Chapter 03
Cinematography basics.
A monitoring app is only useful if you know what you are looking at. This chapter explains the ideas behind every tool in the app, in the order you meet them on a shoot. Each section ends with the tool that handles it.
Exposure: how much light reaches the sensor
Exposure is the amount of light the sensor records. Too little and the image is dark and noisy. Too much and bright areas clip, which means they turn solid white with no detail that can be recovered later. Three camera settings control exposure, and they interact, which is why people call them the exposure triangle.
Aperture (f-number)
The aperture is the opening inside the lens. A low f-number such as f/1.8 is a wide opening: more light and a shallow depth of field, so the background blurs. A high f-number such as f/11 is a small opening: less light and more of the scene in focus. Each full stop (f/1.4, 2, 2.8, 4, 5.6, 8, 11, 16) halves the light. Cinema lenses are often marked in T-stops, which are f-stops corrected for the light the glass actually passes.
For video, aperture is mostly a creative choice about depth of field. Use it to decide how much of the frame is sharp, then set exposure with ISO and ND filters rather than constantly changing the aperture mid-shot.
The Aperture tile on the Controls page, and the aperture chip in the Live View top bar. Tap either to pick a value. See Controls.
Shutter speed and shutter angle
The shutter sets how long each frame is exposed. In video there are two ways to express it. Shutter speed is a fraction of a second, such as 1/50. Shutter angle is borrowed from film cameras, where a rotating disc with an opening of a certain angle exposed each frame. A 180° angle means the frame is exposed for half of the frame interval. At 24 fps that is 1/48 s; at 50 fps it is 1/100 s.
The reason cinematographers prefer angle is the 180-degree rule: a 180° shutter gives the amount of motion blur that looks natural to viewers at any frame rate. If you switch from 24 to 60 fps for slow motion, the angle stays 180° and the blur stays right, while a fixed shutter speed would not. Smaller angles (45° to 90°) give crisp, staccato motion for action or sports; larger angles (270° to 360°) give heavy blur and gather more light in low light.
Shutter also interacts with artificial light. Lights on mains power flicker at twice the mains frequency. Use 1/50 or 1/100 (180° at 25 or 50 fps) in 50 Hz countries and 1/60 or 1/120 in 60 Hz countries to avoid rolling bands.
The Shutter tile shows speed or angle, and the Shutter Mode toggle switches between them on cameras that support shutter angle (ZR, Z8, Z9, Z6III and RED). Angle mode keeps the blur constant when you change frame rate, so most video shooters leave it on.
ISO, base ISO and dual base ISO
ISO is the sensor's amplification. Raising it brightens the image but also amplifies noise. Every sensor has a native (or base) ISO where it delivers its best balance of noise and dynamic range. Many modern cinema sensors have two base ISOs, for example a low base around 800 for daylight and a high base around 6400 for low light. At each base the sensor runs a different analog gain and is cleanest there. ISO values between the two bases are digital pushes of the nearer base, so ISO 3200 can be noisier than ISO 6400 on a dual-base sensor.
Rule of thumb: shoot at a base ISO whenever you can and control brightness with aperture, ND filters and lighting. Move to the high base only when the scene is genuinely dark.
The ISO tile lists only the values the camera accepts in the current codec. On the Nikon ZR in R3D the Base ISO tile switches between the Low and High base; the exposure tools recalibrate their thresholds automatically when you change it. See Controls.
Exposure compensation and metering
In an automatic exposure mode the camera's meter picks the exposure. Exposure compensation nudges that decision brighter or darker in thirds of a stop. Metering mode decides which part of the frame the meter trusts: the whole frame (matrix), the centre, a small spot, or the highlights so they are protected. In manual mode (M) the meter only advises; your settings decide.
The Mode, Exposure comp and Metering tiles on Controls. Most cinema work is done in M so that exposure does not drift during a shot.
Frame rate and slow motion
Frame rate is how many pictures per second the camera records. 24 fps is the cinema standard; 25 fps is used for television in 50 Hz countries and 30 fps (29.97) in 60 Hz countries. Recording at 50, 60, 100 or 120 fps and playing it back in a 24 or 25 fps project gives slow motion: 120 fps played at 24 fps is five times slower than life.
Two frame rates matter: the sensor (or capture) rate, which is what you record, and the project (or timeline) rate, which is what you deliver. RED cameras expose both; Nikon bodies set the sensor rate in the frame size and rate menu and conform in the editor.
The Res/FPS tile sets the frame size and rate. The SLOW pill in Live View jumps to a saved high-frame-rate format and back with one tap. In Playback, the playback speed option plays a 120 fps clip at 24 fps so you can judge the slow motion on set. See Controls and Playback.
White balance, Kelvin and tint
Light sources have a colour. Candlelight is orange, midday sun is neutral, open shade is blue. White balance tells the camera what should look white. It is expressed as a colour temperature in Kelvin: around 3200 K for tungsten lamps, 5600 K for daylight, 6500 K or more for shade. Setting a lower Kelvin makes the image bluer; a higher Kelvin makes it warmer. Tint is the second axis, shifting between green and magenta, and it corrects fluorescent and LED light that no Kelvin value fixes.
For video, set white balance manually and keep it fixed inside a scene. Auto white balance drifts as people move or clouds pass, and a drift in the middle of a shot is very hard to fix in post. When shooting RAW the white balance is only metadata and can be changed later, but the monitor still needs a sensible value so you can judge the image.
The WB, Kelvin and Tint tiles. The tint tile opens a full-screen amber–blue and green–magenta grid. In Multicam the shared exposure rail sets Kelvin and tint on all cameras so they match. The vectorscope shows the colour balance of the frame: a grey card should sit in the centre.
Log, Rec.709 and HLG: why the picture looks flat
A sensor captures far more brightness range than a normal screen can show. The camera has to squeeze that range into the file. Rec.709 (standard video) does it by throwing away the extremes: it looks finished straight out of the camera, but highlights clip early and the shadows are crushed. Log profiles instead compress the range with a logarithmic curve that keeps every stop the sensor captured. The result looks grey, flat and low-contrast on a monitor, but it holds detail that you bring back in colour grading.
Each manufacturer has its own log curve. Nikon uses N-Log, RED uses Log3G10 inside R3D, Apple uses Apple Log. They place middle grey and the clip point at different code values, so a tool that reads exposure must know which curve it is looking at. HLG is a third option, a hybrid curve for HDR delivery that is less flat than log.
The practical consequence: when you shoot log you cannot judge exposure by eye from the flat picture, and you cannot judge it from a LUT either, because the LUT changes the brightness. You need scopes or false color that read the recorded signal.
Every exposure tool is calibrated for the profile in use. It detects N-Log and R3D automatically from the codec and LUT, and the thresholds for zebras, false color, the waveform guides and X-Lights follow. The calibration profile is shown in the scope options. See Exposure tools.
LUTs: looking at log through a lens
A LUT (look-up table) is a file that maps every input colour to an output colour. A technical LUT converts a log image to Rec.709 so it looks normal on a monitor. A creative LUT adds a style on top: a film emulation, a bleach bypass, a warm teal-and-orange look. On set a LUT is for monitoring: the camera still records the flat log file, and the LUT only changes what you see on the phone. That is the safest workflow, because nothing is baked in and the colourist starts from the full-range original.
The important subtlety is that a LUT can hide problems. A contrasty LUT makes a slightly overexposed sky look fine while the log file is clipping. That is why the app analyses the picture before the LUT is applied. The LUT changes what you see; the scopes, zebras and false color read what is being recorded.
The LUT pill switches the look on and off, and LOOK opens the browser with bundled N-Log and RED IPP2 conversions, 32 RED creative looks and your own imported .cube files. WIPE compares the LUT with the ungraded image on a split screen. See LUTs and looks.
Reading exposure with scopes
Scopes are graphs of the picture. They do not care how the screen is set or how bright the room is, which makes them the only trustworthy way to judge exposure in log. The app offers four, and two overlay tools that paint the result straight onto the image.
IRE, code values and middle grey
Scopes measure brightness in IRE, a 0 to 100 scale where 0 is black and 100 is the brightest value the file can hold. Under the hood the file stores 8- or 10-bit code values (0 to 255 or 0 to 1023). Two landmarks matter. Middle grey is an 18% reflectance card, the reference for a correctly exposed scene. In Rec.709 it sits around 40 to 45 IRE; in N-Log around 35 IRE; in Log3G10 around 33 IRE. Clip is where the sensor runs out, 100 IRE in Rec.709 but lower in some log curves. Caucasian skin typically sits one to one-and-a-half stops above middle grey.
Histogram
A histogram counts how many pixels sit at each brightness, dark on the left, bright on the right. A pile-up against the right edge means clipping; against the left edge, crushed shadows. It is fast to read but it does not tell you where in the frame the problem is.
Waveform
A waveform plots brightness vertically against horizontal position in the frame, so a bright window on the right of the picture appears as a bright spike on the right of the scope. It is the cinematographer's standard tool: you can read the level of a face, a sky or a wall directly. The app draws guide lines for clip, highlight, middle grey and shadow at the right levels for the current log curve, so you do not need to remember the numbers.
RGB parade
Three waveforms side by side, one for red, green and blue. When the three traces line up for a neutral object the white balance is right. If the blue trace sits higher than the others on a grey wall, the image is too cool. The parade is the best tool for matching white balance between cameras.
Vectorscope
The vectorscope ignores brightness and plots only hue and saturation. The centre is grey; distance from the centre is saturation; the angle is the hue. Six boxes mark the standard colour-bar targets, and a diagonal skin line shows where all human skin tones fall regardless of ethnicity. If a face sits off the skin line, the white balance or tint is wrong.
Zebras
Zebras paint diagonal stripes on the parts of the picture that reach a chosen level. Classic use: highlight zebras at the clip point, so stripes mean "this is gone". The app also offers zebras for middle grey, so you can put stripes on a grey card or a face and know it is at the right level, and for crushed shadows.
False color
False color replaces the whole picture with a colour per exposure band: purple for crushed black, blue for deep shadow, green for middle grey, pink for skin, yellow and orange for highlights, red for clipped. It is the fastest way to see the exposure of everything in the frame at once. Different cameras and apps use different palettes, so a legend is always shown.
Exposure lights and posts
When the whole scope is too much, a summary helps. X-Lights are three lamps per channel that turn on as highlights or shadows get close to the limit. X-Posts are two meters, one for clipping pressure and one for noise, that you can glance at while watching the picture. They are the "traffic lights" of exposure: green means fine, amber means close, red means act.
All of these are in Live View under SCOPE, ZEBRA, FALSE and EXPO. See Exposure tools for every option and when to use which.
A simple recipe for exposing log
- Set the base ISO
Pick the base nearest to your light. Leave it there and adjust brightness with aperture and ND.
- Protect the highlights
Turn on highlight zebras or false color and bring the brightest thing you care about just under clip. Anything above clip is unrecoverable; anything below can be lifted in post.
- Place the subject
Read a face or a grey card on the waveform. Faces around one stop above middle grey is the usual target. Many log shooters expose a touch brighter than nominal to keep shadows out of the noise.
- Check the shadows
Shadow zebras or the noise post tell you when important dark areas have fallen into the noise floor. Add light or open up if they have.
- Lock it
Keep M mode and a fixed white balance for the scene so nothing changes between takes.
Focus
A small phone screen makes focus hard to judge, and in video a missed focus cannot be fixed later. Two kinds of help exist. Focus peaking outlines the sharpest edges in a colour, so you can see what is in focus without zooming. Punching in magnifies part of the frame so you can check critical focus on an eye. Peaking is quick and works while rolling; punch-in is precise but you lose the overview.
Depth of field is the zone that looks sharp. It gets shallower with a wider aperture, a longer lens and a closer subject. A shallow depth of field looks cinematic but makes focus pulls unforgiving. A focus pull is moving focus from one subject to another during a shot; on a film set a focus puller does it by hand with marks on the lens. Autofocus cameras can do it between two AF points, and the speed of the move is part of the look.
PEAK for peaking, ZOOM for monitor punch and camera focus punch, PULL to set and move between AF points, and MF to drive a manual-focus-mode lens to saved In and Out points with a timed pull. See Focus tools.
Framing, aspect ratios and anamorphic
The sensor records one shape (usually 16:9 or 3:2) but you may deliver another: 2.39:1 widescreen cinema, 1.85:1, 4:3, or vertical 9:16 and 4:5 for social platforms. Frame guides draw the delivery shape over the picture, and a matte darkens the area outside it, so you compose for what the audience will see. Safe areas are inset borders that keep action and titles away from edges that some screens crop.
Composition grids are aids rather than rules: the rule of thirds puts subjects on the lines; the golden ratio and golden triangles place them slightly tighter to centre; a centre cross helps symmetrical framing.
Anamorphic lenses squeeze a wide field of view onto the sensor; a 2x lens records a picture that is half as wide as it should be. Desqueeze stretches the monitor picture back out so you see the real frame. The camera file stays squeezed and is desqueezed again in the edit.
FRME, COMP, DSQ, REF (a ghosted reference frame for matching shots) and LVL (the camera's level sensor). See Framing and guides.
Codecs, RAW and bit depth
The codec is how the camera compresses the video into a file. H.264 and H.265 (HEVC) make small files that play anywhere but tolerate less grading. ProRes 422 HQ is lightly compressed, large, and the editing standard. RAW formats (N-RAW, ProRes RAW, R3D) store the sensor data before white balance and much of the processing, so those decisions can be changed in post. RAW files need a compatible editor and cannot be played by most phones.
Bit depth is how many brightness steps the file stores. 8-bit has 256 per channel and shows banding in skies when graded hard; 10-bit has 1024 and is the minimum for log. 12-bit and above are RAW territory.
The Codec tile lists every format the camera reports. The playback page can play H.264, H.265 and ProRes clips from the card; N-RAW and R3D clips are listed but not playable on the phone. See Playback.
Audio levels
Digital audio is measured in dBFS (decibels relative to full scale), where 0 dBFS is the loudest value the file can hold. Unlike analogue tape, there is nothing above 0: a peak that hits it is clipped and distorts. Dialogue is usually recorded with peaks around −12 to −6 dBFS, leaving headroom for surprises. The camera meters show the current level and hold the peak for a moment so you can see what just happened.
Input gain (sensitivity) sets how much the microphone signal is amplified before recording. Line input expects a strong signal from a mixer; mic input expects a weak one from a microphone. A wind filter cuts low rumble; an attenuator pads a loud source. 32-bit float recording stores so much range that clipping becomes almost impossible and levels can be corrected after the fact; where the camera supports it, it is the safe choice for run-and-gun.
The AUD meters in Live View show left and right levels with a peak hold and let you change the camera's input gain with up and down buttons. The Audio tab on Controls holds the rest. See Controls.
Timecode
Timecode is a clock stamped on every frame, written as HH:MM:SS:FF. Editors use it to line up picture with separately recorded sound and to sync multiple cameras. Rec Run timecode only advances while recording, so each clip continues from the last. Free Run advances continuously like a clock, usually set to the time of day, so every camera that has the same time produces clips that line up. Drop frame is a counting trick for 29.97 and 59.94 fps that keeps the timecode matching real elapsed time.
A jam is setting one device's timecode from another so they agree. Professional sets use hardware sync boxes; the app can jam all connected cameras to the phone's clock over the control link, which is accurate enough to line clips up and then trim by eye or audio waveform.
The Audio/TC tab on Controls sets Rec TC, Rec Run or Free Run, drop frame and reset. Multicam has Timecode Jam-Sync. See Multicam.
Stills: long exposures, timelapse and stacking
Bulb keeps the shutter open while you hold the release; Time opens on one press and closes on the next. Both are for exposures longer than the camera's longest timed shutter, typically for astrophotography or light painting. A timelapse is a sequence of photos taken at a fixed interval and played as video: 300 frames at a 5-second interval is 25 minutes of life compressed into 12.5 seconds at 24 fps. The interval must be longer than the exposure. Focus stacking takes several frames at different focus distances and blends them in software for a depth of field no single frame could have; it is a macro and product-photography technique.
The TIME tool runs timelapses and builds the video on the phone. Focus Stack drives the lens between frames. Both are in photo mode. See Stills, timelapse, stacking.
Where to go next
With these ideas in place, the tool chapters will make sense. If a term in a chapter is unfamiliar, the glossary defines it in a sentence or two.
Something missing or wrong on this page? Email mike@itsmadcreative.com with the page name and what you expected to find.