Douglas H. Quebbeman wrote: Or flux changes per inch perhaps? -- hbp

Don't know about fcpi, but... "Douglas H. Quebbeman" wrote: You can use bpi and cpi interchangeably here: the bpi refers to a single track, while the cpi unit refers to the complete set of all nine tracks. If you use 8-bit characters and a parity bit, as was common at that time, you can store exactly one character in a "rung" (I hope this is the correct English term - over here in Germany, I heard the word "Sprosse" in this context, and "Sprosse" translates to "rung"). For 7-track devices (which were common with the low-end recording densities you mention), the same proportion is true with 6-bit characters (see a previous thread about bytesizes: you will find me mentioning the SIXBIT code there). I think 9 tracks did not appear before the higher longitudinal densities (somewhere around 800 bpi), so when saying 200bpi, you implicitly refer to a 7-track device. You may want to look at bpi as a term more oriented in recording density, while cpi is talking about media capacity, but of course, these are very closely related. For fcpi, the "flux changes per inch" definition mentioned by Hans might well be true: I'd have to look up some old papers for details, but I seem to remember that advanced recording technologies didn't need more than one flux reversal per recorded bit (and were still able to derive proper clock pulses from the read signal). Telling from the digits, it is similar to bpi, because 6250bpi was the latest commonly used density (12500bpi did exist, but was rare). -- Andreas Freiherr Vishay Semiconductor GmbH, Heilbronn, Germany http://www.vishay.com