The general recommendation for dither processing of any music is to place it at the end.
I usually Mixing and Mastering at 48kHz, then use plugin Goodhertz's Good Dither processing, and finally export using Ableton conversion to 44.1kHz.
This doesn't seem to be the best processing order? Should I convert the sampling rate first and then perform Dither?
If Dither must strictly follow the last step, then it would be best for me to export the converted 44.1 32-float(or 16?) first and then Dither it?