How Excel column letters become numbers
Excel column letters are bijective base-26. A is 1, Z is 26, AA is 27, and the last column of an .xlsx sheet is XFD at 16,384. Converting letters to a number is one pass: start at 0, and for each letter multiply by 26 and add its place in the alphabet. Going back subtracts 1 before each division, and that subtraction is the whole trick. The older .xls format stopped at IV, column 256. This page converts either direction, translates A1 references to R1C1 and back, and expands a range like A1:D10 into its 40 cells with the row and column counts. Nothing here rounds anything: every conversion is integer arithmetic and reverses exactly across all 16,384 columns.
Most explanations call this base 26 and then quietly get the boundary wrong. It is not base 26.
A positional number system needs a zero digit. Base 26 would run 0 to 25 and write the value after Z as BA, because B is 1 and A is 0. Excel has no letter for zero: the alphabet is 1 to 26, every digit carries a real value, and the column after Z is AA. Systems built like that are called bijective, and the name matters here because it is the one thing that tells you which answer to expect at 27, 703 and 16,384.
The two limits
XFD is 16,384. Multiply it out: X is 24, so 24 × 676 = 16,224; F is 6, giving 6 × 26 = 156; D is 4. That is 16,224 + 156 + 4 = 16,384.
IV is 256, the last column of the old .xls format. Nine alphabets of 26 is 234, and V is 22, so 234 + 22 = 256. Anything to the right of IV disappears when a modern workbook is saved down to .xls, which is the practical reason to know where that boundary sits.
Past XFD this converter keeps answering and changes what it claims. 16,385 is XFE, which is correct, and there is no column XFE in any spreadsheet. Those are two separate facts and the tool states both rather than choosing the tidier one.
A1 and R1C1
R1C1 style writes both halves of a reference as numbers, so D10 becomes R10C4. It is far easier to read arithmetic in, which is why it survives as an option in Excel decades after A1 won the default.
The catch is the brackets. R1C1 is an address and always means the same cell. R[1]C[1] is an offset — one row down and one column right of whatever cell the formula is written in — so it points somewhere different in every cell that holds it. A converter that answers a relative reference without asking where it lives is inventing the base cell. This one asks.
Ranges
The expander takes A1:D10 and lists all 40 references, in reading order. Write it backwards as D10:A1 and you get the same 40, plus a line saying the range was normalised. A whole-column range like A:D covers the sheet’s full 1,048,576 rows, which is 4,194,304 cells: the counts are reported and the list is not built, because producing four million strings would lock the tab rather than answer anything.
