Donnerstag, 10. Dezember 2015

Linear and non linear audio distortion


Essay from the book "Mastering" (translated by google)

Stefan Noltemeyer: www.mastering-online.com

Distortions

All components of a sound studio can be divided into four groups:
1. transducers (microphones, speakers)
2. amplifiers (preamplifiers, power amplifiers, impedance converter, equalizers, compressors, limiters, effects units)
3. Memory (analog and digital tape machines, computer hard discs)
4. Cable

Excepting the digital storage, all these components produce distortion in varying levels.
Distortions inevitably arise in any type of audi transmission.
Distortion in the audio technology means , the modification or manipulation of an original signal. Distortion may intentionally or unintentionally caused.

One distinguishes between linear and non-linear distortions.

Linear distortions

In linear distortions can turn different delay distortion and amplitude distortion. In contrast to non-linear distortions incurred in linear distortions no additional frequencies or harmonics.

Runtime distortions in each transmission chain. A signal always comes later to the receiver, as it has been sent. If a full-range equally offset in time, the problem is mostly irrelevant. Exceptions are the latency problems in working with computers. The processing time required by a DSP (Digital Signal Processor), audio data can be processed depending on the processing power and quality of the audio card to several hundred milliseconds, respectively. When a musician with his instrument or recording his voice "by the computer" listen, he will feel the delay as disturbing.
In the analog audio technology problems arise when several signals with different delays are switched together.
Delay distortion in which the delay is shorter than the period of a single vibration, called phase distortion or phase shift.

If different frequency bands vary greatly delayed in a transmission chain, may lead to audible effects.
I distinguish two different groups of perpetrators.
Firstly produce individual components in the electronics delay distortions, on the other hand caused delay distortions under certain conditions in acoustics.
In an amplifier stage, there are different times for high and low frequencies by specifications of individual components. A well-known example is the so-called RC member. It consists of a combination of a capacitor and an electrical resistance. Such a RC member, for example, the main component of an equalizer. It shifts the phase angle from high to low tones within an audio signal. Audible this is only in extreme cases. Relevant it is when adding up many small runtime shifts. Details are described in the chapter "Equalizer".

On closer examination of the acoustic aspects of audio engineering, one often encounters phenomena that cause delay distortions.
Phase cancellations make disturbing especially with low- frequency signals.
In a graphic example, we have two sine waves with a frequency of 50 Hz and amplitude equal to 10 milliseconds large relative to one another. This corresponds exactly to a 180 degree phase shift. The vibration is placed in the left channel of the stereo image, the other right. If you switch now the signal to mono, the two vibrations are added to a sum. At the crest of a sine wave is a wave of the other sine wave added. In this case, delete them both completely made Dis sum is zero. Silence !

Since there is practically no music in the pure sine waves, and as a phase shift of exactly 180 degrees will never happen, this is a theoretical example. At a wavelength of several meters for low frequencies (6,80m at 50 Hz) can very well keep in mind that the phenomenon of phase cancellation blatantly changes if, for example, one meter further away from the sound source.

When you record a sound source with two or more microphones,
it comes to phase cancellation, since the sound requires different maturities to the different microphones. It caused interference when you switch the signals from the microphones together. Clear up the sound waves, is known as destructive interference, add up the sound waves it is called constructive interference. In both cases there is a distortion of the original sound. In acoustics it is called comb filter effects.

The frequency spectrum of an instrument e.g. an acoustic guitar or a wing, which is picked up by two microphones, by such a comb filter effect is partly strongly distorted. The decisive factor is each other the volume balance of both microphones.
Are they identical, the effect is strongest.
Is a microphone, for example, one meter further away from a sound source, as a
otherwise, the result is a delay difference of 0.029 sec.

Almost classical problem is during the recording of a piano or a grand piano. The problem arises as follows: It is common that a microphone for the bass and one is used for the treble. Now since both microphones not only the area for which they are intended, to record, but the full range of the wing, there is phase shifts due to different maturities of the two signals. Since the sound is very slow, small run-time differences become a problem already. To illustrate this, let us imagine a single bass string of a wing. If she was injured, the resulting sound from two microphones is recorded. Now these two microphones have a different distance from the sounding string. Suppose the microphone, which is actually intended for the recording of the treble, located one meter further away from the sounding string, as the microphone for recording the bass. The sound needs a 1: 340 th of a second longer to treble microphone. In mixing the signals from both microphones now come together. In the nature of the sound, both are almost identical, after all, they have taken up the same sound, but they are time-shifted slightly against each other. It can be assumed that it now provides a panorama of the microphone to the left and the other to the right. Thus, the problem is actually eliminated. Now if a listener is exactly in the middle of the speakers, which reproduce the piano sound, it will at least be able to hear in the deep tones that there are phase cancellation. If you switch now this stereo signal to mono, the two slightly offset signals add together. Now it comes to cancellations of individual frequencies. Compared with a recording of the same instrument with only a microphone, there will be a lesser amount of bass. How serious are the cancellation depends on the ratio of the volume of both signals. The greater the difference in the volume, the less is the problem. If both signals are approximately identical, arising comb filter effects. The human ear will perceive such a comb filter effect than a changed tone.
It is known that the human ear for delay distortions or phase distortion is much less sensitive than, for example for amplitude distortions. But the ear may very well perceive the resulting errors. You identify it only takes a little practice.

To learn how a phase cancellation effect in a complex signal, one can tentatively change the polarity on one of his speakers. The membrane of a loudspeaker vibrates at a resting point. Now, if the polarity is reversed a box that has the result that the membrane of a box goes to the outside at a signal and at the other box inwards. This has the consequence that a box generates a sound-pressure and the other box, a sound negative pressure.
This can be heard. One gets the impression as if the bass reproduction to be considerably weaker.



In 1988 I began in Frankfurt in a recording studio as an assistant, coffee maker Pizzaholer and studio technicians ,. This studio was set up for "audio file" Sound Recording. It had a big recording room with a grand piano and a control room with incorporated premeasured Studio Acoustic (live end / dead end). We have there recorded and produced many jazz and rock bands. In receiving free times I was the studio often for its own sessions and sound-engineering experiments. Many of my knowledge and experience I have acquired to me there. When shooting our wing I have tried many variations on the one hand the best possible way to capture and secondly to minimize the resulting phase shifts through the "Polymikrofonie" the sound of the instrument. By the way, that has my ability to perceive even small phase shifts, trained.
Apart from the reduced bass playback leaves a phase shift an indefinable, skuriles feeling in the head. A friend described it as follows: It feels like someone driving with a sponge over your scalp.
Recently, I have a plug-in of the company Universal Audio named "Little Labs" discovered. It offers the possibility of the phase of a signal of 0 degrees to 180 degrees to rotate continuously, and it can delay the signal continuously from 0 to 4 milliseconds. With this tool, you can compensate for the phase problems of a stereo signal effectively. As a control, I take a phase correlator for help. The less left of center moves the indicator, the optimal will be the result.


Possible technical causes of phase cancellation:
Often I have XLR cable soldered; An XLR connector has three numbered with numbers Anschüsse, No. 1 is the connection for the shield, No. 2 for the positive signal wire, No. 3 for the negative signal wire. When I 2 and 3 Swap connection to one of the two male ends, I create in conjunction with a second (properly connected cables) a 180 degree phase cancellation.
Modulation effects such as chorus produce phase shifts that can cause cancellations of low frequencies. Therefore, the chorus on a bass is critical.

Amplitude distortion

Amplitude distortion called all changes an original signal, wherein the amplitudes of individual frequencies are amplified or attenuated. Are, for example, raised the high frequencies, the sound will be more brilliant than the original. If the low frequencies increased, you will feel the sound "more powerful" than the original sound.
With an equalizer amplitude distortions are deliberately brought about in order to make a sound in a certain way.
In the German translation the word Equalizer "leveler" means.
Originally an equalizer was used in the broadcast technology to compensate for unwanted amplitude distortions of a transmission chain. It should therefore be "razed" by means of an equalizer, the transmitted sound to the original.
There are a number of polluters of amplitude distortion.
Each cable amplitude distortions produced by the electrical resistance of the copper wire. If all frequencies equally reduced, you will be able to perceive the change only in the volume. Will some frequencies more reduced than others, one perceives a change of tone. The problem is known in a guitar cable. Since the output of an electric guitar is high-impedance, the cable acts as a filter that reduces high frequencies.

Amplitude distortions also arise because an amplifier not all frequencies of the audio spectrum accurately amplifies the same. Especially at very low and very high frequencies that is relevant. A quality characteristic of amplifier is the frequency response (20-20,000 Hz for example)
This means that the amplifier is capable of all frequencies between 20 and 20,000 Hz with (almost) the same volume can be transferred.
Relatively strong arise unwanted amplitude distortion in a loudspeaker or a loudspeaker. In most chains About Conference speakers are the "weakest" element, because it causes the most massive linear distortions.

Nonlinear Distortion
Includes a sound non-linear distortions, the original have been added intentionally or unintentionally overtones. Most is integral multiples of the fundamental frequency. The ratio of the root of the unwanted overtones is called distortion. The louder the harmonics relative to the fundamental, the higher is the THD. These distortions are caused by the non-linear characteristics in the active components (transistors, ICs, amplifier tubes) of an amplifier. So there are the non-linear distortions, which are referred to in common parlance as distortions. Due to the high quality of today's electronic components, one can assume that you have this kind of distortion only perceives when overloaded an amplifier massively. The resulting congestion in the heat also plays a role, because the temperature often influence on the characteristic of an active component.

The most famous use of non-linear distortions, there is in the world of guitarists. What is usually unintentionally, is looking for a guitarist. Originally, the distortion created only if you hired a guitar amplifier for maximum gain and thus overburdened. The guitarist enjoys not only the overtones newfound, but he also uses the fact that the Sustain extended by the distortion, or in other words, that the sound is longer before he stopped. Since this effect is desired to produce only in connection with a brutal volume, the industry has been built since the 60s devices that generate at least this or a similar distortion at low volume. They call themselves Fuzz, Tube Screamer, overdrive or distortion.