My experiences with speech enhancer and suggested features to improve it

[natural_Lark5EE1]

I took some old 2006 podcast recordings that sounded like they were in an echoey auditorium and were just horrible. I tried every speech enhancer available. Adobe Podcast is the only one that was able to make a radio broadcast quality voice identical to the original. I applied dereverb to the original recording and mixed it with the AI voice. PERFECT! I am impressed with the v2!

Now I applied it to an even worse distorted old voice recording that was barely distinguishable at all. Adobe STILL was able to recreate the dialog in a useful way. Although some words were spoken wrong which is understandable considering the quality of the original. So that leads me to a suggested feature.

If it is possible, there should be options for word detection where the AI can further auto correct words that might be badly interpreted. Sometimes a recording is scratchy and horrible. So the AI will speak what it hears. AI could take words it badly detects and change them into (hopefully) the correct words. This should be an option to turn on and off. Just like we have Auto Correct on our phones (which sometimes goes badly), AI could do this to speech.

[Harmony J]

Hi @natural_Lark5EE1, thanks so much for your thoughtful feedback on Enhance Speech. I think this is a great request and will share it with our product team. Appreciate you being part of the community!