{short
                    description of image}
SOUND 2
Michael Drolet -- 2017
Home Audio Recording



INTRODUCTION:

 

Analog Systems

Not too long ago, most audio recording systems were analog.  Although simple to implement, maximising performance required careful design and constant maintenance.

Digital Systems

Digital systems are more complex to implement as they have more components.  The initial design is critical, but they tend to be more stable and require less maintenance.

ANALOG RECORDING:

In analog recording, we would set up an analogy between parameters (variables) of the original sound, and parameters of the recording medium.  We would vary some parameter of the medium corresponding to changes in the sound wave.  The two (sound) parameters usually used were time and amplitude.

 

Vinyl Disk:

In phonograph recording, changes in the amplitude of the original sound were used to cause changes in the width of the groove cut into the surface of the disk.  The louder the instantaneous amplitude of the sound, the wider the groove was cut.
Time was represented as distance along the spiral groove.  As the stylus moved from the outside edge toward the centre; the diameter decreased.  Since the disk rotated a constant speed -- 33-1/3, 45 or 78 rpm -- the linear speed decreased near the centre.
  • Scratches and dirt on the disk surface would disrupt the amplitude analogy; causing noise (pops and crackle) in the playback. 
  • Warping of the vinyl material  would disrupt the time analogy; causing changes in pitch (warble or wow). 
  • Irregularities in the composition of the vinyl would disrupt the amplitude analogy randomly;  causing surface noise.

Film Sound:

In analog photographic (or optical)  film sound tracks, changes in the amplitude of the original sound were used to cause changes in the width of the clear area of the optical track.  The louder the instantaneous amplitude of the sound, the wider the "wiggles" sound track.
Time was represented as distance along the film.  Film speed was constant  -- 18 in./sec for 35 mm. film; 7.2 ips for 16 mm.
  • Scratches and dirt on the film surface would disrupt the amplitude analogy; causing noise (pops and crackle) in the playback. 
  • Shrinkage of the film with age would disrupt the time analogy; causing changes in pitch (warble). 
  • Inadequacies in the photo-chemical process would disrupt both amplitude and time analogies; causing distortion.

Magnetic Recording:

In magnetic recording, changes in the amplitude of the original sound were used to cause changes in the strength of the magnetic field stored on the recording tape.  The louder the instantaneous amplitude of the sound, the stronger the magnetic field.
Time was represented as distance along the recording tape.  Tape speed was constant -- 7.5 or 15 in./sec for reel to reel;  1- 7/8 in./sec for cassette.
Reel to reel
On a professional reel to reel recorder,  the tape moved from the left supply reel, past the tape guides and head block, between the capstan and pinch roller and onto the right take up reel.

In the recording mode, the erase head erased the previously recorded signal.  The record head recorded a new signal on the tape; and finally, the playback  (replay) head played back the recorded signal.  The replay head served to provide confidence playback to ensure the quality of the recording.  Because of the spacing between the record and play heads, there would be a noticeable delay in the confidence playback; dependent on tape speed.

In the play mode, the erase and record heads would not be used.

Tape
The erase, record and playback heads were electomagnets; coils of wire wrapped around pieces of iron (or other magnetic material).  A space in the magnetic material, called the gap; concentrated the magnetic field.  The erase head had a very wide gap to provide the maximum erasure.  The  play head had a narrower gap to maximize response to high frequencies (short wavelengths).  The width of the record head  gap was somewhere in between.
Tape Heads
  • Stray magnetic fields, from loudspeakers, electric motors or computer monitors could demagnetise (erase) parts of the tape; disrupting the amplitude analogy; causing drops in signal level. 
  • Stretching or shrinking of the plastic tape disrupted the time analogy; causing pitch changes (wow and flutter). 
  • Loss of the magnetic coating (oxide shedding) disrupted the amplitude analogy; causing drops in signal level (dropouts). 
  • The random variation in magnetic particle size also disrupted the amplitude analogy; causing tape hiss.

The tape heads, guides, capstan and pinch roller had to be be cleaned regularly with alcohol to remove glue from splicing tape, finger print oils and wax pencil grease.

The moving magnetic field of the tape could magnetise metal parts in the tape path and these would have to be periodically de-magnetised; otherwise noise would be added to the recorded signal during each playback.

DIGITAL RECORDING:

In digital recording, rather than setting up an analogy to represent the signal parameters, we represent their instantaneous values as a series of numbers or digits.  These numbers can be stored on tape or disc, then later used to reconstitute the original signal.

In English we call this a digital system, but this is really a misnomer.  It is in fact a numerical system.  In French -- numérique.

The signal to be recorded is processed by an Analog to Digital Converter (ADC).

We sample or measure the signal often enough to get a smooth representation of its shape.  Sampling theory dictates that we must sample at a minimum of twice the highest frequency we want to represent.  For professional audio, we consider the highest frequency to be 20 kHz; so we have to sample at a minimum rate of 40,000 samples per second.

  • For compact disc, the sampling frequency is 44,1000 times per second or  44.1 kHz. 
  • In digital video systems we use 48 kHz sampling for the sound signal.
Sampling Pulses
This sampling gives a series of pulses whose amplitude changes with the amplitude of the original signal.
PAM
The amplitude of each pulse is measured and converted into a number.  For CD and DAT, these numbers are represented by 16 binary digits or bits. 

A 16-bit number can represent 65,536 different levels or steps of amplitude. Digital signal processors and consoles use 24  or 32 bit numbers internally to minimise errors.

Each bit represents 6dB of dynamic range i.e. 16 bits offer 96dB dynamic range (compared to your ears' 120 dB.)

These numbers are then stored on hard disk or in flash memory.  Extra data may be included to permit detection and correction of errors, and to identify the individual recordings.  On playback, the extra information is stripped off.

On playback, the sampled numbers are passed to a Digital to Analog Converter (DAC) which  reconstitutes the basic shape of the wave. 

As you can see,  the shape is rather jagged, because the individual steps are wider than the original sampling pulses, representing a false time value.  The DAC output is resampled as shown below.
The pulses are passed through a filter which removes all frequencies above 20 kHz.   This restores the smooth shape of the original signal.  This is the signal which appears at the analog output of the recorder or CD player.
The standard settings for CD or "CD Quality", Windows PCM (wav) audio are :
  • stereo (2 channels left/right)
  • 44,100 samples per second/channel (44.1 kHz)
  • 16 bits per sample (65,536 steps -- dynamic range 6dB/step, 96 dB total.)

Digital vs Analog:

Dynamic Range:  the ratio of the level of the loudest  to the softest sound that can be reproduced -- usually expressed in decibels (dB)

In an analog system, the level of the loudest sound is limited by the onset of distortion -- due to the physical limits of the media. The level of the softest sound is limited by the background noise -- scratches and dirt on film or disc, the composition of the magnetic material on tape.
 
In a digital system,  the dynamic range is defined by the number of digital bits used to represent the sampled audio -- each bit is equivalent  to 6dB of dynamic range.
MediumPeak:Noise RatioDynamic RangeEquivalent Digital
Analog Optical  (Film)400:152 dB8 bits
Analog Disc (Vinyl)1,000:160 dB10 bits
Analog Magnetic (Reel to Reel)4,000:172 dB12 bits
Digital (CD)65,536:196 dB16 bits
Real World  (Human Ear/Brain)1,000,000:1120 dB20 bits
Professional  Digital  (Scarlett 2i2/Reaper)16,000,000:1144 dB24 bits
Frequency Response:  the range of frequencies that can be reproduced -- usually expressed by the upper frequency alone.


In an analog system, the upper frequency  is usually set by the speed of the media and the physical dimensions of the stylus, optical slit or magnetic gap.

In a digital system, the upper frequency  is set by the digital sampling frequency -- the upper limit is one-half the smapling frequency.
MediumSampling/Speed(Upper) Frequency Response
Analog Optical  (Film)60 micron slit/ 45.7  cm/s8,000 Hz
Analog Disc (Vinyl)19.8 micron stylus/ 19.8 cm/s10,000 Hz
Analog Magnetic (Reel to Reel)20 micron gap/ 38.1 cm/s20,000 Hz
Digital (CD)44,100 Hz22,000 Hz
Real World  (Human Ear/Brain)n/a20,000 Hz
Professional  Digital  (Scarlett 2i2/Reaper)96,000 Hz48,000 Hz
Notes:
  1. 45.7 cm/s (18 inches/s) -- 35 mm film speed
  2. 38.1 cm/s (15 inches/s) -- 1/4 inch analog master speed
  3. 19.8 cm/s (7.5 inches/s) -- roughly linear speed of 7 inch, 45 rpm vinyl disc