Audio editing

Learning objectives

When you have completed this session, you should be able to

• define term "digital audio",
• dit sound files,
• create sound files.

Digital Audio and Video Technology

Digital video and audio technology helps archiving our records on CDs [ROSSZ LINK!!!] or DVDs. We can publish them on the Internet and we can even create special effects easily that would be very difficult and expensive with analogue equipment.

Due to small broadband the amount of transmitted audio and video files via Internet today is insignificant compared to the amount of pictures. Downloading some better quality video with an average modem at home may even take several hours.

The situation is similar with audio files although some compression methods can result relatively small size files.

The most important concept in case of movies and audio files to compress the size of the files as much as possible. Different algorithms have been developed for this application which will be discussed in the next chapters.

The editing of audio and video files is done by suitable software. The most known video editor is Premier from the Adobe Company. Well known audio programs are Sonic, Soundforge or Audition (earlier called CoolEdit) from the Adobe Company.

Digital Audio Technology

Sound is small amplitude mechanical oscillation expanding usually in air. Sound can be converted into electric signals with the help of microphones and this way it can be processed, transmitted, and stored electronically.

Digitizing

Learning objectives

When you have completed this session, you should be able to

• grab audio CDs.

Digitizing Analog Sound Frequency Signals

Analog sound frequency signal can take any value of amplitude which may also vary constantly in time interval.

For digital sound storing in time units samples must be taken from the sound (time quantization). When amplitude samples, produced this way, are divided into determined number of units the size of the amplitude sample can be represented by a number (amplitude quantization). Thereafter the numbers received can be stored in digital format (coding).

The frequency of sample taking is determined by Shannon’s theorem. According to this the original analogue signal can only be restored from a sequence of impulse without distortion if the frequency of sampling is at least double of the highest frequency occurring in the original analogue signal.

Since the audible frequency range for man is between 20 Hz and 20 kHz according to Shannon’s theorem at least 40 kHz sample taking frequency is needed to produce perfect audio playback.

Higher than 20 kHz frequencies certainly must be filtered with a low pass filter.

The steps of digitizing are the following:

1. low pass filtering (LPF)

2. time quantization

3. amplitude quantization

4. coding

The equipment which executes these steps called analog-digital converter (ADC, A/D converter).

The equipment which converts digital signal to analog signal is called digital-analog converter (DAC, D/A converter).

Sound cards installed in computers usually contain both equipments.

For computer sound processing the parameters of Audio CD quantization is the standard which are the following:

• 44100 Hz sample taking frequency

• decomposition of 16 bit amplitude quantization ( the samples of amplitudes are decomposed to 65536 parts)

Certainly other quantization parameters can also be used. For studio technology 48 kHz sample taking and 20 or 24 bit decomposition is accepted. 8000, 11025, 22050 and 32000 Hz sample taking and 8 bit decomposition is used for lower requirements.

Sound is usually saved in Microsoft Wave (WAV) format with PCM coding. (PCM = Pulse Code Modulation) In case of 44,1 kHz, 16 bit quantization for saving one second stereo audio material 2 x 44100 x 16 bit = 176400 Byte is needed. This is more than 10 MByte per minutes.

For effective storing various compressing methods have been developed. The most important ones are described here.

Compression methods

Learning objectives

When you have completed this session, you should be able to

• differentiate between compression methods.

Compression methods

ADPCM (adaptive differential impulse coding modulation)

The quintessence of this method is that it is not storing the actual value but the difference between the samples taken subsequently. This method works with only 4 bit samples rendering a scale for it and this way it is able to reproduce the wave bends relatively precisely.

A-Law and u-Law

Standard speech compressing algorithms. Apart from PCM amplitude quantization levels are not defined linearly but in a logarithmic way. A-Law is used in Europe; u-Law is used in America and Japan.

VOC

A method developed by Creative Labs. It uses ADPCM compression but removes silent periods like spaces between words and sentences.

MPEG-1 Audio Layer

Originally it was a part of MPEG video compression method. The standard contains three procedures Layer 1, Layer 2, and Layer 3. (Abbreviated as MP1, MP2 and MP3.) From Layer 1 to Layer 3 simultaneously with the efficiency the complexity is also improved. The three procedures are compatible downward with one another.

In case of 128 kbit/s transmission speed MP3 compression hardly any difference can be heard compared to the original, non-compressed signal. (This means a 111 ratio compression.)

MPEG compression has become widely used due to its very high efficiency and excellent quality. MPEG-1 Audio Layer 3 standard was developed for saving music; MPEG-1 Audio Layer 1 was developed for DCC (Digital Compact Cassette). Digital radio (DAB, Digital Audio Broadcasting) uses MPEG-1 Audio Layer 2 standard.

AC3

A multi-channel sound compression method developed by Sonic Foundry. Supports Dolby Digital 5.1 standard therefore it is an ideal sound compression method for home video systems.

Video Tutorials

Exercises

1. Rip (grab) an audio CD with the help of Windows Media Player! Watch the tutorial.

2. Download the free sound editor software Audacity watch the tutorial and try to record your own voice!