Signal processing

How sound and images become numbers, how those numbers are cleaned up and compressed, and which of these ideas machine learning still relies on.

This part builds on Foundations.

1 of 7 topics ready, listed in reading order

Sampling and quantizationNyquist, aliasing, bit depth and quantization noise.Planned

Audio featuresWaveforms, the short-time Fourier transform, spectrograms, mel scales and MFCCs.Planned

Speech enhancementSpectral subtraction and Wiener filtering compared with neural denoising.Planned

Entropy codingRun-length, Huffman, LZW and arithmetic coding.Planned

JPEG from scratchColour transform, DCT, quantization, zigzag and entropy coding, with rate against quality curves.Ready

Perceptual audio codingPsychoacoustic masking and how MP3 and AAC spend their bits.Planned

Video codingFrames, intra and inter coding and motion compensation.Planned