Chapter 3 — Key Takeaways

A one-page reference to screenshot and re-read before a shoot. Chapter 3 is the recording pipeline — how the image is compressed and stored, and why those choices decide what you can do in post. This is the one Chapter 2–3 decision you can never take back.

The core idea

The camera records a compressed approximation of what you saw — and how good that approximation is, and how far you can push it later, is chosen at capture and can never be undone. Compression damage is permanent; latitude is bought on set and spent in the grade. Ask one question before every shoot: what will happen to this footage in post?

Definitions

Codec vs container — the one that confuses everyone

• Delivery codecs (H.264/H.265): small files, hard to edit (interframe/long-GOP — frames rebuilt from neighbors). Great for capture and web.
• Editing codecs (ProRes/DNxHR): big files, smooth to edit (intraframe — every frame complete). The intermediate you cut on (Ch.27, 36).

The flexibility ladder — standard → Log → RAW

Climb the ladder only as far as the look and your pipeline (card, computer, skill, deadline) actually need.

Bit depth & chroma — the pair that decides the grade

• Bit depth → banding. Grading stretches tonal steps apart; 8-bit's 256 open into visible gaps in smooth areas (skies, walls, skin). 10-bit's 1,024 stay smooth.
• Chroma → keying + heavy color. 4:2:0 looks identical when watching (eyes favor brightness over color) but gives soft, blocky edges for green screen (Ch.35) and less room for big color moves.

Log — the three strings (read before you flip it on)

1. Log must be graded. Ungraded, it looks broken. No grade planned? Shoot a standard profile.
2. Log wants bit depth. Stretching flat 8-bit Log to contrast bands → 8-bit Log is a trap. Use 10-bit.
3. Log needs careful exposure — slightly bright, to keep shadow noise down (Ch.5, §5.5).

The one question, four common answers

"What will happen to this footage in POST?"
  ├─ Cut fast, light color, tight deadline
  │     → STANDARD (Rec.709), highest bitrate you can store. 8-bit 4:2:0 is fine.
  ├─ Build a real look / moody grade, you CAN and WILL grade
  │     → LOG (or flat), 10-bit if possible, high bitrate, exposed carefully.
  ├─ Green screen / VFX / fine color
  │     → 10-bit 4:2:2 minimum, clean high bitrate.
  └─ High-end grade, big screen, must fix WB/exposure in post, you have the pipeline
        → RAW (compressed: BRAW / ProRes RAW / etc.).

The four setups → format lean

Bitrate & cards — mind the units

• Bitrate is in megabits (Mb/s). Card write speed is in megabytes (MB/s). 8 bits = 1 byte, so Mb/s ÷ 8 = MB/s.
• 400 Mb/s = 50 MB/s of writing → want a card rated comfortably above (V60/V90), not exactly at it.
• Match card Video Speed Class (V30/V60/V90 = guaranteed MB/s) to your bitrate, with headroom.

Compression artifacts — name them, then avoid them

All three are permanent. The grade amplifies them; it never removes them.

Common mistakes → fixes

Pre-shoot format checklist

• [ ] Asked "what will happen to this footage in post?" and chosen accordingly
• [ ] Profile chosen (standard vs Log) and justified by the post plan — not by what sounds pro
• [ ] Bitrate set to the highest the card sustains (for anything graded)
• [ ] Bit depth / chroma set for the job (10-bit / 4:2:2 if grading hard or keying)
• [ ] Log? → 10-bit, exposed slightly bright, and a grade is actually planned
• [ ] White balance set correctly in camera (unless shooting RAW)
• [ ] Card fast enough (MB/s ≥ bitrate ÷ 8, with headroom) and from a trusted seller
• [ ] Offload plan written: two copies before the card is ever reused
• [ ] Test clip shot, offloaded, and confirmed to import and play cleanly

The card-and-offload rule (the one that saves careers)

Never format or reuse a card until its footage lives in at least two other places — and confirm the copies actually opened. Offload → back up → test import/playback → then wipe. (Formalized as 3-2-1 backup in Ch.37.)

Frame Log — this chapter's seeing prompts

Add one per day. Watch anything and answer one of: - Can I spot banding in a sky, a fade, or a studio backdrop? - Does this look over-compressed (blocky shadows, mushy motion)? What was the likely bitrate? - Is this a built grade (a real "look") or a straight-from-camera image? What tells me? - Did the camera hold both a bright window and a shadowed face, or sacrifice one? (dynamic range) - Where has a film handed its identity to color — and is that color motivated?

The craft principles this chapter carries

• You shoot for the edit. Bit depth, bitrate, and range are captured on set purely to give the grade room — coverage of a different kind, paid in storage.
• Fix it in pre, not in post. A format/card test on a throwaway clip catches a won't-import file or a banding trap before the day that counts.
• Motivate every choice. Reach up the flexibility ladder only as far as the look genuinely needs — never "to be safe" or "to look pro."
• The best format decisions are invisible. No viewer thanks you for enough bit depth; they just feel the video looks right.

This chapter's project step

Production Checkpoint: decide Project 1's capture format (standard vs Log) and card-and-offload workflow, then shoot a short test and confirm it imports and plays back cleanly. For most talking-heads: a good standard profile at the highest sustainable bitrate (8-bit 4:2:0 is fine), white balance set in camera. Log only if you'll genuinely grade it, in 10-bit, exposed carefully.

The one thing to remember

If you take a single line from this chapter into the rest of the book: the recording format sets the ceiling on what post can do — and you set it, blind, the instant you press record. You can re-light, re-frame, and re-cut, but you can never un-compress footage or add information that was never recorded. Choose the format for the edit you intend to have, test the whole chain before it counts, and back up before you wipe. That is the difference between a shooter who commands their footage and one who finds out the ceiling too late.