Tutorial Prompt Gemini AI Foto Bergerak Jadi Video | The Elephant Never Forgets

dailytips.id Juli 31, 2026 0

Tutorial Prompt Gemini AI Foto Bergerak Jadi Video | The Elephant Never Forgets - Artificial Intelligence berkembang sangat cepat dalam dua tahun terakhir. Jika sebelumnya AI hanya mampu menghasilkan gambar, kini teknologi terbaru mampu mengubah sebuah gambar diam menjadi video sinematik yang terlihat seperti hasil produksi studio Hollywood. 

 

Salah satu teknologi yang paling banyak dibicarakan adalah Gemini AI Veo 3, sebuah AI Video Generator dari Google yang mampu menghasilkan gerakan kamera realistis, ekspresi wajah alami, pencahayaan sinematik, hingga efek suara yang sangat mendekati film profesional.

Banyak kreator konten di TikTok, Instagram Reels, hingga YouTube Shorts mulai memanfaatkan teknologi ini untuk membuat cerita pendek yang emosional. Salah satu konsep yang cukup viral adalah kisah berjudul The Elephant Never Forgets, sebuah cerita sederhana tentang seorang pengendara Vespa yang mengira dirinya akan diserang seekor induk gajah. 

Namun, di balik kepanikannya ternyata induk gajah hanya ingin meminta bantuan menyelamatkan anaknya yang terjebak di dalam lubang berlumpur. Cerita tersebut menjadi sangat menyentuh karena mengajarkan bahwa kebaikan yang tulus tidak pernah benar-benar dilupakan.

Melalui artikel ini, Anda akan mempelajari Tutorial Gemini AI Veo 3 secara lengkap. Mulai dari membuat thumbnail sebagai acuan visual, menyusun Prompt Gemini AI, hingga menghasilkan tiga adegan video yang saling terhubung sehingga membentuk sebuah film pendek berdurasi singkat dengan kualitas sinematik.

Mengapa Harus Membuat Thumbnail Terlebih Dahulu?

Kesalahan yang sering dilakukan pemula adalah langsung membuat video tanpa memiliki gambar referensi. Akibatnya wajah karakter sering berubah-ubah, pakaian berganti sendiri, bahkan latar belakang menjadi tidak konsisten.

Cara terbaik adalah membuat sebuah thumbnail terlebih dahulu. Thumbnail berfungsi sebagai "identitas visual" yang nantinya digunakan sebagai acuan saat membuat seluruh adegan video. Dengan cara ini, Gemini AI memiliki referensi yang sama sehingga wajah karakter, pakaian, dan suasana tetap konsisten dari awal hingga akhir.

Pada contoh proyek The Elephant Never Forgets, thumbnail menampilkan tiga elemen utama, yaitu karakter utama, induk gajah bersama anaknya, serta kendaraan Vespa sebagai ciri khas tokoh utama. Komposisi ini akan membantu AI mempertahankan kontinuitas visual pada setiap adegan.

Persiapan Sebelum Membuat Video

Sebelum memulai proses pembuatan video, pastikan Anda telah menyiapkan beberapa hal berikut:

• Akun Google yang memiliki akses ke Gemini AI Veo 3.
• Satu foto referensi karakter utama dengan kualitas tinggi.
• Konsep cerita yang sederhana.
• Thumbnail sebagai acuan visual.
• Koneksi internet yang stabil.

Semakin baik kualitas gambar referensi yang digunakan, semakin mudah AI mempertahankan identitas karakter sepanjang video.

Langkah 1 – Membuat Thumbnail Cinematic

Tahap pertama adalah membuat poster atau thumbnail yang nantinya menjadi referensi visual.

Pada tahap ini, gunakan Prompt AI Cinematic yang menjelaskan:

• wajah karakter utama,
• pakaian,
• suasana hutan,
• pencahayaan Golden Hour,
• induk gajah dan anak gajah,
• kendaraan Vespa.

Setelah thumbnail selesai, simpan gambar tersebut dan gunakan sebagai Reference Image pada seluruh proses pembuatan video.

Untuk membuat Thumnail Ikuti tahap-tahap berikut:  

1. Masuk ke Gemini Ai
2. Upload Foto Anda, pilih buat foto
3. Masukkan prompt di Bawah ini:

Use the attached reference image as the exact appearance reference for the man. Create an ultra-realistic cinematic movie poster. Vertical 9:16. 8K HDR. Photorealistic. Hollywood blockbuster quality. National Geographic wildlife realism. The title of the movie is: THE ELEPHANT NEVER FORGETS Tagline: Kindness Always Returns SCENE Golden hour inside a lush tropical rainforest after light rain. Warm sunlight streams through tall rainforest trees creating beautiful volumetric light rays and soft golden mist. A calm adult Asian elephant stands proudly in the background with its trunk raised high into the warm golden sunlight. A playful baby elephant stands beside its mother looking happily toward the man. On the right side of the poster sits the same young Asian man from the reference image. Maintain exactly the same facial identity. Never change his face. Never change hairstyle. Never change skin tone. Never change age. He wears a brown overshirt over a white t-shirt, black jeans, brown leather boots, black riding gloves, and a black wristwatch. Beside him is the same classic olive-green Vespa scooter parked naturally. The man smiles warmly while looking at the baby elephant. The atmosphere feels emotional, peaceful, and inspiring. The composition must feel like an official Hollywood adventure movie poster. The mother elephant appears majestic but gentle. The baby elephant looks happy and healthy. At the bottom of the poster are two cinematic preview frames with rough movie-frame borders. LEFT PREVIEW FRAME Episode 1 The man kneels beside a deep muddy pit while helping the trapped baby elephant using a long fallen tree trunk. The muddy pit is deep with steep muddy walls. Golden sunlight shines through the forest. RIGHT PREVIEW FRAME Episode 2 The same man rides his Vespa through the rainforest. The mother elephant peacefully stands in front of him on the forest trail. The atmosphere is warm and emotional. A small golden heart icon separates both preview frames. Below the previews place the tagline: A TRUE STORY OF GRATITUDE TYPOGRAPHY Large bold cinematic movie title. Distressed adventure movie font. THE White ELEPHANT Large golden-yellow textured letters. NEVER FORGETS White distressed letters. The typography should resemble a Hollywood theatrical poster. COLOR GRADING Kodak Vision3 Warm golden tones. Deep rainforest greens. Rich cinematic contrast. Natural skin tones. Beautiful golden highlights. CAMERA ARRI Alexa 65 Cooke Anamorphic Lens 24mm cinematic wide composition 85mm portrait realism Natural HDR Ultra detailed. LIGHTING Golden Hour Volumetric sunlight Soft rainforest mist Beautiful rim light Physically accurate lighting. QUALITY Hyper realistic. Photorealistic. Hollywood movie poster. National Geographic realism. Ultra detailed skin texture. Realistic elephant anatomy. Realistic Vespa details. Professional cinematic composition. IDENTITY LOCK Use the attached reference image as the only identity reference. The man's facial identity must remain identical. Never generate another face. Never change hairstyle. Never change body proportions. Maintain identical facial structure. NEGATIVE PROMPT Cartoon. CGI. Anime. Painting. Illustration. Low resolution. Blurry. Extra elephant. Wrong Vespa. Different man. Different face. Facial distortion. Bad anatomy. Oversaturated colors. Watermark. Logo. Subtitles. AI artifacts. Text errors. Misspelled title. Duplicate elephant. Unrealistic forest. Aggressive elephant. Violence. Blood.

Langkah sederhana ini sering diabaikan, padahal sangat menentukan kualitas hasil akhir.

Langkah 2 – Membuat Adegan Pertama

Pada adegan pertama, fokuslah hanya pada satu aksi. Karakter utama sedang mengendarai Vespa di tengah hutan tropis ketika tiba-tiba mendengar suara induk gajah.

Hindari memasukkan terlalu banyak aksi dalam satu prompt. Semakin sederhana adegan yang dibuat, semakin realistis gerakan yang dihasilkan.

Sebagai contoh, Anda cukup meminta AI membuat karakter berhenti, menoleh, dan mendengar suara gajah tanpa langsung memperlihatkan keseluruhan cerita.Pendekatan seperti ini membuat video terasa lebih dramatis dan membangun rasa penasaran.

Cara membuat prompt video langkah-langkahnya sama dengan cara membuat thumbnail,  hanya saja klik pilih buat Viedeo

Prompt Adegan 1 

Use the attached reference image as the exact appearance reference for the man. Maintain identical facial features, hairstyle, skin tone, body proportions, clothing, sunglasses, black riding gloves, black wristwatch, brown overshirt over a white t-shirt, black jeans, brown leather boots, and the same classic Vespa scooter parked beside the forest trail. Ultra-realistic wildlife cinematic film. Vertical 9:16. 8K HDR. 24fps. Photorealistic. Hollywood-quality. BBC Earth documentary realism. National Geographic cinematography. SCENE Golden hour inside a lush tropical rainforest after a light rain. Warm sunlight filters through towering rainforest trees. Beautiful volumetric light rays shine through soft mist. Rainwater drips from large tropical leaves. The forest feels peaceful and alive. The man's classic Vespa is parked beside a narrow dirt trail. He suddenly hears distressed baby elephant cries coming from deeper inside the forest. He immediately walks toward the sound. After pushing through dense bushes, he discovers a deep muddy pit approximately 2.5 to 3 meters deep. The muddy walls are almost vertical and extremely slippery. At the bottom of the pit, a frightened baby elephant struggles repeatedly to climb but slides back into the mud every time. It is physically impossible for the baby elephant to escape by itself. About 12 meters away, the mother elephant stands quietly. She is visibly worried but calm. She carefully watches the man, instinctively sensing that he means no harm. She never charges or behaves aggressively. The man slowly walks around the pit, carefully examining every angle. He notices a massive fallen tree trunk lying nearby. The trunk is approximately 6 meters long and thick enough to support the baby elephant. He grips the trunk with both hands and slowly drags it across the muddy forest floor. Mud splashes naturally as the heavy trunk moves. Leaves and small branches scrape against the bark. The man reaches the edge of the pit. He carefully lowers one end of the trunk into the hole. Using all his strength, he slowly pushes it downward until the lower end firmly touches the muddy bottom of the pit. The opposite end remains securely on solid ground outside the pit. The trunk now forms a stable natural ramp at approximately a 35-degree angle. The camera clearly shows that the trunk extends continuously from the forest floor all the way to the bottom of the pit. Nothing floats. The trunk is fully supported by gravity and realistic physics. The man gently shakes the trunk to ensure it is stable. Satisfied, he steps back and looks toward the baby elephant. The baby elephant watches cautiously. The scene ends before the elephant begins climbing. CAMERA ARRI Alexa 65 Cooke Anamorphic Lens 24mm cinematic establishing shot 50mm medium shots 85mm emotional close-ups Overhead drone shot revealing the full depth of the pit Low-angle shot from inside the pit showing the trunk touching the muddy bottom Tracking shot following the trunk being lowered Natural handheld movement Slow cinematic dolly Smooth gimbal movement LIGHTING Golden Hour Warm sunlight Volumetric lighting Soft mist Natural HDR Beautiful rim light COLOR GRADING Kodak Vision3 Warm cinematic tones Deep rainforest greens Natural skin tones Rich contrast Golden highlights SOUND DESIGN First two seconds contain only rainforest ambience. Birds singing. Gentle wind. Insects. Water dripping from leaves. Distant elephant breathing. No music. After two seconds, soft emotional cinematic music gradually fades in. Gentle piano. Warm acoustic guitar. Ambient strings. No vocals. No EDM. Heavy tree trunk scraping across mud. Leaves rustling. Realistic footsteps. QUALITY Hyper realistic Photorealistic Hollywood wildlife documentary National Geographic realism Realistic mud simulation Physically accurate lighting Realistic elephant anatomy NEGATIVE PROMPT Shallow hole. Small ditch. Floating tree trunk. Tree trunk beside the hole. Tree trunk stopping halfway. Cartoon. CGI. Fake elephant. Aggressive elephant. Jumping elephant. Impossible physics. Text. Watermark. Logo. Subtitles. AI artifacts.

 

 Langkah 3 – Membuat Adegan Kedua

Adegan kedua memperlihatkan tokoh utama mengikuti suara tersebut hingga menemukan bayi gajah yang terjebak di sebuah lubang berlumpur.

Di sinilah pentingnya menggunakan Gemini Veo 3 Prompt yang konsisten. Jangan mengubah deskripsi wajah, pakaian, maupun lokasi.

Jika pada adegan pertama karakter mengenakan overshirt cokelat, maka gunakan deskripsi yang sama pada adegan kedua.

Prompt Adegan 2

Use the attached reference image as the exact appearance reference for the man. Maintain identical facial features, hairstyle, skin tone, body proportions, clothing, sunglasses, black riding gloves, black wristwatch, brown overshirt over a white t-shirt, black jeans, brown leather boots, and the same classic Vespa scooter parked beside the forest trail. Ultra-realistic wildlife cinematic film. Vertical 9:16. 8K HDR. 24fps. Photorealistic. Hollywood-quality. BBC Earth documentary realism. National Geographic cinematography. CONTINUITY REQUIREMENTS This scene is a direct continuation of the previous scene. Nothing in the environment changes. Use exactly the same rainforest. Use exactly the same lighting. Use exactly the same muddy pit from Scene 1. Do not generate a new hole. Keep the identical shape, depth, width, muddy walls, muddy texture, surrounding bushes, rocks, and forest floor. The muddy pit remains approximately 2.5–3 meters deep with steep vertical slippery walls. Use exactly the same fallen tree trunk that the man positioned inside the pit in Scene 1. Do not generate another tree. Do not generate roots. Do not generate an uprooted tree. Do not generate a standing tree. The log is completely detached from any roots. It is the exact same fallen log from Scene 1. The bark, scratches, mud stains, and orientation remain identical. The lower end of the log remains firmly touching the muddy bottom. The upper end remains securely resting on the forest floor. The log has not moved. Maintain perfect visual continuity. The Vespa remains parked in exactly the same location beside the forest trail. SCENE The opening begins with a slow overhead cinematic shot. The camera clearly shows the exact same deep muddy pit from Scene 1. The identical fallen tree trunk is still firmly positioned, extending continuously from the forest floor all the way to the muddy bottom. Nothing has changed. The camera slowly tilts downward into the pit, emphasizing its depth and showing that the tree trunk reaches the very bottom. Only after this continuity shot does the baby elephant begin moving. The man stands quietly several meters away. He remains calm and avoids sudden movement. He gently raises one hand, encouraging the baby elephant without approaching it. The mother elephant stands approximately 12 meters away. She watches carefully but remains completely calm. She trusts the man's intentions. The baby elephant slowly walks toward the bottom end of the tree trunk. It cautiously places one foot onto the bark. The bark is wet, muddy, and naturally slippery. The baby elephant slips slightly. It regains balance. It climbs higher. Mud falls from its feet. Its breathing becomes heavier. It slips once more but never falls off the trunk. The camera follows every step of the difficult climb. The log never moves. The ramp remains completely stable. The baby elephant slowly reaches the upper edge of the pit. Its front legs climb onto solid ground. After one final effort, its back legs leave the pit completely. The baby elephant is finally safe. It immediately runs toward its mother. The mother elephant gently wraps her trunk around her baby. She softly touches the baby's face, ears, shoulders, and body to ensure it is unharmed. They remain together for several emotional seconds. The mother elephant slowly turns toward the man. She raises her trunk high into the warm golden sunlight. She remains still for several seconds as a silent gesture of gratitude. The man smiles warmly. He respectfully places one hand over his chest and gently bows his head. No dialogue. No narration. No text. Only emotional visual storytelling. The final drone shot slowly rises above the rainforest. Below, the reunited elephant family stands peacefully together. The man's Vespa remains parked beside the forest trail. Golden sunlight illuminates the rainforest canopy. The scene ends with a peaceful cinematic fade. CAMERA ARRI Alexa 65 Cooke Anamorphic Lens 24mm cinematic establishing shot 50mm natural perspective 85mm emotional portrait Opening overhead shot clearly showing the identical muddy pit from Scene 1 Low-angle shot from inside the pit showing the tree trunk touching the muddy bottom Tracking shot following the baby elephant climbing the entire length of the log Macro shots of muddy elephant feet gripping the bark Macro shots of elephant eyes Macro shots of the mother touching the baby with her trunk Slow cinematic dolly Smooth gimbal movement Natural handheld movement Final drone reveal LIGHTING Golden Hour Warm sunlight Volumetric light rays Soft rainforest mist Natural HDR Beautiful rim light COLOR GRADING Kodak Vision3 Warm cinematic tones Deep rainforest greens Natural skin tones Rich contrast Golden highlights SOUND DESIGN The first two seconds contain only rainforest ambience. Gentle wind. Bird songs. Insects. Water dripping from leaves. Natural elephant breathing. No music. After two seconds, soft emotional cinematic music slowly fades in. Gentle piano. Warm acoustic guitar. Ambient strings. No vocals. No EDM. Realistic footsteps. Mud sliding. Leaves rustling. Forest ambience. QUALITY Hyper realistic Photorealistic Hollywood wildlife documentary BBC Earth realism National Geographic realism Ultra detailed Realistic elephant anatomy Realistic mud simulation Physically accurate lighting Natural facial expressions NEGATIVE PROMPT No different hole. No new pit. No shallow hole. No small ditch. No different tree. No tree roots. No standing tree. No uprooted tree. No broken tree. No fresh fallen tree. Only the identical fallen tree trunk from Scene 1. No floating log. No moving log. No unstable ramp. No baby elephant jumping out. No impossible climbing. No impossible physics. No aggressive elephant. No attacking elephant. No CGI. No cartoon. No fake elephant. No blurry image. No low resolution. No watermark. No subtitles. No logo. No text. No AI artifacts. No continuity errors.

Instruksi sederhana ini mampu mengurangi perubahan karakter secara signifikan.

Cara Menjaga Wajah Tetap Konsisten

Salah satu masalah terbesar saat menggunakan Gemini AI Foto Jadi Video adalah perubahan wajah karakter.

Untungnya masalah ini dapat diminimalkan dengan beberapa teknik sederhana.

Pertama, gunakan gambar referensi yang sama pada setiap adegan.

Kedua, jangan pernah mengubah deskripsi karakter.

Ketiga, tambahkan bagian Identity Lock pada setiap prompt.

Contohnya:

IDENTITY LOCK

Use the attached reference image as the only identity reference.

Never change facial identity.

Never change hairstyle.

Never change age.

Never change clothing.

Maintain perfect identity consistency throughout the entire video.

Teknik ini menjadi salah satu rahasia para kreator yang rutin menghasilkan video AI berkualitas tinggi.

Teknik Cinematic Agar Video Terlihat Seperti Film

Selain membuat prompt yang baik, kualitas video juga dipengaruhi oleh gaya visual.

Beberapa elemen yang sebaiknya selalu digunakan antara lain:

• Golden Hour Lighting
• Volumetric Light
• Soft Mist
• Natural HDR
• Kodak Vision3 Color Grading
• ARRI Alexa 65 Camera
• Cooke Anamorphic Lens
• Drone Reveal
• Smooth Gimbal Movement
• Natural Handheld Camera

Elemen-elemen tersebut merupakan standar yang sering digunakan dalam Prompt AI Cinematic sehingga video terlihat lebih hidup.

Editing Setelah Video Selesai

Walaupun hasil dari Veo 3 sudah sangat baik, proses editing tetap memiliki peran penting. Karena, kadang hasil prompt itu bisa berubah-ubah, sehingga pengeditan masih dibutuhkan, apalagi ini video panjang.

Gunakan Adobe Premiere Pro atau tool atau aplikasi lain yang bisa di gunakan untuk mengedit Video, untuk menambahkan:

• Slow Motion pada momen emosional.
• Zoom In perlahan menggunakan keyframe.
• Motion Blur menggunakan efek Transform.
• Color Grading hangat.
• Sound Design seperti suara burung, angin, langkah kaki, lumpur, serta suara napas gajah.
• Musik sinematik yang mulai masuk beberapa detik setelah video dimulai.

Perubahan kecil tersebut mampu meningkatkan kualitas video secara signifikan.

Kesalahan yang Sering Dilakukan Pemula

Berdasarkan pengalaman banyak kreator, ada beberapa kesalahan yang paling sering terjadi ketika mencoba Cara Membuat Video AI menggunakan Gemini Veo 3.

Yang pertama adalah memasukkan terlalu banyak aksi ke dalam satu prompt. Misalnya karakter berlari, menemukan gajah, menolong bayi gajah, lalu mendapatkan ucapan terima kasih dalam satu adegan. AI sering kali kesulitan menjaga kontinuitas jika harus memproses terlalu banyak aksi sekaligus.

Kesalahan kedua adalah mengganti deskripsi karakter di setiap scene. Perubahan kecil seperti warna pakaian atau gaya rambut dapat membuat wajah ikut berubah.

Kesalahan ketiga adalah tidak menggunakan gambar referensi yang sama. Padahal referensi visual merupakan kunci agar karakter tetap konsisten.

Dengan menghindari tiga kesalahan tersebut, hasil video biasanya jauh lebih stabil.

Mengapa The Elephant Never Forgets Menarik?

Cerita The Elephant Never Forgets sebenarnya sangat sederhana.

Namun justru kesederhanaan itulah yang membuatnya kuat.

Penonton diajak percaya bahwa tokoh utama sedang berada dalam bahaya. Ketegangan dibangun sejak awal ketika induk gajah berlari mendekat. Setelah itu muncul plot twist bahwa induk gajah sebenarnya meminta bantuan menyelamatkan anaknya.

Akhir cerita ditutup dengan adegan emosional ketika induk gajah mengangkat belalai sebagai simbol rasa terima kasih.

Struktur cerita seperti ini sangat efektif untuk video pendek karena mampu mempertahankan perhatian penonton dari awal hingga akhir.

Hasil Prompt Ai Foto Bergerak The Elephant Never Forgets

 

@dailytips.id_ Prompt Gemini Ai Foto Bergerak Cinematic Tutorial Prompt Gemini Ai Foto Menjadi Video Cinematic Noted: Mengingat proses dan tutorialnya agak panjang, Dailytips.id akan jelaskan lebih detail di website: https://www.dailytips.id Secara jelas disertai gambar akan saya kupas disana.. See You #promptgeminiai #promptfoto #fotobergerak #videoviral #terbaru ♬ Motivation Soundtrack: Celestial Ascension - Nando Pinheiro

Penutup

Perkembangan teknologi AI membuka peluang baru bagi siapa saja untuk membuat film pendek tanpa harus memiliki kamera mahal maupun kru produksi yang besar. Dengan mengikuti Tutorial Veo 3 Indonesia ini, Anda dapat mulai memahami bagaimana menyusun Prompt Gemini AI, membuat Prompt Foto Bergerak AI, hingga menghasilkan video yang terlihat seperti produksi profesional.

Kunci keberhasilannya bukan terletak pada panjang pendeknya prompt, melainkan pada bagaimana Anda membagi cerita menjadi beberapa adegan sederhana yang saling terhubung. Mulailah dengan membuat thumbnail sebagai identitas visual, gunakan gambar referensi yang sama pada setiap scene, tambahkan Identity Lock, lalu fokuskan setiap prompt pada satu aksi utama.

Jika dilakukan dengan benar, Gemini AI Foto Jadi Video mampu menghasilkan karya yang tidak hanya menarik secara visual, tetapi juga memiliki cerita yang menyentuh. Kisah The Elephant Never Forgets adalah contoh bahwa kombinasi antara ide kreatif, Gemini Veo 3 Prompt, dan sedikit sentuhan editing dapat menghasilkan video pendek yang mampu memikat jutaan penonton.

Kini saatnya Anda mencoba sendiri, bereksperimen dengan berbagai Prompt AI Cinematic, dan menciptakan cerita yang mungkin akan menjadi video viral berikutnya. Selamat berkarya!