Rudrabha Mukhopadhyay

Co-founder & CTO of @synclabs

[sync. labs @synclabs]

For the past 3 years we have been researching and learning how humans communicate and interact with each other. We are now at a point where the technology is ready - in production, on your screens. Lipsync is a frontier that will unlock stories to travel freely. Every story deserves an audience.

There is good in AI, but it is our responsibility to find it.

https://t.co/KOzTdLy9rr

[Rudrabha Mukhopadhyay @RudrabhaM]

first formal ad from @synclabs :)

[sync. labs @synclabs]

For the past 3 years we have been researching and learning how humans communicate and interact with each other. We are now at a point where the technology is ready - in production, on your screens. Lipsync is a frontier that will unlock stories to travel freely. Every story deserves an audience.

There is good in AI, but it is our responsibility to find it.

https://t.co/KOzTdLy9rr

[sync. labs @synclabs]

your characters can now talk

drop in any illustration or animated still, add a voice, and image-to-video brings it to life with lipsync

powered by sync-3

[sync. labs @synclabs]

multiple people in frame?

speaker selection lets you pick exactly who speaks while everyone else stays silent

yes, even from a single frame

image-to-video powered by sync-3

[sync. labs @synclabs]

turn a single image into a speaking video

image-to-video is now live for everyone

powered by sync-3

[Rudrabha Mukhopadhyay @RudrabhaM]

@elonmusk @PeterDiamandis You do not need AI to not bet against Elon🤣

[sync. labs @synclabs]

we had the incredible opportunity to speak at SXSW this past march

"from dubbing to belonging: the end of foreign film"

at sync. labs, we believe films should have no barriers. so as our technology evolves, why haven't our movies?

watch as we bring together artists, leaders, and innovators to talk about where filmmaking is heading

[Rudrabha Mukhopadhyay @RudrabhaM]

Interacting with davinci using code opens up so many possibilities!

[M Hadi @mhadifilms]

i built a DaVinci Resolve MCP so claude can help me edit videos 10x faster

it’s a python library, CLI, and locally hosted MCP server that allows developers and agents to control Resolve via API with simple commands

all you need to get started: pip install dvr

here's dvr in action 👇

[Rudrabha Mukhopadhyay @RudrabhaM]

@synclabs https://t.co/VpsDOTk0Da

Finally😂����

[sync. labs @synclabs]

dubbing is now live on sync. studio

upload, dub, sync

one video, a world of possibilities

[sync. labs @synclabs]

sync-3 handles edge cases that break most lipsync models, lipsync-2-pro included

this is what building for the best looks like

[sync. labs @synclabs]

0:28 is the reason our team doesn't sleep

[Rudrabha Mukhopadhyay @RudrabhaM]

@FQY7654 @synclabs @mhadifilms :)

[Rudrabha Mukhopadhyay @RudrabhaM]

Yes, this is a fair point. The model is significantly heavier than our previous models with a whole new architecture which leads to increased costs on our end. We acknowledge that today this limits our reach but we are working hard to make it faster and cheaper without compromising quality.

[Rudrabha Mukhopadhyay @RudrabhaM]

@thesameman @synclabs Hi, we are working on making it easier to select the speaker you want

[sync. labs @synclabs]

not all shots are created equal

its easy to lipsync someone when they are facing the camera directly, but what happens when they are not?

when they are looking to the side, or when they are putting on lipstick? what about when they are speaking in a mirror?

sync-3 handles all that and more with ease, giving you the ability to tell your stories however you wish

[Rudrabha Mukhopadhyay @RudrabhaM]

@thenomadevel @synclabs Fair point! It is heavy, 16x larger than lipsync-2-pro. We needed the scale to get facial motion right, but we know that limits who can use it today. Starting with studios while we work hard on making it faster and more accessible. Lots still to improve!

[Rudrabha Mukhopadhyay @RudrabhaM]

@jmsltnv @synclabs Done! Can you check now?

[Rudrabha Mukhopadhyay @RudrabhaM]

@jmsltnv @synclabs Hey, can you share the job id with me in a DM? Also, we have this feature called speaker selection, which will let you choose the speaker to lip-sync. Otherwise, the model randomly picks one if there are multiple people.

[sync. labs @synclabs]

today we are launching sync-3: the next evolution in lipsync

every model we've ever built has led to this

every story for every audience