Case Study 2: One Light to Three — Building a Talking-Head, Live
This is a production walkthrough: a from-scratch lighting shoot in one of the book's four recurring setups — the talking head at a table. We build the look one light at a time, showing the thinking at each stage, a Described Shot of the result, and what we adjusted before moving on. Everything here is doable with a window, one or two lamps, and a white board; a version of it works on a phone. Follow along by lighting a real subject as you read — this is the exact process behind your Production Checkpoint.
The brief and the constraints
A local ceramics studio wants a 60-second talking-head for its website: the owner, to camera, explaining why they make everything by hand. Warm, trustworthy, human. No agency, no budget, one afternoon.
Here are the constraints, which are probably close to yours:
- The room: the back of the studio — one big window on a side wall, a plain wall behind where the subject will sit, some clutter, a few work lamps and a floor lamp we can move.
- The gear: one camera on a tripod (a phone would do), one white foamboard, a floor lamp and a desk lamp we can borrow, and a couple of clamps. No "real" film lights.
- The subject: the owner — not an actor, a little nervous, more comfortable at their wheel than on camera.
- The clock: we shoot late afternoon so the window is soft, not blasting direct sun. About two hours, start to wrap.
- The goal: a look that says craftsperson, warm, real — soft key, gentle contrast, clear separation, and every source motivated by the room.
The plan is to build the light in four passes — key, fill, rim, motivate — checking the shot after each so we always know exactly what each addition bought us. That is the whole discipline of this chapter, run live.
⚙️ Settings Box — the talking-head lighting build (starting points). Camera exposure (ISO, aperture, ND, judged on the waveform/zebras) comes from Chapter 5; this box is the lighting plan we adjust to the room and the face.
Pass Element Starting point We'll adjust by… 1 Key window, 45° off the lens axis, subject turned into it, raised feel (window is tall) swinging the subject relative to the window 1 Subject-to-window ~4 ft closer = softer + darker background; farther = flatter 2 Fill white foamboard opposite the window, ~2 stops under key slide closer to fill more, back to fill less 3 Rim / back floor lamp behind + opposite the key, high, just enough to edge the hair dim/flag until it only defines the edge 4 Practical desk lamp in the background, in frame, warm position for a pool of light behind the subject all Overhead room light OFF kill it so the window is the only direction all Background let it fall darker than the subject; add a practical pool so the subject separates and the room has depth
The room, before we touch anything
First we frame the shot, because the light has to serve a composition (Chapter 6), not the other way around. We put the subject at a worktable, camera at a ~50mm-equivalent feel, subject on the right third, looking camera-left to an off-lens interviewer, with correct headroom and lead room.
FIGURE CS2.1 — Framing overlay before lighting (16:9, rule-of-thirds)
+---------------------------------------------+
| O O | O = thirds intersections
| |
| ( subject's head ) | head near upper-right third
| O ( eyes on upper O ) | eyes on the upper third line
| ( shoulders ) |
| | lead room open to camera-left
| [ open lead room ] [ subject R ] | (they look/speak into it)
+---------------------------------------------+
Subject on the right third, eyes on the upper third, looking into open lead room
camera-left. Now we light the face — and the lit side should be the lead-room side.
A note that will matter: the subject looks camera-left, into the open lead room. So the lit side of their face should be the camera-left side too — we key from the side they're turned toward, so the light and the composition agree instead of fighting (a flat-lit or wrong-side-keyed face would put the bright cheek away from the lead room and feel off). Conveniently, the window is on that side. Good rooms make you lucky; reading the room is how you get lucky on purpose.
Now, four passes.
Pass 1 — One light: the window as key
We turn off the overhead (its top-down spill is the enemy), and we turn the subject to face roughly 45 degrees into the window. That is the entire first pass: one soft source, placed like daylight, doing both jobs — exposure and direction.
FIGURE CS2.2 — Pass 1: window key only (top-down view)
WINDOW ☀ (soft KEY)
│
│ ↘ subject turned ~45° into the window
▼
( S ) ───────► eyeline to interviewer, just past the lens (camera-left)
│
│ (shadow side = camera-right, currently falling dark)
[CAM] ~50mm-equiv, subject on the right third; OVERHEAD OFF
One source, both jobs. The camera-left cheek is keyed (and it's the lead-room
side — light and framing agree). The camera-right cheek falls into shadow.
We roll ten seconds and look hard at the result.
FIGURE CS2.3 — "Pass 1: keyed, but stuck to the wall" [constructed teaching example]
THE FRAME Medium close-up, subject on the right third, looking camera-left. The window-lit cheek
(camera-left) is bright and soft; the far cheek (camera-right) falls into shadow.
THE MOVE Locked off on the tripod.
THE LIGHT Soft window key from camera-left, ~45°. A clean catchlight high in each eye. Good modeling —
the face has real shape. BUT: no fill, so the shadow side is quite deep (~3+ stops under),
and the plain wall behind is nearly as dark as the subject's dark side, so the shoulder and
hair on the shadow side start to merge with the background.
THE SOUND (audio comes in Ch.14–15; for now, room tone.)
THE CUT Will intercut with a wider shot — so this contrast must be one we can match.
THE EFFECT Already miles better than the overhead-lit "before": there is a face here, with dimension.
But it reads a touch severe for a warm craftsperson, and the subject feels pasted to the wall.
THE LESSON One good key is 80% of the job — shape and exposure in a single move. What's missing is
named and fixable: the shadow is too deep for the mood (fill), and there's no separation (rim).
What we see, and what we'll adjust. The key is working — good shape, good catchlight, and it's on the correct (lead-room) side. Two problems, both from the chapter: the contrast is too high for "warm and trustworthy" (we want ~2 stops, we're at ~3+), and there's no separation from the wall. Pass 2 fixes the contrast; Pass 3 fixes the separation. Notice we are not starting over — we are adding, because we built in the right order.
A phone-first checkpoint: if all you have is a phone and a window, you could stop here and be fine — a window key with the overhead off already beats most talking-heads on the internet. Everything below is refinement, and the first refinement (a free bounce) is the one that matters most.
Pass 2 — Add fill: dial the contrast to the mood
We stand a white foamboard on the shadow side (camera-right), just out of frame, angled to catch the window's spill and throw it back into the shadowed cheek. Then we tune it by moving it: slide it in until the shadow lifts to about two stops under the key — modeled, but not severe.
FIGURE CS2.4 — Pass 2: window key + bounce fill (top-down view)
WINDOW ☀ (soft KEY)
│
│ ↘
▼
( S ) ───────► eyeline camera-left
│ ╲
│ ○ white foamboard (FILL), camera-right, angled to bounce
│ the window's spill back into the shadow cheek (~2 stops under)
[CAM]
The bounce is just the window's own light returned — it can never overpower the
key. Slide it closer to fill more (flatter), back to fill less (moodier).
FIGURE CS2.5 — "Pass 2: the shadow opens, the mood warms" [constructed teaching example]
THE FRAME Same medium close-up; now the camera-right cheek carries gentle detail instead of falling
to black.
THE MOVE Locked off.
THE LIGHT Window key camera-left; a soft bounce fill camera-right brings the shadow to ~2 stops under
the key. Clear modeling, full detail across the whole face — the standard flattering interview
contrast. The face now reads warm and open rather than severe.
THE SOUND Room tone.
THE CUT This ~2-stop contrast is now our LOCKED ratio; every shot of this subject — wide, medium,
close — must match it, or the intercut will flicker in mood.
THE EFFECT The subject looks approachable and human — right for a craftsperson. The bounce did in ten
seconds and zero dollars what a second light would do, and it can't fight the key.
THE LESSON Fill is a dial, not a switch. You set the *mood* by choosing how many stops under the key the
shadow sits — and a moved bounce card is a precise, free way to set it.
What we see, and what we'll adjust. The face is right now. But the subject is still slightly stuck to the wall — the hair and shoulder on the shadow side don't clearly separate from the background. That is Pass 3's entire job. We also make a decision here we'll honor in the edit: we lock the contrast. From now on, we do not move that bounce, because every angle we shoot has to match.
Why we didn't just add a second lamp for fill: we could have. But a bare lamp pointed at the shadow side is a hard source that throws its own shadow — a second, competing set of shadows that muddies the shape the key built. The bounce is soft and weak by nature; it lifts the shadow without announcing itself. When a beginner's "fill light" creates a second nose-shadow, this is why. Bounce first; add a lamp only when you can soften and control it (Chapter 12).
Pass 3 — Add the rim: lift the subject off the wall
Now the separation. We take the borrowed floor lamp, raise it on a stand behind the subject and opposite the key (so behind-camera-right), and aim it to skim the top and side of the hair and the shoulder — pointing back toward the camera. Then we bring it down until it only just draws a bright edge, and we flag its spill so nothing sprays into the lens.
FIGURE CS2.6 — Pass 3: key + fill + rim (top-down view) — three-point
▣ rim/back light (floor lamp on a stand, high,
╲ behind + opposite the key; skims hair + shoulder,
╲ pointed back toward camera; flagged off the lens)
↙
WINDOW ☀ ( S ) ───────► eyeline camera-left
│ ↘ ╱ ╲
▼ ╱ ○ bounce FILL (camera-right, ~2 stops under)
(key) ──────╱
[CAM]
Three jobs, three sources: window KEY (shape), bounce FILL (mood),
floor-lamp RIM (separation). The subject now sits IN the room, not on the wall.
FIGURE CS2.7 — "Pass 3: separation — the subject lifts off" [constructed teaching example]
THE FRAME Same framing; now a thin bright line rides the top of the hair and the shadow-side shoulder,
peeling the subject away from the dark wall.
THE MOVE Locked off.
THE LIGHT Window key (camera-left) + bounce fill (camera-right, ~2 stops under) + a subtle rim from
behind camera-right, edging the hair. The rim is gentle — a definition, not a halo — and no
light flares the lens. Three-point, complete.
THE SOUND Room tone.
THE CUT The rim reads because the wall behind is darker than the bright edge — separation depends on
that difference, so we keep the background from getting too bright.
THE EFFECT Depth. The subject now clearly sits *in front of* the wall, in a real space with air around
them. The shot reads "produced" — and the whole difference from Pass 2 is one dim edge light.
THE LESSON A rim is the cheapest production value there is: a little light on the edge, against a darker
background, turns a flat shot into a dimensional one.
What we see, and what we'll adjust. That's the technical grammar done: shape, mood, depth. One thing still missing for this brief — the background is clean but a little dead, and nothing yet motivates the warm, crafty feeling we promised. Pass 4 is where the look stops being "correct three-point" and starts being this studio.
The common failure right here: the temptation is to crank the rim until the hair blazes. We resist it. We turn the rim off, then on — the right amount is the amount we miss when it's gone but never consciously noticed when it's there. A rim you notice is a rim that's too hot.
Pass 4 — Motivate it: give the room a reason and a soul
The final pass isn't a new light so much as a new idea: make the whole setup feel like a real, warm workspace by adding a practical into the background and letting it justify the mood. We place the borrowed desk lamp on a shelf behind the subject, in frame, switched on — a warm pool of light in the corner of the background. It does three things at once: it motivates the warm feeling (now there's a visible reason the room glows), it adds a pool of depth behind the subject, and it can throw a gentle warm kick that helps the rim.
FIGURE CS2.8 — Pass 4: three-point + a motivating practical (top-down view)
▣ rim ▽ practical (warm desk lamp
╲ ╱ on a shelf, IN frame, glowing
╲ ╱ in the background corner)
↙ ╱
WINDOW ☀ ( S ) ───────► eyeline camera-left
│ ↘ ╱ ╲
▼ ╱ ○ bounce FILL
(key) ──────╱
[CAM]
The practical (▽) is visible in the shot. It motivates the warm mood, adds a
background pool of light, and makes the whole frame read as a real workspace.
FIGURE CS2.9 — "Pass 4: a lit face in a real place" [constructed teaching example]
THE FRAME Same subject, but now a soft warm glow sits in the background behind their shadow side — a
lamp we can see, pooling amber onto the wall and a shelf of finished pots.
THE MOVE Locked off. (In the edit we may add the Ch.8 slow push-in; the light is built to hold up to it.)
THE LIGHT Cool-ish soft window KEY on the face + ~2-stop bounce FILL + a subtle RIM + a warm PRACTICAL
glowing in the background. The cool key against the warm practical pool gives the frame life
and depth; every source has a reason in the room. (Color balance of that cool/warm mix: Ch.12.)
THE SOUND Room tone.
THE CUT A rich, motivated frame that will hold a viewer and cut cleanly against its own coverage.
THE EFFECT The subject looks like a warm, real craftsperson in their own warm, real studio — exactly the
brief. Nobody can see a single "film light," because there wasn't one.
THE LESSON Motivation is the finish. Three-point makes a face correct; a motivated practical makes the
correct face believable — a person in a place, not a subject on a set.
We shoot the interview at this look, and — crucially — we shoot our coverage (Chapter 7) at the same locked contrast: the medium we've been building, plus a wider shot and a tighter close-up, never touching the lights between them. That discipline is what lets the edit work.
The edit pass
Now the payoff. In post (we're previewing Chapter 26's assembly), we cut the 60-second piece, and every lighting decision either helps us or would have hurt us.
FIGURE CS2.10 — The talking-head assembly (edit timeline)
V1 [ MED: intro ###### ][ CU: the key line ######### ][ WIDE: sign-off ###### ]
A1 [ interview sync audio ############################################## ]
A2 [ room tone / soft music bed ................................................ ]
^ cut MED→CU on a natural pause ^ cut CU→WIDE to breathe at the end
Because every angle was lit at the SAME locked ~2-stop contrast, the cuts between
MED, CU, and WIDE hold the same mood — no flicker. The rim and practical keep the
subject dimensional in every size. Light matched on set = an edit that just works.
The key cut decisions, and how the light earned them:
- Intercutting sizes without a mood flicker. We cut medium → close-up → wide, and the subject's mood holds steady across all three, because we locked the ~2-stop contrast and never moved the bounce. Had we drifted the fill between setups, every cut would jolt softer/harder and the audience would feel an unease they couldn't name. Locking contrast on set is what bought a clean intercut in post.
- The close-up carries the emotional line. We saved the tightest shot for the single most heartfelt sentence. Because the light is dimensional (rim + modeling), the close-up reads intimate and rich, not flat — the light does emotional work the words can only hint at.
- The lit-vs-unlit proof. For the client (and for ourselves), we keep the ugly Pass-0 overhead version beside the final. Side by side, the upgrade is undeniable, and it cost nothing but knowledge.
- The push-in holds up. If we add the slow motivated push-in from Chapter 8, the built light — motivated, dimensional, with a background pool — stays believable as the frame tightens, because nothing in it depended on hiding a bad source.
That is the whole method: build in order (key, fill, rim, motivate), check after each pass, lock the contrast, cover the scene at that ratio, and the edit falls together. You shot for the edit, and here you lit for it too.
A wrinkle we hit (and how we handled it)
No real shoot goes clean, and ours didn't. Halfway through covering the interview, a bank of cloud slid across the sun and our window key — our whole key — dropped by more than a stop. Suddenly the medium and the close-up we still needed would not match the wide we'd already shot. This is the price of a free key: available light is generous but unstable, and it changes while you work (a problem Chapter 13 tackles head-on).
We had three options, and it's worth naming all of them because you'll face this exact moment. One: wait for the cloud to pass and shoot in the same light — cheapest, but you're at the sky's mercy and the light may never return identically. Two: ride the change with exposure — open the aperture or lift ISO (Chapter 5) to hold the brightness — but that only fixes exposure, not the softer, flatter quality the cloud brought, so the shots still wouldn't quite match. Three: stop relying on the sky and take control — which is what we did. We brought the borrowed floor lamp around, bounced it off the same wall the window lit from, and used it to supplement the window so our key stayed roughly constant whether the sun was in or out. It cost us ten minutes and a lamp, and it bought us coverage that actually intercut.
The lesson is the chapter's, learned the hard way: a window is a magnificent key and a fickle one. The moment your look depends on light you don't control, you are one cloud away from an unmatched edit. Knowing why the shots wouldn't match — the change in both brightness (stops) and quality (soft/hard) — is exactly what let us fix it in the right way instead of flailing. Read the room; then be ready for the room to change.
Discussion questions
- We built the light in four passes and checked after each. Why is that better than setting up all the lights at once and adjusting the whole pile? What does the ordered build teach you that a finished setup can't?
- At the end of Pass 1 we noted we "could stop here and be fine." When is one light genuinely enough, and what specifically do the later passes buy that a solo creator on a deadline might reasonably skip?
- We filled with a bounce, not a lamp. Give the two reasons a bounce is often the better fill, and name the situation where you'd want a real fill light instead.
- We "locked the contrast" before shooting coverage. Walk through what would go wrong in the edit if we hadn't. Which chapter's craft (coverage) does this protect?
- Pass 4 added nothing to the face — only a practical in the background. Why did that pass arguably matter most for this brief? What did motivation add that correct three-point didn't?
- Compare this build to Case Study 1's Godfather look. Both are "correct," yet opposite. What decisions (key position, fill, motivation) turn the same grammar into warmth here and menace there?
Your turn: your own four-pass build
Light your Project 1 talking-head with the exact method above, documenting each pass.
The brief: a warm, trustworthy 60-second talking-head of one subject, built one light at a time, every source motivated.
Do this:
- Frame first (Ch.6): subject on a third, correct headroom and lead room, keyed from the lead-room side.
- Pass 1 — key: overhead off, window or one soft lamp at 45° and up, catchlight in the eyes. Shoot ten seconds. Note the contrast and the separation problem.
- Pass 2 — fill: add a white bounce, tune to ~2 stops under. Shoot. Lock the bounce.
- Pass 3 — rim: add a subtle edge light behind, opposite the key; flag it off the lens. Shoot.
- Pass 4 — motivate: put a practical in the background and let it justify the mood. Shoot.
- Cover it at the locked contrast: a wide, a medium, a close-up.
- Edit: cut a 30–60-second piece intercutting the sizes. Confirm the mood holds across the cuts.
Keep all four passes as separate clips. Watching them in order is the most convincing lighting lesson you can give yourself — you will see shape, then mood, then depth, then soul arrive, one decision at a time.
Key takeaways
- Build the light in order: key, fill, rim, motivate. Check after each pass so you always know exactly what each addition bought — shape, then mood, then depth, then believability.
- One good key is most of the job. A window at 45° with the overhead off, catchlight in the eyes, already beats most talking-heads. Everything after is refinement.
- Fill with a bounce, and tune it by moving it. ~2 stops under the key is the warm, flattering default; a bounce lifts the shadow without fighting the key or throwing a second shadow.
- A subtle rim is the cheapest production value there is — one dim edge light against a darker background turns a flat shot dimensional. Never let it blaze or flare.
- Motivation is the finish. A practical in the background gives the correct three-point a reason and a soul — a person in a real place, not a subject on a set.
- Lock your contrast, then cover the scene at that ratio. Matched light on set is what makes the intercut hold in the edit. You light for the edit.
- The entire build here used a window, two borrowed lamps, and a white board. The skill was in the order and the placement, and both are free.