It was done across the vocal range to increase the presense slightly, mostly, as I understand it, achieved by allowing the rise through careful crossover filtering. The classic example of this is the earlier 15ohm model of the LS3/5a, which contrary to popular belief were not flat: remember, these were mobile monitors for mixing purposes, and therefore needed this slight presense lift to aid mixing etc. The later 11ohm version of the LS3/5a with it's computer designed crossover did not have (as much of) a lift (though there still was a little I seem to recall). If you feel this would be of merit to you, then fine. If not, just as fine. Depends on what you listen to, the rest of your system & room, and how you listen / what your preference is I suppose. It might appear to add a touch of sibiance to female vocals, i.e it'll be merciless if there is any on the recording, but useful for the recording studio, or if you prefer increased presense in this region.
Best
Scott
Best
Scott
Hmm ... I found another relevant thread within this site, and I now, including the preceding post, have 3 different explanations to choose from:
a) It is a gentle wide dip across the midrange to compensate for uneven response of early phono cartidges, thereby reducing listening fatigue.
b) It is a narrow dip around the crossover point of early BBC speakers to compensate for the increase of off-axis response as the tweeter takes over.
c) It is a rise (not a dip) across the vocal range to increase presence.
Any more votes?
a) It is a gentle wide dip across the midrange to compensate for uneven response of early phono cartidges, thereby reducing listening fatigue.
b) It is a narrow dip around the crossover point of early BBC speakers to compensate for the increase of off-axis response as the tweeter takes over.
c) It is a rise (not a dip) across the vocal range to increase presence.
Any more votes?
Dip.
But I suspect that it was to compensate for the on-axis response of the "midrange" becoming narrower as it neared the xover point to the tweeter, IF that sort of thinking played a role at all.
In reality I think that someone came up with it just to make things sound "better" and without any direct technical research or purpose behind it. If they came up with some sort of technical justification I'd bet it was after someone decided it "sounded better"!
Pretty sure the LS3 had a dip built in, not a rise.
_-_-bear
But I suspect that it was to compensate for the on-axis response of the "midrange" becoming narrower as it neared the xover point to the tweeter, IF that sort of thinking played a role at all.
In reality I think that someone came up with it just to make things sound "better" and without any direct technical research or purpose behind it. If they came up with some sort of technical justification I'd bet it was after someone decided it "sounded better"!
Pretty sure the LS3 had a dip built in, not a rise.
_-_-bear
We're probably looking at different things. There's both. The dip is typified by designs such as Spendors etc. The rise, on the other hand comes from the infamous LS3/5a. It does / did exist. I quote Alan Shaw of Harbeth (knew I could find the article somewhere -been looking for it for an hour) Hifi News, November 2004.
'The computer optimised LS3/5a (i.e. the 11ohm version) reshaped the frequency response to that of the original 1974 BBC specification except in high frequency output above 10KHz. In the 1970s, pre-digital, a considerable lift at the top end was considered a desirable characteristic to 'pre-emphasise' the contribution of hiss from analouge tape to a recording and hence to alert the sound engineer how best to maximise the dynamic range. This characteristic, however, is completely inappropriate for digital recordings, for which the MK 2 LS3/5a is a much better match.
'A peak in output at around 1-2KHz was not a characteristic of the original BBC Design Report RD 1976/29 -although the HF boost is evident - and is probably not present on the very first production LS3/5As, but is in evidence with a vengence by the mid/late 1980s, if not rather earlier. It is typically +6db above the average pass band output... That the "raw" bass unit is at its most efficient in that region was taken into account in the design of the network at the BBC, and energy was sucked-out between 1-2KHz by judicious use of a series notch filter comprising coil L2, capacitor C5 and damping resistor R2 in the crossover feed to the bass unit.'
There you have it, from a man who probably knows more about BBC speakers than anyone else alive. So it depends which one (the dip many 'BBC' designs, or the lift the LS3/5a had) you were refering to.
'The computer optimised LS3/5a (i.e. the 11ohm version) reshaped the frequency response to that of the original 1974 BBC specification except in high frequency output above 10KHz. In the 1970s, pre-digital, a considerable lift at the top end was considered a desirable characteristic to 'pre-emphasise' the contribution of hiss from analouge tape to a recording and hence to alert the sound engineer how best to maximise the dynamic range. This characteristic, however, is completely inappropriate for digital recordings, for which the MK 2 LS3/5a is a much better match.
'A peak in output at around 1-2KHz was not a characteristic of the original BBC Design Report RD 1976/29 -although the HF boost is evident - and is probably not present on the very first production LS3/5As, but is in evidence with a vengence by the mid/late 1980s, if not rather earlier. It is typically +6db above the average pass band output... That the "raw" bass unit is at its most efficient in that region was taken into account in the design of the network at the BBC, and energy was sucked-out between 1-2KHz by judicious use of a series notch filter comprising coil L2, capacitor C5 and damping resistor R2 in the crossover feed to the bass unit.'
There you have it, from a man who probably knows more about BBC speakers than anyone else alive. So it depends which one (the dip many 'BBC' designs, or the lift the LS3/5a had) you were refering to.
Dip or rise depends what your point of reference is, but dip if you believe google.
Link for story of BBC speaker development. Seems like they had supplier problems with mid bass units. Looks like they ran it (mid bass) with a elliptical LPF and a fair amount of BSC.
http://www.ls35a.com/
Link for story of BBC speaker development. Seems like they had supplier problems with mid bass units. Looks like they ran it (mid bass) with a elliptical LPF and a fair amount of BSC.
http://www.ls35a.com/
The last link provided, to the unofficial LS3/5a web site, is the best information you'll find on the BBC design method which led to the "BBC curve". Dip or rise is dependent upon how you look at it since the "curve" exhibited both.
First, there is no excessive rise in the high end of the BBC designs. While the KEF tweeter used in the 3/5a was capable of extension beyond 20kHz, the high frequencies were intentionally rolled down above 15kHz. (http://www.stereophile.com/standloudspeakers/361/index10.html)
There was a slight rise in the presence region and a most noticeable rise around 110Hz. This midbass response gave the impression of deeper bass than the small package of the 3/5a could actually manage. (Though its in room response and soundstage size remain remarkably impressive for a speaker no bigger than a shoebox.) It also made the speaker difficult to mate with a subwoofer. Those 3/5a users who tried various methods of extending the bas response to mate with the ourstanding soundstaging of the speaker were often disappointed. The best match was generally considered the AB1, a sub designed by Rogers, which also used the same KEF B110, a 4" driver which the 3/5a employed. Placed in a bandpass enclosure, the AB1 was either the thrill of a lifetime for those 3/5a users who wanted some bass extension or the bane of a boomy box combined with the taughtness of the sealed box monitor.
The original 3/5a package (along with its now famous curve) was never intended as a commercial product and became such only after BBC employees began buying parts to assemble units for themself and friends. Several companies where eventually liscenced to manufacture the 3/5a though any speaker from any manufacturer had to use the same cabinet and drivers. (The design of the cabinet is anachronistic but a gold mine of insight into how to greate an inert enclosure with minimal diffraction without the assistance of a computer. The whole project at the BBC was done with slide rules and good ears.) The crossover was designed by the BBC and could be assembled by the manufacture during the period of the 15 Ohm system. The crossovers are fairly complex and employed plug ins to match the response of the system(the tweeter's response) to the 0.5dB spec. During the run of the speaker KEF went off in another direction and the drivers used in the BBC designs began to suffer. A revised design was initated in 1987 and the speaker became an 11 Ohm unit. Later a bi-wired version was introduced. The 3/5a remained in production until the late 1990's.
One of the most interesting facts of the speaker is that any 3/5a produced by any manufacturer at any time must substitute as a single unit for any other 3/5a with drop in ease. Therefore there never was a revision which altered the speaker's frequency response or other performance characteristics. Even the 11 Ohm system must sub for a 15 Ohm with perfect maching side to side. The revision was made to bring the KEF drivers back to spec. Too many were being rejected by the manufacturers to remain in spec. When the change was made from 15 to 11 Ohms, the various manufacturers were finally allowed to produce their own crossovers which still had to match the BBC spec. Although the crossover is rather complex, the 3/5a is still considered one of the most tube friendly speakers ever produced.
If you look at the measurements page in Stereophile, you will notice more and more small monitors today are employing very much the same frequency curve as the 3/5a. There is generally greater frequency extension at the top end which can be viewed any way you prefer. Most often the high frequencies are boosted to make a speaker jump off the shelf. The bass response is still slightly boosted and the mids have the same presence rise. The overall response of the Wilson MAXX, at $45k, shows some signs of following the traditional BBC curve.
In case anyone wonders, I've sold the BBC designs and still use a pair of 1979 Rogers 3/5a's as my main speakers running off a pair of McIntosh tube amplifiers. So I've remained a dedicated BBC fan and have responded many times to questions concerning their nature. The link to the Unofficial Site has many interesting factoids about the development of the speakers. Most important to your question would be that the original 3 series speakers were never intended as commercial speakers and the 3/5a's were designed for mobile broadcast monitoring. This created a very nearfield lisetning situation and hence the BBC curve.
First, there is no excessive rise in the high end of the BBC designs. While the KEF tweeter used in the 3/5a was capable of extension beyond 20kHz, the high frequencies were intentionally rolled down above 15kHz. (http://www.stereophile.com/standloudspeakers/361/index10.html)
There was a slight rise in the presence region and a most noticeable rise around 110Hz. This midbass response gave the impression of deeper bass than the small package of the 3/5a could actually manage. (Though its in room response and soundstage size remain remarkably impressive for a speaker no bigger than a shoebox.) It also made the speaker difficult to mate with a subwoofer. Those 3/5a users who tried various methods of extending the bas response to mate with the ourstanding soundstaging of the speaker were often disappointed. The best match was generally considered the AB1, a sub designed by Rogers, which also used the same KEF B110, a 4" driver which the 3/5a employed. Placed in a bandpass enclosure, the AB1 was either the thrill of a lifetime for those 3/5a users who wanted some bass extension or the bane of a boomy box combined with the taughtness of the sealed box monitor.
The original 3/5a package (along with its now famous curve) was never intended as a commercial product and became such only after BBC employees began buying parts to assemble units for themself and friends. Several companies where eventually liscenced to manufacture the 3/5a though any speaker from any manufacturer had to use the same cabinet and drivers. (The design of the cabinet is anachronistic but a gold mine of insight into how to greate an inert enclosure with minimal diffraction without the assistance of a computer. The whole project at the BBC was done with slide rules and good ears.) The crossover was designed by the BBC and could be assembled by the manufacture during the period of the 15 Ohm system. The crossovers are fairly complex and employed plug ins to match the response of the system(the tweeter's response) to the 0.5dB spec. During the run of the speaker KEF went off in another direction and the drivers used in the BBC designs began to suffer. A revised design was initated in 1987 and the speaker became an 11 Ohm unit. Later a bi-wired version was introduced. The 3/5a remained in production until the late 1990's.
One of the most interesting facts of the speaker is that any 3/5a produced by any manufacturer at any time must substitute as a single unit for any other 3/5a with drop in ease. Therefore there never was a revision which altered the speaker's frequency response or other performance characteristics. Even the 11 Ohm system must sub for a 15 Ohm with perfect maching side to side. The revision was made to bring the KEF drivers back to spec. Too many were being rejected by the manufacturers to remain in spec. When the change was made from 15 to 11 Ohms, the various manufacturers were finally allowed to produce their own crossovers which still had to match the BBC spec. Although the crossover is rather complex, the 3/5a is still considered one of the most tube friendly speakers ever produced.
If you look at the measurements page in Stereophile, you will notice more and more small monitors today are employing very much the same frequency curve as the 3/5a. There is generally greater frequency extension at the top end which can be viewed any way you prefer. Most often the high frequencies are boosted to make a speaker jump off the shelf. The bass response is still slightly boosted and the mids have the same presence rise. The overall response of the Wilson MAXX, at $45k, shows some signs of following the traditional BBC curve.
In case anyone wonders, I've sold the BBC designs and still use a pair of 1979 Rogers 3/5a's as my main speakers running off a pair of McIntosh tube amplifiers. So I've remained a dedicated BBC fan and have responded many times to questions concerning their nature. The link to the Unofficial Site has many interesting factoids about the development of the speakers. Most important to your question would be that the original 3 series speakers were never intended as commercial speakers and the 3/5a's were designed for mobile broadcast monitoring. This created a very nearfield lisetning situation and hence the BBC curve.
soufiej said:
It's all there in the unofficial LS3 -5a link. Thanks for your additional interpretation and insight.
It should be noted there is about 4-5 dB bass boost via the shelving in the midbass crossover, as is often termed on this site as BSC. If employed as near field monitor it should be well away from walls.
FYI the LS3-5a is still available as V2 with updated drivers. Still licensed by the BBC. see link http://store.acousticsounds.com/browse_detail.cfm?Title_ID=16618&keywords=NL
The HF lift went with the 11 ohm version. For the most comprehensive measurements and analysis, see the article in Hifi News, August 1988, which compared the earlier type and the later like-for-like.
They did have major driver problems, particularly with the KEF B110 as some of the materials & construction changed. While the lab models, and the BBC prototypes etc. worked to spec, they were generally using hand-assembled and matched parts in the drivers, lovingly crafted. The drivers off the production-line had far greater variation. The story is well enough known of course, and as was observed above, it's important to remember that the LS3/5a, whatever it's merits or demerits, was not designed as a commercial speaker, but as a nearfield monitor for broadcasting use, which has a different set of requirements. However, let's not turn this into another LS3/5a debate -my apologies and I hold my hand up for starting this one.
Final semi 'analysis', there's a BBC dip, and a less-well-known BBC lift. The latter is pretty specific to one speaker, the LS3/5a, and then only with the earlier 15ohm model, according to the bloke who basically designed the MK2 computer optimised 11ohm variation for them at Harbeth which appeared from 1988. A slight rise in the HF is useful if mastering analogue, or if you like a bright sound. The former is what is perhaps most commonly associated with speakers under the BBC banner, and if you find a slight easing of presence in the vocal region useful to kill any sibilance, then this can also be a good move.
So we're all right.
They did have major driver problems, particularly with the KEF B110 as some of the materials & construction changed. While the lab models, and the BBC prototypes etc. worked to spec, they were generally using hand-assembled and matched parts in the drivers, lovingly crafted. The drivers off the production-line had far greater variation. The story is well enough known of course, and as was observed above, it's important to remember that the LS3/5a, whatever it's merits or demerits, was not designed as a commercial speaker, but as a nearfield monitor for broadcasting use, which has a different set of requirements. However, let's not turn this into another LS3/5a debate -my apologies and I hold my hand up for starting this one.
Final semi 'analysis', there's a BBC dip, and a less-well-known BBC lift. The latter is pretty specific to one speaker, the LS3/5a, and then only with the earlier 15ohm model, according to the bloke who basically designed the MK2 computer optimised 11ohm variation for them at Harbeth which appeared from 1988. A slight rise in the HF is useful if mastering analogue, or if you like a bright sound. The former is what is perhaps most commonly associated with speakers under the BBC banner, and if you find a slight easing of presence in the vocal region useful to kill any sibilance, then this can also be a good move.
So we're all right.
sreten - I have to say I cannot remember ever seeing a measurement of a 3/5a which had the rising response. I don't even remember the original LS3/5 (no "a") having a rising response. I notice your chart is stated to be a "subjective" measurement. I assume that means subjective/in room (possibly with no real measuring device) which would explain the slight rise and why flat response at the highest frequencies might not be the best approach for all speakers. I'm not able to find where you retieved this chart. Can you help?
Sorry to turn this into another 3/5a discussion. However, it and the Spendor BC1 and BC2 seem to be the best representatives of the classic BBC sound. That sound has always meant, to my ears, a very lazy sinusoidal wave which mainly features the slight lump in the midbass and the gradual rolloff at the high frequencies. While not truly represented in the measurements, I would also add that most people I know who are familiar with the BBC "sound" would claim it has to have precise clarity and correct timbre through the mids. The now classic admonition, "first, get the mids right", seems to have come from the period when the BBC designs dominated a certain segment of speaker design. I think that quality alone, more than the "dip", "rise" or "curve" of the frequency response truly represents the BBC sound.
Those cynical few out there would say the "classic" BBC sound is best represented by the dead silence of a little, tiny driver with a scrawny, little voice coil that couldn't tolerate a real he-man American made Ampzilla or Phase Linear 700 unloading all its wattage into the speaker when they dropped the stylus or played "Tank".
Sorry to turn this into another 3/5a discussion. However, it and the Spendor BC1 and BC2 seem to be the best representatives of the classic BBC sound. That sound has always meant, to my ears, a very lazy sinusoidal wave which mainly features the slight lump in the midbass and the gradual rolloff at the high frequencies. While not truly represented in the measurements, I would also add that most people I know who are familiar with the BBC "sound" would claim it has to have precise clarity and correct timbre through the mids. The now classic admonition, "first, get the mids right", seems to have come from the period when the BBC designs dominated a certain segment of speaker design. I think that quality alone, more than the "dip", "rise" or "curve" of the frequency response truly represents the BBC sound.
Those cynical few out there would say the "classic" BBC sound is best represented by the dead silence of a little, tiny driver with a scrawny, little voice coil that couldn't tolerate a real he-man American made Ampzilla or Phase Linear 700 unloading all its wattage into the speaker when they dropped the stylus or played "Tank".
http://www.stereophile.com/standloudspeakers/361/index1.html
Hi,
there was also a letter in HiFi News after the "LS3/5a shootout"
from the guy that has run Harbeth for years debunking the myth
that the original version is the "best".
He described a number of problems in the originals addressed by
later updates, the HF emphasis being one of the things mentioned.
He should know, he's made the things for years.
/sreten.
Fig.2 Rogers LS3/5a, 1978 sample, spatially averaged, 1/3-octave response in JA's room.
Anyone familiar with room curves can see the top end emphasis.
Hi,
there was also a letter in HiFi News after the "LS3/5a shootout"
from the guy that has run Harbeth for years debunking the myth
that the original version is the "best".
He described a number of problems in the originals addressed by
later updates, the HF emphasis being one of the things mentioned.
He should know, he's made the things for years.
/sreten.
Fig.2 Rogers LS3/5a, 1978 sample, spatially averaged, 1/3-octave response in JA's room.
Anyone familiar with room curves can see the top end emphasis.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.