Reading VU Meters (Sept. 1976)

Home | Audio Magazine | Stereo Review magazine | Good Sound | Troubleshooting

by C. E. Motile

The VU meter was introduced in 1939 as a standard meter to avoid the confusion that existed in the broad casting industry at that time. Audio level meters up 'til then had a bewildering array of reference levels, various speeds of movement, slow, medium or fast, and some people even chose to ignore the fact that they were dealing with a non-symmetrical wave shape. The meters still do vary on cheap equipment where they were meant to solve these problems. How ever, they can do very well, on the technical specifications'. But now there seems to be more confusion, because of the different ways of reading the meter and also expressing what is read. The problem has recently become more acute, now that the use of the meter is not confined to broadcasting, but has found ready acceptance in sound recording, both for commercial and private uses. Even a novice to the hi-fi market looks for a VU meter on each channel of his new equipment.

But why have a VU meter? Perhaps its need should be justified before de tailing the reasons for confused reading. The use of one piece of equipment confined to one person would seem to be the occasion for least use of a level indicator-for the low level passages, listening to a speaker will let you know when the level is getting too low and into noise, and the speaker can still be used to get a rough assessment of "when it's too loud, it distorts." Even so, with the best of equipment, where no compression of volume is required, a starting point for setting the volume control is difficult without the aid of a level meter. Then the meter has a further use, to enable a repeat of the same maximum volume level which has been found to be satisfactory before-the hearing of a speaker won't be accurate enough after only an hour or so. The advantages are more obvious when a change of personnel is involved, or when there is a change of tapes or programs.

A Standard Meter

Not the least of the claims for the use of a VU meter, compared with other level indicators, is that its dynamic characteristics are rigidly specified and controlled, such that the speed of movement happens to correspond very closely with the effect of music and speech on the ear, i.e. the loudness effect. Combined with the background color on the scale, the meter is easy to read and yet easy on the eyes. The same cannot be said for some "peak" reading meters; they need extra equipment to drive them, they look odd with their wild upswing and slow return and still never quite make the peaks.

Let's have the "loudness" type meter as long as provision is made for the peaks that are there above the VU reading. Most VU meter users allow 8 or 10 dB above the +4 VU or +BVU, or whatever standard they use, but peaks of 14 dB above the meter reading have been observed on unusual pro gram material. Also it is rare to find an operator that can aim and maintain the pointer deflection just up to the "0" mark most of the time, without getting, say, 2dB over. If 2dB is allowed for the human error and 10 dB for the peaks, then 12 dB seems to be the appropriate allowance to make for overload and distortion tests, e.g. +20 dBm for a +8 VU circuit. This could be the first of the confused ways of reading a VU meter. How many tape recorders are tested for distortion, with a continuous tone only deflecting the meter to the "0" mark? The VU meter will stand 14 dB above this mark continuously without any ill effect, so should the rest of the equipment. Why not specify and check it at the right level (not 5 dB above,')? Even if the VU meter had no other advantages, it would still be worth using, just because it is standard, and so valid comparisons can be made be tween various pieces of equipment and between different locations in a network.


It was the network folk who probably helped to conceive the idea of a standard meter with rigid specifications', but the advice on how to read was then slanted towards network operators, with observations being made at a transmitter, repeater or a switching center. This was good at the time because each location was staffed and was presumed to have a complete VU meter, i.e. a meter with an attenuator in 2 dB steps, with the recommendation that the attenuator be adjusted until the majority of peaks reached within + or-1 dB of the "0" mark on the meter. Then the measurement was expressed as the sum of the meter reading and the attenuator-easy, if the meter deflects to the "0" mark, just read off the attenuator setting and there it is, so many VU.

But the position has changed, there are fewer transmitter, repeater and switching locations that are staffed, and at the places where programs originate, and certainly on most tape recorders, there is no attenuator visible, and no marks on the meter to say +4, +8, +10, or whatever standard is used. Furthermore, the instructions as to where the needle should deflect are either too brief, such as "up to zero," or too ponderous, "the reading is determined by the greatest de flections occurring in a period of about a minute for program waves, or a shorter period (e.g. 5 to 10 seconds) for message telephone speech waves, excluding not more than one or two occasional deflections of unusual amplitude."[5] There appears to be 4 different interpretations given to these instructions:

1. Keep the needle above the "0" mark, most of the time.

2. Never let the needle get above the "0" mark.

3. Look for some mythical average of the deflections and adjust until this average is at the "0" mark.

4. There must be a correct way. "In average program material (speech and light music), peaks occurring at the rate of 6 per minute are regarded as the most significant, and should deflect to the zero mark, allowing for an occasional over-swing; classical music will need to be observed over a longer period."

Perhaps this last interpretation is more significant when it is seen how some people have twisted it around, and said that you should see a peak every 10 seconds. This is quite wrong, because music and speech are not performed with such regularity; it would be terribly monotonous if it were so. But it does raise the importance of elevating the monitoring speakers or phones to equal status with the VU meter-how can you make a valid observation without ears to tell you what type of program material is present? Even without a knowledge of music, most people can tell from the way the music is being played whether it should be loud or soft.

Using a VU Meter

The instructions should be simplified further and included with each piece of equipment using a VU meter.

The advice to novices, just starting with their first tape or cassette recorder could be: get some prerecorded material, either news type speech or light music with a rhythm, i.e. some thing with only small variations in volume and, while it is playing, adjust the volume control until it is easily seen that the majority of peak readings come to the same point on the scale, adjust the volume control again until these peaks reach the 100% mark, then you have correct level. A little more time is needed to get correct level on classical music or unusual program material, but what to call it? There is neither the time or the need to call it anything in some situations. If the concern is only with your own tapes, then it needs only to be called "correct level," and the figures on the meter are superfluous.

The manufacturers of the equipment may tell you in the instruction book that "0" on the meter is +4 VU, +8 VU, or whatever they choose, but they have also allowed a margin of 8 dB to 14 dB above the VU meter reading for peaks. The "medium-fi" market is probably best served by what some of the equipment is now using; i.e. a meter with a black arc up to about 70% of the scale, then the arc continues with red-no figures or letters to confuse the user; just the direction that most of the peaks should come up to the division between black and red.

Something better is needed for the professional and broadcast equipment, but not the present "A" scale with its confusing -20 to +3 VU scale.

Certainly this scale has its merit when used in a system that requires line-up tone tests, but for the majority of users, concerned with the origin of a program, and with no meter attenuator in sight, perhaps the best solution would be to have the meter marked +4 or +8 at the present "0" mark. The marking chosen would depend on the reference being used, and then as the meter is only going to be read between the present -1 and +1 VU marks, all that is needed in addition is +3 and +5 markings for +4 VU equipment. Similarly, for +8 VU equipment, the scale would read +7, +8, +9 VU under the arc, and on top, retain the 0 to 100 to emphasize the idea of 100% utilization of the equipment. This could be called a "C" scale to distinguish it from the present "A" and "B" scales. Now the operators and technicians could abandon such phrases as "it's zero on plus eight," "ifs zero level" (this was supposed to be buried long ago) or "it's peaking at minus five." This last expression is the most confusing of all: is it -5 VU, -1 VU, +3 VU or -5 on the scale or what? Wouldn't it be better to be able to say "this program is +4 VU (or whatever the standard is in use)? No more is required, but it would imply three things:

1. An approved VU meter, correctly calibrated, is being used.

2. Program material is being ob served, not tone or peaks.

3. The progress has been observed for a long enough time to make a re liable statement.

As starters for a new scale meter, how about only 3 markings for VU and the 0 to 100% in 20% steps on top? For those interested in a composite program of speech and music, the percentage scale is very useful for setting the level for the less important parts of the program, e.g. themes to introduce a speech session or bridging music between two scenes in a play. Most broadcasting organizations look for this refinement in the presentation aspect of operating, and issue instructions worded something like "keep the less important portions of the program down to -5 on the scale, or down 4 dB or down to 60%. The % mark is easier to see and is another reason for retaining the 0 to 100% markings.

It is worth noting that most motion pictures produced over the last 15 years appear to be using this refinement, combined with the use of VU meters. The effect on the ears is certainly better than what is still evident on older films, where peak type meters were used and the music intervals always sounded 5 to 10 dB louder than the speech. The intelligent use of a VU meter can assist in making a very well presented program.


1. "A standard volume indicator and reference level," Communications, April, 1939.

2. "A new standard volume indicator and reference level," Chinn, Gannet & Morris, Proceedings I.R.E., Vol. 28, No. 1, January, 1940.

3. Review of cassette recorder, Audio, April, 1974, page 58.

4. "The measurement of audio volume," H.A. Chinn, Audio, Sept. & Oct., 1951.

5. Radiotron Designer's Handbook, Langford-Smith, Fourth Edition, Page 824.

6. "Uses and Abuses of the VU meter," Oliver Berliner, Audio, November, 1955.

(Source: Audio magazine; Sept. 1976)

Also see: How We Hear (May 1977)

= = = =

Prev. | Next

Top of Page    Home

Updated: Tuesday, 2018-04-10 14:50 PST