You should be aware of the BBC's position on this and the other LS speakers.


The selection of loudspeakers for BBC Radio & Music.


R. Walker, BBC Research and Development Department.





Apart from the obvious need for high quality reproduction, the large number
of studios and control rooms (more than 600 in the period around 1980) meant
that the quality and character of the audio reproduction had to be
maintained over a large number of units, over a range of loudspeaker sizes
and for periods of many years. One of the main reasons for requiring
consistency was (and still is) production practices that require operational
staff to move between rooms for different jobs and for programme audio to be
processed consecutively in different facilities (especially in television).
Both of those factors require the sound monitoring in different rooms to be
as similar as possible, otherwise much time and effort can be spent in doing
and re-doing changes to the recorded material. That is unlike a commercial
recording studio or small radio/TV station, where the technical staff work
almost all the time in the same few rooms and can become familiar with their
eccentricities.



Those considerations led, over the years, to a succession of BBC-designed
loudspeakers that also achieved substantial national and international
recognition. The BBC-designed loudspeakers introduced several new
developments, such as the use of polyethylene/ polypropylene (PE/PP)
co-polymers for the diaphragm and Kapton® voice coil formers. The unit costs
of BBC-designed loudspeakers were also significantly lower than equivalent
commercial products, at that time.



The last successful BBC-designed loudspeakers were the small LS3/5A (c.
1975), the large LS5/8 (c. 1982) and the medium LS5/9 (c. 1984). Those
loudspeakers were the main ones in use from the dates of their original
designs to the present. Very many are still in use at the present time
(2004). In design, they may be well past their time and are, in any case,
now irreplaceable. This author suggested in about 1992 that work should be
started on their successors. At that time, the loudspeaker designs were
between 8 and 17 years old. They are now between 20 and 29 years old.

Three main factors led to the demise of the loudspeaker development work.
With the coming of a more commercial imperative to the BBC as a whole, it
was no longer considered worthwhile to develop loudspeakers internally.
.........

A second factor has been the availability of improved commercial
loudspeakers. Over the period of about the last 40 years there is no doubt
that commercial loudspeakers have improved dramatically. Even over the last
20 years since the design of the LS5/9, developments in commercial
loudspeakers have led to improved properties, perhaps most of all in
consistency.



BBC Radio and Music (BBC R&M) considered that the scale of the project made
it necessary to carry out a study to select a range of loudspeakers with the
objective of setting up a Framework Agreement for the procurement of
'standard' loudspeakers. They invited News and World Service to take part,
so that they could take advantage of the results if they wished. The size of
the requirement made it necessary to go through a formal tendering process.
It also exceeded the EU threshold, which meant that the EU Procedure
(Restricted) process had to be applied.

In order to include the quality of the loudspeaker as a selection parameter
it was necessary to carry out formal listening tests. Even without that
potential legal requirement, it was thought desirable to have a uniform
loudspeaker installations throughout the new development.
........................



3 Test methodology.

3.1 The Parameters of an Ideal Test.

A rigorous subjective test should consist of double blind, individual
assessments using a suitable number of descriptive axes by a suitable number
of test subjects. Probably five to ten subjective axes would be used. That
is likely to be higher than the true number of different quality attributes,
but would allow some scope for subsequent statistical factor reduction.
Other tests doing similar things have commonly used around 20 descriptors.
If the set of descriptors is too limited then test subjects feel constrained
whilst having too many leads to confusion and overload of the test subjects.



It was thought that not less than 10 test subjects would be needed, if they
were highly skilled and could demonstrate a reasonable degree of
consistency. For less able subjects, at least two or three times that number
would have been necessary. The subjects were also being asked to assess many
different aspects of loudspeaker quality at the same time. Less expert
subjects would have found that difficult. Because of the background to the
tests, it was thought that the test subjects should also be, in some sense,
'chosen' by their colleagues, so that the tests would be seen as reasonable
and the final results more readily accepted by the majority of production
staff who had not taken part directly.

This is not the whole text of this report but it can be found at the BBC's
Research web site. http://www.bbc.co.uk/rd/pubs/reports/index.shtml



You will need to download a 23 page pdf file and look for the above title.