← 14 Day Smartphone FilmmakerAll programsHomeSearch
14 Day Smartphone Filmmaker·Day 6: Lighting + Day 7: Audio·2:26:00

Days 6-7: Direction, Intensity and Quality of Light; Three-Point Setups on a $150 Budget; and the Seven Steps to Audio That Does Not Lose the Viewer

Anthony Gallo Instructor - teaches all ten lessons, builds three lighting setups live and walks a real client spec ad through sound design · Courtney Sweeney On-camera subject for several lighting demos and the lead in the Merrell hiking-boot spec ad used as the sound-design case study · Shannon Model for the three-point lighting demonstration

The short version

  1. Lighting reduces to three variables you can actually control, and Anthony calls it 'the number one differentiator between amateur and skilled content creators'. Direction: put the key 'in front, above, into the side' of the subject, because light straight-on flattens the face and kills depth. Intensity: judged against the background, not in isolation. Quality: hard or soft, and almost every light is hard out of the box until you put something in front of it.
  2. Three-point lighting is 'the bread and butter of commercial filmmaking' - key for shape, fill opposite to soften without erasing the shadow, backlight to peel the subject off the background. The positioning check is a visible Rembrandt triangle on the shadow cheek. His own taste runs to less fill: 'I typically err on the side of more shadow... more depth, which means more cinematic quality.' Many commercial setups are really two lights plus a bounce.
  3. Mixed colour temperature is the amateur tell. Window light at about 5,600K fighting warm household bulbs at about 3,200K 'doesn't look professional at all'. Either kill one source or lean into the clash deliberately - a daylight key against warm practicals is the orange-and-teal look - but do not leave it to chance.
  4. Softness is size. A softbox, a silk, a five-in-one reflector or a cloud all work the same way: put a second medium in the path and make the source effectively bigger. Moving a diffused light further back makes it softer still. A honeycomb grid does the opposite job, narrowing spill so the light lands on the subject and not the whole room.
  5. Audio has one dominant variable and it is not the microphone. 'Audio is half of the viewing experience', and the mic wants to be six to twelve inches from the mouth. After proximity comes the room - hardwood, high ceilings and bare walls put echo into everything, and a hung blanket beats a better mic.
  6. Levels are a safety margin: peak between about -6 and -12dB. Clipping past 0dB 'is going to sound like a distorted guitar' and is as unrecoverable as a blown highlight, whereas quiet audio can always be raised later. Record a test clip and listen on headphones before the real take.
  7. The connector problem catches everyone: professional gear is TRS, phones and computers are TRRS, and the fix is a two-dollar adapter in whichever direction you need.
  8. Sound design sorts into three buckets - the physical action you can see, the ambience that sets the place, and 'perceived movement' where the brain expects a sound that nothing actually made, like a whoosh on a fast camera move. Build a named library so you find effects instead of re-sourcing them. And decide how each one enters: a J-cut lets the sound arrive before the picture and pulls the viewer forward; a hard cut lands it on the frame for impact.

At a glance, three clicks deep

Skim here first: the closed row is the glance, open is the study card with the key points and timestamps, and the ↓ link drops to that concept's full write-up below.

01Direction, intensity, quality - the three things you controlDirection (front, above, to the side) + intensity (relative to background) + quality (hard vs soft, i.e.

Direction (front, above, to the side) + intensity (relative to background) + quality (hard vs soft, i.e. diffused or not).

Key light in front, above, into the side (p2154654879 0:02)

Flat or broad lighting removes shadow and kills depth (p2154654879 0:02)

Intensity is judged against the background, not alone (p2154654879 0:03)

Nearly every light is hard until diffused (p2154654879 0:05)

Midday sun is avoided because all three go wrong at once (p2154654879 0:06)

↓ Full write-up of this concept

02Key, fill, back - and why the back light is the one to buy secondKey (shape) + fill (soften, often bounced) + back (separation);

Key (shape) + fill (soften, often bounced) + back (separation); check with the Rembrandt triangle.

Key in front, above, to the side; too high gives raccoon eyes (p2154654880 0:02)

The Rembrandt triangle is the positioning check (p2154654880 0:03)

Fill softens, never erases - erasing gives flat broad light (p2154654880 0:06)

Fill can be a bounce off a wall, sheet or panel (p2154654880 0:08)

Backlight separates subject from background, same side as key (p2154654880 0:09)

Many commercial setups are two lights plus a bounce (p2154654880 0:12)

↓ Full write-up of this concept

03Mixed colour temperature is the amateur tellEliminate one source, or contrast them on purpose (daylight key, warm practicals);

Eliminate one source, or contrast them on purpose (daylight key, warm practicals); set white balance to the dominant source or deliberately between them.

Window light and household bulbs are different temperatures and fight (p2174627733 0:03)

Simplest fix: turn the conflicting source off (p2174627733 0:03)

Deliberate mix: 5,600K key against warm practicals = orange and teal (p2174627733 0:16)

Overpower one and the scene looks like 'living inside a blueberry' (p2174627733 0:16)

Mixed indoor/outdoor may want an off-preset value, e.g. ~3,450K (p2154303671 0:07)

↓ Full write-up of this concept

04Softness is size - diffuse it, or move it backDiffusion enlarges the apparent source;

Diffusion enlarges the apparent source; larger and closer = softer; a grid narrows spill instead.

Soft rolls off and hides imperfections; hard emphasises them (p2154654879 0:05)

A softbox or panel spreads the rays and enlarges the source (p2154654879 0:06)

Cloud cover or a handheld panel is the outdoor equivalent (p2154654879 0:07)

A honeycomb grid narrows spill onto the subject only (p2174627733 0:12)

Moving a diffused light back increases effective size and softness (p2154653334 0:08)

↓ Full write-up of this concept

05Seven steps to audio, in the order that actually mattersCare -> right mic -> 6-12in -> kill ambient noise -> treat reverb -> peak -6 to -12dB -> test clip on headp…

Care -> right mic -> 6-12in -> kill ambient noise -> treat reverb -> peak -6 to -12dB -> test clip on headphones.

Four mic types: lavalier, onboard, shotgun, condenser (p2155017343 0:02)

A mic and a recorder are not always the same device (p2155017343 0:01)

Lav is the default for flexibility (p2155017343 0:04)

Six to twelve inches from the source (p2155017343 0:04)

Hard surfaces add reverb; foam, rugs, artwork or a hung blanket fix it (p2155017343 0:07-0:08)

De-reverb in post supplements treatment, never replaces it (p2155017343 0:08)

↓ Full write-up of this concept

06Peak at -6 to -12dB, because clipping cannot be undonePeak -6 to -12dB;

Peak -6 to -12dB; never touch 0dB; verify with a headphone test clip before the real take.

Target peaks between about -6 and -12dB (p2155017343 0:09)

Clipping past 0dB is permanent; quiet audio can be boosted (p2155017343 0:10)

Gain looks different per device: dial, scale, or slider (p2155017343 0:10)

Most phones auto-level; Filmic Pro gives manual control (p2156411441 0:02)

Test clip on headphones - some mics clip on hard consonants (p2156411441 0:04)

↓ Full write-up of this concept

07TRS versus TRRS, and going wirelessTRS = pro gear;

TRS = pro gear; TRRS = phones and computers; adapters bridge either way; wireless removes the cord and the post-sync.

TRS for professional gear, TRRS for phones and computers (p2156411441 0:09)

Cheap adapters convert either direction (p2156411441 0:09)

PoP Voice Pro ships TRRS and needs an adapter for a camera (p2156411441 0:10)

Wireless solves mobility, not quality (p2156411441 0:07)

A wireless receiver bakes audio into the recording - no post-sync (p2156411441 0:08)

↓ Full write-up of this concept

08Sound design: three kinds of effect, and how each one entersPhysical action + ambience + perceived movement;

Physical action + ambience + perceived movement; a named library; J-cut to lead, hard cut to hit.

Physical action sounds match something visible (p2154994798 0:02)

Ambience establishes place without being seen (p2154994798 0:03)

Perceived movement adds weight to motion that makes no sound (p2154994798 0:03)

Foley substitution is fine if it matches emotionally (p2154994798 0:18)

J-cut leads the picture; hard cut lands on the frame (p2154994798 0:27)

Keep a named, searchable library (p2154994798 0:06)

↓ Full write-up of this concept

09Three setups: a window, one light, or a roomWindow light -> one diffused light plus a bounce and practicals -> keyed studio with grid, accents and XLR…

Window light -> one diffused light plus a bounce and practicals -> keyed studio with grid, accents and XLR audio.

No budget: phone, ~$30 tripod, window or door light, composition for depth (p2154303671)

Under $150: one YN300 II at about $70, diffusion, $20 reflector, practicals (p2154653334 0:04-0:11)

Studio: Godox SL60, 85in softbox, honeycomb grid, accent lights (p2174627836 0:05-0:07)

Audio scales with it: lav into phone, then Zoom H1n, then NTG3 over XLR (p2154653334 0:12; p2174627836 0:01)

Each tier buys control, not legitimacy (p2174627733 0:00-0:16)

↓ Full write-up of this concept

The concepts in full

01

Direction, intensity, quality - the three things you control

'In front, above, into the side of their subject.'

Once white balance is handled, everything else about a light is direction, intensity or quality. Direction decides whether a face has shape: light straight on is flat or broad lighting, which removes the shadows that carry the sense of three dimensions, so the key goes in front, above and off to one side. Intensity is a relationship rather than a number - what matters is the subject against the background, not how bright the subject is on its own. Quality is hard or soft, and almost every source is hard 'straight out of the box' until something diffuses it. Midday sun is the worst case because all three are wrong at once: overhead direction, brutal intensity, undiffused quality.

Why it matters

Three variables replace an unbounded amount of gear anxiety.

Light and sound: the three light variables, and the audio chain in the order that decides quality LIGHT DIRECTION front, above, to the side INTENSITY vs the background QUALITY hard until diffused THREE-POINT key / fill / back Check: a Rembrandt triangle on the shadow cheek. Fill softens the shadow. It never erases it. SOUND 1 PROXIMITY - 6-12 inches 2 THE ROOM - rug, panels, blanket 3 LEVELS - peak -6 to -12dB 4 TEST CLIP - on headphones Clipping past 0dB is as unrecoverable as a blown highlight. Quiet is always fixable. Mixed colour temperature is the amateur tell - kill one source or contrast them on purpose
Left: key placement and the three-point setup. Right: the audio chain in the order that actually decides quality - proximity, room, levels.
Lighting is the number one differentiator between amateur and skilled content creators.
02

Key, fill, back - and why the back light is the one to buy second

Look for the Rembrandt triangle on the shadow cheek. That is the key light in the right place.

The three-point setup is 'the bread and butter of commercial filmmaking'. The key does the work and makes the shape; too high and you get raccoon eyes, so the check is a small triangle of light on the shadow side of the face. Fill sits opposite and softens the shadow rather than erasing it - erase it and you are back to flat broad lighting - and it does not have to be a light at all, since bouncing the key off a wall, a bedsheet or a panel does the same job. The backlight, ideally on the same side as the key, separates the subject from the background and is the highest-value third element. His own preference is deliberate: 'I typically err on the side of more shadow... more depth, which means more cinematic quality.' Many real commercial setups are two lights and a bounce.

Why it matters

One setup that covers interviews, talking heads and product, and scales down to one light plus a bedsheet.

I typically err on the side of more shadow, which means more depth, which means more cinematic quality.
03

Mixed colour temperature is the amateur tell

Window light is about 5,600K. The bulb over your head is about 3,200K. One of them has to go.

Daylight through a window and warm household bulbs are different colours, and letting them fight 'doesn't look professional at all'. The simplest fix is subtraction: switch off the ceiling lights and shoot on window light. The sophisticated version is to make the clash deliberate - a 5,600K key on the subject with warm practicals behind gives the orange-and-teal separation that reads as intentional. What you must not do is overpower one temperature and leave the scene looking, as he puts it, like living inside a blueberry. When indoor and outdoor light genuinely mix, white balance can sit deliberately between the presets - around 3,450K in one demo - to split the difference.

Why it matters

The single most common reason home-shot video looks wrong and nobody can say why.

04

Softness is size - diffuse it, or move it back

A cloud is a softbox. So is a bedsheet.

Soft light rolls off gradually and forgives skin; hard light throws sharp-edged shadows and points at every imperfection. The mechanism is the same in every case: put a second medium in the path - a softbox, a silk, a five-in-one reflector - and the source becomes effectively larger, which is what softness actually is. Outdoors, cloud cover or a panel held by an assistant does the job. Two refinements come up in the budget build: moving a diffused light further back increases its effective size and softens further, and a honeycomb grid does the inverse job, narrowing the spill so the light lands on the subject rather than washing the whole room.

Why it matters

Understanding it as size, not as a product, means a bedsheet works when the softbox does not exist yet.

05

Seven steps to audio, in the order that actually matters

how-to

'Audio is half of the viewing experience.'

The sequence is: decide audio matters, pick the right tool, get close, handle ambient noise, handle room reverb, set levels, and record a test clip. Four mic categories cover the ground - lavalier, onboard, shotgun, condenser - and a useful clarification is that a microphone and a recorder are not always the same object; a lav needs something to plug into. His default is the lav, for flexibility across commercial, adventure and course work. Above all else sits proximity: six to twelve inches from the mouth, because every inch further multiplies room and background noise. Then the room itself - hard floors, high ceilings, bare walls and big windows - answered with foam, rugs, artwork, furniture or a hung blanket as a 'pop-up studio', with a de-reverb effect in post as a supplement and never a substitute.

Do it in this order
Why it matters

Bad sound loses viewers inside a single clip; this is the order to spend attention in.

Audio is half of the viewing experience.
06

Peak at -6 to -12dB, because clipping cannot be undone

Clipped audio 'is going to sound like a distorted guitar' - and like a blown highlight, it is gone.

Levels are a safety margin, not a target. Peaks belong between about -6 and -12dB, well under the 0dB ceiling where clipping starts. The asymmetry is the whole argument: quiet audio can be raised in post, clipped audio cannot be repaired at all. Every device exposes the control differently - a physical dial on a Zoom H1n, a 0-20 scale on a Sony A7S III, on-screen sliders in Filmic Pro - and most phones auto-level by default, which is convenient and takes the control away. The habit that catches problems is the test clip: record a few seconds, listen on headphones, and check that hard consonants are not spiking, because some mics clip on them long before the meter suggests.

Why it matters

The one audio setting that is genuinely unrecoverable if you get it wrong.

It's going to sound like a distorted guitar.
07

TRS versus TRRS, and going wireless

The mic will not work and the cable is not broken. It is one ring.

Professional audio gear terminates in TRS - tip, ring, sleeve. Phones and computers expect TRRS, with an extra ring for the microphone channel. Plug one into the other and you get silence or a very quiet recording, and the fix is a cheap adapter in whichever direction you need. The PoP Voice Pro lav, for instance, ships as TRRS ready for a phone and needs an adapter to reach a professional camera. When the cord itself becomes the constraint - a subject who has to walk around a room - a wireless system such as the RODE Wireless GO II moves the problem: a body-worn transmitter, a receiver into camera or, through a TRS-to-TRRS cable, into a phone, with the audio baked into the recording so there is nothing to sync afterwards.

Why it matters

Five minutes of understanding that saves an afternoon of thinking the gear is faulty.

08

Sound design: three kinds of effect, and how each one enters

A whoosh on a fast camera move is a sound nothing made. Add it anyway.

Effects sort into three buckets. Physical action - a splash, a footstep, a zipper - matches something visibly happening. Ambience - birds, wind, crickets - establishes a place even when its source is never on screen. And perceived movement, where the brain expects a sound that nothing physically produced, which is why a fast camera move wants a whoosh. Foley substitution is legitimate: a clip named for a ticking clock can carry a shot of crawling insects if it matches emotionally. Two habits make it fast. Build a named, searchable personal library so effects get found rather than re-sourced. And decide how each one arrives - a J-cut lets the sound start before the picture cut and pulls the viewer into the next scene; a hard cut lands it exactly on the frame for deliberate abruptness.

Why it matters

Sound design is the cheapest production-value upgrade in an edit and almost nobody does it.

09

Three setups: a window, one light, or a room

Tier one is a window, a phone and a $30 tripod. It is a real answer, not a consolation prize.

The same talking-head shot is built three times at three budgets. No budget: phone, cheap tripod, and window or door light, with the depth coming from composition and distance to the background rather than from anything switched on. Under $150: add one continuous light such as a Yongnuo YN300 II at about $70, diffuse it, add a $20 five-in-one reflector as a bounce, put warm practicals in the background, and run a lav into a phone or a Zoom H1n. Full studio: a Godox SL60 into an 85-inch softbox with a honeycomb grid, accent lights, a shotgun over XLR into a recorder, and the rest of the 2024 gear tour. The progression is the argument - the first tier already looks professional, and each step buys control rather than legitimacy.

Why it matters

Gives a client three honest price points instead of one quote.

Tools referenced

ToolCoverageMomentContext
GodoxdemonstratedSL60 as the studio key, into an 85in softbox with a honeycomb grid
WestcottdemonstratedSolix Apollo (~$400) as key and a panel light as back in the three-point demo
RODEdemonstratedNTG3 shotgun over XLR; Wireless GO II for mobile subjects; Smart Lav in the budget build
ZoomdemonstratedH1n recorder with a manual gain dial; H4N Pro converting XLR to line level; Zoom software for livestreaming
Filmic ProdemonstratedManual audio levels and exposure lock on the phone
Adobe Premiere ProdemonstratedSound-effect layering, J-cuts and hard cuts
Epidemic SounddemonstratedMusic and effects for the spec ad
SonymentionedA7S III as the studio A-camera
SanDiskmentionedExtreme SSD for active edits

Session materials

Archived locally on V: — click to open. Companion pages link to the LMS.

Action items

Resources mentioned

Resources
  • docAudio test exercise (three levels)
  • docMy YouTube Setup (2024 version)
  • docMerrell hiking-boot spec ad
  • docEpidemic Sound and free effect searches

Extraction notes

This page was built from an auto-generated transcript, which garbles product and people's names. Those were corrected silently in everything above and logged here for transparency. The warnings flag claims that were true on the recording day but change fast.

Transcript corrections applied

The transcript saysThe trainer actually means
Momon MT12Moman MT12 teleprompter
Feel world Ultra Bright 4KFeelWorld monitor
GatorWorks desk mountGator Frameworks microphone desk mount
Sigma 16-28 mini millimeterSigma 16-28mm f/2.8
Yongnuau / Yongnuow YN300Yongnuo YN300 II
completely like Diagocompletely diagonal
Wall E stand / Kairo BVL tripods / LAXI laptop standunidentified brands - left as heard, unverified

True on recording day — verify before relying