Shift JIS
A Japanese character encoding that writes the kanji and kana of JIS X 0208 in two bytes, used on Windows as cp932 with NEC and IBM extensions.
Shift_JIS is a Japanese character encoding that uses one byte for Latin letters, digits and half-width katakana, and two bytes for the kanji and full-width kana of JIS X 0208.
Windows uses Windows-31J (code page 932, cp932), which adds NEC and IBM extension characters. The circled digit "①" and the kanji "髙" exist only in cp932, and the same two bytes 81 60 read as "〜" in Shift_JIS but as "～" (U+FF5E) in cp932. The browser encoding standard (WHATWG Encoding) treats shift_jis and windows-31j as labels for the same encoding.
UTF-8 is a different scheme that writes every character in one to four bytes, with most kanji and kana taking three. In "写真2026" the digits 2026 are the same bytes in both, but "写真" takes 4 bytes in Shift_JIS and 6 in UTF-8. So when a ZIP whose names were written in cp932 by Japanese Windows is read as UTF-8, the file names come out garbled. How to spot and fix this is covered in Why ZIP File Names Turn into Garbled Text and How to Fix It.