Camera Intermediate

Introduction

A director has eight cameras and one second to choose. Five of them are showing usable pictures. One is still zooming, one is searching for a presenter who walked out of frame, and one looks slightly greener than the rest. The operators of the first five will be booked again.

Camera Basics teaches the camera and the job. This book teaches what makes an operator's camera the one the director can always use: a picture set up deliberately rather than accepted, exposure measured rather than guessed, a recording chosen to survive the edit, a box lens driven at the speed of a live show, and a camera that matches the others at every cut.

It is built on published sources: the international guide to HDR production practice, the European recommendations on signal levels and camera quality, camera and lens makers' own manuals, and the published specifications for log curves, false colour and recording formats. Where the book gives a number, it says where the number comes from.

How to Use This Book

Read Camera Basics first. This book assumes everything in it and does not repeat it. Camera Theory sits alongside: it explains why the picture behaves as it does, and this book points to it where a reason would take more than a sentence.

Then read in order. The first chapters are about the picture: how the camera is set up, how it is exposed, how focus and colour are held, and how it is recorded. The later ones are about the show: operating at speed, working with the director, the different outputs a camera feeds, the specialist cameras, and matching.

Practise with a camera and a monitor

Most of this book is best learned with a camera that has a waveform, zebras and a menu you can explore. Change one setting at a time, watch what it does to the picture and the waveform, and set it back. A grey card and a white card cost very little and make every exposure chapter practical.

A note on the words

A "system camera" is a broadcast camera connected to a camera control unit, whose picture is controlled by a vision engineer, also called a shader or racks. "Painting" a camera means adjusting its picture settings. Percentages of signal level are on the waveform's scale, where 0 is black and 100 is nominal white. "SDR" and "HDR" are standard and high dynamic range.

Beyond the Basics

Taking camera operation from knowing to doing well.

From basics to craft

Camera Basics introduced the camera — its parts, how it works, and basic operation. Camera Intermediate develops the craft of operating it well: real control over exposure, focus, colour, and framing, and the smooth, consistent operation that distinguishes a good camera operator. The goal moves from getting an image to getting a good, controlled, professional image, reliably. This step — from knowing the camera to operating it with craft — is what this book builds.

This matters because professional camera work requires more than basic operation — it requires controlling the image deliberately (exposed right, focused sharp, coloured correctly, framed well) and operating smoothly and consistently. These are the skills that make a camera operator employable and good.

Camera Basics teaches the camera and the job: the parts, the exposure tools, focus, framing, support, preparing for a show, working to a director and looking after media. If you have not read it, read it first, because this book does not repeat it.

Camera Intermediate covers what an operator needs next. How to set up the camera's picture itself, through its scene files, gamma, knee, detail and colour settings. How to expose by measurement, including log and HDR, where judging by eye fails. How recording formats and media work in enough depth to choose them. How to operate a box lens at the speed of a live show. How to work ahead of a director rather than behind. How IMAG, broadcast and streaming each change the shot. How to run the specialist cameras that now appear on most shows. And how to help a set of cameras match.

Camera Theory sits beside this book. It explains why the picture behaves as it does, and this book refers to it wherever a reason would take more than a sentence.

The operator's control

An intermediate camera operator controls the key image variables deliberately: exposure (how bright the image is), focus (what's sharp), white balance (correct colour), and framing (the composition). Rather than leaving these to automatic or chance, the operator sets and controls each to get the intended image. This deliberate control over the image is the foundation of good camera work. The operator who controls exposure, focus, colour, and framing makes good images on purpose; one who relies on auto and luck doesn't.

This book develops control over each of these (exposure, focus, colour, framing) plus the operating craft (smooth moves, consistency). Mastering deliberate control is what turns basic camera use into professional operation.

Who controls what on a show

Before touching any setting, know whose it is. On a system camera, a broadcast camera connected to a camera control unit, the operator owns the framing, the focus and the zoom, and often the ND or colour filter wheel. The vision engineer owns the picture from a remote control panel: iris, black level, gain, white balance, knee, detail and matrix. Changing those from the camera without asking breaks the match the engineer is holding.

On a camcorder, a cinema camera or a PTZ camera without a vision engineer, the operator owns everything, including the settings this book's chapters on picture setup and exposure describe. Many shows mix the two. Ask at the start of every job which settings are yours, and leave the rest alone.

Practice it

You arrive on a show with four Sony system cameras on pedestals, a vision engineer at a row of remote control panels, and two camcorders for roaming shots.

Setting Up the Picture

Every camera turns light into a picture through a chain of settings: the curve, the black level, the knee, the detail and the colour matrix. Knowing what each one does to the picture is what lets you set a camera up rather than accept whatever it arrived with.

Scene files, and why you load rather than build

A camera's picture settings are stored together as a scene file, sometimes called a picture profile or a look. Sony's camcorders, for example, keep scene files in internal memory, offer factory presets, and save files to and load them from a memory card. System cameras keep theirs in the camera control unit and the vision engineer's panel.

On a multi-camera show the scene file is how matching begins. Somebody, usually the vision engineer or the director of photography, builds one file for each camera model, and every camera of that model loads it. An operator who builds their own from scratch on show day has made a camera that will not match the others. Load the show's file, confirm it loaded, and change nothing in it without agreement.

Still learn what is inside it, because you will be asked to build one, and because when a picture looks wrong the cause is usually one of the settings in the sections that follow.

Gamma, the camera's curve

Camera Theory explains why video signals are curved: the camera applies a transfer curve, called gamma, so that the shadows get enough of the signal's steps. In a system camera's paint menu the curve is a number and a table. Sony's HDC3300R, for example, sets its standard gamma at 0.45 and labels the matching table as equivalent to ITU-R BT.709, the HD standard.

Changing the gamma number changes the middle of the picture. The signal is roughly the light raised to that power, so a lower number, such as 0.40, lifts the mid-tones: the picture looks brighter and flatter, with lighter shadows. A higher number, such as 0.50, pushes the mid-tones down for a darker, contrastier picture. Neither changes the black or the white points.

Many cameras also offer hypergamma or cine curves, which bend the top of the curve so that much brighter highlights fit inside the normal signal. The HDC3300R's hypergamma tables, for instance, squeeze light up to 325 or 460 percent of normal white into the range up to 100 or 109 percent. They hold more highlight detail without a knee, at the cost of slightly flatter whites.

Black level and black gamma

Master black, also called pedestal, raises or lowers the whole picture's black level. Raise it and blacks turn grey and the picture looks washed out; lower it and the darkest shadows crush to solid black and lose their detail. Sony's camcorders set it from minus 99 to plus 99, with separate red and blue blacks for correcting a coloured tint in the shadows.

Black gamma changes only the shape of the curve in the shadows. It can open up detail in dark clothes and dark scenery without lifting the black itself, or press the shadows down for a stronger look while black stays where it is.

Black level is one of the first things to match between cameras, because a difference in black shows at every cut. On system cameras it belongs to the vision engineer.

The knee, and saving the highlights

A standard curve would clip everything brighter than normal white, and a show is full of things brighter than white: lamps, reflections, a sunlit window, a white shirt in a spotlight. The knee is a bend near the top of the curve. Below the knee point the signal is unchanged. Above it, the signal rises much more slowly, governed by the knee slope, so a range of bright highlights is squeezed into the space left below the clip.

Sony's PXW-Z300 sets the knee point anywhere from 75 to 109 percent, with a default of 90 percent, and switches the knee on by default in SDR and off in HDR. Lower the point and more highlight detail survives, but everything above it is compressed, so whites look greyer and flatter. That has a practical limit: faces on a studio show sit in the seventies of the SDR scale, as the next chapter shows, so a knee set much below 80 percent starts to flatten skin.

Above the knee sits the white clip, a hard limit on the maximum level. EBU R 103, the European recommendation on signal levels, advises that camera clippers for live production be set to its preferred range limit, which allows the signal to reach 105 percent. Highlights above that are lost however the knee is set.

Bright saturated colours have their own problem at the knee: as the signal is compressed, colour can bleach out and change hue. Knee saturation, on cameras that have it, restores colour above the knee. Sony describes it as keeping natural skin tones under strong lighting.

Detail, which is sharpening

Detail is electronic sharpening. The camera finds edges in the picture and adds a thin dark and light line on each side of them, so they look crisper. Camera Theory explains that this adds no real detail; it raises the contrast of edges that already exist.

Set well, detail makes a picture look clean and sharp on a monitor. Set too high, it makes outlines look drawn on, with bright halos around dark objects and dark halos around bright ones, and it makes skin look rough. Set too low, the picture looks soft next to other cameras.

The detail menu has several controls that shape where sharpening is applied. Level sets how much. Crispening sets a threshold so the smallest variations, which are mostly noise, are not sharpened; without it, detail sharpens noise as eagerly as edges. Level dependence reduces detail in dark areas, where noise lives. The limiter caps how bright or dark the added edges can get. And the H/V ratio balances sharpening on horizontal and vertical edges.

Skin detail is the one most worth knowing on events. It picks out a chosen range of hues, usually skin, and turns the detail down in that range only. Faces soften while hair, eyes, clothes and the background keep their sharpness. Sony's system cameras let three hues be set independently, which matters on a stage with presenters of different skin tones.

The colour matrix

White balance makes white neutral. The colour matrix decides how every other colour is reproduced, by mixing the sensor's red, green and blue signals into the output. Cameras offer preset matrices for the different standards, such as ITU-709 and EBU, and a user matrix that mixes each channel into the others.

The multi-matrix goes further. It divides the colour circle into sixteen axes and lets the hue and saturation of each be adjusted on its own: the reds of a jacket without the skin, the greens of a pitch without the blues of the sky. Sony's documentation calls it useful in colour matching among multiple cameras, which is where it is mostly used, as the chapter on matching explains.

Matrix work is easy to overdo, because every change affects everything of that colour in every shot. On a show it belongs to whoever is matching the cameras, and it is changed against a chart, not by eye on one face.

Log and HDR looks

Modern cameras offer a choice of base looks, which replace the whole curve and colour. Sony's PXW-Z300 lists S-Cinetone, ITU709 and 709tone for standard dynamic range, and HLG Live, HLG Mild, HLG Natural and S-Log3 for high dynamic range.

They suit different jobs. ITU709 is the broadcast standard and looks right on any HD monitor straight away. HLG, hybrid log-gamma, is the HDR curve built for live television: it can be shown directly on HDR screens and still looks acceptable on SDR ones. S-Log3 is a recording curve meant to be graded afterwards; it looks flat and grey until a conversion is applied, and keeps the most range for the grade.

On a live show the look must match whatever the rest of the system expects. A camera sending S-Log3 into a switcher of ITU709 cameras will look washed out at every cut, unless a conversion is applied in between. Ask which curve the show runs before choosing.

How to do it

1. Ask which scene file, base look and recording format the show runs.
2. Load the show's scene file from the card, and confirm the camera reports it loaded.
3. Check the settings that most often drift: knee on or off, detail level, black level and the matrix setting.
4. Point the camera at the same chart as the other cameras and compare on the waveform.
5. Report any difference rather than adjusting the file yourself.

What goes wrong

Practice it

A camcorder's knee point is set to 75 percent with a steep slope. Presenters' faces sit at about 72 to 78 percent on the waveform.

Exposure

Master the iris/shutter/gain triangle.

The exposure triangle

Exposure — how bright the image is — is controlled by three things, the exposure triangle: iris/aperture (how much light the lens lets in, which also affects depth of field — how much is in focus front-to-back), shutter (the shutter speed/angle, which affects exposure and motion blur), and gain/ISO (the sensor's sensitivity, which affects exposure and noise/grain). The operator balances these three to get correct exposure with the desired look (depth of field, motion, noise). Understanding and using the exposure triangle is the core of exposure control.

Each control has a trade-off beyond exposure: iris affects depth of field (wide iris = shallow focus; small iris = deep focus), shutter affects motion (slow = blur; fast = crisp), gain affects noise (high gain = noisy). The operator chooses the combination that exposes correctly and gives the look they want. Mastering this balance is central to camera craft.

Exposing correctly

Exposing correctly means setting the image brightness right — not too dark (crushed shadows) or too bright (blown-out highlights), with the important parts (faces, the subject) well-exposed. Operators judge exposure using tools (the monitor, and especially scopes/aids like the waveform, zebras, or histogram that show exposure precisely) rather than just the eye. Correct exposure makes a usable, good-looking image; poor exposure ruins it (and is hard to fix later).

This means understanding what correct exposure looks like (and what the scopes show), and using the exposure triangle to achieve it for the subject and conditions. Good operators expose accurately and consistently; poor ones guess and get inconsistent or wrong results.

Zebras set to a level that means something

Camera Basics teaches zebras as a warning. Used deliberately, they become a measuring tool. Most professional cameras offer two zebras. On Sony's cameras, zebra 1 defaults to 70 percent and marks a band around that level, its width set by the aperture or width setting, 10 percent by default. Zebra 2 defaults to 100 percent and marks everything above it, which makes it a clipping warning.

Set zebra 1 to the level where you want the most important thing in the picture to sit, and expose until it shows the stripes. The level is not fixed. It depends on the curve the camera runs, the subject, and the show's practice, which is what the next sections give you.

Where faces actually sit

There is no single correct level for a face, and the evidence says so. ITU-R Report BT.2408, the international guide to HDR production practice, includes a Japanese survey of 713 faces in studio programmes. In standard dynamic range the faces averaged 74.6 percent: 71.8 percent for men and 77.6 percent for women. The report notes that practice differs between regions, mainly because of long-standing production habits.

Skin itself varies far more. BT.2408's indicative ranges in HLG, the live HDR curve, run from 55 to 65 percent for the lightest skin types, 45 to 60 percent for medium, and 25 to 45 percent for the darkest, with the note that the data for the lightest and darkest types is less complete. Exposing every face to the same level would make dark skin look grey and washed out, and light skin look dull.

So expose for the person. Set the scene so that a neutral reference sits where it should, and let each face fall where its own reflectance puts it. Then check that dark skin keeps its richness and light skin keeps its modelling, rather than forcing both to one number.

The grey card as a reference

An 18 percent grey card is the most dependable reference, because it is defined, neutral, and in the middle of the range. Camera Theory works out where it sits on the plain BT.709 curve: about 41 percent. On a real camera, gamma, black and knee settings move it a little, and BT.2408 warns that the grey card's level can differ significantly once a camera has been painted.

In HDR the reference levels are published. BT.2408 puts the grey card at 38 percent in both HLG and PQ, and the HDR reference white, a 100 percent white card, at 75 percent in HLG and 58 percent in PQ, which on a reference display is 203 candelas per square metre. Graphics are meant to sit at that same white.

In S-Log3, Sony's own technical summary puts 18 percent grey at 41 percent, 10-bit code value 420, and 90 percent white at 61 percent, code value 598, with black at 3.5 percent. Those are the numbers to expose a log camera to. Exposing log by the same zebra you use for faces in BT.709 would overexpose it badly.

False colour

False colour replaces the picture with flat colours that show exposure zones, so the whole frame can be read at a glance. ARRI's published specification is a clear example. Green marks 18 percent middle grey. Pink marks one stop above it. Yellow marks two thirds of a stop below clipping, and red one third of a stop below. Blue marks the edge of shadow detail, and purple the noise floor. Everything else shows in grey.

Two cautions from the same document apply to every camera. The colours have no colour meaning: they are labels, not a picture. And the zones are worked out for a particular curve and exposure index, so the levels behind each colour change when those settings change. Different makers use different colours and different zones, so learn the key for the camera you are using before trusting it.

Headroom and legal levels

Digital video does not use every number available. In 10-bit signals, EBU R 103 puts nominal black at code 64 and nominal white at 940, and allows a preferred range of 20 to 984, which is minus 5 percent to plus 105 percent. The space above and below is headroom for the things the recommendation names: a camera moving, the sun coming out, specular reflections.

The same recommendation explains why the clip matters. A signal clipped hard rings at its edges and compresses badly, and black clipped below zero removes the sub-black that engineers use to line up monitors. Codes 0 to 3 and 1020 to 1023 are reserved on SDI and never carry picture.

When the lighting keeps changing

Live shows change their lighting constantly, and an operator who re-exposes for every lighting state makes the faces jump in brightness at every change. Expose for the key light on the faces, which usually stays steady, and let the stage behind them go bright or dark with the lighting design.

In HDR, change exposure gently. BT.2408 warns that rapid exposure adjustments made while shading in SDR should be avoided, because the change is more visible in the HDR picture.

How to do it

1. Find out which curve the camera is running: BT.709, HLG, PQ or a log curve.
2. Place an 18 percent grey card in the presenter's position, under the show's key light.
3. Expose until the card sits at the level for that curve, using the waveform or a zebra set to that level.
4. Set zebra 2 at 100 percent, or your camera's clipping level, as a highlight warning.
5. Check faces: dark skin should be rich, light skin should keep its modelling, and no face should touch the clipping warning.

Practice it

Three cameras are set up for a hybrid show: one in BT.709, one in HLG and one recording S-Log3 for a later edit.

Focus, Colour & Framing

Sharp focus, correct colour, good composition.

Focus control

Focus determines what's sharp in the image, and controlling it is a key camera skill. The operator sets focus on the subject (sharp where it matters) and holds or adjusts it as needed — including following focus (keeping a moving subject sharp by adjusting focus as they move, or as the camera moves). Sharp, well-controlled focus is essential — a soft (out-of-focus) shot is usually unusable. Especially with shallow depth of field (where only a thin plane is sharp), focus control is demanding and critical.

This requires technique (focusing accurately, pulling/following focus smoothly) and attention (keeping focus through the shot as things move). Good operators keep their shots sharp; poor focus ruins otherwise good shots.

White balance and framing

Two more controls: white balance — setting the camera's colour reference so whites are white and colours are correct (matching the light's colour temperature), so the image isn't too orange or blue. And framing/composition — how the shot is composed (the subject's placement, headroom, the rule of thirds, leading the eye), which makes shots look good and professional. The operator sets white balance for correct colour and frames each shot well.

White balance keeps colour accurate (and consistent across shots and cameras); framing makes shots compelling and correct. Both are deliberate operator choices that separate professional images from amateur ones. Good operators white-balance correctly and frame with a strong eye; poor ones get colour casts and weak compositions.

Focus in 4K on an HD viewfinder

Many cameras record or output 4K while their viewfinder shows HD or less. A viewfinder with a quarter of the picture's detail cannot show focus errors smaller than its own pixels, so a shot that looks sharp in the viewfinder can be visibly soft on a 4K screen or a large IMAG wall.

Peaking, which outlines sharp edges in colour, helps, and it is not proof. It responds to contrast, so it marks a high-contrast edge that is slightly soft and misses a low-contrast face that is sharp. In low light, gain noise can make peaking sparkle all over the picture. The reliable check is magnification, often called expanded focus, which shows part of the picture at full resolution. Focus on the eyes with magnification, then return to the full frame before the shot is needed.

Colour through a changing show

Show lighting changes colour. A warm walk-in state, a cooler presentation state and a saturated music state can all happen in the first ten minutes, and LED walls behind the stage add their own colour. Faces lit by the same key light should look the same throughout, while the background is free to change.

That is why the white balance is set for the light on the faces, not for the room. On system cameras the vision engineer rides it through the show. On camcorders, decide in advance who changes it and when, because one operator rebalancing mid-show makes their camera stop matching the others. The chapter on matching covers how to do it together.

Practice it

You are shooting a 4K recording of a conference on a camera with an HD viewfinder. The stage is dim and the presenter wears a dark suit.

Recording and Media

Camera Basics covers handling cards and handing them over. This chapter covers choosing what goes on them: the codec, the bit rate, and the timecode that ties it all together.

Intra-frame and long-GOP

Every recording format compresses the picture, and there are two broad ways to do it. EBU R 118, the European recommendation on camera quality, describes them. Intra-frame codecs process each frame on its own. Their bit rates are generally much higher, but every frame stands alone, so editing and further processing are easy. Long-GOP codecs, from group of pictures, store a full frame only every so often and record the frames in between as changes from their neighbours. They give the most reduction in bit rate, but the recording needs more processing power downstream.

Long-GOP formats include H.264, H.265 and Sony's XAVC Long. Intra formats include Apple ProRes, Avid DNxHD and Sony's XAVC Intra. The practical rule: long-GOP for long recordings where space matters, intra for anything that will be edited heavily or must survive further processing.

Bit rates, and what is enough

Bit rate is how much data each second of recording uses, in megabits per second. More data means less compression damage, until the point where the picture is clean and more data buys nothing.

EBU R 118 publishes minimum bit rates for broadcast-acceptable cameras, for 10-bit standard dynamic range. For HD on its long-form tier, it asks for at least 100 megabits per second intra at 25 frames and 200 at 50. Long-GOP H.264 can be much lower, from 25 to 50 megabits depending on frame rate and structure, and H.265 lower still, though the recommendation notes its H.265 figures are under review. For UHD it asks for 250 and 500 megabits intra.

Real formats sit around those figures. Apple publishes target rates for ProRes: at 1920 by 1080 and 25 frames, ProRes 422 is 122 megabits per second, ProRes 422 HQ is 184, and ProRes 4444 is 275. Sony's PXW-Z300 records XAVC Intra at about 112 megabits at 1080 and 25 frames, and XAVC Long at 35 to 50 megabits at 1080 and 50 frames.

How much card you need

Bit rate converts to storage simply. Megabits per second times 3,600 seconds gives megabits per hour; divide by eight for megabytes, and by a thousand for gigabytes. In short, megabits per second times 0.45 gives gigabytes per hour.

ProRes 422 HQ at 1080 and 25 frames, 184 megabits per second, comes to about 83 gigabytes an hour, which is the figure Apple publishes. The same codec in UHD at 50 frames runs at 1,475 megabits and fills 664 gigabytes an hour. A four-hour show recorded that way on four cameras needs over ten terabytes before anything is backed up.

Work this out before the show, for every camera and recorder, and add a margin for overruns and for the backup copy. Running out of card is one of the few failures that is entirely predictable.

Bit depth and colour in the recording

Two settings decide how much picture information survives: bit depth and colour sampling, which Camera Theory explains. EBU R 118 requires 10-bit 4:2:2 for its top HD tier. ITU-R BT.2408 notes that 10-bit is adequate for HDR, at bit rates similar to standard dynamic range.

Eight-bit recording is still common on smaller cameras. It is fine for a picture that will not be changed much, and it bands when it is graded, keyed or converted. Record 10-bit for anything that will be graded, keyed, converted to HDR, or shot in a log curve.

Timecode beyond record run and free run

Camera Basics explains record run and free run. Cameras offer more than that. Sony's camcorders, for example, have a mode setting as well as a run setting: preset, which starts from a value you set; regenerate, which continues from the timecode at the end of the previous clip so a card has no repeated numbers; and clock, which follows the time of day.

At 29.97 frames per second there is one more choice, drop frame or non-drop frame. The frame rate is slightly slower than 30, so timecode that counts 30 frames to every second drifts behind the clock, by 3.6 seconds every hour. Drop-frame timecode corrects this by skipping two frame numbers at the start of every minute, except every tenth minute. No pictures are dropped; only labels are skipped. Drop-frame timecode is usually written with a semicolon before the frames, as 10:00:00;00. At 25 and 50 frames there is no drift, and no drop frame.

Every device on a show must use the same choice. A camera in drop frame and a recorder in non-drop frame will disagree by tens of seconds over a long day. Cameras also carry user bits, a few spare characters in the timecode often used for a date or a camera number, which help editors keep cards apart.

Recording on a multi-camera show

Larger shows record in several places at once: the programme from the switcher, each camera as an ISO, and often a backup of each. The camera's own card may be the backup, or the only recording. Find out before the show which recordings your camera is part of, so you know whether starting and stopping your recording is your job or somebody else's.

Many cameras can also record a small, low-bit-rate copy alongside the main recording, called a proxy. Apple's ProRes Proxy, for example, runs at 38 megabits per second at 1080 and 25 frames. Proxies let an editor start cutting quickly on a laptop while the full-quality files are copied.

What goes wrong

Practice it

A two-day conference will be recorded on three cameras in ProRes 422 HQ at 1080 and 25 frames, eight hours a day.

Operating Craft

Smooth is the start. On a live show the shot also has to arrive on time, at the right size, on a subject that will not wait for you.

Smooth, consistent, professional operation.

Smooth operation

Beyond the image settings, professional camera work requires smooth operation — controlled, steady camera moves (pans, tilts, zooms) that look intentional and professional, not shaky or jerky. The operator moves the camera smoothly and deliberately (starting and stopping moves cleanly, keeping the frame steady, executing moves at the right speed), so the footage looks polished. Smooth operation is a craft skill built through practice and good technique (and good support — tripods, heads, rigs).

Shaky, uncontrolled camera work looks amateur and is often unusable; smooth, controlled operation looks professional. The operator develops the technique and steadiness to operate smoothly. Smooth operation is a hallmark of a good camera operator.

Consistency and reliability

Professional camera operation is also consistent and reliable — delivering good images consistently (matched exposure, colour, and framing shot to shot and across cameras), and reliably (getting the shot right every time, not hit-or-miss). In multi-camera work especially, the cameras must match (so cutting between them looks seamless), which requires consistent settings and operation. Consistency and reliability are what make an operator trustworthy on professional work.

This means setting up correctly and matching to the other cameras/shots, and operating reliably so the director and team can count on every shot. A consistent, reliable operator is valuable; an inconsistent one is a liability. Combining exposure, focus, colour, framing, smooth operation, and consistency is intermediate camera operation.

Try this. With any camera (even a phone with manual controls), practise controlling exposure, focus, and framing deliberately, and try a smooth slow pan. Notice how much more professional a controlled, well-exposed, sharp, smoothly-moved shot looks than an automatic, shaky one.

Zoom and focus demands on a box lens

A studio or sports camera with a box lens is not operated from rings on the lens. The lens is driven by motors, and the operator drives the motors from two controls clamped to the pan bars: a zoom demand under the right thumb and a focus demand under the left hand. Camera Basics introduces the parts. This section is about how they behave, because they do not behave like rings.

The zoom demand is usually a rate control. Push the rocker a little and the lens zooms slowly; push it further and it zooms faster; let go and it stops. The position of your thumb sets the speed, not the focal length. That is what makes smooth zooms possible, and it is why a nervous thumb makes jerky ones.

The focus demand is usually a position control. Each position of the wheel corresponds to a focus distance, so turning the wheel a quarter turn always moves focus by the same amount, as a ring would. That lets an experienced operator find a mark by feel.

Both demands have settings, usually on the demand itself or in the lens menu: the maximum zoom speed, a response curve that makes small movements gentler, and the direction of travel. Operators who move between cameras on a show set these the same on every position, or set them to their own taste at the start of each day. Finding the zoom reversed on air is the most avoidable surprise in live camera work.

Most box lenses also store positions. A frame preset or shot memory records a zoom position, and sometimes a focus position, and returns to it at a chosen speed when a button is pressed. Used off air, it gets you from a wide to a tight close-up in a fraction of a second. Used on air, it moves at a fixed speed that rarely suits the moment, so most operators keep presets for repositioning between shots.

Fujinon's manual for its XA99x8.4 sports lens shows what these lenses are. It zooms from 8.4 to 832 millimetres, a 99 times range, and its built-in 2x extender doubles that to 1,664. The zoom servo can cover the whole range in about one second, the focus servo in about 0.8. Its aperture is f/1.7 up to 341 millimetres and falls to f/4.15 at the long end: the ramping Camera Theory describes, more than two and a half stops darker on the tightest shot. With the extender in, every f-number doubles, f/3.4 to f/8.3, which is two stops more.

The same maker's zoom demand shows how much of the feel can be set. It offers three response curves: one that starts faster near the centre of the thumb ring, one that starts slower and accelerates at large movements, and one between, each adjustable. It has a dead zone setting for how far the ring moves before anything happens, zoom limits at the wide and tele ends, and presets that recall zoom and focus together or separately. Its switches can be assigned to presets, an intercom channel, or F-number hold. Settings for several operators can be stored as users and locked.

Zoom speed, and arriving on time

A zoom on air is a move, and like any move it needs a start, a speed and an end. Camera Basics teaches the reason for a move. The intermediate skill is landing it: the zoom should ease in, travel at an even speed, and ease out exactly on the final frame, without overshooting and creeping back.

Long lenses make this harder. At the long end of a box lens the view is so narrow that any overshoot or wobble on arrival fills the screen, and the operator is watching a picture that is itself magnified a hundred times. Experienced operators judge the zoom by how the picture changes, not by how far the thumb has moved, and ease the demand off well before the final frame so the last part of the move settles rather than stops.

Arriving on time also means arriving before the call. A director who says "standby two, tight on the guitarist" will cut in a few seconds. If the zoom is still moving when the take comes, the shot arrives moving. Make the size change off air, quickly, and spend the remaining time on the small things: headroom, focus, the edges of the frame.

Following action you cannot predict

A presenter on a marked path is easy. A keynote speaker who wanders, a dancer, a footballer or a guitarist who suddenly walks to the edge of the stage is not, and on a long lens they can leave the frame in a moment.

The working rules are the same across sport and shows. Frame a little looser than ideal on anything that might move, because a slightly loose shot that keeps the subject is better than a perfect one that loses them. Leave more space on the side the subject is facing or moving toward. Pan with the subject's body, not their hands or head, because the body moves more predictably. And keep both eyes working where the camera allows it: one on the viewfinder, and an awareness of the room, so you see the subject start to move before they reach the edge.

When the subject does leave the frame, widen first and then find them, rather than swinging a tight shot around the stage looking for them. A wide shot that catches up is usable. A tight shot searching is not.

Watching the event, not only the frame

The operators directors rely on most are not the ones with the smoothest hands. They are the ones who see what is about to happen. The drummer counting in before the solo, the presenter glancing at the next slide, the award winner's family standing up, the striker shaping to shoot. Every one of those is a shot the director will want, and the operator who is already there when the call comes looks like a mind reader.

Anticipation is mostly preparation. Read the running order, so you know what comes next. Watch rehearsal, so you know who moves where. Learn the music, so you know when the solo is coming. And watch people, because a body turns before it walks.

How to do it

1. Check the direction of both demands: push the zoom rocker and turn the focus wheel, and confirm the lens moves the way you expect.
2. Set the maximum zoom speed and the response curve to suit the show.
3. Zoom fully in on the furthest subject you will shoot, focus, then zoom out and confirm the shot stays sharp. If not, call the vision engineer to check back focus.
4. Store presets for the shots you will need fast, such as the lectern close-up and the stage wide, and test each one at the speed you set.
5. Note which way the extender switch sits, and remember that the extender costs light.

What goes wrong

Practice it

You are on a box lens at the back of a conference hall. The running order says the chief executive speaks next, then walks to the centre of the stage for a product reveal.

Working Ahead of the Director

Camera Basics teaches you to answer the director's call. This chapter is about having the shot before the call is made.

Seeing the show the way the director does

A director watching a row of monitors is choosing between pictures, and can only choose from what the cameras are showing. Every camera that sits on a usable shot is an option. Every camera that is moving, searching or soft is not. The single most useful thing an operator can do is keep their camera an option as much of the time as possible.

That means the time off air matters as much as the time on it. When your tally goes out, the director has not finished with you. They are about to need the next shot, and the quicker yours is steady, sharp and different from the camera on air, the more likely it is to be used.

Directors also build a programme from contrast. Cutting from a close-up on one camera to a close-up of the same person on another, slightly different, looks like a mistake. Cutting from a wide to a close-up, or from the speaker to a listener, looks like a decision. So an operator who knows what the camera on air is showing offers something that contrasts with it.

Offering shots

On most live shows operators are expected to offer shots without being asked. When you are clear, find a good shot that the programme might need, frame it, hold it steady, and leave it there long enough for the director to see it and decide. A shot offered for two seconds and then abandoned is not an offer.

Good offers are the shots nobody else is covering: a reaction in the audience, the second presenter who is about to speak, the guitarist's hands, the screen the speaker is pointing at. Watch the programme monitor or the return feed so you know what is already on air, and offer what is not.

Two things make offering go wrong. Offering shots that break the show's rules, such as a camera crossing the line or showing an audience member who did not agree to be filmed, wastes the director's attention. And moving to a new offer just as the director takes the last one puts a moving shot on air. Once you have offered, hold until you are told or until the moment has clearly passed.

Your camera's job in the coverage

Every camera on a multi-camera show has a role, usually written on its camera card: the wide, the lectern close-up, the handheld on stage, the audience reverse. The role tells you what to protect. The wide camera's job is to always have a clean wide the director can cut to in an emergency, so it does not wander off to offer close-ups. The close-up camera's job is to have the speaker, so it does not leave the speaker to find a reaction unless the director asks.

Knowing the roles of the other cameras matters as much. If camera three is the audience camera, your audience shot is a duplicate. If camera one has the wide, you can go tighter than you otherwise would, because the safety shot is covered.

Recorded cameras, and the shot that is always used

Many shows now record every camera separately, so an editor can cut a better version later. Each separate recording is called an ISO recording, or just an ISO, for isolated. On an ISO-recorded show nothing your camera does is ever really off air. A wandering search, a whip pan or a moment of soft focus between shots is on the recording, and an editor looking for a reaction may find that instead.

The habits are simple. Move between shots deliberately, not in a hurry. Hold every shot long enough to be usable, at least a few seconds, even when the director has moved on. And if you know a moment on your camera was unusable, mention it in your camera log so the editor does not have to find out.

The vision engineer is on your side

On a show with system cameras, the operator frames and focuses, and a vision engineer, also called a shader or racks, controls the picture from a remote control panel: iris, black level, gain and colour. The operator who understands that relationship works with it rather than around it.

Tell the vision engineer what is coming. If you are about to pan from a dark audience to a brightly lit stage, or from a presenter to a window, a quick word on the engineering channel lets them ride the iris with you. If your picture looks wrong in the viewfinder, say so rather than guessing, because the viewfinder may be set differently from the engineer's monitor. And never change a setting the engineer owns without asking, because on a matched show one camera changed is a camera that no longer matches.

When there is no director

Not every job has a director. A single-camera recording of a conference, a camera feeding IMAG with nobody switching, or a livestream with one operator doing everything all put the director's decisions in your hands.

The same principles apply, run in your head. Know what the audience needs to see: the person speaking, what they are showing, and the reaction to it. Move between shots when nothing important is happening, never during a key sentence. And hold longer than feels natural, because without a director to cut away, every reframe is seen.

What goes wrong

Practice it

You are camera two, the lectern close-up, on a five-camera awards show. Camera one has the wide, camera three is the audience. The presenter is reading the nominees.

IMAG, Broadcast and Stream

The same camera can feed a ten-metre screen in the room, a television at home and a phone on a train. Each one wants a slightly different picture.

Image magnification

IMAG, image magnification, puts the camera's picture on large screens in the same room as the event. Its audience can see the real person on stage and the screen at the same time, and that changes everything about how it is shot.

The audience already has the wide shot, because they are sitting in it. IMAG is there to show what they cannot see from their seats: the face, the expression, the hands on the guitar. So IMAG is mostly close-ups and mid shots, steady and well held. A wide shot on IMAG shows the audience a smaller version of what they are already looking at.

Delay matters more than anywhere else. The audience hears the voice directly and sees the lips on the screen, so every frame of delay through cameras, converters, switchers and processors puts the picture further behind the sound. IMAG systems are designed to keep that delay to a few frames, and anything that adds processing, such as a frame-rate converter or a streaming encoder, is kept out of the IMAG path.

The screen is enormous, so every fault is enormous. A soft focus that would pass on a monitor is plain on a ten-metre screen, and so is a small wobble in the frame. IMAG operators frame and focus to a tighter standard than they would for a recording.

IMAG also shows things to people who are in the room with the subject. A close-up of a presenter's nervous hands, or of an audience member who did not expect to be on the big screen, is seen by everyone, including that person. The usual rule is to show the stage and people who are part of the show, and not to linger on the audience unless the event wants it.

Broadcast

A broadcast audience has nothing but the screen. They do not know the room, so the programme has to explain it: a wide to show where we are, close-ups to show who is speaking, reactions to show how it landed. That makes the wide shot far more important than on IMAG, and the editing rhythm slower.

Broadcast is also delivered to a technical specification. Signal levels, colour and safe areas for graphics are all checked, and the vision engineers keep every camera within them. The operator's part is to leave room for the graphics that will be added, keep the camera's settings where the engineers set them, and frame for a screen that might be anything from a large television to a small one.

Streaming

A stream goes through an encoder that compresses the picture heavily before it reaches viewers, and many of them watch on a phone. Both facts change the shot.

On a small screen, wide shots turn into tiny people, so streams favour closer framing and simpler compositions with one clear subject. Text on a slide that is readable in the room may be unreadable on a phone, so a shot of the screen needs to be tight enough to read.

Compression struggles with anything that changes a lot from one frame to the next. Fast pans, whip zooms, flashing lights, haze, confetti and the grain from high gain all cost data, and when the encoder runs out of data the picture breaks into blocks. A clean, well-lit picture moving smoothly streams far better than a noisy, busy one. That is one more reason to light the room rather than add gain.

Streams usually run several seconds behind the event, sometimes much more, so delay that matters on IMAG does not matter here. What matters is that the stream's sound and picture stay together, which is the encoder's job.

Shooting for all three at once

Many events feed IMAG, a recording and a stream from the same cameras. The demands conflict: IMAG wants close-ups with no delay, the stream wants close framing and calm movement, and the recording wants a wide to explain the room.

Larger shows separate them, with their own switchers or even their own cameras for IMAG and for the programme. Smaller shows compromise: the close-up cameras serve IMAG and the programme together, and the wide camera stays on the programme. Ask at the production meeting which outputs your camera feeds, because that decides how tight, how fast and how steady you should be.

What goes wrong

Practice it

A product launch feeds two eight-metre IMAG screens, a livestream and a recording, from three cameras and one switcher.

Specialist Cameras

Beside the cameras on tripods and pedestals, most shows now carry a few that work differently: remote heads, gimbals, wireless cameras, tiny point-of-view cameras and cinema cameras pressed into live service.

PTZ and robotic cameras

A PTZ camera, for pan, tilt and zoom, has its head built in and is driven entirely from a controller. Panasonic's AW-UE160, Sony's BRC series and Canon's CR-N series are typical. A robotic head, from makers such as Vinten or Ross, does the same job for a full broadcast camera and lens. One operator at a control panel may run four or six of them.

Camera Basics covers recalling presets off air. The intermediate skills are the ones that make remote cameras look operated rather than parked. Build presets during rehearsal, named for what they show, and check them again after the set is dressed, because a moved lectern makes every preset wrong. Set preset speeds slow enough that a recall caught on air looks like a deliberate move. Use the joystick's speed scaling so small corrections are small. And watch every remote camera on a multiviewer, because there is nobody at the camera to see that a presenter has stepped out of frame.

Remote cameras are usually connected over a network, carrying control, and often the picture itself, by NDI, SRT or SMPTE ST 2110, with power over the same Ethernet cable. Each of those adds delay, from a frame or two to much more, which matters if the camera feeds IMAG.

Gimbals and stabilisers

A gimbal, such as a DJI Ronin, uses motors on three axes to hold a camera level and steady while the operator walks. It gives the smooth moving shots that once needed a Steadicam, which does the same job with a counterbalanced arm and a vest instead of motors.

A gimbal is only as good as its balance. The camera must be balanced on every axis before the motors are switched on, so the motors hold it rather than fight it. A badly balanced gimbal drains batteries, vibrates, and drifts. And every change, a different lens, a new cable, a filter, changes the balance.

The motors smooth rotation, not up-and-down movement. The bounce of each footstep still reaches the picture, and the cure is in the walk: knees soft, steps rolled heel to toe, the camera held close to the body. Gimbals also offer modes that decide which axes follow the operator's movement and which stay locked. Know which mode you are in before you go on air.

Wireless cameras

A wireless or RF camera sends its picture by radio, from a transmitter mounted on the camera to receivers around the venue. It is what lets a handheld camera follow a presenter into the audience, or a camera ride on a stage lift, without a cable.

Wireless brings three things a cabled camera never has to think about. Delay: the picture is encoded and decoded, which adds time, so a wireless camera feeding IMAG can look late against the others. Coverage: the signal can fade or break up behind obstructions, so walk the route in rehearsal and find the dead spots. And frequencies: every radio system in the venue, from microphones to intercom, has to be coordinated so they do not interfere, which is the RF coordinator's job and must be booked in advance.

Wireless cameras are also often painted remotely through a data link, and many carry tally and talkback by radio too. Check each of those in rehearsal: a camera with a picture but no tally is a camera whose operator cannot tell when they are live.

Point-of-view and locked-off cameras

Point-of-view cameras are the tiny ones mounted where no operator could stand: on a drum kit, inside a lectern, on a guitar headstock, above a stage. Many have small sensors and fixed or remotely controlled lenses, and some can be painted from a control panel like a full system camera.

Small sensors give deep focus, which suits a camera nobody can refocus, and they also give more noise in low light, which a dark stage will show. Mount them so they cannot move, cable them so performers cannot catch the cable, and frame them for the moment they are for. A drum camera is wanted for the solo, so frame for the solo, not for the whole kit.

Cinema cameras on a live show

Events that want a filmic look increasingly put cinema cameras, such as Sony's FX6, FX9 or Venice, or large-sensor system cameras such as Sony's HDC-F5500, into the live camera line-up. Their large sensors give the shallow depth of field and soft backgrounds that a 2/3-inch broadcast camera cannot.

That look has costs, and the operator pays most of them. Shallow depth of field makes focus far less forgiving, which on a moving subject often means a second person pulling focus with a wireless follow-focus. Cinema lenses are usually primes or short zooms, so changing shot size means moving or changing lens. The camera may need converters to fit a broadcast system, each adding delay. And its colour must be matched to the broadcast cameras around it, which the next chapter covers.

What goes wrong

Practice it

A concert uses four system cameras, two PTZ cameras on the balcony, a wireless handheld on stage, a point-of-view camera on the drums and a Sony FX6 on a gimbal.

Matching the Other Cameras

Viewers never see one camera. They see the cut between two, and any difference between them shows at that moment.

What has to match

When a director cuts from one camera to another, the viewer compares them instantly. A face that turns warmer, a black that turns grey, a highlight that goes from detailed to flat white, or a picture that suddenly looks sharper or noisier: each of these reads as a jolt. On a well-run show none of them happens, because the cameras have been matched.

The things that must agree are brightness, black level, colour, contrast and highlight handling, sharpness, and noise. On system cameras most of these are set and held by the vision engineer, as Video Engineering teaches. The operator's part is to make matching possible, and to not undo it.

The operator's part

Start the same. Every camera on a matched show should run the same frame rate, the same shutter, the same picture profile or scene file, and the same recording format, unless someone has decided otherwise for a reason. A camera set to a different shutter will show motion differently, and one with a different detail setting will look sharper or softer at every cut.

Balance together. White balance on camcorders and cinema cameras is done by the operators, so do it at the same time, under the show's actual lighting, on the same chart or card, held where the presenter will stand. Balancing each camera on whatever white thing is nearest gives as many slightly different whites as there are cameras.

Choose the same way. Where the operator controls exposure, agree what the faces should sit at on the waveform, and expose every camera to that, rather than each operator judging by eye on a different viewfinder. Where ND and iris are the operator's, agree on them too, because different apertures give different depth of field, and a background that is soft on one camera and sharp on another is a mismatch as well.

Frame alike. Matching is not only technical. Two close-ups of the same person on two cameras should be about the same size, with about the same headroom, so the cut between them looks intentional. Camera cards and a quick look at the other cameras on the multiviewer keep framing consistent.

Different cameras, different colour

Cameras from different makers, or even different models from the same maker, see colour differently, because their sensors, filters and processing differ. A Sony system camera, a Panasonic PTZ and a cinema camera can be balanced to the same white and still show a different red jacket and a different skin tone.

The cure is in the colour matrix, which vision engineers adjust to bring the cameras together, and in choosing cameras that match for the important positions. The operator's part is to report what they see. If the PTZ on the balcony makes the presenter look greener than camera two does, say so before the show, not after.

Checking a match

The quickest check is the one the audience makes: cut between the cameras. During rehearsal, ask for each camera to be cut against the main close-up camera on the same subject, and watch the cut, not the cameras side by side. Differences that are invisible side by side jump out at a cut. A chart held in the presenter's position and shot by every camera in turn shows on a waveform and vectorscope exactly where they differ.

How to do it

1. Agree the frame rate, shutter, picture profile and recording format, and set every camera to them.
2. With the show lighting on, hold a white or grey card where the presenter will stand, and white balance every camera on it.
3. Agree where faces sit on the waveform, and expose each camera to it on the same person or chart.
4. Agree ND and iris where depth of field must match.
5. Ask for each camera to be cut against the main close-up camera, and note any differences in colour, brightness or sharpness.
6. Report remaining differences to whoever can correct them, before doors.

What goes wrong

Practice it

A conference uses three Sony FX6 cameras run by their operators and two Panasonic PTZ cameras. There is no vision engineer.

Toward Professional Operation

Building real camera craft and a video career.

The intermediate operator

Pulling it together, Camera Intermediate develops real camera craft — mastering exposure (the triangle), focus (including following focus), white balance, and framing, and developing smooth, consistent, reliable operation. The intermediate operator controls the image deliberately and operates with craft, delivering good, professional, consistent images. This is the level that makes a camera operator genuinely capable and employable, and it is the foundation for advanced and specialized camera work (broadcast, film, live). Camera Intermediate mastered is operating a camera with real professional craft.

This builds on Camera Basics toward professional camera operation across the video world (broadcast, live events, film, corporate). A capable intermediate operator can shoot good footage reliably and is ready to grow into specialized and senior camera roles. Developing intermediate camera craft is a key step in a video career.

Developing your camera craft

Developing your camera craft means shooting a lot (practice is everything in camera work), mastering the controls (exposure, focus, colour, framing) until they're second nature, developing smooth operation and a strong eye, and learning from good footage and operators. From here you can specialize (broadcast camera, film camera/AC, live/IMAG, studio) and advance (camera operator to senior operator to DP/director of photography in some paths). The best operators have total control of the image, smooth reliable operation, and a great eye, built through much shooting. Deliberately developing these is the path forward.

From here, growing as a camera operator means shooting more and more demanding work, specializing in a camera discipline, and building toward senior camera and cinematography roles. Camera work is a large, varied video field with many paths. The skills this book brings together — exposure, focus, colour, framing, smooth consistent operation — are the foundation; mastery comes from much shooting and developing your eye. Keep shooting, keep mastering your controls and craft, and you grow into a professional camera operator.

Try this. Pick a camera discipline that interests you (broadcast, live events, film, studio) and consider what camera skills it emphasises. Then practise the fundamentals — exposure, focus, framing, smooth moves — that underpin all of them.

What separates a senior operator

Senior camera operators are rarely the ones with the most impressive single shot. They are the ones whose cameras are always usable. Their settings are right before the show, their shot is steady when the director looks at it, they see the moment coming, they say the right thing on comms and nothing else, and they leave every camera they touched as matched as they found it. Directors book them because those things are rarer than talent.

They also know the system around the camera: what the vision engineer is doing, how the signal reaches the switcher, which outputs the camera feeds, and what the recording is for. That knowledge lets them prevent problems instead of reporting them.

Practising on purpose

Shooting a lot builds craft only when the practice is aimed. Four drills cover most of what this book teaches. Land a zoom: from a wide to a tight close-up on a fixed point, ten times, easing in and out without creeping back. Follow the unpredictable: have someone walk randomly while you keep them framed on a long lens. Find focus by feel: learn how far your focus wheel turns between common distances. And offer: during any multi-camera show you can watch, decide which shot you would have offered next, and see whether the director cut to it.

Practice it

You want to move from operating handheld cameras on small events to operating a box lens on large broadcast shows.

Glossary

Where to Go Next

Video Engineering takes over where the matching chapter stops: the camera control unit, painting a set of cameras, and the signal system around them. Camera Direction and IMAG Direction are the other side of the headset. And Visual Aesthetics for AV is about how the picture should look, now that you can make it look the way you choose.

The first ninety days

Five things. Open your camera's paint or picture menu and change each setting in this book one at a time, watching the waveform. Expose a grey card in every curve your camera offers and write down where it sits. Work out the card space for your most common job from the bit rate. Set up the box lens demands on the next show you work, both directions, speed and presets. And on every multi-camera show, count how often your camera was a usable option when the director looked.

What the gear actually looks like

Every device named in this book, in one place. Recognition is the whole point of Chapter 4 — use this page to attach a picture to each name before you meet it on a dock at six in the morning.

Audio

Dynamic microphone

Condenser microphone

Wireless handheld

Radio rack

Mixing console

Stage box / snake

Power amplifier

Main loudspeaker

Subwoofer

Stage monitor wedge

In-ear monitors

DI box

Video

Camera

Video switcher

Scaler / processor

Projector

LED wall panel

Confidence monitor

Lighting & control

Conventional fixture

Moving-head fixture

Lighting console

Dimmer

Show controller

Media server

Networking & infrastructure

Network switch

Patch bay

Equipment rack

Power distribution

Cable tester

A properly coiled cable

Connectors, close up

XLR

TRS / TS ¼-inch

SDI on BNC

HDMI

Cat5e / Cat6 (RJ45)

Fiber optic

DMX (5-pin XLR)

AC power — not signal