Travelling with Sound
This is a serie of 4 articles that will cover in deep sound in a PC.
Part One : The basics and theory of sound
Part Two : Actual implementation of the Sound Processing at a Hardware level
Part Three : The various standards in Sound ( Cables, Standards, Hi Definition )
Part Four : Various topics and issues about PC Sound ( EAX, HDCP, PAP, Software issues )
Part One : The basics and theory of Sound
The signal path :
The path the sound uses through the process can be thought off in three parts :
1. Source input ( and eventually Analogic to Digital Conversion )
Soundwawes have to be of numeric form to exist in a PC world. There is two cases then :
. Either the source is numeric ( a CD for an exemple ), then the PC will directly process this
. Either the source is analogic ( a microphone or an analog line in plugged into the PC), then
the PC will have to digitalize it ( to make it into a numeric format ) : this is named ADC ( Analog to Digital Conversion ). Then, the PC will process this digital content.
2. Signal processing
The numeric signal will be processed in different ways by the PC ( decodig logics, formats conversion, mixing, special effects ). Exemples of decoding logics are Dolby DIgital 5.1 or EAX.
3. Source output and Digital to Analogic conversion
After this numeric content has been travelling and processed in the PC, it will have to be made into an analog format, before being amplified and sent to a speaker. This is named DAC ( Digital to Analogic conversion ). There are two cases here too :
. Either the PC makes the Digital to Analogic Conversion ( DAC ) itself, and the analogic signal is sent out to a sound amplifier, using an analog cable plugged into the PC. This will use the analog outputs of the PC
. Either the PC sends the numeric signal directly out of the PC to an amplifier that has numeric inputs using a numeric cable plugged into the PC. The amplifier will make the Digital to Analogic Conversion ( DAC ). This will use the numeric outputs of the PC : Coaxial SP/DIF, Optical SP/DIF, HDMI.
This final Analogic signal is then amplified, and sent to the speakers. This is the whole signal path.
Sound caracteristics :
Digital sound has a number of characteristics that are important to understand clearly.
They are bit depth, sampling rate, and encoding.
Analogic sound, as well as ADC and DAC conversion, have characteristics that are important to understand too : SNR, Frequency response, Dynamic range, THD, signal level.
1. Digital Sound characteristics
This is the picture of a basic digital sound :
( image from wikipedia )
The red line is the digital signal. The grey wave is the Analogic signal it corresponds to.
The more vertical subdivisions there are on the vertical axis, the more accurate the digital signal will stay to the Analogic signal : this is Bit Depth
The more horizontal subdivisions there are on the horizontal axis, the more accurate the digital signal will stay to the Analogic signal : this is the Sampling Rate.
The higher these two values, the more precise and accurate the sound will be.
For an example, a CD is 16 bits ( bit depth ) and 44,1 kHz ( sampling rate ) sound.
An analogy to Video signal quality would be bit depth = number of available different colors and sampling rate = number of FPS ( Frames per Second, ie Images per Second ).
A third caracteristic of a digital signal is its encoding. While the above exemple is a simple digital signal ( that is named PCM or LPCM ), several digital signals can be compressed and/or coded into a specific bundle : MP3 ( compression ), Dolby Digital ( Multichannel +
compression, etc ... )
Bit Depht and Sampling Rate are important in ADC and DAC conversion too, as they decide the conversion quality.
( image from wikipedia )
2. Analogic Sound characteristics
The Analogic parts of the signal path have some very important caracteristics too. They are :
SNR : Signal to Noise Ratio. This is the ratio between the actual sound and the background sound ( white noise, ... ) that is added to the signal by a particular device. To hear this noise, start playing a CD, pause it, and turn the volum high : the hissing sound you hear is this added noise. The SNR is expressed in decibels. The higher the number in dB ( decibels ), the less noise will be added to the original sound.
Sound devices always create some noise : Amplifier, Sound Card, ...
Frequency response : The width of the frequency band a device can reproduce. Human beings have an average hearing range from 20 Hz to 20 KHz. The device should be able to reproduce as much as this range, and with the lowest volume modifications from the original signal depending on frequency. The Frequency response is expressed in Hz ( band width ) and Db ( volume accuracy ).
A close to ideal Frequency response would be 10Hz-30KHz ( band width ) 0 dB ( volume accuracy ). The highest the first number, and the lowest the second, the more accurate the sound will be.
Dynamic range : Dynamic Range is the ration between the lowest volume and the highest volume a device can process. It is expressed in dB. The higher the value, the better. An analog cassette has a Dynamic Range of 60 dB, a CD 100 dB.
THD : Total Harmonic Distortion, expressed as a percentage. It is the level of harmonics added to the original signal. The lower is the better.
Signal level : Analog line signal has to have a good voltage value, so that the signal is carried strong and clean. A good signal level helps avoid noise to be added during the signal travel, either trhough the device, or from device to device. It is expressed in V (Volts ) and has to be high.
Other values: There are some others numeric values sometimes quoted :
Stereo crosstalk : noise and sound induced between channels
IMD : ( Inter Modulation Distortion : Distortion not harmonically related, expressed as a percentage.
THD+N : Total Harmonic Distortion + Noise, expressed as a percentage. It is the level of harmonics+noise added to the original signal. The lower is the better
IMD+N : IMD + Noise. Distortion not harmonically related + noise added to the original signal. Expressed as a percentage,the lower is the better
These last four values have to stay has low as possible.
Amplifiers also state output power, that is the highest sustainable non-clipping non-distorded sound an amplifier can send to the speakers. This value is expressed in Watts. The higher the better.
The sound path is a complexe and long travel through digital and analogic devices. It needs a clear understanding of both this path and of the numerics used to evaluate it.
commenter cet article