The editing is fast paced because video models prior to Seedance 2.5 do best on lengths under 10 seconds and decohere with intention frequently.
A typical "scene" might cost thousands of dollars in generation expenses depending on shot complexity. Directors will shoot the same shot dozens of times if they're not getting what they want, which in a way is not unlike physical production. That's not counting the experimental first passes.
The editing is fast paced because video models prior to Seedance 2.5 do best on lengths under 10 seconds and decohere with intention frequently.
A typical "scene" might cost thousands of dollars in generation expenses depending on shot complexity. Directors will shoot the same shot dozens of times if they're not getting what they want, which in a way is not unlike physical production. That's not counting the experimental first passes.