So I am working on a large 24P project with a lot of footage, filmed with a Canon C100 and editing in Premiere. Partway through my project I added an Odyssey recorder to my workflow. Unfortunately, for several of my early shoots I failed to correctly set the record menu to remove pulldown from the camera HDMI output (I just had it set to follow the framerate of the camera, which is not the right way to set up for my camera). So I have a bunch of footage that needs to be converted to 24P with 3:2 pulldown.
As far as I know, it's best to do this before I start editing so that all my clips are in one framerate. I did a bunch of experiments with different ways of doing the framerate conversion, and by far the easiest and best looking has been by using the interpret function in AE, changing the field order and telling it to guess 3:@ pulldown. This is a smooth and easy process and the converted footage looks great. I am an inexperienced AE user, but this workflow is simple enough for me (basically similar to this: https://www.artbeats.com/assets/articles/pdf/pulldown.pdf)
BUT I just noticed a big problem, which is that when I render these new 24P clips, my two separate mono audio tracks (using two separate mics that need to be kept separate for editing!) are mashed into one mixed down audio track. This is not going to work for me, as I need to be able to work with these two tracks separately. I also tried using the audio from my original unconverted clips with the new 24P video, but–not surprisingly–the audio drifts due to the mismatched framerates.
Is there a workaround so that I can both fix the framerate / pulldown on my clips and preserve the original audio information for editing?