>>2250938It's a bit hard to say since videos can vary in bitrate/file size so much. I also haven't really been keeping specific notes; so far I've just been doing it based mostly on guesswork and gut feeling.
That said, if the clip is short, like 1-15ish seconds at 720p, I pretty much don't use two-pass. If the clip is long, like 30+ seconds, I'm much more likely to do two-pass to begin with, to save myself some time and guesswork inherent in the normal method.
If the clip contains a lot of detail and movement, and you just want to be sure to squeeze the value out of every bit, two-pass is justified.
It's also justified if you've already tried the normal method and the webm was still too big. Two-pass will get you the right file size guaranteed, without having to tweak the b:v and crf values a bunch of times.
The normal method with -b:v and -crf, and the two-pass method will both generate files of the same quality, given a certain output file size. You're just making a choice between the 100% reliable, but slightly more cumbersome two-pass method, and the simpler but more prone to getting the output size wrong normal method.
I hope that was clear enough. Those are just my thoughts, I'm not perfect with the process yet myself.
