is it better to apply the unsharp mask on each frame before composing with HF, or is it better to postpone it on the final image?
Does it affect the precision of the algorythm?
HF will be able to merge the pictures even without sharpening first. I think that you would find it to be less work if you only sharpen the final product.