Digital Audio Fundamentals

In this video, we take the first step at the process of converting a continuous signal into a discrete signal for processing within the digital domain, and take a stab at the Nyquist Shannon Sampling Theorem which dictates how often a continuous signal needs to be sampled to accurately recreate it back. There is a lot of misconceptions regarding sample rate and its implications, and this video should hopefully help in clear them up by looking at sampling rates more objectively, or at least give you a starting point for more exploration.

Content:

0:00 Continuous vs discrete signals

2:35 Nyquist Shannon sampling theorem

5:53 Bandlimiting using low pass filter

6:40 Sampling examples in Audacity

12:30 Re-conversion of digital signals to analog signals

16:43 Aliasing artifacts

19:20 Practical sampling rate and outro

In this video, we take what we learnt from the previous video on the Nyquist Shannon Sampling Theorem and apply it to real world audio signals. We discover some of the commonly used sample rates used in audio consumption and transmission and the advantages and disadvantages of using them. We look at the difference between bandwidth in the context of signal and the context of transmission, and discuss why lower rates are ideal for consumption, and higher rates are only required during signal processing.

Content:

0:00 Intro to sampling theory

2:19 Common audio sample rates

3:04 Bandwidth 5:04 8kHz

7:39 Sampling with Audacity

11:26 16kHz

12:00 44.1kHz

12:47 48kHz

13:16 Higher sampling rates

In this video, we demistify and take a stab at understanding aliasing in audio signals. We start off the discussion by looking at how aliasing could occur in video signals by looking at the wagon wheel effect and how the aliasing pattern that you visually see can be applied to understand the concept of aliasing in audio signals. We force aliasing to occur in the digital domain by using the Nyquist programming language, draw a pattern, and derive a formula for predicting the aliased frequencies. We illustrate the process of sampling and see how a low sampling rate coupled with high frequencies can cause aliasing. We take the discussion towards real world observations of aliasing in square waves, and how to counter them manually using oversampling and downsampling.

Content:

0:00 Wagon wheel effect

2:08 Temporal aliasing

4:08 Forcing aliasing in Audacity

9:22 Sampling illustrated

11:24 Aliasing from square waves

13:56 Solutions to avoid aliasing

In this video, on our quest to create a discrete signal out of a continuous signal, we will begin the discussion on how amplitude values of each sampled signal is represented and stored. We'll discuss how the determination of resolution of a sample is lossy when compared to sampling. Finally, we'll look at the real world effects of quantization and bit depth on digital audio - namely noise and dynamic range.

In this video, we'll explore the concept of binary representation of states as bit, and how bit depth applies to representing audio amplitude levels. We'll examine the real world affects and artifacts of low resolution audio file and hear the effects of quantization noise or error. We'll look at signal to quantization noise ratio (SQNR) as a metric for measuring the dynamic range of signals.

Content:

0:00 Binary system

2:03 Base-2 vs Base-10

3:15 Bit vs Byte

4:44 4 bit audio demo

8:32 Properties of quantization error

9:58 8 bit audio demo

12:27 Signal to Quantization Noise Ratio (SQNR)

16:03 Nature of noise

In this video, we'll explore the concept of dithering and why we need it. Dithering is the process of intentionally adding noise to a signal during quantization to preserve low level information and prevent correlated distortion. The video tries break down this definition through image and audio illustrations.

Content: 0:00

Dithering intro

0:45 Image dithering

2:59 Dithering to preserve information

8:10 Closer look

9:45 Dithering to prevent distortion

13:39 When and where to dither

In this video, we'll explore the concept of the different types and techniques involved in dithering. We'll look at 3 different probability functions - rectangular, triangular and gaussian. We'll take a closer look at each ne of these with the help of input output characteristics graphs which measure the linearity of a quantization system, and conclude on the best one to use. We will also briefly touch upon subtractive dithering as an extension of non-subtractive dithering to reduce the noise floor.

Content:

0:00 Overview

0:40 Probability density function

1:25 Rectangular PDF

1:47 Triangular PDF

2:23 Gaussian PDF

3:40 IO characteristics

6:17 Reaper demo

8:09 Subtractive dither

Noise shaping is a technique used alongside dithering to spectrally shape the noise associated with the process of quantization. Noise shaped audio has a higher dynamic range. The quantization noise is redistributed to less sensitive frequency bands so that it's less perceptible to our ears. In this video we'll talk about the absolute threshold of hearing curve, a holistic view of the filtering and shaping process, the POW-R (Psychoacoustically Optimized Wordlength Reduction) algorithms and a small anecdote on the history of analog dither.

Content:

0:00 Noise shaping

2:10 Absolute threshold of hearing

3:54 Simple noise shaping algorithm

5:25 Noise shaping schematics

7:16 POW-R

9:34 Story on analog dither

Pulse Code Modulation is an encoding mechanism, a way of representing digital data for the purposes of transmission and storage. But what form does this encoding take, and why is it needed? In this video, we'll explore the the concept of encoding and modulation as a corner stone of digital audio. We'll start from first principles and discover different modulation strategies and their advantages, and naturally arrive at why PCM is so resilient and has survived the test of time.

Content: 0:00

Encoding 1:21

Frequency Modulation

4:28 Pulses - Digital encoding

6:06 Pulse Width Modulation

6:25 Pulse Position Modulation

7:59 Pulse Amplitude Modulation

9:05 Pulse Code Modulation

9:50 Bandwidth of PCM

11:30 Overview of ADC

Multiplexing is the combination of 2 or more signals for the purpose of transmission. Time division multiplexing is predominant in digital audio. Though it may be a hardware component, the software abstraction of multiplexing is found when we interleave data samples or frames when we save the audio to a file. Error correction, is quite simply correcting errors that may occur due to degrading storage media or noisy transmission. We'll talk briefly about redundancy as a cornerstone of error detection and correction, and provide the realization that not all errors can be corrected. We can only build resilience to errors.

Content:

0:00 Multiplexing

3:48 Error Correction