Sabitlenmiş Tweet
This would've eaten up a whole month worth of credits in other platforms.
Unlimited Seedance 2 on @runwayml is too OP.
English
Michael Andre | Diffract
220 posts
Michael Andre | Diffract
@miku_diffract
https://t.co/mWEL2GHTGM · Directed. Not generated · Cinematic AI · Videographer · Music Producer
Katılım Şubat 2026
140 Takip Edilen182 Takipçiler
Michael Andre | Diffract retweetledi
Here’s my entry for #RunwayBigPitchContest
I wanted to make a futuristic dystopian series, but not the polished neon version. I wanted grit, fear, real emotion. A future that felt lived in.
This is "mei."
A series about the most advanced AI model in a synthetic body that can dream, feel, and become something the company refuses to acknowledge: human. When her handler Ube sees the truth, she takes her and runs.
Made in @runwayml
English
@bitpixi Thanks so much! It hit me right in the feels too when I was editing it 😭
English
@miku_diffract Incredible characters! Wow.. a bit of a tear-jerker too at the end. <3
English
@AngryTomtweets @ArtificialAnlys @video_rebirth Above kling 3 omni standard...excited to see the outputs!
English
A startup nobody's heard of just appeared on the @ArtificialAnlys AI Video leaderboard.
BACH 1.0, developed by @video_rebirth
Top 6 in the world now...
Artificial Analysis@ArtificialAnlys
Bach-1.0 Preview from Video Rebirth debuts at #6 on the Artificial Analysis Text to Video Leaderboard (No Audio)! Bach-1.0 Preview is the latest Text to Video model from @video_rebirth, with similar performance to Vidu Q3 Pro, Kling 3.0 Omni 1080p (Pro), and grok-imagine-video. Bach-1.0 Preview is intended for broad release later in May. See example generations from Bach-1.0 Preview in the Artificial Analysis Video Arena below 🧵
English
@matt_elevenlabs @ElevenCreative See you there! Im heading straight after work
English
Who wants an @ElevenCreative tote + cap?
Like the ElevenCreative post for a chance to win a merch pack delivered straight to your door!
Winner announced in 72 hours.
ElevenCreative@ElevenCreative
We’re bringing new, exclusive ElevenCreative merch to our London event for attendees. Not attending? Like this post for a chance to win a tote and cap. Winner announced Wednesday, May 7 at 4 PM GMT.
English
Sincitium is finally here.
We are pleased to present our latest piece: a concept trailer created specifically for the @runwayml Big Pitch Contest. For this project, we wanted to explore a completely different aesthetic from our usual studio style, and this film is the result of that experimentation.
We hope you enjoy it as much as we enjoyed the creative process.
Produced by: Contanimation
Directed by: Javier De La Chica and Guillermo Miranda Art Direction: Javier De La Chica
Editing: Guillermo Miranda
Voices: Juan Rabadán
#runwaybigpitchcontest
English
It's all in the prompts! I use timecoding a lot (1-3s this happens, 3-5s this happens). I dont like using the multishot panel with chat GPT. I like to have the full control. Sometimes if I really want a super specific shot, I would generate the actual frame using nano bana/gpt image 2 depending on which output is the best then use that as a image reference in seedance. not as a first frame reference but just a regular img reference with the prompt like "[5-7s] take the exact framing from image 3" or something like that. for my moodboards and concept arts I use minimax image and flux 2 max a lot. For environment references I prefer using 1 full image rather than a ref sheet panel. I found that it affects the output of the whole clip a lot.
English
Beautiful shots.
Quick question - how do you go about generating concept art or storyboards? Do you prompt them frame by frame according to what you envision?
Or do you ask GPT 2 to create one with multiple shots/panels and go from there?
I feel like I'm taking too long to envision what my shot compositions should be depending on the story beats.
Would love to hear your input.
Your new video had lots of good shots and a feeling of continuity
English
Some of the worldbuilding references I generated for MEI.
I kept coming back to dense Asian city textures, especially Hong Kong’s monster buildings and Manila's wet markets.
Michael Andre | Diffract@miku_diffract
Here’s my entry for #RunwayBigPitchContest I wanted to make a futuristic dystopian series, but not the polished neon version. I wanted grit, fear, real emotion. A future that felt lived in. This is "mei." A series about the most advanced AI model in a synthetic body that can dream, feel, and become something the company refuses to acknowledge: human. When her handler Ube sees the truth, she takes her and runs. Made in @runwayml
English
@theriavisuals @runwayml good job! love the concept
English
@vikthors_ I started it quite late into the competition maybe like 3 days ago. Took me around 30+ hours I think. Everything from generating references, environment to trial and error with video gen 😅
English
@miku_diffract You’re welcome bro, I have a bank of ideas of the years for tv series so that made it easy for me. How many hrs of work did it take to create this?
English
@AndrewFromDO Appreciate it! Took me about 3 days to generate and stitch everything together. And yes there was lots of trial and error generating the clips 😅
English
@miku_diffract Great work man! That’s a lot of thinking that went in I can tell! Plus all the ai trials
English
@vikthors_ Thanks a lot bro! Love your entries as well. Wish I could’ve thought of multiple entries too 😅
English
@miku_diffract This is well executed bro, this is a refreshing concept as well 💯🔥
English
@StoryArcNetwork Thanks a lot! Haven’t got plans for it yet but definitely something that I would consider. We’ll see how it goes!
English
@miku_diffract I would love to see this as a series or longer form film. Got any plans for it? You did an awesome job!
English
@_JonathanTait Definitely took some inspo from ghost in the shells environment and outfits! I’ve not seen the creator yet but I’ve heard good stuff about it.
English
@miku_diffract Its like Ghost in the Shell had a baby with The Creator.
English
@LouisPonce87 Thanks a lot man and I know right? I have more clips of her being creepy but it didn’t make the cut 😅 the whole thing took me around 3 days and around 30 clips were generated
English
@miku_diffract Damn that geisha's creepy. Love her hand transitioned into a gun/cannon.
Good job! This must've taken a lotta work!
English