1. Your First Usable Shot

You type a sentence. A few minutes later a computer hands you five seconds of moving picture that nobody filmed. There is a woman, a street, a grey morning, and she sets a parcel down on a step. No camera was carried anywhere. No actor was hired. And when you watch it closely, her hand goes half an inch into the parcel at the moment she puts it down.

That gap — between the shot you asked for and the frames you actually got — is what this chapter is about.

The thing making those frames is a video generation model. It is a program trained on an enormous amount of existing film and footage, which has learned to produce new footage from a description. You give it words. It gives you a short clip. Some of these models run on a computer you own, which needs a powerful graphics card and some patience with software. Others run on somebody else's computers and you reach them through a web browser, the same way you reach your email. Those are the hosted services, and that is where we are starting, because you can be generating today without installing anything.

Hosted generation sits between two older ways of getting moving pictures. On one side is filming: cameras, locations, people, permission, time. It gives you total control over what is in the frame and it costs real money and real days. On the other side is animation: total control again, and enormous labour per second. Generation is fast and cheap by comparison, and in exchange you give up a lot of control. You cannot tell the model to move the actor's hand two inches left. You can only describe, look at what came back, and describe again. That trade is why it is useful to a person working alone, and why the working habit in this chapter matters more than any clever wording.

So here is what we will do. We will take the smallest brief that still counts as a shot. We will open one hosted service, and I will tell you exactly how you get in and what it will and will not do for you. We will build a prompt out of named parts, so you know why each word is there. We will generate. Then we will spend real time looking at what came out. We will make one targeted change, generate again, and compare. And we will finish with a clip saved where you can find it, next to a written record of how it was made.

Let me give you the brief first, because everything else serves it.

There is a courier named Mara. She rides a bicycle around a city delivering packages. Early one morning she leaves a small paper parcel on the stone step of a bakery that has not opened yet. That is the whole shot: one person, one plain physical action, one place, about five seconds long.

Mara is invented. The bakery is invented. The story she belongs to is invented, and I made it up for teaching, not because it exists somewhere. I will keep coming back to her, because a small story with one character and one clear action makes problems visible. When her jacket changes colour between shots, you can see it. That is the only reason she exists.

Notice how little is in that brief. One action. No dialogue. No second person. Nothing that requires the model to understand a relationship or a joke. When you are learning, a small brief is not a compromise. It is the instrument you are measuring with.

Now let us get you into a tool.

The service we will use is Runway. You reach it in a web browser at its web app — that is, you log into a site rather than installing a program — and its documentation recommends Google Chrome. There is no local installation, no graphics card requirement, and no code. You register with an email address, or by signing in with a Google or Apple account, and no credit card is required to start.

The exercise uses Gen-4.5, which requires a Standard plan or higher. You can register for free, but the free account does not unlock this model. Each generation spends credits, including takes you reject. Before starting, check the charge shown for a five-second clip and budget for three attempts. If you are not ready to pay, follow the example first and decide afterwards whether to try it.

Now, what does the service actually ask you for?

The input is a text prompt: a plain-language box where you describe what appears in the frame and how it moves. Keep it clear enough that you can find and change one instruction. There is no magic word count. A longer prompt can be useful, but extra detail can also introduce competing requests.

Gen-4.5 can also animate an uploaded image. That is a different starting point: the image supplies the appearance and composition, and the prompt describes the movement. Today we are using text-to-video, so leave the image input empty.

Describe what you want to see. Write static camera rather than no camera shake. Put camera movement in the prompt too: a slow push-in, an orbit, or a camera that stays still. For this exercise, static is enough.

For this exercise, choose a five-second widescreen clip. Gen-4.5 supports durations from two to ten seconds; five gives our small action room to finish without making the first attempt too expensive. We will use the standard seven-twenty picture and the default MP4 download. Leave upscaling and other delivery formats until you have a shot worth keeping.

We are judging the picture first. Leave sound design for a later pass, when you know which take you will keep. A convincing image can make you imagine the paper crunch and the birds before you have added either.

In Runway’s Tool Mode, select Video and choose Gen-4.5 from the model selector. Leave the image input empty and enter the text prompt. Set the duration to five seconds, keep the widescreen frame and default MP4 output, and check the credit charge before generating.

Now the prompt. I want to build it out of named parts, because from here on those names are how you and I will talk about a shot.

A shot is one continuous run of camera, from the moment it starts recording to the moment it stops. Cut, and you have a new shot. A scene is bigger: it is the action in one place and one stretch of time, and it is usually made of several shots. Mara leaving the parcel at the bakery is a scene. The five seconds of her hands putting it down is a shot. Everything a generation model gives you is a single shot. That is worth sitting with, because it means a scene is something you assemble, not something you ask for.

The parts of the shot, then. First the subject: who or what the frame is about. Mara, a young bike courier in a grey rain jacket. Second the action: the one physical thing that happens. She sets a small paper parcel down on a stone step. Third the framing: how much of the subject the frame contains. A medium shot takes a person from roughly the waist up. A close shot is head and shoulders, or a pair of hands. A wide shot puts the whole figure in its surroundings. Framing is a decision about what the audience is allowed to notice.

Fourth, camera movement: whether the camera holds still or travels. Still is called static, and for a first shot it is the right choice, because a moving camera gives the model one more thing to get wrong. Fifth, light: where it comes from and what it feels like. Soft light from the left, early morning. Naming a direction gives the model something to be consistent about. Sixth, duration, which we have set at five seconds. Seventh, aspect ratio, which we have set at widescreen.

Here is the prompt I typed, at forty-four words:

Medium shot of a young bike courier in a grey rain jacket. She crouches and sets a small brown paper parcel down on a stone bakery step, then straightens up. Empty city street at dawn, wet pavement, soft light from the left. Static camera.

Each sentence has a job: subject and framing, action, setting and light, then camera movement.

Before we generate, one thing about that action, because it is the difference between a nice clip and a useful one. A shot can look handsome and still be no use to whoever cuts the film together, and often that is because the action does not finish. If Mara begins to crouch and the clip ends while she is still bending, an editor has nothing. They cannot cut away from a half-finished movement without it looking like a mistake. They need to see the parcel land. That is why the action is simple: crouch, place, straighten. A complete action in five seconds means one small movement, not three. Simplicity here is not modesty. It is what makes the footage cuttable.

Press Generate.

Now the part that decides whether any of this was worth the credits: look at what came back.

The temptation is to watch it once, feel pleased or disappointed, and move on. Do not. Watch it at least three times, and give each pass a job. This is a habit, and it is the habit that separates directing from gambling.

First pass, watch the action only. Does it start and does it finish inside the five seconds? In my clip, Mara crouches at about half a second, the parcel is on the step by three and a half, and she is upright again by four and a half. There is half a second of nothing at the end. That is fine. A little air at the end is useful.

Second pass, watch the contact points. Where a body touches a thing, generation models are weakest. Do the fingers close around the parcel, or does the parcel simply travel with the hand? Do her feet meet the pavement, or float a few pixels above it? In my clip her right hand passes into the corner of the parcel as she lets go. Not badly. But if you froze the frame and pointed, someone would see it. That is a real flaw and I am going to name it: broken contact at the moment of release.

Third pass, watch the face. Generated faces drift. Does she look like the same person at second one and second four? Mine holds until she straightens, and then her face changes — the jaw narrows and the eyes sit differently. That is called identity drift, and it matters more than it looks, because a story needs the same person to keep being the same person.

Then check the light. Does it keep coming from one direction? Mine does, mostly, though the puddles brighten oddly halfway through as if a second sun arrived. And check for text. Models love to invent lettering, especially on signs and shopfronts. There is a smear of not-quite-letters across the bakery window in mine — not words, just the shape of words. In a finished film that is a distraction. Then listen once, so you know what sound work the saved clip will need.

So one generation, five flaws noticed, none of them fatal. Here is the attitude I want you to have about that. Those seconds you decide not to use are not failure. They are rejected work, and rejected work is a normal cost of producing one accepted shot. A working production counts credits and seconds spent per usable shot, not per generated shot. If you throw away four clips to keep one, your cost for that shot was five clips. That is a number, not a shame. People who hide their rejected work cannot estimate anything, and they end up believing every generated second is footage.

Now, one change. Exactly one.

The flaw I am going after is the broken contact, because that is what an audience notices without knowing why. And I am changing one element of the brief to address it: the framing. I will go from a medium shot to a close shot on the hands and the parcel. The reasoning is that a close frame gives the model far more pixels for the thing that has to be right, and takes the face — the other flawed thing — out of the frame entirely.

I am changing the framing and keeping the action, light and camera movement as close as possible to the first request. The new generation may change other things too. This comparison will tell us whether the new take works; one pair of clips cannot prove that framing alone caused the improvement.

The revised prompt, at thirty-nine words:

Close shot on the hands of a young bike courier in a grey rain jacket. Her fingers set a small brown paper parcel down on a wet stone step and release it. Soft light from the left. Static camera.

Generate again.

The second clip fixes the thing I aimed at. The fingers open, the parcel sits, the shadow under it lands where the light says it should. The face problem is gone because there is no face. And the revision brought in a new problem: in this version the parcel is already resting on the step when the clip begins, and her hands only let go. The placing is not there. As a piece of storytelling it is weaker, because a hand releasing something already at rest does not read as delivery.

I want to be straight with you: a targeted revision can come back worse, and often the way it comes back worse is exactly this — you fix what you named and lose something you were not watching. When that happens you have three honest options, and which one you pick is a budget decision as much as a craft one. You can accept the earlier clip, flaws and all, if its flaws are small enough to survive on screen. You can run the revision once more with the same single change plus one sentence protecting what you lost — in my case, adding that the hands lower the parcel onto the step. Or you can accept the new clip and adjust the brief around it, deciding that this shot is about release and something else will carry the placing.

I ran the third version: close shot, hands lowering the parcel onto the step and releasing it, same light, static camera. That one gave me the full movement, clean contact, and consistent light, with a faint wobble in the paper's edge in the last few frames. That is the one I am accepting. Three generations, fifteen seconds produced, five seconds kept.

One more thing about returning to a result: keep the accepted clip and the rejected takes together. Save the prompt and settings for each, then add your reason for keeping or rejecting it. Those notes let you resume the work without reconstructing your decisions. They do not promise an identical new generation.

So here is the finish line for this chapter, and it has two pieces sitting next to each other.

The first is the clip. Download it with the download arrow on the preview pane; it comes down as an MP4. Put it in a folder you will find again, with a name that says what it is rather than what the service called it — something like Mara, bakery step, close on hands, take three.

The second is a short written record beside it, in a plain text file or a note. Write down the service and the model version. Write down the prompt, exactly as you typed it. Write down the duration, aspect ratio and generation cost. Then one line on what you rejected and why: two earlier takes, first for a hand passing through the parcel and a face that changed, second because the parcel started already on the step.

Six lines. Two minutes of typing. And that record is the first item in a library you will keep adding to, because the moment you want this shot again, or a shot that rhymes with it, the note is the only thing that lets you start from where you finished instead of from a blank prompt box.

Where generated sound is arriving

We left sound design for later in this exercise. Other workflows combine picture and sound during generation, so when you compare services, check which exact model produces the audio and what you can control.

Hailuo AI released a model called MiniMax H3 at the end of July twenty twenty-six. Its headline is what they call an omni-pass: stereo audio generated in the same text pass as the picture, including room tone, footstep-and-object sound, and dialogue. It also runs single clips up to fifteen seconds at two-K resolution, across six frame shapes from very wide to upright phone. The production task that changes is delivery of a standalone shot — no separate sound step, no separate upscale. The smallest useful thing you could do with it today: pick MiniMax H3 in the Hailuo web console, set fifteen seconds at two-K, and add sound words to the end of your prompt the way you added light words — wind howling, gravel crunching — then listen to what comes back and ask whether the crunch lands on the frame where the foot lands.

Kling AI has done the same thing on a different schedule. Its VIDEO three-point-zero Turbo, updated through the middle of June twenty twenty-six, has native synchronised audio, clips up to fifteen seconds at ten-eighty or two-K, and it reads camera paths directly out of your wording. It has also scheduled its older models, versions one through two-point-one, for full retirement on the fifteenth of September twenty twenty-six, which matters if you have prompts you rely on there. The smallest next action: switch the model selector to three-point-zero, turn the audio toggle on, and write your camera move as an explicit direction — a slow push-in, a low-angle tracking shot — rather than hoping.

We used Runway’s Gen-4.5 for text-to-video. Its Gen-4 Turbo workflow starts from an image instead. If you try that next, choose the starting frame deliberately: it supplies the composition you were describing in words today.

Every one of those tests is the same loop you just ran. Small brief, one change, and look hard at what actually came out.