When FAVE-align processes the .txt transcript and spits out unknown.txt it's reading one of the lines as ?unicode or something else and gives an output file like the picture below:
When I then put in the pronunciations, delete all except columns A & B and feed it back through as input.txt FAVE can't recognise the first phone so errors straightaway.
My guess is that at some point the .txt file is being read as unicode and converted and so doesn't work from then on, but I don't know enough to fix it, anyone experienced this before/ have any suggestions?
Thanks,
Caitlin