Regarding scaling while maintaining ratio, some codecs will give an error if you try to put a size that isn't a multiple of n, for example if you put -2 you'll

Regarding scaling while maintaining ratio, some codecs will give an error if you try to put a size that isn't a multiple of n, for example if you put -2 you'll actually have it calculate the appropriate multiple of 2.