I must test the DSL 5.0, regarding my sloping audiogram and what I have just read: "(…) In practical terms, loudness equalization [like NAL] procedures usually require less gain for soft sounds than loudness normalization procedures [like DSL]. In a steeply sloping loss, for example, the equalization method typically provides less audibility of soft speech for the high frequencies. There hasn’t been a lot of definitive research on the topic but, in general, it’s been found that true normalizat…
I am encouraged also by what I have read.
You’re ignoring this member’s posts.