/

AUDIO engineering society ,

Containing the Activities and Papdn of tho Society, and pub-
lished monthly as a part of AUDIO ENGINEERING Magazine
OFFICERS
w
Audio Enginbring Society,
..
John D. Colvin . . . . . . President Bob Hugh Smith Western Vice.-Pres.
C. C. McProud Exekutive Vice-Pres. Lawrence Shipley Central Vice.-Pres.
Box F, O s e r n ~ ,N. Y. . ..
Norman C. Pickering . . . . Secretary Ralph A. Schlegel . . . Treasurer

The Measurement of
Audio Volume
lj. A. CHINN"
Part II-A comprehensive discussion of the problems involved and the instruments em-
ployed to indicate program level and sine-wave tones in broadcast and recording circuits.
- is calibrated so as to read zero vu on a
A THOROUGH COMPREHENSION of the brating voltage, inasquch as a volume
connotations of the term "reference sine wave power of, say, one milliwatt in indicator is generally a high-impedance,
volume" is fundamental to any a stated impedance, a speech or program voltage-responsive device. A reference
studio engineering endeavor. , Unfortu- wave in the same impedance whose in- Eve1 could conceivably be established
nately, experience has shown that this tensity is such as to give also a reading based on voltage and the unit of
subject is often completely misunder- of zero vu will have instantaneous peaks measurement might be termed "volume-
stood. I t is hoped that the following of power which are several times one volts." However, volume 'measurements
will dispel the vague understanding that milliwatt and an average power which are a part of the general field of trans-
sometimes surrounds this simple subject. is only a small fraction of a milliwatt. mission measurements, and the same
I t is important to appreciate that I t is therefore erroneous to say that reasons apply here for basing them on
reference volume is a practical and use- reference volume is one milliwatt. power considerations as in the case of
ful concept, but one which is quite arbi- Moreover, it should be emphasized ordinary transmission measurements us-
trary and not definable in fundamental that although it is convenient to measure ing sine waves. If the fundamental con-
terms. As already mentioned, it cannot the performance of amplifiers and sys- cept were voltage, apparent gains or
be expressed in any single way in terms tems by means of single frequencies losses would appear wherever impedance
of the ordinary electrical units of power, there is no ex&f universaP relatiofiship
potential, or current. Reference volume between the single-frequency load-carry-
is describable only in terms of the elec- ing capacity indicated by such measure-
trical and dynamic characteristics of an ments, and the load-carrying capacity
instrument, its sensitivity as measured by for speech and program waves expressed
its single-frequency calibration, and the in terms of volume level. This relation-
technique of reading it. I n other words, ship depends upon a number of factors
reference volume may be defined as that such as the rapidity of cutoff at the over-
level of program which causes a standard load point, the frequency bandwidth be-
volume indicator, when calibrated and ing transmitted, the quality of service
used in the accepted way,, to read zero to be rendered, and similar factors.
vu. The question may well be raised why
The sensitivity of the standard volume reference volume has been related to a
indicator is such that reference volume calibrating power rather than to a cali-
corresponds to the indication of the in-
strument when it is bridged across a I
Courtesy General Electric Co.
600-ohm resistor3 in which is flowing
one milliwatt of sine-wave power. Fig. 3. The "B" type standard yolume indicator
It is especially cautioned that refer- scale emphasizes the percentage scale. This
ence volume should not be confused with scale is used extensively for program trans-
the single-frequency power used to cali- mission applications. /

brate the zero volume setting of the

volume indicator. If a volume indicator transforming devices (such as trans-
formers) occur in a circuit. This diffi-
* Columbia Broadcasting System, 485 , culty is avoided by adopting the power
Madison Ave., New York 22, N. Y. concept, making suitable corrections in
a A standard impedance of 600 ohms was the readings when the impedance is
chosen originally since, keeping in mind other than 600 ohms.
the telephone plant, there was more audio
equipment designed to this impedance than Volume Measurement Terminology
to any other.
Courtesy General Electric Co. (a) VU. The terminology that is used
to express volume measurements was
Fig. 2. The "A" type standard volume indicator created to avoid confusion as to the type
scale. emphasizes the VU markings and has an of volume indicator used and 'the refer-
inconspicu~~isvoltage scale. This type of scale
ence level. The term "vu" (pronounced
is commonly used for transmission measuring
sets. "vee-you") is used; the number of vu ,
'24 AUDIO ENGINEERING OCTOBER, 1951
 time, the length of time depending upon
the program material. For speech a 5 to
10 second period of observation may be
sufficient whereas for symphonic music
1to 2 minutes may be necessary. During
this time the adjustable attenuator,
which is a part of the volume indicator,
is adjusted so that the extreme deflec-
tons of the instrument needle just reaches
Fig. 4. The basic circuit for the standard volume indicator has an input impedance of 7500 ohms. the reference point; i.e., a scale reading
For maximum sensctivity the loss in the adjustable attenuator is reduced to zero, but in order to of zero on the vu scale or 100 on the
maintain proper dynamic characteristics the indicating instrument itself must always fqce approxi- percent voltage scale (see Figs. 2 and
mately 3900 ohms. 3). The volume level is then given by
being numerically the same as the num- the "vu" level used for program trans- the designations @umberedon the attenu-
ber of db above or below the reference misson peaks and the "dbm" level used ator. If, because of the coarseness of the
volume level. The use of this term is for system measurements. In practice it adjustments ,provided or for other
restricted to the ASA standard volume has been found that with typical pro- reasons the deflections cannot be brought
indicator described herein. A volume gram material of a given crest amplitude, exactly to the 0 vu or 100 per cent mark,
level reading can be correctly expressed the standard volume indicator reaches the reading obtained from the setting of
in terms of vu only when it has been an indication 8 to 14 db below that the attenuator may, if desired, be cor-
made with an instrument having the reached with a steady tone of the same rected by adding the departure from 0
electrical and dynamic characteristics crest amplitude. T o nominally take into shown on the vu scale of the instrument.
described. account this 8 to 14 db difference in In the interests of accuracy the steps on
(b) DBM. For steady-state measure- response, the established practice is that the adjustable attenuator should not ex-
ments a reading in "vu" would denote a
specific single-frequency audio power, * 390QnAdjusjobb
for dynamic program indications "vu"
denotes only a volume level. This dual
meaning of "vu" is avoided by the use Copper
of the term "dbm" for all steady-state
measurements. As defined, a reading ex- Instrument
pressed in "dbm" a t once indicates the
power level of a steady, single-frequency 1-
signal where the number o f "dlirn" is Fig. 6. Volume indicator circuit having low input impedance which terminates source and also
equal to the number of decibels above or provides increased volume indicator sensitivity.
below a reference Dower of 1 milliwatt.
(c) DBM vs. VU. It is to be noted performance requirements must be met ceed 2 db. so that the departure from the
that a "vu" reading can be made only at a single-frequency test-tone level that reference point never need exceed this
on a standard volume indicator whereas is at least 10 db higher than the normal amount. Particular attention is called
sine-wave power measured with the program peaking level (for example, in to the fact that, unlike almost any other
standard volume indicator or with any a system that is to transmit program electrical indicating instrument, the voE
other suitable a.c. instrument can be material at + 8 vu, all single-frequency ume indicator reading is determined pri-
expressed in "dbm"." measurements would be made at +I8 dbm marily from the setting of the associ-
DBM is a unit of finite power whereas test-tone level). This procedure reason- ated range switch and, in effect, only a
"vu~'is a measure of volume level and, ably insures that system performance is secondary correction is obtained by
as already discussed, has no connotation within standards under normal operating observing the deviation of the needle
of filcite power level. Thus no direct re- conditions. from exact coincidence with the refer-
lationship between "dbm" and "vu" can ence point on the scale. t

be established. Reading the Volume Indicator

From a practical standpoint, however, Since program material is of a rapidly Features of the Standard Volume Indicator
some relationship is desirable between varying nature, a reading of a volume The volume indicator that has been
indicator cannot be obtained instantly. described has the statute of an American
H. A. Chinn, AUDIOENGINEEBINC,
Vol. Rather, the gyrations of the needle must Standard.= I n the many years since it
32, No. 3, p. 28, March 1948 be watched for an appreciable period of was first placed into service there have
been no ihanges, whatsoever, in either
the fundamental requirements or the spe-
cific features of the instrument that was
developed to meet the basic needs. Be-
cause of the importance and the wide-
spread use of the instrument some of the
detailed characteristics that are of con-
cern to the design engineer are pre-
sented below.
(a) Response vs. Frequency Character-
istic. The sensitivity of the volume indi-
cator instrument shall not depart from
that a t 1000 cps by more than 0.2 db
between 35 and 10,000 cps nor more than
0.5 db between 25 and 16,000 cps.
(b) Input Impedatzce. The impedance of
the volume indicator arranged for
bridging across a line is about 7500

I 10
CIRCUIT
100
ITERATIVE IMPEDANCE - OHMS
I000

5 Volume
[Continued ON $a&e 481

measurements of electrical
Fig. 5. Volume indicator correction factors (to be added to volume indicator reading) for use speech and program waves, American
when instrument is bridged across circuits having iterative impedances other than 600 ohms. Standards Association C16.5-1942

AUDIO ENGINEERING OCTOBER, 1951

 stitute a complete volume indicator but
AUDIO VOLUME must have certain simple circuits asso-
ciated with it. The basic form which this
MEASUREMENT circuit takes is illustrated in Fig. 4. This
is a high-impedance (7500-ohm) ar-
[ f r o m page 261 rangement intended for bridging across
low impedance lines. As noted above,
ohms when measured with a sinusoidal about 3600 ohms of series resistance has
voltage sufficient to deflect the pointer been removed from the instrument and
to the 0 vu or the 100 per cent scale must be supplied externally in order to
marking. Of this impedance, 3900 ohms provide a point where the impedance is
:s in the meter and about 3600 ohms the same in both direction, for the in-
must be supplied externally to the meter, sertion of an adjustable attenuator. A
this value of series resistance being re- portion of the series resistance is made
adjustable as shown by the slide wire
quired in order to meet the above d ~ - in the diagram. This is for the purpose of
namic characteristics. facilitating accurate adjustment of the
(c) Sensitivity. The application of a sensitivity to compensate for small dif-
1000-cps potential of 1.228 volts r.m.s. ferences between instruments and any
(4 db above 1 milliwatt in 600 ohms) to slight changes which may occur with
the instrument in series with the proper time.
external resistance (3600 ohms) causes The maximum sensitivity possible
a deflection to the 0 vu or 100 point on with this, the simplest circuit, is + 4 vu
the scale. The instrument, therefore, has for indications at the 0 vu or 100 per
only sufficient sensitivity at its normal cent mark when placed across 600-ohm
reference point (0 vu or 100) to indi- line. The maximum sensitivity occurs, of
cate a volume level of + 4 VU. I t has not
been found possible to design. more course, when the loss in the adjustable
sensitive instruments while meeting attenuator is zero. The upper limit to the
other requirements. range of measurement is limited only
There should be no confusion because by the amount of loss introduced by the
the instrument deflects to a scale mark- adjustable attenuator, its power handling
ing of 0 vu when a level of + 4 VU is capacity and that of the two series re-
applied to it., The 0 point on the vu sistors.
scale is merely an arbitrary point at The volume indicator of Fig. 4 can
which it is intended nominally to read be bridged across circuits of other than
.the instrument, and the rest of the vu 600 ohms, of course, but when this is
scale represents deviations from the 0 done a correction factor must be applied
point. The volume level is read, not from in order to determine the true level.
the scale, but from the indications on the Figure 5 shows the magnitude of the
associated sensitivity control when the correction factor. I t is to be noted that
latter is set so as to give a scale de-
flection of zero (as detailed in a fore- the basic volume indicator circuit ( F i g .
going section). 4) has an input impedance of 7500 ohms
(d) Overload Cajmcity: The instrument and should not be bridged across circuits
is capable of withstanding, without in- of appreciably higher than about one-
jury or effect on calibration, peaks of tenth of this impedance if undue loading
10 times the voltage equivalent to a de- of the circuit is to be avoided. O n the
flection to the 0 vu or 100 scale point other hand, when used across circuits of
for 0.5 second and a continuous overload less than the normal 600 ohm value, the
of 5 times the same voltage. sensitivity of the instrument is reduced,
(e) Presence of Magnetic Material. I t as indicated by Fig. 5.
should be cautioned that the presence of Figure 6 shows an arrangement in
magnetic material near the movement which, by adding a transformer, the
of the instrument may affect its calibra- sensitivity has been increased a t the
tion and dynamic characteristics. This
is because, to obtain the desired sensi- expense of decreasing the input imped-
tivity and dynamic characteristics, it ance to a low value. The circuit is de-
has been necessary to employ more pow- signed so that the impedance facing the
erful magnets than usually required for instrument itself is the same as in the
such instruments, and any diversion of basic circuit ( F i g . 4). Thus the correct
flux to nearby magnetic objects effec- dynamic characteristics are obtained.
tively weakens the useful magnetic field The input impedance, on the other hand,
beyond the point where these character- is low, hence the device cannot be
istics can be met. The instruments bridged across a through line but must
should not, therefore, be mounted on be used to terminate the circuit. I n prac-
steel panels. (The effect is only slight, tice, approximately a 10 db increase in
however, if they are mounted on 1/16-in. sensitivity may be obtained by this ar-
steel panels with the mounting hole cut rangement.
away as far as possible without extend- In high-fidelity audio systems the use
ing beyond the face of the meter case.) of a 150-ohm circuit impedance is be-
In the instruments as now available, coming common p r a c t i ~ e . ~However,
the deviation of the sensitivity with tem- when the basic volume indicator circuit
perature is less than 0.1 db for tempera- is used with 150-ohm circuits. a loss in
tures between 50' F. and 120' F., and sensitivity of 6 db results. Therefore, on
is less than 0.5 db for temperatures as a 150-ohm circuit the instrument is ca-
low as 32O F.
The instrument by itself does not con- Monroe and Palmquist, Proc. I.R.E.,
Vol. 36, No. 6, p. 786 (June 1948).
AUDIO ENGINEERING OCTOBER, 1951
pable of reading levels only down to veloped that on the average, listeners
c 10 vu. The loss in sensitivitv in situa- prefer to hear broadcast music and
tions such as this can be avoided by speech a t about the same peak levels
utilizing a circuit similar to Fig. 6 ex- ar read on a standard volume ilzdicator.
cept that the step-up ratio of the trans- Furthermore, listeners like to hear
former is made only great enough to broadcast music and speech a t the same
overcome its own losses and that which
results from the existence of lower relative levels, regardless of the absolute
voltage (for a given volume level) on sound level that is individually pre-
a circuit of lower impedance. I n the ferred.
particular instance cited a step-up ratio The Columbia Broadcasting System
of slightly more than 1 : 2 would be is following this practice with consider-
necessary if the transformer were with- able success. The measure of success in
out losses and slightly greater than this this instance being the almost complete
in a practical case.
The input impedance of the volume absence of listener complaints concern-
indicator (7500 ohms in the basic cir- ing the relative loudness of speech and
cuit) will be reduced under these cir- music. A few complaints continue to be
cumstances by a factor equal to the received of course, but when specific
square of the turns ratio. This is not cases are investigated it is almost al-
a serious disadvantage, however, since ways found that either (a) the recom-
the impedance of the circuit being mended transmission practices were
bridged has also been reduced by es- violated'because of some special circum-
sentially this same factor. Therefore the stance or (b) that the program origin-
ratio of the circuit impedance to the
bridging impedance remains approxi- ated on another network.
mately the same. For the particular ex-
ample cited, namely a 150-ohm circuit,
practical volume indicator transformers,
when terminated in a standard volume
indicator, have input impedances of
about 1700 ohms.
Transmission Level Practices
The volume levels that are used for
the transmission of speech and program
waves may be determined upon in a
number of ways depending upon the
communications service involved. For
example, in a system employed for voice
communication, where loss of natural-
ness of the speaker's voice is not a
factor, the relative audio levels might
be maintained as high as possible with-
out loss of intelligence. Another cri-
terion might be the transmission of
voice and music a t the same loudness.
On this basis, using the standard volume
indicator, voice would be peaked 2 or 3
db below music.l Still another possible
basis for determining relative trans-
mission level is the appearance of aural
distortion caused by overloading of the
particular facility involved. Determina-
tions of this kind are usually made on
an A-B test basis1 and the results will
depend to a large extent upon the shape
of the overloading characteristic of the
system involved. Transmission practices
based upon this criterion ensure the
maximum use of the facility concerned
while avoiding detectable aural distor-
tion. For this reassn it has great appeal,
from a purely technical standpoint, for
high-fidelity broadcasting and sound
recording applications. However, it does
not take into consideration the listener's
preferences.
P r o g r a m transmission practices,
where listening for pleasure is con-
cerned, may well be determined on the
basis of the average listener's wishes.
A study7. made with this criterion de-
Chinn, Gannett and Morris; PTOC.
I.R.E., Vol. 28, No. 1, p. 9, Jan. 1940.
1Loc. cit, pg. 4.
7 Chinn and Eisenberg, Proc. I.R.E., Vol.
35, No. 12, p. 1547 (Dec. 1947).

AUDIO ENGINEERING OCTOBER, 1951