Chapter 2 · Watch, then practise
Sound / Audio System
Connect sampling, channel count and sample representation to digital-audio storage. Distinguish the time spacing of samples from the precision used to encode each sample.
3 questions · 3 with related videos. Matches are based on playlist titles; broader background matches are labeled.
What to study
- Sampling and aliasing
- PCM bit depth
- Channels and storage
Chapter playlists
Choose a playlist
Notes
Calculating File Size: Audio
Eddie Woo · 10:33
Supplementary worked audio-file-size lesson; use this question’s rate, depth, channels and duration.
1. Raw audio size
Calculate one minute of uncompressed stereo PCM at 48,000 samples/s per channel and 16 bits per sample, excluding headers.
Rate = 48,000 × 16 × 2 = 1,536,000 bit/s. For 60 seconds, size = 1,536,000 × 60 / 8 = 11,520,000 bytes, or 11.52 MB using decimal units. Compression and container overhead change the stored-file size.
aLec39 NyquistTheorem
Jonathan Valvano · 17:09
Supplementary Nyquist-theorem lecture from the embedded-systems course.
2. Sampling a band-limited signal
An ideal signal contains frequencies up to 8 kHz. What sampling rate is needed to avoid aliasing?
The theoretical boundary is twice the highest frequency, 16 kHz. Choose a rate above that boundary in practice and filter frequencies above the allowed band before sampling. Sampling rate limits represented frequencies; increasing sample precision does not fix aliasing.
Calculating File Size: Audio
Eddie Woo · 10:33 · Background lecture
Audio-size and sampling background for relating sample frames, rate and duration.
3. Duration from samples
A mono audio buffer contains 96,000 sample frames at 48,000 Hz. What is its duration, and would stereo change it?
Duration = 96,000 / 48,000 = 2 seconds. A stereo buffer with the same number of sample frames and sample rate is also 2 seconds; each frame contains values for both channels, so it contains more sample values, not more playback time.