AutoSense switching can cause volume fluctuations and so much program shifting that it confuses users.
What about just spending more time with programming Autosense subprograms?
FWIK, changing Autosense subprograms isn’t abrupt even with the shortest transition time (10s - my case). There is continuous gradual change e. g. from Calm to Comfort in Noise (the most audible differences in my case).
YMMV, but automatic change from Calm to Speech in Noise without engaging my hand is very comfortable and makes less mental burden.
You’re ignoring this member’s posts.