The network for creativity
Join 1.25M professional creatives like you
Connect with clients, get discovered, and run your business 100% commission-free
Creatives on Contra have earned over $150M and we are just getting started
Back to feedPost
Some recent tests using Seedream 5 Pro through Magnific Spaces. The goal was to see how far I could push character consistency and photography direction using a node-based workflow.
Skin rendering is still my number one enemy and there are aspects that don't quite look natural, but the node language is starting to click for me, and that opens up a lot more control over the final result!
Sharing some of the examples below 😊
The network for creativity
Join 1.25M professional creatives like you
Connect with clients, get discovered, and run your business 100% commission-free
Creatives on Contra have earned over $150M and we are just getting started
Trending
Claude
Claude has entered the design space. How are you using it?
Contra University
Learn from expert creatives how to earn more using next-gen AI tools.
Brand Design
The best brand designers are on Contra. Scroll to see what's trending in brand design. What are you building?
creativeaiflow
Creative AI workflows are evolving. What tools do you use, and what are their strengths and weaknesses?
freelancerlife
Freelancer life is wins, pivots, and everything in between. What’s yours right now?
Related posts
I use FigmaWeave as my go to tool for the creativeflow, by the recent introduction of Gen Effects it seems more easier to adopt AI in Post-Production as well.
The node view makes the move into post-production easy to follow. Have Gen Effects been more useful for polishing one finished clip or for trying several versions earlier?
Hello! Just launched my second Framer template -
Multify - a premium template for AI startups.
Every AI template out there is dark, cold and identical. So I built the opposite: warm, editorial, and made to feel trustworthy, because in AI, looking credible is the sale.
With 13 pages, 4 CMS collections, this template is just right for businesses looking to publish their site fast.
Check out the live preview here.
Cool stuff
Ai artists manifest in Seedance 2.5
Pushed the new model to its limits with this dynamic AI music video.
▾Few takeaways:
Prompt adherence has improved significantly.
Audio artifacts are almost completely gone.
Generating a full 30-second video in one pass is still risky — control and smaller details start to fall apart.
Omni Reference reinterprets the input audio instead of preserving it exactly. For precise synchronization, separate shorter generations are still the safer workflow.
I couldn’t stop the model from repeating parts of the lyrics. Adding the lyrics to the prompt actually made it worse, sometimes creating duplicated lines and extra voices.
Local editing tools probably won’t solve that particular issue.
So be careful with your credits when generating 30-second clips.
Still, the result with the BytePlus is kind of staggering — especially the physical consistency, camera dynamics, character performance, and environmental motion.
Made in CapCut with GPT-2 for references.
The 30-second failure point is useful because it changes how you plan the whole edit. I'd keep one visual detail fixed in each shorter shot so there's something stable to match when you stitch them together.