Comparing audio changes at matched levels

LucasEa

I am planning a short workshop where students compare generated music exports. Students should describe what they hear before discussing the source or processing method. Neutral labels and matched playback levels seem essential for the exercise. The versions would share the same passage, avoiding differences in composition. Students could note the location of a distraction before suggesting a cause. The group includes people with different listening experience, so technical vocabulary should be optional. I would collect individual notes before discussing them together, and let students replay the excerpt when they need to check whether their first impression holds up. I found a resource on ai audio artifact removal that might inform an optional later comparison. I have not evaluated it and would avoid presenting it as a demonstrated solution. The exercise needs room for the original version to be preferred. How would you structure the notes so expectations do not determine the conclusion?