Grandma Moon Speaks: Pulse Code Modulation (PCM)

Topics Covered:

  • Pulse-Code Modulation
  • Linear PCM

No; we have been as usual asking the wrong question. It does not matter a hoot what the mockingbird on the chimney is singing. The real and proper question is: Why is it beautiful? – Bertrand Russell

PCM
Image Courtesy: Ktims

Modulation is the art of changing something. As we have seen in Signal we can change the amplitude, frequency, phase or any of the characteristics that go into making a signal.

From Professor Sampler’s notes on Sampling, you will remember how audio signals are cut up into little chunks to beget their digital brethren. This, too, is one form of modulation. You are changing an analog signal into a digital one.

There are many ways one can sample, and in the case of audio signals, the most widely used standard is the Pulse-Code Modulation (PCM) sampling system. This system relies on two main properties:

  • Sampling Rate
  • Bit Depth

The above diagram shows how a sine wave is sampled at regular intervals. Nothing new here. The signal is encoded in digital bits, binary of course. Hence the term ‘Code’.

And ‘Pulse’ is from the nature of these signals – which look like pulses. You must realize that all this science used to go closely with electrical transmission lines and early telephone and telegraph systems.

The device that converts the analog audio waves into digital signals is an Analog to Digital Converter (ADC). The principles of the PCM system are simple and follow the Nyquist Frequency criterion.

Why is PCM important? It is the single-most widely used system in the world, even today. From telephones to CDs to DVDs to Cameras, analog audio signals are converted to digital signals using the simple PCM method.

Linear PCM (LPCM)

When you see a digital camera or audio device referring to PCM, it is probably referring to Linear Pulse-Code Modulation, or LPCM.

How is it different from PCM?

PCM with linear quantization is LPCM. What this means, is that when the amplitude is sampled and written as a binary number (digitization), there is a proportional relationship between the two. E.g., the relationship is a linear one: y = ax + b.

As you probably know, a non-linear equation involves a ‘power’ (xn) or a root or a logarithm, and so on. Linear is as simple as it gets.

Just like PCM, LPCM shares similar properties:

  • Sampling Rate
  • Bit Depth
  • Number of Channels

There are a few others, but we’ll focus on just these three for the moment.

LPCM is used commonly for uncompressed audio, and is meant to be lossless. From audio CDs to PCs to DVDs to Blu-rays to DV to AVCHD to HDMI, all of them support LPCM.

The most common sampling bit depths (sometimes called ‘resolution’) for LPCM are 8, 16, 20, 24, and 32 bits per sample (channel).

The most common sampling frequencies are 44.1 KHz and 48 KHz, and high-end equipment easily crosses 96 KHz.

In the next chapter, we’ll look at channels.

Takeaways:

  • Pulse-code modulation is a method used to digitize analog audio signals.
  • Linear PCM is PCM where the digital signal is quantized (calculated) linearly.

Links for further study:

Next: Coming Soon
Previous: The Bit Depth of the Human Ear