Case Study 2: Covering a Two-Person Scene — a Kitchen-Table Shoot-Along
This is a production walkthrough — a from-scratch coverage shoot you can copy this weekend with one phone and two willing friends. Where Case Study 1 took apart a famous scene, this one builds an ordinary one from nothing and proves that the pieces cut together. Everything here is an illustrative shoot (a constructed teaching example); the Described Shots render results, not real footage. Read it with your camera nearby.
The brief and the constraints
Here is the whole job: shoot a fifteen-second, two-person exchange with full coverage, on one side of the line, and cut it into a scene that reads. The content is deliberately trivial, because the content is not the point — the coverage is.
The scene: two people at a kitchen table. One (we'll call them the giver) slides a set of house keys across the table to the other (the taker) and says a short line about it — a handoff, a small moment of trust. The taker replies. That's it. A beginning (the keys arrive), a middle (the exchange), an end (the taker takes them).
The constraints — the same ones a phone-only reader faces:
- One camera (a phone is perfect), shot in several passes, not several cameras at once.
- Available light only — a kitchen window and whatever's on. No lights bought.
- Two non-actors — friends, not professionals. The performances will be plain; the coverage will carry it.
- One location, thirty minutes. This is a coverage drill, not a film shoot.
The goal at the end is concrete and testable: drag the pieces into an editor and confirm they cut into a legible fifteen-second scene — the geography clear, the eyelines matched, the exchange readable — and that you could cut it a second, different way. If you can, you shot coverage. If you can't, you'll see exactly which piece you missed.
Gear and the starting settings
You need almost nothing: a phone, something to brace or mount it (a small tripod, a stack of books, a windowsill), and — if you have one — any external microphone for the dialogue, because sound is half the picture and a kitchen has echo. If you have no mic, get the phone close for the sound and record in the quietest moment you can.
⚙️ Settings Box: a two-person table scene in available light (a starting point, not a recipe).
Setting Starting point Why Orientation Horizontal (16:9) A two-person scene needs width for both people and the room Resolution / frame rate 1080p or 4K at 24 fps 24 fps for the filmic look; 4K gives you room to reframe in the edit Shutter ~1/50 s (180° at 24 fps) Natural motion blur (the Chapter 2 shutter rule — not the line rule!) Aperture / depth Moderately shallow if you can (phone: portrait or main lens) Soft background separates faces from a busy kitchen ISO As low as the light allows (native first) Clean image; raise only if the window light is weak White balance Set for the window (~5600 K) or one custom setting, then lock it So every angle matches for the edit — don't let it auto-shift Mic External if possible, placed close and out of frame Clean dialogue; a kitchen is echoey Framing MCU for singles, room for captions at the bottom The talking-head standard; leave clearance for subtitles The one non-negotiable for coverage: lock your exposure and white balance before you start and don't touch them between passes. If shot 1 (the master) and shot 4 (the close-up) are different brightnesses or color, they will not cut together cleanly — and matching that in post is a chore you can avoid for free on set. (Full shot-matching is Chapter 31; the cheapest fix is to not create the problem.)
The setup: find the line, pick a side
Before a single frame, do the most important thing in this whole case study: find the line and pick a side. The two people at the table are your two subjects; the axis of action runs straight through both of them. Choose one side of that line and commit every camera position to it.
FIGURE CS2.1 — The kitchen table: the line and the shooting side (top-down)
[ WINDOW / soft key ] (sun)
|
v soft daylight rakes across the table
( GIVER ) - - - - - - axis of action - - - - - - ( TAKER )
faces right > (the line through both) < faces left
| |
[------------------------ the SHOOTING side ------------------------]
[ CAM 1 ] [ CAM 2 ] [ CAM 3 ]
MASTER wide MS/OTS favor giver MS/OTS favor taker
(both + table + room) giver looks right taker looks left
+ the CU pass + the insert
The window is camera-left, so it keys the near cheek of whoever faces it and models both faces.
Every camera stays below the line on this diagram -- the shooting side. The giver always faces/looks
screen-RIGHT, the taker screen-LEFT. Cross to the far (window) side and they'd swap and the scene breaks.
Two things to notice. First, the window is on the shooting side's left, so it lights the faces rather than backlighting them into silhouette (a lesson from Chapter 5 — don't put your subject between the camera and the bright window). Second, the line runs along the table between the two people, and everything you shoot lives on the near side of it. Mark your side mentally as a wall you won't cross. Now shoot.
The shoot, phase by phase
We shoot in the professional order: the master first, then tighter and tighter. Each pass runs the whole fifteen-second exchange, so the action overlaps across every size — which is what will let the pieces cut together on any word.
Phase 1 — the master wide (the spine)
The thinking: get the whole scene, once, in one continuous wide take. This is your safety net and your establishing shot. If everything else goes wrong, you still have the scene; and its wide framing will ground the audience before you go tight.
FIGURE CS2.2 — "The master" (whole scene, one take) [illustrative shoot]
THE FRAME Wide: both people at the table, the giver camera-left, the taker camera-right, the window and a
slice of kitchen behind. Both are small-ish; the table and the keys between them are visible.
THE MOVE Locked off on a mini-tripod at eye-level. It holds the whole exchange without moving.
THE LIGHT Window light from camera-left rakes across the table; both faces are modeled, the far cheeks a
stop darker. Nobody bought a light.
THE SOUND Both voices, plus kitchen room tone; we record the full exchange and hold two extra seconds of
silence at the end for room tone.
THE CUT This is the establishing shot and the re-establishing shot; the edit will open here and return
here whenever the geography needs resetting.
THE EFFECT We read the space instantly -- who's where, the table, the keys -- so the coming close-ups can
isolate faces without ever losing us.
THE LESSON Shoot the master first and shoot it complete. It is the one piece you cannot rebuild from the
others, and every tighter shot is safer because it exists.
What was adjusted: on the first take the giver's face fell into shadow because they'd turned too far from the window. Fix — angle both chairs a few degrees toward the window so the light catches both near cheeks. Re-shot. This is why you run the master a couple of times.
Phase 2 — the medium / over-the-shoulder favoring the giver
The thinking: now come closer for the giver's half of the exchange. Frame over the taker's shoulder (an OTS) so we feel the taker's presence and the giver looks past the near shoulder, camera-right, into the taker's eyeline. Run the whole exchange again.
FIGURE CS2.3 — "Favor the giver" (OTS medium) [illustrative shoot]
THE FRAME Medium over-the-shoulder: the taker's near shoulder and cheek soft in the foreground camera-right,
the giver sharp beyond, framed on the left third, looking camera-RIGHT toward the taker.
THE MOVE Locked off, same side of the line as the master, just moved in and around to the taker's shoulder.
THE LIGHT Same window key -- unchanged, so this cuts with the master. The giver's near cheek catches the light.
THE SOUND The giver's lines are now closer and cleaner (mic moved in); the taker's reply is still captured for
reference but will be replaced by their own single's audio.
THE CUT Cuts against the master and against the taker's reverse; the near shoulder anchors the right side of
frame, matching the taker's screen-right position from the wide.
THE EFFECT We're now inside the exchange, close on the giver, but the shoulder keeps us aware that someone is
listening -- the scene feels like a real two-way moment, not a monologue.
THE LESSON An over-the-shoulder keeps the listener present. Frame it so the near shoulder matches that person's
side of the frame from the wide, and the space stays perfectly consistent.
What was adjusted: the first framing put too much shoulder in the frame, blocking the giver. Fix — drop the camera slightly behind the shoulder so the shoulder is a soft edge, not a wall. Also confirmed the giver was looking at the taker's actual eyes (the real eyeline), not vaguely at the camera.
Phase 3 — the reverse, favoring the taker
The thinking: the reverse of Phase 2 — same idea, flipped. Move around (staying on the same side of the line) to shoot over the giver's shoulder onto the taker, who looks camera-left. This is the other half of the shot-reverse-shot.
FIGURE CS2.4 — "Favor the taker" (the reverse OTS) [illustrative shoot]
THE FRAME Medium OTS reversed: the giver's near shoulder soft in the foreground camera-LEFT now, the taker
sharp beyond on the right third, looking camera-LEFT toward the giver.
THE MOVE Locked off, still on the shooting side of the line -- we did NOT cross to the window side.
THE LIGHT Same window key. Because the taker faces the giver (and the light), their near cheek is lit; matches.
THE SOUND The taker's reply, now close and clean.
THE CUT The matched reverse to CS2.3: giver looks right, taker looks left, so cutting between them reads as
the two of them looking straight at each other.
THE EFFECT The eyelines lock. Cut CS2.3 and CS2.4 together and two separately-filmed shots become one
conversation -- the illusion the whole chapter promised.
THE LESSON The reverse only works if you stayed on one side of the line. Same side = opposite, matching eyelines
= a real exchange. Cross the line here and the taker would suddenly face the wrong way.
What was adjusted: nothing major — because the line was respected, the reverse "just worked." That is the payoff of the discipline: when you keep the camera on one side, the reverse is almost automatic.
Phase 4 — the close-ups (the beats)
The thinking: get a clean close-up of each person for the moments that matter — the giver's face on the line about the keys, the taker's reaction as they receive them. These are your emphasis, saved for the beats that earn them.
FIGURE CS2.5 — "The reaction" (the taker's close-up) [illustrative shoot]
THE FRAME Close-up: the taker's face fills the frame, eyes on the upper third, looking camera-LEFT (matching
their eyeline). The kitchen is a soft blur behind.
THE MOVE Locked off, same side, moved in close (or a step back on the tele lens for a flattering compression).
THE LIGHT Window key on the near cheek; the face is the brightest thing in frame, so the eye locks to it.
THE SOUND A breath, then a quiet "thanks." The room tone drops under the beat.
THE CUT Cut IN to this on the moment the keys land in the taker's hand; hold a beat on the reaction, then
cut out. This is the emotional peak, so it gets the tightest size and the longest hold.
THE EFFECT Because the frame excludes everything but the face, we read the small flicker of feeling as the taker
understands what the keys mean. Size makes the moment.
THE LESSON Save the close-up for the beat that matters and hold it a touch longer than the surrounding rhythm.
Withheld and then spent, the close-up carries the scene's meaning.
What was adjusted: the taker's eyes drifted to the camera in the first take (non-actors do this). Fix — the giver crouched just off the lens on the correct eyeline so the taker had a real person to look at. Instantly better.
Phase 5 — the insert (the escape hatch)
The thinking: grab one detail shot — an extreme close-up of the keys sliding across the table into a hand. This is your cutaway: the piece that lets you compress time, hide a trim, or smooth any awkward join. It is the shot beginners forget and editors beg for.
FIGURE CS2.6 — "The keys" (the insert) [illustrative shoot]
THE FRAME Extreme close-up: the keys on the wood, a hand entering to slide them, the other hand receiving.
Shallow focus, the table filling the frame; no faces.
THE MOVE Locked off, or a tiny reframe to follow the slide.
THE LIGHT The same window light skimming the table surface, catching the metal of the keys.
THE SOUND The scrape of keys on wood, close and specific -- a sound effect that sells the handoff.
THE CUT Drop this over any point in the exchange to compress time or to bridge two shots that wouldn't
otherwise cut; it also gives the "keys" their own moment of importance.
THE EFFECT A concrete, tactile detail that both advances the story (the handoff) and hands the editor total
freedom to trim around it.
THE LESSON Always grab the insert. It is the cheapest shot on set and the most useful in the timeline -- the
difference between a cut you can make and one you can't. (B-roll and inserts get their own chapter, 20.)
What was adjusted: shot it twice — once wide enough to see both hands, once tight on just the keys — because an insert with options is even more useful. Two seconds of work; endless value in the edit.
The edit pass: proving the pieces cut
Six setups, shot in half an hour, all on one side of the line. Now assemble them and watch the coverage pay off. Here is the timeline and then the assembled scene.
FIGURE CS2.7 — The assembled scene: the timeline (video over audio)
V1 [ MASTER wide ][ OTS giver ][ OTS taker ][ CU taker ][ insert:keys ][ MASTER wide ]
A1 [ giver + taker sync dialogue ........................................................ ]
A2 [ kitchen room tone / ambience ....................................................... ]
^establish ^the exchange, cut on the volley ^the beat ^hide trim ^re-establish/exit
The dialogue (A1) runs continuously underneath while the picture (V1) cuts between sizes -- so the words
carry across the cuts and the scene feels continuous even though the picture is changing. The insert (keys)
sits over the join where we compress the pause before the taker's reply. We open and close on the master.
And the assembled cut, as a Described Sequence — the proof that the pieces fit:
FIGURE CS2.8 — The kitchen handoff, covered and cut [illustrative shoot]
# | Shot (size / subject) | Dur | Audio | Cut to next
---+-----------------------------------+------+---------------------------+-------------------------
1 | MASTER wide: both at the table | 3 s | room tone; "Here -- take | cut IN to the giver
| (giver left, taker right) | | these." (giver) |
2 | OTS/MS favor GIVER (looks right) | 3 s | "...you've earned it." | cut to the reverse
3 | OTS/MS favor TAKER (looks left) | 2 s | "You're sure?" | cut to the insert
4 | ECU insert: keys slide to a hand | 2 s | scrape of keys on wood | cut to the reaction
5 | CU TAKER (looks left) | 3 s | a breath; "...thanks." | cut back wide
6 | MASTER wide again: taker pockets | 2 s | room tone returns | out (end of scene)
| the keys | | |
Read it as an editor. Shot 1 establishes the space and the two sides of the line. Shots 2 and 3 are the shot-reverse-shot — giver looking screen-right, taker screen-left, matched eyelines, so it reads as one exchange. Shot 4, the insert, both advances the story (the keys move) and lets us compress the beat before the reply. Shot 5 saves the close-up for the emotional peak — the taker's reaction — held a beat longer. Shot 6 returns to the master to re-establish and exit. Fifteen seconds, four sizes, one side of the line, and it reads.
Now the real test of coverage: cut it a second way. Open on the insert (the keys, mysterious, before we know the context), then the master, then the reverse-shot, ending on the giver instead of the taker — a version that's about the giver's generosity rather than the taker's gratitude. You can do this only because you covered it. Same six pieces, a different scene. That freedom is what this whole chapter was for.
The four cut decisions that made it work
An assembly is a sequence of small decisions; naming the four that mattered here teaches you to make them on your own footage:
- Where to open. We opened on the master, not a single, because the audience needed the geography first — who is where, the two sides of the line. Open on a close-up and you'd have a floating face with no context. (The second cut breaks this rule on purpose, opening on the mysterious insert — proof that once you know the rule, you can bend it for effect.)
- When to leave the wide. We held the master only long enough to establish and hear the first line, then cut in on the giver. Staying wide any longer would go flat; cutting in sooner would skip the grounding. The wide's job is done the instant we understand the space.
- Where to hide the compression. Real life has a pause between the offer and the reply; the finished scene doesn't want it. We dropped the insert (the keys) over that pause, so the cut compresses real time invisibly — the viewer's eye is on the keys, not on the missing seconds. That is the single most useful trick an insert performs.
- Whose face ends the scene. We ended on the taker's reaction, making the scene about receiving the trust. Ending on the giver (the second cut) makes it about offering it. Same footage, opposite meaning — decided entirely by which reaction you hold last.
🎒 Gear Note: what a second camera would (and wouldn't) buy you. A pro shooting this with two cameras at once would get the giver's and taker's coverage in a single performance, which helps with non-actors (you capture the one good, spontaneous take from two angles instead of asking them to repeat it). But notice what a second camera does not change: you still have to respect one line, still frame matching eyelines, still get the insert separately, still lock exposure so the two cameras match. The grammar is identical whether you shoot one camera in six passes or two cameras in three. This is why a phone-only reader is not shooting a lesser version of this scene — just a more patient one. Coverage is a method, not a gear tier.
✂️ In the Edit: what the coverage bought. Everything you got to choose in the edit — where to open, whose reaction to end on, how much to compress, whether to hide the pause under the keys — was a choice you created on set by shooting the pieces. Under-cover this scene (shoot only the master) and you'd have exactly one flat version, un-trimmable without visible jumps. The half hour of extra passes is what turned the edit from a place of constraints into a place of options. This is you shoot for the edit, made completely concrete.
What went right, and the honest mistakes
Three things carried the shoot: locking exposure and white balance so every size matched; finding the line first and never crossing it (which made the reverse "just work"); and running the full exchange on every pass so the action overlapped and any cut was possible.
Three predictable mistakes, and their fixes, so you can skip them:
- Non-actors looking at the lens. Fix: put a real person on the correct eyeline just off-camera for every single and close-up. People look natural when they have someone real to look at.
- The light shifting between passes. A cloud moved and the window dimmed mid-shoot, so an early close-up didn't match a later one. Fix: shoot your coverage in one continuous window of time, and if the light changes, re-shoot the odd one out. (Or, next time, add a bounce to stabilize it — Chapter 13.)
- Forgetting the room tone. The first assembly had a dead, silent gap under a cut. Fix: always record thirty seconds of the room doing nothing, so you have a continuous ambience bed to lay under the whole scene (Chapter 15 makes this a law).
Discussion questions
- The scene's content (handing over keys) is trivial, yet the shoot is a complete coverage exercise. Why does the triviality of the content make it a better teaching scene, not a worse one?
- The reverse shot (CS2.4) "just worked" with no adjustment. What earlier decision made that possible, and what would have happened if it had been ignored?
- The insert (the keys) is called "the escape hatch." Describe a specific edit problem it solves that no other shot in the coverage could.
- You were asked to cut the scene a second way, ending on the giver instead of the taker. How does that single change alter the meaning of the scene — and what does that tell you about the power of coverage?
- The Settings Box insists you lock exposure and white balance before you start. Trace exactly what goes wrong in the edit if you don't.
- Compare this shoot-along to the famous scene in Case Study 1. What did the professional production have that yours didn't — and what did the two share completely?
Your turn: cover, then complicate
The core assignment: run this exact shoot with two friends and any object (keys, a book, a mug, an envelope). Six setups: master, OTS each way, a close-up, an insert. One side of the line. Cut it. Then cut it a second, different way. Confirm the eyelines match and the scene reads.
Then extend it, one step at a time:
- Add a third size. Shoot one person in a wide-to-close range — a medium and a tight close-up — and let your cut tighten as the moment intensifies, the way Case Study 1's coverage escalated.
- Add a reaction you didn't plan. Shoot an extra close-up of one person reacting to something not said — a glance away, a small doubt. See if it gives your edit a new option.
- Break the line on purpose, once. Deliberately shoot one reverse from the wrong side. Cut it in. Watch it break. Now you'll never do it by accident — and you're ready for Chapter 9, where crossing the line becomes a tool you can control.
Key takeaways
- Coverage is a method, not a talent: master first, then tighter sizes, each pass running the full action so everything overlaps and any cut is possible.
- Find the line and pick a side before you roll. Respect it absolutely, and the reverse shot — the hardest-seeming part — becomes almost automatic.
- Lock exposure and white balance so every size matches; matching shots you never mismatched is free.
- Always grab the insert. It is the cheapest shot on set and the most useful in the edit — the difference between a cut you can make and one you can't.
- Save the close-up for the beat that matters and hold it a touch longer than the surrounding rhythm.
- The proof of coverage is a second cut: if you can build two different-feeling scenes from the same pieces, you shot for the edit — and that freedom was created on set, for free, with one phone.