The short answer
ASCII is a set of 128 characters, the English letters, digits and punctuation, each stored in one byte. Unicode is a catalogue that gives a number, a code point, to every character in every script, more than 150,000 of them. UTF-8 and UTF-16 are encodings: ways of storing those code points as bytes. UTF-8 uses one to four bytes per character and is identical to ASCII for ASCII text; UTF-16 uses two or four.
Look inside any character
Type anything. The code points come from Unicode; the bytes from each encoding. The emoji needs two UTF-16 units, a surrogate pair, which is why JavaScript says "😀".length is 2.
Side by side
| ASCII | UTF-8 | UTF-16 | |
|---|---|---|---|
| Covers | 128 characters | All of Unicode | All of Unicode |
| Bytes per character | 1 | 1 to 4 | 2 or 4 |
| A | 41 | 41 | 00 41 |
| é | not in ASCII | C3 A9 | 00 E9 |
| 😀 (U+1F600) | not in ASCII | F0 9F 98 80 | D8 3D DE 00 |
| ASCII text unchanged? | Yes | Yes | No: every character takes 2 bytes |
| Mostly used in | Old protocols, as a subset | The web, files, Linux, most APIs | JavaScript and Java strings, Windows APIs |
Why UTF-8 won
UTF-8 is backward compatible with ASCII: an ASCII file is already valid UTF-8, byte for byte. It has no byte-order question, it never produces a zero byte except for the character U+0000, and a program can always find where a character starts. Those properties made it the default for the web, for file formats such as JSON, and for most of the internet's protocols.
UTF-16 survives where it was built in early: JavaScript and Java strings, and Windows' own APIs, all designed when 65,536 characters looked like enough. When Unicode grew past that, characters above U+FFFF had to be split across two 16-bit units, a surrogate pair.
When it goes wrong
Text saved in one encoding and read in another turns into garbage such as ’ for an apostrophe. That's mojibake, and the mojibake fixer repairs the most common kind. The UTF-8 decoder explains bytes that aren't valid UTF-8 at all.
Questions
Is UTF-8 the same as Unicode?
No. Unicode is the list of characters and their numbers; UTF-8 is a way to store those numbers as bytes.
Is ASCII a subset of UTF-8?
Yes. The 128 ASCII characters have the same single-byte values in UTF-8.
Why is the length of an emoji 2 in JavaScript?
JavaScript strings are UTF-16, and characters above U+FFFF, emoji included, take two 16-bit units.
Should I use UTF-8 or UTF-16?
UTF-8 for files, the web and anything that leaves your program. UTF-16 mostly where a platform already uses it internally.
Sources
Added . What's new