Articles Comments

KARAGYOZOV.COM » Sound theory » Methods of measurement of the sound. Measurement scales

Methods of measurement of the sound. Measurement scales

Methods of measurement of the sound. Measurement scales



We can say that nowadays few employees in a production and media departments are aware of what they measure, how they measure, in which indicators and standards. Regularly there are questions such as:


  • Why in some scales there are plus values, and??in others – not?

• What metering device indicates what?

• Why the same material placed on different measuring equipment shows different values??

  • Why a sound material brought to a certain level returns from next instance (studio, media house, media) with a resolution of nonstandard levels?


Generally by level indicators of any type we acquire visualization of the signal level tha

Screen Shot 2014-04-02 at 8.20.44 AM

t we are registering,  respectively broadcasting or magnifying.

We need it, at first, to meet the technical requirements in terms of maximum and minimum levels permitted in the sound tract and on the other to help our auditory sensation in the balance of different audio elements in a program.


Historically, the measuring equipment, which has established itself in the wake of sound recording and Broadcasting is a standard voltmeter that actually is to measure the electrical signal level. That, given in decibels acoustic energy shows how strong would be the sound:

VU-meter concept comes from volume units. It was used from BELL Laboratories and subsequently by many other company to measure since the time of the telephone signals.

The typical VU scale is from -20 to +3 dB, often with LED indicating the transition above the maximum value.

What is the ratio of the zero reading of the device and the voltage? That is, if our device shows 0dB, what voltage will we receive if we measure the output of our recorder, mixer or other audio device?


0db usually correspond to 0.775 V, +6 dB to 1.55 V, +4 to 1.23V, and 0.25V at -10


It is good to know though that there is a difference in the voltages, respectively in the level of signal to the various professional and semi-professional devices. Adopted in most cases voltages are +4 dBu, while in the domestic and semi- pro equipment there is a standard of 10dBv, where the level, which is measured as zero, is significantly weaker as voltage.

This means that if we connect two devices, one of which is +4, and another at -10, the zero indication of the output of the one will be not a zero level at the input of the other. This often leads to overloading and distortion of the input in certain situations or in the opposite direction – a very weak signal. Therefore it is better to check the standard voltage of devices that will connect. Some of them have a switch of +4 / -10 on the back near to the analog input-output connectors, which is good to be used as intended, while manipulating the input and output voltage of the device to that of the whole system. So the zero of all devices will be the same as a level of the signal.

Back to the volt-meter.


Screen Shot 2014-04-02 at 8.21.17 AM

Fundamental mistake of many people is that they believe that under the VU-meter is understood arrow indicator. Nothing of the kind, in historical plan it is so, but it can, and many times look like this:

(in this case by LEDs),


which is the same with respect to the measuring scale, but also in terms of something very important, which characterizes it as a VU-meter, namely:

How exactly it measure the sound.


This indicator is relatively slow, and it is not responding to peak levels, unless they continue for a a longer period of time. It detects average density of the sound – ie. the RMS (Root Mean Square) level.


This is very useful in monitoring the sound texture because it is similar to the way human hearing averaged the sound levels and the perceived volume. Standard for VU-meter is considered that on which he needs 300 ms to reach the required level, after receiving the relevant voltage and 300 ms, to return after an interruption of submission of the voltage, ie it is his, figuratively said inertness.


This leads us to two very important parameters characterizing the testimony of a meter – the time to reach setpoint value (or a certain percentage of it) – ie. integration time, and the time of return (or reducing with a certain decibel ratio) – ie. fallback time, decay time or return time. This settings in the  VU meter are about 300 ms in both directions, making it a good tool for measuring average levels of sound.

How such a device would show peaks in the audio material? The answer is – very imprecise and approximate. In most cases, he never would reach the instantaneous peak values ??that are real, because they are shorter than the integration time of the VU-meter.


Because the VU meter misses many peaks in situations of need for monitoring of peak signal in broadcasting or recording, we need another type of meter – PPM – Peak Programme meter. It is tuned for fast integration time – from zero to a few tens of milliseconds, ensuring proper accounting of the peaks in the signal.

Where necessary, PPM is irreplaceable – for records, especially digital, where each excess of the the maximum levels would lead to distortion and degradation of sound in the media, where compliance with the peak levels is essential for unproblematic flow of the process, etc.

To allow time for the eye to account for such a rapid peaks, as the PPM-meter, it needs a significantly slower return time – in one of it’s modifications the indication is reduced by 20 dB for 1.7 seconds, and in the other – by 24 dB for 2.8 sec .



   In other words the PPM indicator is a quick indicator to measure the peak values with a relatively slow decay, and VU meter is a slow indicator to take account of the average level of the density of sound.


How PPM indicator shows an average level? The answer is – Absolutely inadequate. Due to the response to the peak values is obtained the situation where  the phonograms with quick and sharp peaks are accounted by the PPM meter such as much stronger than those with constant and equal density, while in the same moment the ear reported opposite.

In other words in terms of the density of sound the Peak meter is lying to us, but in terms of the peaks – the VU meter.

This is especially important nowadays when most colleagues working on a limited budget in media and small project studios rely primarily on software ultrafast peak meters. They actually sacrifice their proper reporting of RMS density and after the actual peaks only, something which is very wrong when preparing the music program in the media or producing a record in the studo

What else should we know about Peak meter? First, what is its scale? In VU-meter it is from -20 to +3 dB. But here?

There are different scales for measuring peak signal.

   First, because of the large variations in the level when we measure the peaks, the scale must be wider as at the bottom, and in its upper part.

   Second, geographically and historically different scales and methods of measurement were formed :


Screen Shot 2014-04-02 at 8.21.32 AM


The DIN scale – the values are vi
sible-from-between 50 +5 dB-integration time – 5 to 10 ms;




Screen Shot 2014-04-02 at 8.21.37 AM


Nordic scale – more “flattened” then the DIN scale, a slightly narrower range, the same time constants;




Screen Shot 2014-04-02 at 8.21.42 AM


BBC scale – as usual, with Brits everything is not like the others – the levels are measured in values ??without dimension – from 0 to 7, as the 4 corresponds to 0dB, the maximum level is 6 and 7 is the overload. Integration and return time are slower;




Screen Shot 2014-04-02 at 8.21.48 AM


EBU scale – scale to measure when exchanging international programs, the same scale like the English, with the same speed of reaction, but  with decibel values.



The point here is this:


Zero dB in itself is something quite relative and depends on the way of measurement. In this case the important thing is:


  • First, what is the decibel value as a physical value compared to the levels, which in different scales are set as benchmark – the level of “alignment” and the level of maximum permissible value. Only if a person wiith a comparative table showing what corresponds to as the physical decibel level to the level of the corresponding scale can understand where he is in the scale as a real volume. Only through it he can also “translate” decibel values ??from one scale to another, ie 0.775 V or 1.55V how many decibels at the appropriate scale meet. I will add to this table;
  • Second, the integration time of different PPM meters. Let’s talk about this.


We talked that there is a slow measure of the relative density of the sound by VU-meter and quick – with rapid response to attacks – by PPM or Peak meter. Yes, but the very fact that the reaction of the PPM meter to peak is not instantaneous, but with a delay, goes to show us that it has some inertness, which in turn means what?

Well, very clearly – he will miss “around his ears” some peaks. A measuring instrument according to DIN scale has 5 ms integration time, which means that for 5 ms, it reaches 80% of the actual value of the peak, which should be displayed. In 10 ms – 90%. In less than 5 ms – well below the true value.


In other words this type of PPM meters are inertial just like the VU-meter, just a lot faster, which does not mean instant. For this reason they are often called QPPM – Quasi Peak Programme meter.

A type of indicator without inertia as the software indicators, which is showing the real values ??is called TPPM – True PPM.

Its integration time = 0.

Actually in most software applications the measurement is not completely accurate because they can scan only the levels of individual samples, and, the digital system “can not understand” what happens between them. This actually  is what most digital level indicators built into audio software are showing. This is called SPPM-Sample PPM.

To become SPPM to a TPPM – ie to display the correct values, we apply an oversampling – lifting in four times the sampling frequency of the wave to reconstruct the system peak and to take them into account. Thus we have and Over-Sampling PPM

Why we pay so much attention to these tiny differences in the measurement of the high levels that have to talk about QPPM, TPPM, SPPM and called? What has changed that we have the need for such an accurate measurement of the peak levels without any deviation?


The digital sound


So far, speaking of the scales of measuring, we never touched a digital one. All previous have both minus and pluses values, which means that the peaks in the sound image can move freely within a certain range on the red portion of the scale. This is because the music is not a sine wave with a specified height and strength, it is changed dynamically, respectively the peaks fall at different rates.

Screen Shot 2014-04-02 at 11.43.57 AM


Numerical scale is not the same. It can only reach up to the highest quantization level – 0 dB and no more part of decibel. Accordingly the scale in a 16-bit sound starts from the lowest quantization level reported – – 93, -96 dB and reaches to 0. This is the numerical scale dBfs – DeciBel full scale. In the measurement of  the levels a standardized scale is from -60 to 0 dB – Digital Scale:


As can be seen, it is also logarithmic, such as others.




Let me clarify some things…




In the analog audio the measurement of signal chain is done by the readings of VU-meter with 0dB RMS level. This is considered a so-called. alignment level – control level, a kind of point of reference of sound.

The maximum sound levels reach the +9 dB, ie 9 dB above the level alignment. This level is defined as the permitted maximum level (PML).

Along with this, there is a so called. headroom – a kind of safe zone, covering an additional 9 dB above the maximum level. This is an area in which it is considered that if they are occasional peaks, they will be reproduced with a small distortion within the limits. Thus, the analog signal is provided the possibility of exceeding the levels above the permitted without consequences. How does this correspond with the digital scale?


As here any level above zero is equivalent to distortion here these safe zones are below zero, respectively the zero goes down. For alignment level is taken-18dB, and for the maximum – permitted maximum level – 9dB. Levels above -9 to 0 are reserved for headroom – occasional peaks, which are going above the -9 dB.


Here is a graph showing the relationship between the main analogue scales and digital scale dBfs:


Screen Shot 2014-04-02 at 1.34.04 PM


How to translate this? What is seen as a test or measurement level – alignment level – according to DIN scale is -9, and on the digital scale it corresponds to-18dB. On BBC standard it is 4. Other words, if we use standard analog peak meter, showing -50 to +5 dB in DIN, the volume which will display, will be -9, but the indicator of a professional digital recorder or a sound card will be -18 dB!


We see that the same decibel level is not reported as the same on the analog and on the digital scale!


For this reason, the recommendations of the EBU is present permitted maximum level of -9 dB, such as the area above this is generally undesirable. When submitting materials in digital format and measured at dBfs scale, these recommendations must be followed in order to align the levels between studios, production and media houses.


What happens in practice and what is happening in the sound recording and music postproduction?

In principle in the field of modern popular music headroom levels above the maximum level virtually don’t exist. Why?

Because the use of brickwall limiters and highly dynamic compression on musical material lead to that all the peaks are aligned digitally by not allowing some of them are stronger or weaker (waveform style “English Meadow”). In this situation from  the usually called “headroom” nothing exist, and the PML is reduced to -0.1 dB….

Screen Shot 2014-04-02 at 1.55.24 PM



Leveling and raising all the maximum levels with the use of digital limiter and maximizer.


The result is a very strong sound as the average levels are rising once with 9 dB due to the lack of headroom, and a second time by a few dB due to the compression, which is preceding the limiting.

While listening to such phonogram, it is much stronger than an analog musical material. Why? Because people (and unfortunately many colleagues) think that 0 dB on the analogue scale are equal to 0 dB in the digital, and as we have seen, this is not so.


Filed under: Sound theory

Leave a Reply