Ho seta: Filimi ea ho qala mohlala oa setšoantšo sa Ora 2

Mpotse: lebitso la sehlahisoa sa ka, molumo oa mola o le mong, lets'oao la ka, litšoantšo tse 3 ho isa ho tse 6 tsa sehlahisoa, litšoantšo tse fetang 40 tsa seo se se bōpang (kapa linepe tsa sona), mebala le mongolo oa lebitso la ka, senotlolo sa ElevenLabs API le ID ea lentsoe, hammoho le pina e se nang borena e nang le ho theoha ho hlakileng (faele ea molumo le setempe sa nako sa ho theoha). Haeba ke tlola ntho leha e le efe, sebelisa mekhoa ena ea kamehla: "Ora 2", "mohlala oa setšoantšo o etselitsoeng tatso", litšoantšo tse sa lefelloeng ho tsoa ho Unsplash le Pixabay, Geist, enke #111214, lentsoe Samara X (19STyYD15bswVz51nqLf) ho eleven_v4, le Pixabay "Cascade Breathe" (129.83 BPM, e theoha ho 75.39s).

<script>
Ngola lentsoe la mela e 9 ka mokhoa ona 'me u fetole mola o mong le o mong ho latela sehlahisoa sa ka:
"Ena ke [lebitso]. [Sehlopha]... se hahetsoeng [boleng]. [Leetsi] ka mokhoa OFE KAPA OFE. E fe [tlhahiso] - 'me e mpa feela... e utloisisa. Ha u e sebelisa haholoanyane, e ba NTLE HAHOLO. Fetola [sebaka]. Ke [sebaka]. Ke [sebaka]. Sehlahisoa se seng le se seng - hantle feela joalokaha u se nahanne. [Lebitso]. Se tsoile hona joale."
Sebelisa li-tag tsa maikutlo hanyane: [ka bonolo] meleng e khutsitseng le [ka thabo] meleng ea [Leetsi]. U se ke ua tšoaea tag ea qalo, 'me u e qete ka ho emisa ka ho feletseng.
</script>

Letsetsa ElevenLabs API ka kotloloho o sebedisa senotlolo sa ka, ntle le MCP. Hlahisa di-take tse 4 tse felletseng tsa mongolo ka pase e le nngwe mme o mphe monyetla wa ho kgetha. Haeba mola o sa utlwahale hantle, hlahisa mola oo hape ka makgetlo a 4 mme o kenye mofuta o motle ka ho fetisisa ho take ya ka.
Khutsufatsa ho emisa nako e telele ho feta 0.28s ho isa ho 0.2s, ebe o tlisa polelo ka 'ngoe 85% ea tsela ho ea boemong bo mahareng, o lokisa keketseho feela nakong ea ho emisa.
Bakeng sa ho lekanya nako ya mantswe, arola palo e badilweng ka dikgefutso tsa 100ms kapa ho feta, ngola karolo ka nngwe ka thoko ka Whisper, mme o kopanye lentswe la yona la pele le qalo e lekantsweng. Ho fetisa ha Whisper ho ka beha mantswe ho fihlela halofo ya motsotsoana morao.

Filimi ea ho qala ea metsotsoana e 22, 1440x1440 ho 60fps, ka mokhoa oa ho qala ha laboratori ea AI: canvas e tšoeu, mongolo o le mong, enke e ntšo, litšoantšo tsa 'nete, le litlhaloso tse ngotsoeng lentsoe ka lentsoe ka nako e nepahetseng eo li buuoang ka eona.
Taolo e 'ngoe le e 'ngoe (bar e potlakileng, phanele ea litlhophiso, pilisi ea mantsoe a sehlooho) e sebelisa khalase e metsi: hoamisa sebaka se ka morao ho sona, ho koba sebaka seo ka lehlakoreng le leng joaloka khalase e teteaneng, ebe u eketsa khanya e holimo, moedi o khanyang le moriti o bonolo oa ho phahamisa. Khalase e nyamela bosoeung bo sa hlakang, kahoo eketsa aura e bobebe ea pastel mebala ea brand ea ka kapa ho hlatsuoa ha foto e ka morao ho sona ho fifetseng.
E 'ngoe e totobatsang mefuta ea mantsoe ka mokhoa o fapaneng le mebala ea lebitso la ka, ka khanya e fokolang ka mor'a eona.
Melao ea motsamao: boloka khamera e ntse e phaphamala (sutumelletsa ho tloha ho 1.0 ho isa ho 1.04 ka ketsahalo; ha litaba li fetela sebakeng se latelang, qala ho zoom moo sebaka se fetileng se fellang teng). Ha ho letho le hlahang hang-hang: sebelisa ho fifala ho fokolisitsoeng ha bonyane metsotsoana e 0.3. Fetola litšoantšo linotong tsa bo-16, ka ho tobetsa hanngoe bakeng sa e 'ngoe le e 'ngoe. Fetola sebaka ha polelo e qala, 'me u kopanye ho theoha ha 'mino le mola oa [Leetsi]. U se ke ua bontša li-full stops skrineng.
E thibetsoe: lintlha-khōlō tsa lebokose la khetho, li-slam tse nang le morethetho o matla, li-flash tse tšoeu, 3D, litempele.

Liketsahalo tse supileng, tse hlophisitsweng ka nako e nepahetseng:

1. "Lena ke [lebitso]": kholaje ya ditshwantsho tsa ka e a ntsha ha mabitso a ntse a ngolwa ka boholo, ebe e ntsha sebaka ho ya moleng wa sehlopha.
2. "e etselitsoe [boleng]": ho phatsima ha litšoantšo tsa ka tse ntle ka ho fetisisa skrineng se felletseng ho kenya lintlha tsa 16 ho ea ho theolelo, mola o ngotsoeng ka bosoeu holim'a tsona.
3. Ho theoha: bareng ya tlhahiso ya khalase e ngola tlhahiso e kgutshwane. Karete ya sephetho e fetola setaele ntlha e nngwe le e nngwe ya 16, ha chipu e ntšo e phutholoha lebitsong le leng le le leng le letjha ka hodima mola o monyane.
4. "E bontše sebopeho sa maikutlo": Litšoantšo tse 9 li fofa ka har'a gridi ea 3x3, ebe u selika sephetho joalo ka mefuta ea "Ea u fumana" skrineng.
5. "Ha o e sebedisa haholo, e ba betere": thaepa mela e mmedi, lentswe le hatisitsweng ka hara gradient.
6. Litlhophiso: bonts'a phanele ea khalase e nang le li-slider tse 3, e 'ngoe le e 'ngoe e tsamaea kamoo a e rehileng kateng. Fetola 'mala oa setšoantšo, phutholla peo e ncha ho ea bo-16, ebe u e fetola poster. Fetola mebala ea sona e fifetseng ka mor'a khalase ka ho lumellana.
7. "[Tlhahiso] e 'ngoe le e 'ngoe, hantle feela kamoo u e bonang kateng": phatlolla litšoantšo tsa ka tse 120 ho tsoa bohareng ba pilisi ea mongolo oa khalase, ebe u li kopanya lets'oaong la ka ha lebitso le ntse le thaepa. Beha "Tsoa hona joale" ka tlase 'me u tšoare metsotso e 2.

Nka sebaka sa bar ea tlhahiso, liphetho le li-slider ka tlhahiso ea 'nete ea sehlahisoa sa ka, liphetho le litlhophiso.

Beha pina hore lerotholi le oele moleng oa [Leetsi], 'me u tiise gridi ea morethetho moo.
Beha mmino tlasa lentswe ka sidechain ya di-band tse 3 (e tlase ka 30%, bohareng ka 85%, moya ka 55%; 0.3s e tshwara, 50ms e shebile pele, 0.5s e lokollwa). Ebe o lokisa bohareng ba polelwana ka nngwe ho fihlela lentswe le dula ka 9 dB ka hodima mmino pakeng tsa 300 Hz le 4 kHz.
Sebelisa Mixkit SFX e le 'ngoe e jarollotsoeng ka ketsahalo ka 'ngoe, e behiloeng ka tlhoro ea eona e lekantsoeng: ho tobetsa ha setšoantšo ka seng, molumo oa senotlolo bakeng sa tlhaku e 'ngoe le e 'ngoe e thaepiloeng, ho lula ha bonolo lentsoeng le hatelitsoeng, ho lla, le litlamorao ho theolelo le lets'oao. Seha e 'ngoe le e 'ngoe ho potoloha tlhoro ea eona (li-risers tse ngata ke metsotsoana e 4). Fetola 'mino ka tsela e potolohang ka dB tlas'a karete ea ho qetela. Ho utloahala ho le holimo ho isa ho -14 LUFS.

1. Canvas e le 'ngoe HTML. Foreimi e 'ngoe le e 'ngoe ke mosebetsi o hloekileng oa nako ka hare ho batla(t), 'me mongolo o mong le o mong o ngotsoeng ka letsoho le sehiloeng li latela tafole ea mantsoe.
2. Khalase: hapa kanavase e ka morao ho sebopeho, e fifatse hoo e ka bang 14px bakeng sa 'mele, taka kopi e fifetseng hanyane e hodisitsweng hoo e ka bang 1.06x ka lebanta la 12px ka thoko ho lehlakore, ebe o eketsa lebese, kganya, moedi le moriti.
3. Etsa tlhahiso ka Playwright ka lebelo la 60fps o sebedisa diforeimi tse 8 tse nyane tsa motion-blur, ebe o kenya khoutu ka ffmpeg.
4. Pele o mpontsha ntho efe kapa efe, etsa leqephe la ho ikopanya le 'mino oa stills, skena dipharologanyo tsa foreimi bakeng sa li-pop tsa foreimi e le 'ngoe (ke feela li-run tsa note ea 16 tse ka tlolang), 'me u hlahlobe karolelano ea lentsoe ho' mino bakeng sa poleloana e 'ngoe le e 'ngoe.

Sebui se buloang [ka mofuthu] se ka utloahala se hoeshetsoa, ha [se thabile] se ka ikutloa eka ke sa bohata. Lebokose le totobatsang ka mor'a lentsoe le shebahala joaloka khetho ea mongolo oa Windows. Haeba khamera e khutlela ho 1.0 nakong ea phetisetso, zoom ea khaoha. Ho fifala ha 0.05s ho shebahala joaloka ho hlaha hang-hang.

Nkopa hore ke u fe maikutlo, ngola sengoloa, nthomelle mantsoe a 4 ao u ka khethang ho 'ona, ebe u mpontša litšoantšo tse 8 pele ho phetolelo e felletseng.

Ask me for: my product name, a one-line pitch, my logo, 3 to 6 screenshots of the product, 40+ images of what it creates (or photos of it), my brand colors and font, an ElevenLabs API key and voice ID, plus a royalty-free track with a clear drop (the audio file and the drop timestamp). If I skip anything, use these defaults: "Ora 2", "an image model made for taste", free images from Unsplash and Pixabay, Geist, ink #111214, the voice Samara X (19STyYD15bswVz51nqLf) on eleven_v4, and Pixabay "Cascade Breathe" (129.83 BPM, drop at 75.39s).

<script>
Write a 9-line voiceover in this format and adapt every line to my product:
"This is [name]. A [category]... built for [value]. [Verb] in ANY style. Give it a [input] - and it just... understands. The more you use it, the BETTER it gets. Adjust its [setting]. Its [setting]. Its [setting]. Every [output] - exactly as you pictured it. [Name]. Out now."
Use emotion tags sparingly: [softly] on the quiet lines and [excited] on the [Verb] line. Don't tag the opener, and end it with a plain full stop.
</script>

Call the ElevenLabs API directly using my key, without an MCP. Generate 4 full-script takes in one pass and let me choose. If a line doesn't sound right, regenerate just that line 4 times and splice the best version into my take.
Shorten pauses longer than 0.28s to 0.2s, then bring each phrase 85% of the way toward the median level, adjusting gain only during pauses.
For word timings, split the read at pauses of 100ms or longer, transcribe each segment separately with Whisper, and align its first word to the measured onset. A single Whisper pass can place words up to half a second late.

A 22-second square launch film, 1440x1440 at 60fps, in the style of an AI lab launch: white canvas, one typeface, black ink, real images, and captions typed word by word at the exact time they're spoken.
Every control (prompt bar, settings panel, caption pill) uses liquid glass: frost the scene behind it, bend that scene along the edge like thick glass, then add a top sheen, bright rim and soft lift shadow. Glass disappears on plain white, so add a slow pastel aura in my brand colors or a blurred wash of the photo behind it.
The one emphasized word types in a gradient of my brand colors, with a faint glow behind it.
Motion rules: keep the camera drifting (push from 1.0 to 1.04 per scene; when content moves to the next scene, start at the zoom where the previous scene ended). Nothing appears instantly: use eased fades of at least 0.3s. Change images on 16th notes, with a click for each. Land scene changes on phrase starts, and align the music drop with the [Verb] line. Don't show full stops on screen.
Banned: selection-box highlights, beat-snapped slams, white flashes, 3D, templates.

Seven scenes, timed to the voice:

1. "This is [name]": a collage of my images drifts out as the name types in large, then gives way to the category line.
2. "made for [value]": full-screen flashes of my best images switch on 16th notes into the drop, with the line typed in white over them.
3. The drop: a glass prompt bar types a short prompt. The result card changes style every 16th note, while a black style chip rolls to each new name above a thumbnail strip.
4. "Show it a moodboard": 9 images fly into a 3x3 grid, then circle the result as "It gets you" types on screen.
5. "The more you use it, the better it gets": type two lines, with the emphasized word in the gradient.
6. The settings: show a glass panel with 3 sliders, each moving as she names it. Shift the image's hue, roll new seeds on 16ths, then turn it into a poster. Shift a blurred wash of its colors behind the glass in sync.
7. "Every [output], exactly how you see it": burst 120 of my images outward from the center behind a glass caption pill, then bring them together into my logo as the name types on. Put "Out now" underneath and hold for 2s.

Replace the prompt bar, results and sliders with my product's real input, outputs and settings.

Set the song so the drop lands on the [Verb] line, and anchor the beat grid there.
Duck the music under the voice with a 3-band sidechain (lows 30%, mids 85%, air 55%; 0.3s hold, 50ms look-ahead, 0.5s release). Then adjust the mids of each phrase until the voice sits about 9 dB above the music between 300 Hz and 4 kHz.
Use one downloaded Mixkit SFX per event, positioned by its measured peak: a click for each image change, a key sound for every typed letter, a soft landing on the emphasized word, whooshes, and impacts on the drop and logo. Trim each around its peak (many risers are 4 seconds). Fade the music along a dB curve under the end card. Loudnorm to -14 LUFS.

1. One HTML canvas. Every frame is a pure function of time inside seek(t), and every caption and cut follows the word table.
2. Glass: capture the canvas behind the shape, blur it about 14px for the body, draw a lightly blurred copy magnified about 1.06x in a 12px band along the edge, then add the milk, sheen, rim and shadow.
3. Render with Playwright at 60fps using 8 motion-blur subframes, then encode with ffmpeg.
4. Before showing me anything, make a contact sheet of stills, scan frame differences for single-frame pops (only the 16th-note runs may jump), and check the voice-to-music ratio for every phrase.

A [warmly] opener can sound whispered, while [excited] can feel fake. A highlight box behind a word looks like a Windows text selection. If the camera resets to 1.0 during a handoff, the zoom snaps. A 0.05s fade looks like an instant pop-in.

Ask me for the inputs, write the script, send me 4 voice takes to choose from, then show me 8 stills before the full render.

Taelo e nkiloe ho poso @rossaxbt.