Was the original design of the LS3/5a blazingly hot at 20kHz or was it not?
Oh, yes, can you also answer some of my other questions. Most particularly, the one in post #36.
A simple note when you post a graph would be sufficient rather than expecting all of us involved in this truth seeking adventure to search through all the 3/5a files in existence. Could we agree that such notation is important, please?
Now on to the graph.
I repeat, what am I supposed to take away from this chart?
What you have shown, in case you missed my point, is a response curve which shows the speaker is flat at 20kHz when compared to a relative level at 1kHz. This is in contradiction to your previous posts which indicated a rising response which was significantly up at 20kHz when compared to a relative level at 1kHz. If I have misrepresented your case or misread the graph (you're so right, I am so baaaaad at interpreting charts), please instruct me where I've gone wrong without getting snippy about the whole affair. We are trying to find some truth here, not what is relative to anyone person's point at any one moment.
The 15 Ohm version did drift from the original spec by a significant amount over the course of the more than one decade during which the speaker was produced under BBC liscence. That is not in dispute.
Is it?
Does anyone here disagree about that issue?
That problem was the cause for the revision made in the late 1980's. But, this rising response in the 3kHz region you point to is not a rising response in the highest frequencies. That was your contention previously; was it not? +6dB at 20kHz. Correct? That is what you implied if not out right stated? That is why you posted the graphs showing a subjective and in-room response which mirrored your previous contention; is it not?
The rising response in the region just North of 1kHz was the result of the B100 midbass driver going out of spec. Is that not correct? The change was made to correct the individual variations from driver to driver and centered mostly on the surround material, did it not? (That the crossover changed is not, I think, relevant to the discussion at hand.) The midbass, though a very good driver, would seem to have little effect on the speaker system's response above, say, 10kHz. Correct? So, along with my previous questions, which you've avoided, please explain how a revision to the midbass driver would affect the repsonse at 20kHz. That still is the area we are debating; is it not? Or, has your argument changed to fit the present facts? By the way; what exactly is your argument now?
Sir, I assure you, I am not making this tough on you just to embarras you. I am truly in need of some education regarding the LS3/5a. Iam looking for information which I might be lacking. If you haven't noticed, I have great respect for the BBC engineers who designed this and many other simlarly voiced products. I would truly like to know the history of the product as best possible. I turn to those who seem to have significant knowledge of the who, what, when, why and how of the design. Can you help me sort out the contradictions I notice in your previous posts? Was the original design of the LS3/5a blazingly hot at 20kHz or was it not?
Oh, yes, can you also answer some of my other questions. Most particularly, the one in post #36.
Hi,
Its 5:30 pm here and I'll soon be off to the pub for some beer.
I had assumed you had read the complete article you linked to.
All the measured in room responses in the article point to the
final octaves being too hot (flat is too hot for in-room responses).
They also show the early 11 ohm replacement was similar in treble
balance to the old version but also that the Harbeth certainly isn't.
I think Harbeth may have ploughed their own furrow on this.
The treble excess was never officially corrected as far as I know.
Also in the article is the fact the Harbeth version had the tweeter
wired in opposite polarity to the KEF version so go figure that one !
FWIW as far as I'm concerned (ignoring idiot KK in Hifi-News) it
was common knowledge the LS3/5a was bright at the top end.
In the days of vinyl only combining it with a cartridge that does
have a large +6db peak at ~18 to 20KHz - quite a few did, many
did if the capacitive loading was incorrect, gives a recipe for ear
ripping treble, and an even bigger reputation for being too bright.
/sreten.
Its 5:30 pm here and I'll soon be off to the pub for some beer.
I had assumed you had read the complete article you linked to.
All the measured in room responses in the article point to the
final octaves being too hot (flat is too hot for in-room responses).
They also show the early 11 ohm replacement was similar in treble
balance to the old version but also that the Harbeth certainly isn't.
I think Harbeth may have ploughed their own furrow on this.
The treble excess was never officially corrected as far as I know.
Also in the article is the fact the Harbeth version had the tweeter
wired in opposite polarity to the KEF version so go figure that one !
FWIW as far as I'm concerned (ignoring idiot KK in Hifi-News) it
was common knowledge the LS3/5a was bright at the top end.
In the days of vinyl only combining it with a cartridge that does
have a large +6db peak at ~18 to 20KHz - quite a few did, many
did if the capacitive loading was incorrect, gives a recipe for ear
ripping treble, and an even bigger reputation for being too bright.
/sreten.Of course I've read the article but it has been several years since I've done so; and I am perfectly willing to admit I don't know nor remember everything on those I've learned. Sometimes I find myself thinking I know something only to find out I hadn't learned such a thing at all. You know how that goes, I'm sure. A simple notation on any graph you drag over here would have been of assistance was my only point. So, can we just agree to that bit of politeness, if nothing else?
This is the caption under the graph you've presented; is it not?
"Fig.6 Harbeth LS3/5a, spatially averaged, 1/3-octave response in JA's room. "
If this is true, and the response is at the same relative level at 1kHz and 20kHz, how can that be considered too bright for in room response? You have no other information regarding the speaker's performance to go on You don't know the conditions of the room. You are making a rather bold statement to draw attention from not answering my question, sir. What your subjective opinion of flat in-room response amounts to isn't of any concern to me. I am trying to sort out what you claimed in earlier posts concerning excessive high frequency response but that you cannot seem to back up now.
Did you not claim a rising high frequency response of as much as +6dB in your earlier posts? Would not a flat response in-room indicate a rolled off high frequency response in an anechoic situation? A response just as I've suggested and shown is the anechoic response graphs I've linked to. Is not the anechoic response the only thing the designer has real control over?
I had already guessed your opinion of Kessler, but you seem willing to site sources which tend to support your opinion d'jour and discount any which disagree with what you wish to be the truth. Could you please answer a few of my questions from post #46? And the question from post #36 still begs an answer. You have bounced around and evaded the facts, sir. Now you wish to say the Harbeth speaker is not really a 3/5a and somehow justifies your claims. You claim in your most recent post, "The treble excess was never officially corrected as far as I know." This seems to contradict your earlier statement, "Yes there was (a rising response, my words) in the early versions of the LS3/5a, later removed." Which is it, sir? Official or make believe? Please, I am trying to figure out what to accept as truth and you have made this whole affair more confusing with each subsequent post. I know you think me just another idiot such as Kessler, but I am trying to learn. However, the shifting sands of your posts has left me not knowing what exactly your position is. Is it possible you have no leg to stand on, sir?
This is the caption under the graph you've presented; is it not?
"Fig.6 Harbeth LS3/5a, spatially averaged, 1/3-octave response in JA's room. "
If this is true, and the response is at the same relative level at 1kHz and 20kHz, how can that be considered too bright for in room response? You have no other information regarding the speaker's performance to go on You don't know the conditions of the room. You are making a rather bold statement to draw attention from not answering my question, sir. What your subjective opinion of flat in-room response amounts to isn't of any concern to me. I am trying to sort out what you claimed in earlier posts concerning excessive high frequency response but that you cannot seem to back up now.
Did you not claim a rising high frequency response of as much as +6dB in your earlier posts? Would not a flat response in-room indicate a rolled off high frequency response in an anechoic situation? A response just as I've suggested and shown is the anechoic response graphs I've linked to. Is not the anechoic response the only thing the designer has real control over?
I had already guessed your opinion of Kessler, but you seem willing to site sources which tend to support your opinion d'jour and discount any which disagree with what you wish to be the truth. Could you please answer a few of my questions from post #46? And the question from post #36 still begs an answer. You have bounced around and evaded the facts, sir. Now you wish to say the Harbeth speaker is not really a 3/5a and somehow justifies your claims. You claim in your most recent post, "The treble excess was never officially corrected as far as I know." This seems to contradict your earlier statement, "Yes there was (a rising response, my words) in the early versions of the LS3/5a, later removed." Which is it, sir? Official or make believe? Please, I am trying to figure out what to accept as truth and you have made this whole affair more confusing with each subsequent post. I know you think me just another idiot such as Kessler, but I am trying to learn. However, the shifting sands of your posts has left me not knowing what exactly your position is. Is it possible you have no leg to stand on, sir?
sreten said:
I'm surprised by your views Sreten
1) ("Flat is too hot for in room responses") News to me, never known this to be true. Maybe if the room is an echo chamber. Hopefully the speaker/ room interactions regarding late reflections should be at least 10-20 dB down from direct HF sound in an audio reviewers (JA) room for the measurements. Even if they didn't gate the measurment, which they most certainly did.
2) Haven't seen a curve yet, showing other than flat or rolled off HF response. Citing "common knowledge" instead of data is not very useful here.
3) With MM cartridges the HF responses are rolled off if the source inductance is not peaked with sufficient load capacitance. Most preamps are too low in C right out of the box, because it's easier to add the C later. Sounds like youre grasping straws now, blaming the sound sources for perceived HF peaking you said was designed in the small monitor speaker.
sreten - While we're all here and getting along so fabulously, might I ask a question? Have you ever heard a LS3/5a? If so, which version(s)? What system was driving the speakers? I ask because your reliance on "common knowledge" seems to be more valuable to you than any knowledge of who, what, where, when, why and how. Dare I say, you border on misrepresenting the facts of what actually happened at the BBC. I was looking forward to discussing this with someone who valued those bits of knowledge but I don't believe I've found that person here. Oh, well, maybe I'll just read Kessler, he seems to stick to one story which he can reliably footnote. Possibly Scottmoose will come to the rescue. I don't want to place any extra burden on you, Scottmoose, but you seem to be the only sane voice left from your side of the ocean.
Oh, also, sreten, don't let this question deter you from answering those others I've asked and you've ignored. Especially the one from post #36. Thanks in advance.
Oh, also, sreten, don't let this question deter you from answering those others I've asked and you've ignored. Especially the one from post #36. Thanks in advance.
Just for kicks, here's a response I got from the Yahoo LS3/5a forum when I entered "BBC rise" into the search function. These folks really know the 3/5a.
Re: [ls35a] Re: Subwoofers corrupt the LS3/5a magic
Perhaps I should mention the Zanzibar fallacy as told to me by Percy Wilson
as I drove him back to his hotel on one of his last trips to the USA if we
want to use stories to defend the LS3/5a, a speaker that doesn't need any
defense, just understanding of what it does and what it doesn't do.
----- Original Message -----
From: "Geraint Smith" <geraint.smith@...>
To: <ls35a@yahoogroups.com>
Sent: Monday, July 04, 2005 5:28 PM
Subject: Re: [ls35a] Re: Subwoofers corrupt the LS3/5a magic
The measurements at which you are looking do not come close to
refutation. They are, in fact, barely relevant. Mr Colloms was using a
set up that, with the particular speakers he is testing, produces this
particular pattern. That does not rule out the fact that using the same
speakers with other combinations of equipment may produce different
results. Forgive me if I have misunderstood you, but it seems that you
are saying that because such measurements exist it is therefore never
possible to produce anything else. Further, you are also, apparently,
saying that other measurements showing that these observations at which
you are looking are not universal cannot exist. Would you explain the
logic of that, please? It looks very like a couple of false premises to
me. "This animal is black. Therefore all animals are black. Therefore
that white animal cannot exist."
And, incidentally, the only person whom I have seen using the word
"perfect" in relation to the LS3/5A is you. Allow me to quote myself:
"The LS3/5A is a great mini monitor. It is not a fairy godmother."
GOS
On 4 Jul 2005, at 19:00, Allen wrote:
> I am looking at issue 21 of Hi Fi Choice on speakers by Martin Colloms
> and
> the bass is rated at -6 db at 59 hz and then it rises smoothly to 150
> hz and
> the drops smoothly noticeably to about 450 hz and rises to a flat
> portion
> starting at 1 khz which is below the level of the 150 hz bump - not
> flat.The
> LS3/5a is a great speaker, not a perfect one. It has flaws some of
> which
> were deliberately designed in by the BBC so it would do its job for
> the
> BBC(not for us) better.
>
> ----- Original Message -----
> From: "Geraint Smith" <geraint.smith@...>
> To: <ls35a@yahoogroups.com>
> Sent: Monday, July 04, 2005 11:48 AM
> Subject: Re: [ls35a] Re: Subwoofers corrupt the LS3/5a magic
>
>
> Sorry. Not true. Flat response from 16 Hz to 300 (plus or minus 3dB)
> with Rogers 15 Ohm on Tallis rosewood stands, and using Quad 606 power
> amp and (perfectly integrated) REL V (a sealed box, incidentally,
> which, I think might be the crucial difference). No measurable or
> audible bass hump. That means exactly none. At all. In any way.
> Period.
>
> No veiling. No muddying. Subjectively accurate reproduction of a wide
> range of music, confirmed with comparison with good quality headphones
> (Sennheiser HD650s, for interest).
>
> That might be the fourth time I've said that. Perhaps I should set my
> mail app to send it on a continuous loop. Perhaps I should just email
> a
> Realplayer track of someone screaming it very loudly on a continuous
> loop. Jo Ki could do the same concerning his more elaborate set up,
> which is clearly extremely accurate. Repeating it is getting very
> very
> boring. And still all that comes back is contradiction, not
> refutation.
> (Perhaps because it can't be refuted).
>
> No need for further elaborate costly crossovers. No need for Wilsons
> (the closest I have heard in quality to the combination, although
> their
> mid-range seems coarse in comparison with my Rogers LS3/5A and even
> more so in comparison with Stirling V2s). No need even for Quad 989s
> (which don't have quite the physical weight of bass 'feel' of the
> sub-woofer combination, although otherwise the quality of the sound
> throughout the musical scale seems to me to be better).
Re: [ls35a] Re: Subwoofers corrupt the LS3/5a magic
Perhaps I should mention the Zanzibar fallacy as told to me by Percy Wilson
as I drove him back to his hotel on one of his last trips to the USA if we
want to use stories to defend the LS3/5a, a speaker that doesn't need any
defense, just understanding of what it does and what it doesn't do.
----- Original Message -----
From: "Geraint Smith" <geraint.smith@...>
To: <ls35a@yahoogroups.com>
Sent: Monday, July 04, 2005 5:28 PM
Subject: Re: [ls35a] Re: Subwoofers corrupt the LS3/5a magic
The measurements at which you are looking do not come close to
refutation. They are, in fact, barely relevant. Mr Colloms was using a
set up that, with the particular speakers he is testing, produces this
particular pattern. That does not rule out the fact that using the same
speakers with other combinations of equipment may produce different
results. Forgive me if I have misunderstood you, but it seems that you
are saying that because such measurements exist it is therefore never
possible to produce anything else. Further, you are also, apparently,
saying that other measurements showing that these observations at which
you are looking are not universal cannot exist. Would you explain the
logic of that, please? It looks very like a couple of false premises to
me. "This animal is black. Therefore all animals are black. Therefore
that white animal cannot exist."
And, incidentally, the only person whom I have seen using the word
"perfect" in relation to the LS3/5A is you. Allow me to quote myself:
"The LS3/5A is a great mini monitor. It is not a fairy godmother."
GOS
On 4 Jul 2005, at 19:00, Allen wrote:
> I am looking at issue 21 of Hi Fi Choice on speakers by Martin Colloms
> and
> the bass is rated at -6 db at 59 hz and then it rises smoothly to 150
> hz and
> the drops smoothly noticeably to about 450 hz and rises to a flat
> portion
> starting at 1 khz which is below the level of the 150 hz bump - not
> flat.The
> LS3/5a is a great speaker, not a perfect one. It has flaws some of
> which
> were deliberately designed in by the BBC so it would do its job for
> the
> BBC(not for us) better.
>
> ----- Original Message -----
> From: "Geraint Smith" <geraint.smith@...>
> To: <ls35a@yahoogroups.com>
> Sent: Monday, July 04, 2005 11:48 AM
> Subject: Re: [ls35a] Re: Subwoofers corrupt the LS3/5a magic
>
>
> Sorry. Not true. Flat response from 16 Hz to 300 (plus or minus 3dB)
> with Rogers 15 Ohm on Tallis rosewood stands, and using Quad 606 power
> amp and (perfectly integrated) REL V (a sealed box, incidentally,
> which, I think might be the crucial difference). No measurable or
> audible bass hump. That means exactly none. At all. In any way.
> Period.
>
> No veiling. No muddying. Subjectively accurate reproduction of a wide
> range of music, confirmed with comparison with good quality headphones
> (Sennheiser HD650s, for interest).
>
> That might be the fourth time I've said that. Perhaps I should set my
> mail app to send it on a continuous loop. Perhaps I should just email
> a
> Realplayer track of someone screaming it very loudly on a continuous
> loop. Jo Ki could do the same concerning his more elaborate set up,
> which is clearly extremely accurate. Repeating it is getting very
> very
> boring. And still all that comes back is contradiction, not
> refutation.
> (Perhaps because it can't be refuted).
>
> No need for further elaborate costly crossovers. No need for Wilsons
> (the closest I have heard in quality to the combination, although
> their
> mid-range seems coarse in comparison with my Rogers LS3/5A and even
> more so in comparison with Stirling V2s). No need even for Quad 989s
> (which don't have quite the physical weight of bass 'feel' of the
> sub-woofer combination, although otherwise the quality of the sound
> throughout the musical scale seems to me to be better).
"It looks very like a couple of false premises to
>me. "This animal is black. Therefore all animals are black. Therefore
>that white animal cannot exist."
Putting LS3/5as aside for a moment
A joke involving an astronomer, a chemist, and a mathematician on a bus
during their first visit to Scotland. They see a black sheep grazing alone
in a pasture as they drive by.
The astronomer excitedly exclaims, "Ah, this shows Scottish sheep are
black!"
The chemist didactically corrects him: "No, no, it just shows some Scottish
sheep are black."
The mathematician then says, "Actually, we can only be sure there is at
least one Scottish sheep of which at least one side is black"
>me. "This animal is black. Therefore all animals are black. Therefore
>that white animal cannot exist."
Putting LS3/5as aside for a moment
A joke involving an astronomer, a chemist, and a mathematician on a bus
during their first visit to Scotland. They see a black sheep grazing alone
in a pasture as they drive by.
The astronomer excitedly exclaims, "Ah, this shows Scottish sheep are
black!"
The chemist didactically corrects him: "No, no, it just shows some Scottish
sheep are black."
The mathematician then says, "Actually, we can only be sure there is at
least one Scottish sheep of which at least one side is black"
"Re: [ls35a] Re: Subwoofers corrupt the LS3/5a magic
Love it! Very nice. Also very true. Of course, had there been a
statistician, he would have said:
We know that one side of one sheep is black. However, we can also see
that several thousand sheep are at least half white. Therefore, the
occurrence of black sides on a sheep is so close to zero as to be
statistically insignificant. Therefore that sheep does not exist. (He
will then, as Douglas Adams said, go on to prove that black is white,
and get killed on the next zebra crossing)."
"Re: [ls35a] Re: Subwoofers corrupt the LS3/5a magic
As someone with art as my main discipline I'd have said that it looks
like that sheep has been rolling around in the mud. So maybe Scottish
sheep are actually yellow?"
"Re: [ls35a] Re: Subwoofers corrupt the LS3/5a magic
Then I would say to you that you will forever be what you are and will never be
more than you could possibly be.
If I say to you this: A speaker that doesn't need any defence, just
understanding of what it does and what it doesn't do; respect it and add what it
doesn't have whilst take nothing away from what it already have; would you go
against it?"
Love it! Very nice. Also very true. Of course, had there been a
statistician, he would have said:
We know that one side of one sheep is black. However, we can also see
that several thousand sheep are at least half white. Therefore, the
occurrence of black sides on a sheep is so close to zero as to be
statistically insignificant. Therefore that sheep does not exist. (He
will then, as Douglas Adams said, go on to prove that black is white,
and get killed on the next zebra crossing)."
"Re: [ls35a] Re: Subwoofers corrupt the LS3/5a magic
As someone with art as my main discipline I'd have said that it looks
like that sheep has been rolling around in the mud. So maybe Scottish
sheep are actually yellow?"
"Re: [ls35a] Re: Subwoofers corrupt the LS3/5a magic
Then I would say to you that you will forever be what you are and will never be
more than you could possibly be.
If I say to you this: A speaker that doesn't need any defence, just
understanding of what it does and what it doesn't do; respect it and add what it
doesn't have whilst take nothing away from what it already have; would you go
against it?"
"The BBC design data says differently.
> The T27 has one major quirk-ie a prominent and peaky 10-20khz range
> elevated by 4-5dB.The "zobel network" is an attempt to address
> this,more succesfully in the revised crossover.Look at these graphs
> if you need further proof.
> http://www.harbeth.co.uk/images/35acurve1b.jpg
No,I do not agree. You should rather go back to BBC's own curve for
the 15ohm version, with standard crossover.
http://www.bbc.co.uk/rd/pubs/reports/1976-29.pdf
Look at the on-axis respons on page 9. This shows that there isn't
any "quirk" to speak of. And remember, this is the crossover with
virtually no "quirk-compensation". You can argue that over-all there
is a tendency to peaking close to 20khz, esp taken the off-axis
response into consideration, but this is common to many such
tweeters. (cased by breaking up tendencies in the membrane at these
extrem frequencies). If it can be heard at all (usually the
recordings have not much overtones in them over 15khz anyway), it is
not as an unevenness in the tonal balance of the speaker (as you
probably know), but at most as a very slight tendency add a "metallic
sheen" to the sound, and to make treble distortion from the source
sound a tad worse. The thing is very similar to what you used to have
with phono cartridges, peaking in this area. (in a way these frequncy
plots are lying to the eye: It looks like there is a lot going on
over 10-12 khz, while there is not. Some airiness is added, and it is
important enough, but many people would not notice if this was
suddenly filtered away, if they were not listening attentively."
**************************************************
"> > > The BBC reference you quote says quite clearly in
> the "equaliser /crossover" paragraph that "R4 and C6 serve to
adjust
> the frequency response at the upper end of the band".This is
> definitely "quirk compensation"
I would not put it so strongly. And maybe it is a printing error!
There are errors even in the Bible. Maybe they meant to
say "impedance", not "response"!.
If laziness not prevails, I will consider posting sound examples (to
the files section), made with and without the zobel net, to show that
any change (presumably) is not noticeable. These components were
usually added to avoid amplifier hf ocillation.
In the cases when there is a considerable resistance in series with
the tweeter, one will sometimes perceive a slight effect at the very
top end. (Our baby shows little resistance).
> The Harbeth curves prove that the real world T27 is quite peaky in
> the top end,much more so than modern
> tweeters.
Or may be they picked up a substandard 15ohm speaker without knowing.
After all there was a rich variety of speakers from the "problematic
years" to choose among at this point.
>Indeed it has been consistently criticised over the years
> for a hard, fizzy, sibilant sound
Wait a minute! A couple of dB adjustment at 20KHz (assuming that you
and the BBC are right about the zobel) would not save a tweeter with
those awful characteristics. The original LS3/5a became a legend, and
the charming, uncoloured sound of the early year's units was the
reason behind its success. This would have been impossible, had it
come with a nasty or unpleasant sounding tweeter!
Despite your good points in the discussion, I still think one can say
that the high pass section of the original crossover in itself is
straightforward and conventional, when we look away from the unusual
level adjustment thing.
Some quirks and quarks may have entered onto the scene afterwards,
that may also be true."
> The T27 has one major quirk-ie a prominent and peaky 10-20khz range
> elevated by 4-5dB.The "zobel network" is an attempt to address
> this,more succesfully in the revised crossover.Look at these graphs
> if you need further proof.
> http://www.harbeth.co.uk/images/35acurve1b.jpg
No,I do not agree. You should rather go back to BBC's own curve for
the 15ohm version, with standard crossover.
http://www.bbc.co.uk/rd/pubs/reports/1976-29.pdf
Look at the on-axis respons on page 9. This shows that there isn't
any "quirk" to speak of. And remember, this is the crossover with
virtually no "quirk-compensation". You can argue that over-all there
is a tendency to peaking close to 20khz, esp taken the off-axis
response into consideration, but this is common to many such
tweeters. (cased by breaking up tendencies in the membrane at these
extrem frequencies). If it can be heard at all (usually the
recordings have not much overtones in them over 15khz anyway), it is
not as an unevenness in the tonal balance of the speaker (as you
probably know), but at most as a very slight tendency add a "metallic
sheen" to the sound, and to make treble distortion from the source
sound a tad worse. The thing is very similar to what you used to have
with phono cartridges, peaking in this area. (in a way these frequncy
plots are lying to the eye: It looks like there is a lot going on
over 10-12 khz, while there is not. Some airiness is added, and it is
important enough, but many people would not notice if this was
suddenly filtered away, if they were not listening attentively."
**************************************************
"> > > The BBC reference you quote says quite clearly in
> the "equaliser /crossover" paragraph that "R4 and C6 serve to
adjust
> the frequency response at the upper end of the band".This is
> definitely "quirk compensation"
I would not put it so strongly. And maybe it is a printing error!
There are errors even in the Bible. Maybe they meant to
say "impedance", not "response"!.
If laziness not prevails, I will consider posting sound examples (to
the files section), made with and without the zobel net, to show that
any change (presumably) is not noticeable. These components were
usually added to avoid amplifier hf ocillation.
In the cases when there is a considerable resistance in series with
the tweeter, one will sometimes perceive a slight effect at the very
top end. (Our baby shows little resistance).
> The Harbeth curves prove that the real world T27 is quite peaky in
> the top end,much more so than modern
> tweeters.
Or may be they picked up a substandard 15ohm speaker without knowing.
After all there was a rich variety of speakers from the "problematic
years" to choose among at this point.
>Indeed it has been consistently criticised over the years
> for a hard, fizzy, sibilant sound
Wait a minute! A couple of dB adjustment at 20KHz (assuming that you
and the BBC are right about the zobel) would not save a tweeter with
those awful characteristics. The original LS3/5a became a legend, and
the charming, uncoloured sound of the early year's units was the
reason behind its success. This would have been impossible, had it
come with a nasty or unpleasant sounding tweeter!
Despite your good points in the discussion, I still think one can say
that the high pass section of the original crossover in itself is
straightforward and conventional, when we look away from the unusual
level adjustment thing.
Some quirks and quarks may have entered onto the scene afterwards,
that may also be true."
sreten - Is it possible you have merely superimposed the sound of a conventional T27 onto what you believe to be the sound of the original LS3/5a? The KEF speakers which used this driver during the same time period did not sound identical to the BBC design and possibly you have confused the issue of which speaker did what.
I have not heard the speaker in question. However, a comparison of two graphs presented, one graphically on the forum and another through a link originally given by pinkmouse, would tend to bear out Sreten's observations that the speaker is a liitle bright sounding with a rising top end.
Below is the 1976 frequency response, from pinkmouse's link. I added some colored lines, via MSPaint, at 67.5 dB, 70 dB, and 72.5 dBFrom this, we can see that the early-1976-version of the speaker had a swiftly rising top end starting around 15kHz or so. We can also see that the midrange, from 200 to 900 Hz or so, is suppressed about 3 dB compared to the midpoint, while from 1,000 Hz up the level varies between 0 dB and 1 dB above the midpoint. that's a 3 or 4 dB difference, not even mentioning the quick rise at the very top end. So this would seem to be a bright sounding speaker, any way you look at it.
Below is the 1976 frequency response, from pinkmouse's link. I added some colored lines, via MSPaint, at 67.5 dB, 70 dB, and 72.5 dBFrom this, we can see that the early-1976-version of the speaker had a swiftly rising top end starting around 15kHz or so. We can also see that the midrange, from 200 to 900 Hz or so, is suppressed about 3 dB compared to the midpoint, while from 1,000 Hz up the level varies between 0 dB and 1 dB above the midpoint. that's a 3 or 4 dB difference, not even mentioning the quick rise at the very top end. So this would seem to be a bright sounding speaker, any way you look at it.
Attachments
Now, here is the graph from Stereophile, which I believe are for later versioins of this speaker. As you can see, the response from 200 to 900 would again seem to be a couple of dB below the midpoint, while from 1,000 Hz on up it is somewhat above the midpoint. the major difference seems to be that quick rise from 15 kHz or so. If these are in fact the later models, then Sreten's statements would appear to be true-that quick peak got tamed later.
As to why they did not change the number-I dunno, you'd have to ask the BBC. But if I can hazard a guess, the 1/3 octave response on the later speakers seems fairly similar to the response for the 1976 speaker save for that quick jump at 15kHz, and possibly the BBC decided that 15kHz is so far up there, many people cannot hear it, that the change did not materially affect the sound enough to merit a model number change.
I think the more significant response factor is the fact that from 1,000 Hz on up the speaker is 3 dB or more louder than from 200 to 900 Hz.
As to why they did not change the number-I dunno, you'd have to ask the BBC. But if I can hazard a guess, the 1/3 octave response on the later speakers seems fairly similar to the response for the 1976 speaker save for that quick jump at 15kHz, and possibly the BBC decided that 15kHz is so far up there, many people cannot hear it, that the change did not materially affect the sound enough to merit a model number change.
I think the more significant response factor is the fact that from 1,000 Hz on up the speaker is 3 dB or more louder than from 200 to 900 Hz.
Attachments
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.