Models for extracting Dialogues vs ME-track from OLD MONO video

[adrjork]

Hi everyone, absolute newbie here. Just two questions:

1. In my case I have an old VHS (that I digitalized) with mono audio: which model is the best to extract dialogues vs all-the-rest (i.e. music&effects)?
2. Same question but with a clean stereo DVD version of the same video: should I use a different model?
   Thanks in advance for your kind help.