Chapter 30 — Key Takeaways

The one-page reference to screenshot and re-read before you import a card or set up your library. The images are made elsewhere; this is the card that keeps you from losing them.

The one rule that prevents the most loss

A file is never allowed to exist in only one place for longer than it takes to copy it. Corollary: never format a card until a verified second copy exists. Corollary: one copy is zero backups.

The pipeline at a glance (§30.1)

[ CARD ] → INGEST → [ LIBRARY ] + [ BACKUP ] + [ OFFSITE ] → CULL → KEYWORD → DEVELOP → DELIVER + ARCHIVE
   1 copy   copy·verify·         3 copies = safe            rate    tag       finish    export·freeze
   DANGER   rename·back up                                  keepers  keepers

Ingest checklist (run every shoot — §30.1)

• [ ] Copy, don't move, off the card (leave originals until verified + backed up)
• [ ] Verify the copy — file count matches (487 on card = 487 in library), or checksum confirms
• [ ] Rename to your convention on the way in (not later)
• [ ] Back up — a second copy to a different drive, ideally in the same import action
• [ ] Then and only then format the card (copy #2 must demonstrably exist first)

The two-pass cull (§30.2)

Rating shorthand: ★★★ solid keeper · ★★★★ strong, worth developing, portfolio candidate · ★★★★★ best of the shoot (rare; the one you'd print). On a phone: the Favorite heart is a one-bit keeper flag.

Why passes beat one sweep: rejecting (fast, reflexive pattern-match against failure) and selecting (slow comparative judgment) are different mental jobs that poison each other if mixed. Split them.

Findability: keywords & metadata (§30.3)

Principle: folders store, keywords retrieve. A photo lives in one dated folder but carries many keywords, so it surfaces in every relevant search.

Keyword in layers (cheap → expensive):

Caption = alt text. A good caption ("Red maple in river fog, north trail, late afternoon") is searchable and a screen-reader description at once — findability and inclusion in one field.

The 3-2-1 backup rule (§30.4) — the most important table in the chapter

Make it survive contact with reality:

• [ ] Automatic, not "when I remember" — a manual backup is the one you'll forget at the worst moment
• [ ] Offsite is the cloud's job — the cloud is your offsite leg, not your whole plan
• [ ] Verify 2–3× a year — actually restore a file from the backup; an untested backup is a hope
• [ ] Archives obey 3-2-1 too — they sit untouched for years, exactly long enough for a lone drive to die

Folders & naming that scale to 10,000 (§30.5)

Photography/
├── 2026/
│   ├── 2026-10-14 park-maple-fog/
│   └── 2026-10-18 intersection-saturday/
└── 2027/ …

File:  2026-10-18_intersection-saturday_0042.dng
       └── DATE ──┘ └─── shoot slug ───┘ └seq┘ ext

Rule: choose one convention and never deviate. A consistent "good enough" scheme beats a perfect one you abandon.

The catalog: map vs. territory (§30.5, from Ch.26)

• The catalog = the map (pointers to originals + edit recipes + ratings + keywords). The folders = the territory (the actual files).
• Never move the territory without the map. Reorganizing photo folders in the file browser outside the catalog breaks every link → "missing" photos (files are fine, pointers dangle).
• Do all moving/renaming inside the catalog. If links break, use locate/relocate — never re-import (that makes duplicates and orphans your ratings).

Delivery & archiving (§30.6)

Archive package (self-contained): originals + edit recipe/sidecars + full-res finals (TIFF/JPEG, opens in anything) + delivered exports + a plain-text _README.txt (what/when/for whom/the detail future-you forgets). Store offsite.

Top mistakes → fixes

Portfolio Checkpoint (this chapter)

Cull, rate, keyword, and back up everything so far — turn the growing pile of keepers into a managed, findable collection:

• [ ] Cull the whole body of work (two-pass); portfolio candidates rise to ★★★★, the best to ★★★★★
• [ ] Organize & name into a clean YYYY/YYYY-MM-DD shoot/ tree with consistent file names
• [ ] Keyword every image by chapter/technique, subject, and location (so Chapter 40 assembles by search)
• [ ] Back up to 3-2-1 — this collection is now the most valuable thing on your drives

Not a new frame — a real collection. A portfolio you can't find or could lose is a liability, not a portfolio.

First-defined terms (this chapter)