I got and adapted the questions from a survey about these issues but I've been trying to get the question right as we tend to use language which presupposes things.

I like this one as it asks to form an image and asks about clarity. The image is about hearing. So it crosses both parts.

Another example would be to form an image of a telemarketer on the telephone. Change the accent. Introduce line distortion.

For me it's in the middle. Happy birthday is clear to visualise and I can follow the tune. it's not very vivid but it's like I'm singing to myself with my mouth shut. I don't have a mental "visual" image of the song by default. When changing it the imagery appears a bit more but the focus is on the sound. I can easily change it to children's voices, the unix greybeards is more difficult as it requires me working out what they would sound like including spatial echoes from the auditorium. I find the resulting image (which is clearer than the children) is amusing.