I was initially planning to only upload a test of episode 1 and then go to five episode batches, but I ended up changing so much stuff with episode 2 that I felt a need to get it out for more feedback. The first change is that I've now started using sushi to automatically retime the dialogue to fit. This seems to have worked well, much thanks to govna. Second change is that I've manually retimed all of the signs that sushi couldn't handle because there were no audio cues (I only checked one sign in episode 1, I checked them all here). Hopefully this didn't mess anything else up. Third change is that I shifted both the English and Japanese audio forward by 40ms as they seemed desynced compared to all of the other releases I have. The third change is the one that prompted me to make this upload. 40 milliseconds was an educated guess and I'm still not confident that the audio is as it should be, although it's definitely better than it was at least. Any suggestions for a more objective way of figuring this out are welcome.