原文要点
原文完整内容
So I’ve been creating mostly ASMR videos and some other advertisements using VEO 3 and I know how to use JSON prompts and character consistency.
Now I’m going to start learning Sora 2 today because I’ve seen some funny and impressive videos there.
Have any of you created videos with both VEO 3 and Sora 2? Please let me know what the major differences are if any.
Thank you
原帖关键讨论
Veo 3 sounds impressive with its strong physics and character consistency. That’s so important for controlled storytelling, especially for complex scenes. Sora 2’s natural language approach seems like it could cut down on a lot of learning time. I’m curious if real-world users find one easier than the other in daily workflows
sora 2 is gong to shut down
Veo 3's physics engine and character consistency strong for controlled storytelling. Sora 2's natural language prompting easier learning curve. Freepik doesn't fit video but stock integration could help B-roll mixing. Real differences: Veo handles technical specs (JSON, motion control) better. Sora more intuitive for narrative flow. ASMR content benefits from Veo's precision control. What's your actual priority control precision or ease of use?
Создат игра
来源:https://www.reddit.com/r/VEO3/comments/1o5q9v0/veo_3_vs_sora_2/
Sora's character consistency in dialogue heavy scenes is worth the learning curve though. Either way run your thumbnail stills through Magnific before posting, ASMR audiences are surprisingly detail oriented about visual quality.