{"course":{"slug":"ai-video","title":"Video with AI","url":"https://addaly.com/learn/ai-video","level":"intermediate","track":"create"},"lesson":{"slug":"start-from-a-still","title":"The still frame is where the decisions belong","url":"https://addaly.com/learn/ai-video/start-from-a-still","position":11,"of":66,"minutes":9,"answer":"Every decision you can make in a still costs a hundredth of what it costs to make in a clip, so fix the frame before you ever spend on motion.","keyIdea":"Every decision you can make in a still costs a hundredth of what it costs to make in a clip, so fix the frame before you ever spend on motion."},"body":{"format":"markdown","text":"## Two ways in, and they are not equal\n\nText-to-video: you type a description and the model invents the framing, the faces, the wardrobe, the colour, the light and the lens, then animates its own invention. You are gambling on a dozen decisions simultaneously, at video prices.\n\nImage-to-video: you supply the first frame. The model's only job is motion.\n\nAlmost everyone doing this professionally works the second way, and the reason is economic before it is aesthetic. A still costs a cent or two and takes seconds. An eight-second clip costs fifty to a hundred times that and takes a minute or two in a queue. Image-to-video moves every hard decision out of the expensive loop and into the cheap one.\n\n> Figure: What one attempt costs, by where you make the decision. Bars (cost of one attempt, where making one still is 1): Make or remake a still 1; Draft clip, low resolution 9; Eight-second final clip 60. Repairing a still by hand is not on this chart because it costs nothing at all, and it is deterministic besides. Every decision you can move out of the right-hand bar and into the left one is money you keep, which is the whole argument for the still-first workflow.\n\nThere is a second reason that matters more. A still is not only cheaper to regenerate: it can be repaired by hand. You can photograph it. You can composite the client's real product into it in Photopea, which runs in a browser tab, opens PSD files and needs no install. You can fix a bad hand in GIMP or Krita. You cannot do any of that inside a video model. Whatever you fix in the still is fixed for the whole clip.\n\nFor making the still, see [ai-images](/learn/ai-images). For repairing it, [image-editing](/learn/image-editing).\n\n## The first frame is a contract\n\nEverything in that frame persists and then degrades. Six fingers in the still means six fingers for eight seconds, getting stranger. A wobbly sign in the still becomes a melting sign. A face you are not quite happy with in the still is a face you will hate by second five.\n\nThe frame also fixes the style. The model will not restyle mid-clip: it animates what it is given. So the still is where you decide whether this is a photograph, a gouache painting or a 3D render, and that decision is then locked in a way no prompt word can override.\n\nThe practical rule: do not animate a still you would not be happy to publish as a still.\n\n## Match the aspect ratio and the resolution\n\nA quiet, common waste of money. You compose a still you like at 3:2. The video model works at 16:9 and either crops it, pads it or stretches it, and the framing you chose is gone.\n\nGenerate the still at the ratio the video model outputs (16:9 for landscape, 9:16 for vertical short-form, 1:1 where a platform wants it) and at or a little above the model's native resolution. Most generate at 720p or 1080p. Feeding in a 4000-pixel still does not buy you detail; it gets downscaled on the way in.\n\n## Start and end frames\n\nSeveral tools let you supply both the first and last frame: Luma's keyframes, Kling's start-and-end frames, and similar controls elsewhere. This is the most underused control in the whole category. It lets you decide where the shot lands, which is what a real camera operator does.\n\nUse it for a reveal, for a match cut into the next shot, or for arriving exactly on a product hero frame.\n\nThe failure mode is worth knowing: if the two frames are too different, the model does not invent a camera move between them, it invents a morph. Keep them the same scene, same light, same lens, with a plausible path between. A push in works. A cut from a kitchen to a beach does not.\n\n## What text-to-video is still good for\n\nIt is not useless. When nothing is decided yet, text-to-video is a fast way to find a look, generate options and see what the model is good at before you commit. Concept phase, not production phase.\n\nThere is also a real trade-off on sound. Some models generate synchronised dialogue and ambience along with the picture, and on several of them that native audio path is tied to text prompting rather than to an image you supply. If you need the model's own audio, you may have to give up the control of a fixed first frame. Decide which matters more for that shot.\n\n## The free path\n\n- Stills: several image tools give a daily free allowance, and open models run in a browser through a Hugging Face Space with no install. See [hugging-face](/learn/hugging-face).\n- Fixing stills: Photopea in a browser handles layers, masks and PSDs on a laptop that cannot run Photoshop. GIMP and Krita if you can install. On a phone, Snapseed for tone and a browser tab for the rest.\n- Video: most hosted tools give a small allowance that refreshes. Spend it on image-to-video, at the lowest resolution, on real shots from your list. You learn far more per free credit that way than by rolling text-to-video prompts.\n\n---\n\nToday: take one shot you want. Make the still. Fix one thing in it by hand: the label, the hand, the sign. Then animate it, and compare that clip to one generated from text alone."},"check":{"question":"A drinks brand's product label comes out as gibberish in every generated clip. The team keeps rewriting the video prompt with more precise wording about the typography. What is the better fix?","options":[{"text":"Add explicit negative-prompt terms for warping and blur so the letterforms stop degrading across frames.","correct":false,"tempting":"Treats a representational limit as an artefact-suppression problem that a parameter can clean up."},{"text":"Generate at the highest available resolution so the letters have enough pixels to resolve correctly.","correct":false,"tempting":"Assumes text failure is a sampling and detail problem rather than a symbolic one."},{"text":"Composite the real label into the first frame in an image editor, then animate that frame, because the video model propagates what it is handed rather than rendering text it never had.","correct":true},{"text":"Switch to text-to-video so the model designs the whole bottle and label coherently from scratch.","correct":false,"tempting":"Moves a decision the team can fully control into one they cannot control at all."}]},"prev":{"slug":"running-video-models-yourself","title":"What it takes to run one on your own card","url":"https://addaly.com/learn/ai-video/running-video-models-yourself"},"next":{"slug":"composition-that-survives-motion","title":"Compose for the frame you will end on","url":"https://addaly.com/learn/ai-video/composition-that-survives-motion"},"authorship":"Founder and series editor: Dr Mustafa Mun. Written by AI models to a written standard. Exam keys checked blind by a second model: 3,665 of 3,666 agreed. Academic review invited.","review":{"status":"Draft, open for academic review"},"cite":"Addaly (2026) 'The still frame is where the decisions belong', in Video with AI. Edited by M. Mun. Addaly. Available at: https://addaly.com/learn/ai-video/start-from-a-still (Accessed: 31 October 2026).","licence":{"name":"CC BY-NC-SA 4.0","url":"https://creativecommons.org/licenses/by-nc-sa/4.0/"},"reader":{"schema":1,"title":"The still frame is where the decisions belong","courseTitle":"Video with AI","module":{"index":2,"slug":"the-first-frame","title":"The first frame, and everything decided in it"},"answer":null,"keyIdea":"Every decision you can make in a still costs a hundredth of what it costs to make in a clip, so fix the frame before you ever spend on motion.","hero":{"kind":"figure","figure":{"t":"figure","label":"Figure 2.1","figure":{"type":"bars","unit":"cost of one attempt, where making one still is 1","title":"What one attempt costs, by where you make the decision","series":[{"label":"Make or remake a still","value":1},{"label":"Draft clip, low resolution","value":9},{"label":"Eight-second final clip","value":60,"highlight":true}],"caption":"Repairing a still by hand is not on this chart because it costs nothing at all, and it is deterministic besides. Every decision you can move out of the right-hand bar and into the left one is money you keep, which is the whole argument for the still-first workflow."},"alt":"Figure: What one attempt costs, by where you make the decision. Bars (cost of one attempt, where making one still is 1): Make or remake a still 1; Draft clip, low resolution 9; Eight-second final clip 60. Repairing a still by hand is not on this chart because it costs nothing at all, and it is deterministic besides. Every decision you can move out of the right-hand bar and into the left one is money you keep, which is the whole argument for the still-first workflow."}},"blocks":[{"t":"h2","id":"two-ways-in-and-they-are-not-equal","runs":[{"text":"Two ways in, and they are not equal"}]},{"t":"p","runs":[{"text":"Text-to-video: you type a description and the model invents the framing, the faces, the wardrobe, the colour, the light and the lens, then animates its own invention. You are gambling on a dozen decisions simultaneously, at video prices."}]},{"t":"p","runs":[{"text":"Image-to-video: you supply the first frame. The model’s only job is motion."}]},{"t":"p","runs":[{"text":"Almost everyone doing this professionally works the second way, and the reason is economic before it is aesthetic. A still costs a cent or two and takes seconds. An eight-second clip costs fifty to a hundred times that and takes a minute or two in a queue. Image-to-video moves every hard decision out of the expensive loop and into the cheap one."}]},{"t":"p","runs":[{"text":"There is a second reason that matters more. A still is not only cheaper to regenerate: it can be repaired by hand. You can photograph it. You can composite the client’s real product into it in Photopea, which runs in a browser tab, opens PSD files and needs no install. You can fix a bad hand in GIMP or Krita. You cannot do any of that inside a video model. Whatever you fix in the still is fixed for the whole clip."}]},{"t":"p","runs":[{"text":"For making the still, see "},{"text":"ai-images","href":"https://addaly.com/learn/ai-images"},{"text":". For repairing it, "},{"text":"image-editing","href":"https://addaly.com/learn/image-editing"},{"text":"."}]},{"t":"h2","id":"the-first-frame-is-a-contract","runs":[{"text":"The first frame is a contract"}]},{"t":"p","runs":[{"text":"Everything in that frame persists and then degrades. Six fingers in the still means six fingers for eight seconds, getting stranger. A wobbly sign in the still becomes a melting sign. A face you are not quite happy with in the still is a face you will hate by second five."}]},{"t":"p","runs":[{"text":"The frame also fixes the style. The model will not restyle mid-clip: it animates what it is given. So the still is where you decide whether this is a photograph, a gouache painting or a 3D render, and that decision is then locked in a way no prompt word can override."}]},{"t":"p","runs":[{"text":"The practical rule: do not animate a still you would not be happy to publish as a still."}]},{"t":"h2","id":"match-the-aspect-ratio-and-the-resolution","runs":[{"text":"Match the aspect ratio and the resolution"}]},{"t":"p","runs":[{"text":"A quiet, common waste of money. You compose a still you like at 3:2. The video model works at 16:9 and either crops it, pads it or stretches it, and the framing you chose is gone."}]},{"t":"p","runs":[{"text":"Generate the still at the ratio the video model outputs (16:9 for landscape, 9:16 for vertical short-form, 1:1 where a platform wants it) and at or a little above the model’s native resolution. Most generate at 720p or 1080p. Feeding in a 4000-pixel still does not buy you detail; it gets downscaled on the way in."}]},{"t":"h2","id":"start-and-end-frames","runs":[{"text":"Start and end frames"}]},{"t":"p","runs":[{"text":"Several tools let you supply both the first and last frame: Luma’s keyframes, Kling’s start-and-end frames, and similar controls elsewhere. This is the most underused control in the whole category. It lets you decide where the shot lands, which is what a real camera operator does."}]},{"t":"p","runs":[{"text":"Use it for a reveal, for a match cut into the next shot, or for arriving exactly on a product hero frame."}]},{"t":"p","runs":[{"text":"The failure mode is worth knowing: if the two frames are too different, the model does not invent a camera move between them, it invents a morph. Keep them the same scene, same light, same lens, with a plausible path between. A push in works. A cut from a kitchen to a beach does not."}]},{"t":"h2","id":"what-text-to-video-is-still-good-for","runs":[{"text":"What text-to-video is still good for"}]},{"t":"p","runs":[{"text":"It is not useless. When nothing is decided yet, text-to-video is a fast way to find a look, generate options and see what the model is good at before you commit. Concept phase, not production phase."}]},{"t":"p","runs":[{"text":"There is also a real trade-off on sound. Some models generate synchronised dialogue and ambience along with the picture, and on several of them that native audio path is tied to text prompting rather than to an image you supply. If you need the model’s own audio, you may have to give up the control of a fixed first frame. Decide which matters more for that shot."}]},{"t":"h2","id":"the-free-path","runs":[{"text":"The free path"}],"role":"free"},{"t":"ul","items":[[{"text":"Stills: several image tools give a daily free allowance, and open models run in a browser through a Hugging Face Space with no install. See "},{"text":"hugging-face","href":"https://addaly.com/learn/hugging-face"},{"text":"."}],[{"text":"Fixing stills: Photopea in a browser handles layers, masks and PSDs on a laptop that cannot run Photoshop. GIMP and Krita if you can install. On a phone, Snapseed for tone and a browser tab for the rest."}],[{"text":"Video: most hosted tools give a small allowance that refreshes. Spend it on image-to-video, at the lowest resolution, on real shots from your list. You learn far more per free credit that way than by rolling text-to-video prompts."}]]},{"t":"rule"},{"t":"p","runs":[{"text":"Today: take one shot you want. Make the still. Fix one thing in it by hand: the label, the hand, the sign. Then animate it, and compare that clip to one generated from text alone."}]}],"check":{"question":[{"text":"A drinks brand’s product label comes out as gibberish in every generated clip. The team keeps rewriting the video prompt with more precise wording about the typography. What is the better fix?"}],"options":[{"text":[{"text":"Add explicit negative-prompt terms for warping and blur so the letterforms stop degrading across frames."}],"correct":false,"tempting":[{"text":"Treats a representational limit as an artefact-suppression problem that a parameter can clean up."}]},{"text":[{"text":"Generate at the highest available resolution so the letters have enough pixels to resolve correctly."}],"correct":false,"tempting":[{"text":"Assumes text failure is a sampling and detail problem rather than a symbolic one."}]},{"text":[{"text":"Composite the real label into the first frame in an image editor, then animate that frame, because the video model propagates what it is handed rather than rendering text it never had."}],"correct":true,"tempting":null},{"text":[{"text":"Switch to text-to-video so the model designs the whole bottle and label coherently from scratch."}],"correct":false,"tempting":[{"text":"Moves a decision the team can fully control into one they cannot control at all."}]}],"order":[3,0,2,1],"why":null},"prev":{"slug":"running-video-models-yourself","title":"What it takes to run one on your own card","minutes":10},"next":{"slug":"composition-that-survives-motion","title":"Compose for the frame you will end on","minutes":9}}}