How to Read a Model Number
A field cryptanalysis
How to read a model number
Every machine in this issue announces itself in a short burst of letters and digits, and after a decade of monitoring the traffic the department can report: some of it is plaintext, some cipher, some pure noise transmitted with great confidence. Herewith the codebook — four classes of scheme, graded in the intelligence style: verified, inferred, or resisting decryption entirely.
Method first. A model number has at most three working parts: a prefix (the letters), a payload (the digits that mean something), and ceremony (the digits that don’t). The craft is knowing which is which before the dealer does. Four classes cover the industry; memorise them and burn this page — or laminate it, which in this household amounted to the same devotion.
Class I — the honest declarers
These brands print the speed on the nameplate and dare you to check. Toshiba’s BD line is the clean intercept: first two digits are the copies per minute — 2060, 3560, 4550, 9050 — verified almost exactly against a period EPA catalogue. (“BD” itself — “Beauty Duplicator,” says dealer folklore — remains unverified lore.) Ricoh’s FT series transmits from the other end: last two digits are the cpm, verified for the FT4527 and FT5535 in the service manual, inferred across the rest.
Minolta’s Di digitals barely bother encrypting: the number is the ppm — Di620, 62 a minute, confirmed at the flagship, the rest following the pattern. Panasonic’s FP speaks honesty in two dialects: leading digits in the xx80 family (FP-1780, 17; FP-4080, 40), tail digits in the Medallion 77xx (7735, 7742, 7750 — 35, 42, 50). Same truth, different word order.
Class II — the dynasty-namers
Here the prefix is the message and the digits are house business. Canon runs three dynasties: NP (“New Process,” patented 1965 — the letters commemorate the physics), CLC for the colour court, PC for the cartridge commoners. Trust the prefix, audit the digits: the NP-6030 does 30 cpm and the NP-6050 does 50, but the NP-9800 does 83. Mita’s DC dynasty runs from the DC-15 of 1979 to the DC-9285 — and the department files a minority report: no doctrine is declared in the record, yet the EPA dataset shows the 90s fleet’s leading digits tracking speed almost perfectly (1820 at 18, 9285 at 92). A speed-declarer on dynasty papers. Kodak’s Ektaprint numbers are pure heraldry: the 95 runs 60 cpm, the 235 about 85. Seniority, not speed.
Class III — the era-coders
Sharp is read the way a geologist reads a cliff: the prefix is a stratum. Z is the personal-cartridge bed; SF — the Sharpfax lineage, 1972 — the analog office layer; SD the high-volume iron; AR, from 1994, the digital cap-rock. Date any Sharp by its first letter. Within the SF-20xx bed the tail digits then declare speed — 2022 at 22, 2050 at 50 — stratigraphy with a bonus cipher.
Class IV — the enigmas
Some traffic is not meant to be read. Xerox ran overlapping series like agents on parallel covers: the 10 Series Marathons; the 50 Series, where 5018 means 18 cpm and 5028 means 28 right up until the 5090 means 135; then DocuTech, plainly named for its speed (135, 90, 180) the moment the office line stopped being decodable; then Document Centre, named for a mood. Océ defeats the department entirely: the 2400 runs 60 cpm, the 2500 runs 100; the digital 3165 of 1997 runs 62 while the analog 3100 of 1998 runs 100. Even the chronology decrypts wrong. We admire it, as one admires any unbroken code, from a distance.
Gradings follow the source discipline of this issue: verified means a period EPA catalogue or service manual states the figure; inferred means the convention filled a gap the record left open; unverified stays unverified. The stars overleaf grade the scheme’s reliability as a cipher — not the machine, which was usually more trustworthy than its name.
The decoder ring — all schemes, graded
| Brand | Scheme | Worked example | Reliability |
|---|---|---|---|
| Toshiba BD | First two digits = cpm | BD-4550 → 45 cpm. It is. | ★★★★★ |
| Ricoh FT | Last two digits = cpm | FT5535 → 35 cpm (manual-verified) | ★★★★☆ |
| Minolta Di | Di + the ppm, entire | Di620 → 62 ppm (confirmed) | ★★★★☆ |
| Mita DC | Prefix = dynasty; leading digits track cpm in practice (undeclared) | DC-9285 → 92 cpm, 78 in duplex | ★★★★☆ |
| Panasonic FP | Leading digits ≈ cpm (xx80 family); tail digits in the 77xx | FP-2680 → 26 cpm | ★★★☆☆ |
| Sharp | Prefix = era (Z→SF/SD→AR); SF-20xx tail = cpm | SF-2050 → analog office era, 50 cpm | ★★★☆☆ |
| Canon | Prefix = dynasty (NP / CLC / PC); digits moody | NP-6030 → New Process; 30 cpm by grace, not doctrine | ★★☆☆☆ |
| Xerox | Overlapping series; DocuTech alone named by speed | 5028 → 28 cpm; 5090 → 135. Quite. | ★★☆☆☆ |
| Kodak | Ektaprint + seniority number | Ektaprint 95 → 60 cpm | ★☆☆☆☆ |
| Océ | None detected | 3165 → 62 cpm; 3100 → 100 | ★☆☆☆☆ |
Field drill
Presented with an unfamiliar nameplate: (1) read the letters — they date and dynasty the machine; (2) read the digits twice, once from each end — the payload hides at either extremity, never the middle; (3) state your confidence aloud before lifting the platen cover. A cryptanalyst who says “45 cpm, inferred” and is right has done intelligence work. One who says “45 cpm” and is wrong has done marketing.