Bubbly, distorted, honky voices are generally caused by overuse of Noise Removal, or possibly just using the tool wrong.
Noise Reduction isn’t an open-ended tool. You don’t turn it loose on your show and go to lunch. It’s a balancing act and shows can fall off.
It’s super important that your first step, the profile step is a segment of the show that’s Noise Only. Audacity will try to remove whatever you put in that sample from the show. If you put a little voice in there by accident or maybe breathing or clothing noises, that will drive Noise Reduction nuts.
Then select the whole clip or show and apply Noise Reduction starting with very gentle settings. If you’re reading for an AudioBook, I like 6, 6, 6. Those settings have worked for a number of people. If that’s not enough, increase the first number. 12, 6, 6. If you make it to 18, 6, 6, the show is going to start sounding funny much like what you have now. That can mean a permanently damaged recording.
That’s when you have to fix the recording equipment or environment. This isn’t magic. You can’t set a microphone up in a noisy, bare-floor apartment or next to a street and walk away with an ACX AudioBook sound product.
Are the raw capture clips posted anywhere we can hear them? We know what the corrected, honky ones sound like. We don’t need to hear those.
If you feel like it, record and post a simple, 10-second voice clip according to this recipe.
It’s a very newbie error to immediately destroy the original capture sound files. Big Mistake. You may need to go back to those if anything happens to either the edit or the posting. There should never be a reason to announce or perform twice.