The text column
Bytes, characters, and encodings
A byte is a number. A character is an agreement about which number means which mark. The agreement is the encoding. The hex column does not use an encoding. The text column does, and a good editor lets you choose which one.
ASCII
ASCII assigns characters to the numbers 0 through 127. It is the encoding you are looking at in most older formats, and in the English parts of newer ones. A handful of values are worth recognizing on sight:
20is a space.30through39are the digits 0 through 9.41is A.61is a. The alphabet runs on from there, so B is42and b is62.0Ais a line feed, the usual end of a line on Linux and macOS. Windows text often uses0D 0A, a carriage return followed by a line feed.
00000000 48 65 6C 6C 6F 20 0A Hello .The dot
When a byte is not a printable character in the encoding you selected, the editor still has to put something in the text column or the rows would no longer line up. The usual stand-in is a dot, or a small middle dot. In the dump above, the line feed became that dot. The byte did not become a period. A real period is the byte 2E.
UTF-8
UTF-8 keeps every ASCII character as a single byte, which is why English text looks the same in an ASCII column and a UTF-8 column. Other characters take two, three, or four bytes. An accented letter or an emoji is several hex pairs. In an ASCII view those pairs look like odd characters or dots. Switch the text column to UTF-8 and, if the file really is UTF-8, the character comes back.
The hex does not change when you switch encodings. You asked for a different interpretation of the same numbers.
EBCDIC
Some files, mostly from IBM mainframes, store text in EBCDIC rather than ASCII. The letter A is C1, not 41. In an ASCII column that text looks like noise even though it is ordinary words. If the hex is full of values in the C0–F0 range where you expected a sentence, try an EBCDIC text column before you conclude the file is binary.
Numbers are not text
A value such as “the length is 13” is often stored as an integer, not as the characters 1 and 3. The four bytes 00 00 00 0D can be the number 13. The text column will not show 13; it will show a dot, a dot, a dot, and a carriage-return stand-in, because none of those bytes is a digit character. Digit characters would be the bytes 31 and 33.
Order matters, and it is a property of the format. The same number 1 can be stored as 00 00 00 01 or as 01 00 00 00. The first is called big-endian, the second little-endian. The editor can show both readings once you select the bytes. It cannot know which reading the format intended.