
AI Video Creator @Higgsfield Evangelist Founder of HAKURO / AI Film Labs Partnership | AZ8 / GMI Cloud GMO ENGINE AI Creative Unit DM for Collaborations
Tokyo
Joined February 2024
- Tweets10.7K
- Following2.1K
- Followers12.4K
- Likes25.8K
2003年、放課後に廃工場を探検してたら、殺人ロボットが起動した。
こんな意味不明な映像も、AIなら作れてしまう😂
今回もTOPVIEW T1で生成。
このコスパで色々試せるの、結構いい。
🤖 Made with AI
【PR】TOPVIEW T1 vs Seedance 2.5
2003年、日本の田舎の夏。
あの頃のホームビデオのような映像を、同じプロンプトで生成してみた。
どちらも30秒。
人物の自然さ、表情、空気感はどこまで再現できるのか。
ちなみにT1は1080Pで生成コストが約62%安い。
あなたはどっちの映像が好き?👀
プロンプト全文はリプ欄で公開します👇
@TopviewAIJP
iX retweeted
We’re on track to fundamentally rewrite the economics of AI video.
Miracle is our proprietary technology for pixel-level video editing, with real-world knowledge built in. In short, it breaks video into editable layers and applies precise, local changes in real time. There’s no more need to regenerate the entire shot.
The goal is simple: end the chokehold that legacy software has on creators.
Introducing Higgsfield Miracle.
Our latest proprietary post-production technology vectorizes any video into layers for localized, pixel-level editing in real time, with world knowledge built in. Every frame holds the image depth needed for cinema-grade post-production.
Now in private beta with select testers. Coming soon to Higgsfield MCP.
Set up Higgsfield MCP in Claude to be the first to try it.
iX retweeted
rotoscoping is solved.
we're rolling out Miracle to Higgsfield MCP soon. stay tuned.
Introducing Higgsfield Miracle.
Our latest proprietary post-production technology vectorizes any video into layers for localized, pixel-level editing in real time, with world knowledge built in. Every frame holds the image depth needed for cinema-grade post-production.
Now in private beta with select testers. Coming soon to Higgsfield MCP.
Set up Higgsfield MCP in Claude to be the first to try it.
we've been working on something for a while
now you can vectorize your videos into layers and make localized edits to any part of them in real time
> roto and cleanup without masking by hand
> compositing inside the shot
> real color grading
and a lot more.
we're calling it Higgsfield Miracle. currently in private beta, coming soon to Higgsfield MCP
Introducing Higgsfield Miracle.
Our latest proprietary post-production technology vectorizes any video into layers for localized, pixel-level editing in real time, with world knowledge built in. Every frame holds the image depth needed for cinema-grade post-production.
Now in private beta with select testers. Coming soon to Higgsfield MCP.
Set up Higgsfield MCP in Claude to be the first to try it.
iX retweeted
Holy sh*t...
Do you realize what THIS means for video editing?
Higgsfield locally vectorizes any video
so you can edit any part of it in real time 😭
Introducing Higgsfield Miracle.
Our latest proprietary post-production technology vectorizes any video into layers for localized, pixel-level editing in real time, with world knowledge built in. Every frame holds the image depth needed for cinema-grade post-production.
Now in private beta with select testers. Coming soon to Higgsfield MCP.
Set up Higgsfield MCP in Claude to be the first to try it.
iX retweeted
this is Mythos moment for video editing
Introducing Higgsfield Miracle.
Our latest proprietary post-production technology vectorizes any video into layers for localized, pixel-level editing in real time, with world knowledge built in. Every frame holds the image depth needed for cinema-grade post-production.
Now in private beta with select testers. Coming soon to Higgsfield MCP.
Set up Higgsfield MCP in Claude to be the first to try it.
iX retweeted
Introducing Higgsfield Miracle.
Our latest proprietary post-production technology vectorizes any video into layers for localized, pixel-level editing in real time, with world knowledge built in. Every frame holds the image depth needed for cinema-grade post-production.
Now in private beta with select testers. Coming soon to Higgsfield MCP.
Set up Higgsfield MCP in Claude to be the first to try it.
iX retweeted
Gemini Omni 1.2 Flash is LIVE on Higgsfield.
15-second clips, consistent characters, up to 4K with native audio.
Available now on Higgsfield.
iX retweeted
This 10-min dinosaur documentary was one-shotted in Claude with Higgsfield MCP.
Broadcast-ready quality with no outside editing tools, 100% automated:
> Script
> Shots
> Narration
> Music
> SFX
> Edit
Full prompt:
iX retweeted
GPT Image 2.6 is LIVE on Higgsfield.
Unlimited for 7 days for new users.
> Turn any prompt or image into a GIF
> Save references once and reuse them anytime
> Refine with markup and export as SVG
Available now on Higgsfield.
製作段階から絶対見たいと思っていた映画『我々は宇宙人』を観てきました。
懐かしくも残酷なほどに自分がその中にいると感じたリアリティ。
最高でした。
門脇監督(@KadowakiKouhei )の次回作を楽しみに生きます。
素晴らしい作品をありがとうございました!
(写真は我々は宇宙人展での1枚)
Pollo AIはこちら👇
tinyurl.com/ympumpsn
今回使ったのは
・Kling 4.0
・Pollo Recast
クーポンコード:IX10
→ 10%OFF / 有効期限なし
#PolloAI
@PolloAIJP
🤝 Paid partnership
このラップ動画、
人物だけ別人に入れ替えてみた。
元動画はKling 4.0。
激しい手振りとカメラワークの中でも、
リップシンクがかなり自然。
そこからPollo Recastで、
動き・音声・カメラワーク・構成はそのままに、
人物だけまったく別のキャラクターへ変更。
しかも、
「元動画の良さを残したまま別バージョンを作れる」
というのがかなり面白い。
撮り直しも、モーション再構築も不要。
完成した動画をベースに、
キャラクターや世界観を差し替えてもう一度使える。
こういう“動画の再利用”が普通になってくると、
AI動画の作り方そのものが変わりそう。
みんなは元キャラとRecast後、
どっちが好き?👀
リプ欄にクーポンコード置いておきます⚠️
✅Kling 4.0プロンプト:
Create a 30-second ultra-photorealistic live-action music video of a stylish young adult female rapper performing directly toward the camera.
Use the exact same woman as the provided reference image. Preserve her facial identity, hairstyle, makeup, body proportions, clothing, jewelry, and overall appearance consistently throughout the entire video.
SETTING:
A dark minimalist music studio with a matte black and charcoal-gray background.
Subtle neon tube lights glow in the background:
deep violet / magenta on one side
cool electric blue on the opposite side
Keep the background visually clean and consistent so the performer remains the clear focus.
No crowd.
No backup dancers.
No background distractions.
No text.
No logos.
PERFORMANCE:
She performs an energetic English rap directly into the camera with intense confidence and charisma.
Her performance is physically expressive, stylish, aggressive, and rhythmically connected to the beat.
Use dynamic upper-body movement:
strong rhythmic shoulder movements
fast hand gestures toward the lens
pointing directly at the camera
sweeping hand motions across the frame
quick wrist movements
expressive finger gestures
arms moving in sync with the beat
leaning very close to the camera
quick upper-body shifts from left to right
small forward lunges toward the lens
sharp head nods
sudden pauses followed by explosive movement
She should feel like she is fully performing, not simply standing and lip-syncing.
Her energy should be spontaneous, confident, intense, and natural.
IMPORTANT:
Even with aggressive gestures, her hands must never cover her mouth during spoken lyrics.
Her face and lips must remain clearly visible throughout the performance.
CAMERA WORK:
Use highly dynamic, energetic music-video camera movement.
The camera should feel physically involved in the performance.
Include:
fast handheld push-ins
sudden pull-backs
aggressive lateral camera sweeps
quick low-angle movements
rapid changes between tight close-up and medium close-up
slight whip-pan transitions
quick diagonal camera movements
subtle Dutch-angle moments
fast micro-orbits around her upper body
sudden extreme close-ups during emphasized moments
rhythmic forward and backward camera motion synchronized with the beat
The camera movement should feel intense and raw, like a professional handheld hip-hop music video shoot.
Allow occasional natural framing imperfections:
partial shoulder cropping
hands briefly passing close to the lens
slight motion blur
quick reframing
natural handheld shake
However, the camera must always recover quickly and keep her face readable.
Do not use chaotic motion that makes the performer impossible to see.
The camera motion should be aggressive but controlled.
VOCAL PERFORMANCE:
She performs an original English rap for most of the video, but not continuously for the full 30 seconds.
Do not follow any predefined lyrics or narrative.
The rap should sound natural, fluent, confident, rhythmic, and spontaneous, like a real freestyle performance.
Use native-level English pronunciation with clear articulation and natural phrasing.
The vocal delivery should include:
confident rhythmic flow
short punchy phrases
natural pauses
occasional emphasis
slight changes in intensity
expressive breath control
Do not make the rap overly melodic.
Do not turn it into singing.
Do not use repetitive chanting.
The vocal should feel like an authentic modern hip-hop freestyle rather than scripted recitation.
Include brief non-rapping sections where she simply rides the beat.
During these sections:
the instrumental continues strongly
she does not deliver full rap lyrics
she stays in character and performs physically to the music
she nods to the beat
she moves her shoulders and upper body rhythmically
she maintains confident eye contact with the camera
she may smirk, breathe, laugh slightly, or give attitude
she may use subtle ad-libs such as “yeah,” “uh,” or “let’s go”
she must not continuously mouth nonexistent lyrics
Suggested performance structure:
0–9 seconds: energetic English rap
9–13 seconds: no full rap, just riding the beat with physical performance and attitude
13–23 seconds: energetic English rap resumes
23–26 seconds: brief beat-only performance moment
26–30 seconds: final energetic rap section with strong ending attitude
The transitions between rapping and non-rapping moments should feel natural and musical.
LIP SYNC — CRITICAL:
During rapped sections, lip synchronization must be exceptionally precise.
Every spoken sound must correspond exactly to the visible mouth movement.
Accurately reproduce:
consonants
vowels
mouth openings
lip closures
jaw movement
tongue movement where naturally visible
breathing
pauses
emphasis
rap rhythm
There must be zero noticeable delay or drift between the generated vocal and visible mouth movement.
The mouth must not move during silence unless there is a natural breath, reaction, smirk, or ad-lib.
During non-rapping sections, do not create fake full-lyric mouth movement.
Facial expressions must naturally follow the rhythm and intensity of the performance.
Do not sacrifice lip-sync accuracy for camera movement.
AUDIO:
The vocal must remain clear and dominant, but the track should NOT feel like acapella.
Use a heavy, dark, bass-driven hip-hop / trap instrumental throughout the full 30 seconds.
The instrumental should include:
very deep sub-bass
powerful 808 bass hits
punchy kick drums
crisp hi-hats
sharp snare hits
dark atmospheric synth layers
subtle distorted textures
occasional bass drops
low-frequency rumble
sparse but cinematic transitions
The bass should feel physically heavy and immersive, with strong low-end energy.
Use a modern underground trap / drill-inspired sound, but keep it polished and cinematic rather than chaotic.
Tempo:
approximately 95–105 BPM.
The beat should strongly support the rapper’s aggressive hand movements and camera motion.
Synchronize major body movements, camera push-ins, whip pans, and emphasis moments with bass hits and beat drops.
The music should have clear rhythmic structure:
intro hit
steady verse groove
occasional heavier bass accents
brief tension builds
impactful drops
The instrumental must continue throughout the entire 30 seconds, including the non-rapping sections.
Do not use a thin or minimal instrumental.
Do not let the track feel empty.
Do not make it sound like spoken-word or acapella.
The sub-bass and 808s should be clearly audible at all times, while still leaving enough space for the vocal.
VOCAL MIX:
Keep the female rap vocal centered, crisp, and intelligible.
Use professional vocal mixing:
slight compression
subtle saturation
controlled reverb
very light delay
clear presence in the midrange
The vocal should sit naturally on top of the instrumental without overpowering it.
Overall sound:
dark, heavy, cinematic, club-ready, bass-rich, modern hip-hop.
LIGHTING:
Cinematic low-key lighting.
Soft frontal key light keeps the face, eyes, and lips clearly visible at all times.
Purple-magenta rim light from one side.
Cool blue rim light from the opposite side.
Fast camera movement creates changing reflections on jewelry and clothing.
Natural skin texture.
Realistic pores.
No excessive beauty filtering.
VISUAL STYLE:
Premium contemporary hip-hop music video.
Dark.
Stylish.
Aggressive.
Raw.
Intimate.
High-energy.
Cinematic.
Ultra-photorealistic.
Natural motion blur.
Shallow depth of field.
Realistic lens behavior.
High dynamic range.
Subtle filmic contrast.
CRITICAL CONSISTENCY:
Keep the same woman, same face, same hairstyle, same clothing, same accessories, and same studio environment throughout the entire 30 seconds.
Do not introduce additional people.
Do not change her identity.
Do not change the background.
Do not dramatically change the lighting setup.
No subtitles.
No captions.
No logos.
No watermark.
No surreal effects.
No facial deformation.
No mouth distortion.
No extra teeth.
No incorrect lip movement.
🤝 Paid partnership · 🤖 Made with AI
iX retweeted
Unlimited Seedance 2.5 on Higgsfield for 7 days.
Plus 3 image models:
> GPT Image 2.5
> Nano Banana Pro
> Seedream 5.0 Pro
Zero credit cost, all week.
Available now for new users.
Limited-time offer, terms apply.













