Version HTML de base
20
“Uncertainty-noise” Le Mans
Acoustique
&
Techniques n° 40
according to the GUM, leaving it to the users of a certain
standard to develop their own empirical knowledge to derive
the total uncertainty of a measurement.
Another apparent problem, though essentially of an editorial
nature is the terminology related to uncertainty. Terms such
as “random and systematic errors”, “level of confidence”,
“accuracy” in quantitative statements, etc. should be avoided
but are still in common use. Based on the same background,
the former rules of “error propagation” are often applied. This
seems to be a question of education and will certainly quickly
be overcome.
On each of the circulated documents, the Committee Chairman
submitted detailed comments mostly offering specific
proposals for amendments. Even minor editorial points were
raised to avoid the risk of poor examples being copied. In the
majority of documents however, some technical changes are
necessary. In most cases, the given advice is well accepted
and efforts are visible to improve the respective sections.
Comments of an editorial nature have usually been immediately
incorporated in the next version.
It was noted that comments received from ISO Member Bodies
on these documents touch the subject of uncertainty only in
exceptional cases indicating that knowledge in this area is not
yet very widespread at the national level.
Policy of CEN/TC 211
At its first meeting in 1989, CEN/TC 211 adopted a resolution
in principle to develop most of its standards through the
corresponding international committees, ISO/TC 43 and ISO/
TC 43/SC 1. Since that time, this policy has been followed
and all its standards with one exception have been adopted
by parallel voting, in conformity with the Vienna Treaty. While
CEN/TC 211 has not taken a resolution specifically on policy
in relation to statements of measurement uncertainty, it has
by adoption of all the relevant ISO standards taken a de facto
position of following the ISO practice. This remains the position
to date.
Conclusions
ISO TC 43 “Acoustics” and its Subcommittee 1 “Noise” recently
adopted a specific policy on the treatment of measurement
uncertainty in newly developed or revised standards dealing
with measurement and prediction of sound. This policy
could not be expected to be fully implemented immediately.
Considerable improvements were, however, achieved in
less than two years since its adoption. Most of the Working
Groups concerned are quite eager to meet the new challenge
and it is certain that further progress will be made if more
acceptable documents become available. A Working Group
having once reached consensus on a certain concept and
wording will quickly transfer this to other documents under
their responsibility. However, considering the large number
of approximately 80 sound measurements standards issued
by TC 43 and its SC 1 so far, it will certainly take some years
before each of these standards is updated with respect to
measurement uncertainty.
The efforts of the standardization bodies should be considered
as a contribution towards a more transparent description of
noise measurement results facilitating their assessment. It is
hoped that they will be acknowledged by the customers.
■
Treatment of Measurement Uncertainties in International and European Standards on Acoustics