Nice article, however it seems the author started with a lot of punch, explaining Frequency Domain and it's lossy compression in high detail, but then got tired/bored to write rest of the article. It's like a movie that starts very slow at the beginning and compensates at the end by just glossing around things.
I particularly would have enjoyed a detail explanation of the Chroma subsampling, with the same level of detail (graphics, examples, etc) as the first third of the article.
I particularly would have enjoyed a detail explanation of the Chroma subsampling, with the same level of detail (graphics, examples, etc) as the first third of the article.