Normalizing Audio in Premiere Pro is a miserable experience
I should be able to select all clips in a sequence and use the "Normalize All Peaks to" feature. But I can't. Even when I use the workaround of waiting for "Peak Amplitude" to finish calculating before clicking on any of the other UI elements, if I click "OK" after "Normalizing All Peaks to" 0dB, most clips have lost all audio and are silent. So that feature is unusable for me.
What I'm doing now is to manually select each and every clip in the sequence, and for each one I set "Normalize Max Peak to 0dB" manually. It takes a loooong time. On occasion I have to cut a clip into multiple pieces to make the audio sound even.
My ultimate goal is to have my voice have roughly the same volume across the entire video. Using the "Loudness" balancing feature in "Audio Essentials" isn't sufficient to make that happen. "Loudness" by itself doesn't get my voice volume anywhere close to 0dB or even 3dB. So I have to use "Loudness" as a secondary feature after I've gone through the painstaking process of manually applying "Normalize Max Peak" to every single clip. :( I'm super unhappy about that.
And no, I don't own a license for Adobe Audition. I refuse to pay extra money for a feature that supposedly exists in Premiere Pro already!!! I might start using Audacity to normalize the audio somewhat before I start using it in Adobe Premiere. Just saying that out loud makes me feel sad inside. :(
Matt Stegner commented
The Auto-Match feature, and the Normalize feature will end up canceling each other out if I am following what you are describing above. Both features analyze the audio content of the clips on the timeline and adjust the Gain.
The analysis portion of the two features work very differently though. The Peak just looks at the instantaneous volume level and adjusts according to that. That isn't how our ears work though, we hear in an averaged fashion. The Auto-Level feature scan the audio and calculates the perceived volume for each clip. This is more akin to how we, as human beings, hear audio.
"My ultimate goal is to have my voice have roughly the same volume across the entire video"
Normalization is not what you want to be using then. I see a lot of misinformation about normalization as it is used a lot in tutorials when it probably shouldn't. The auto-match feature is designed to set all of the audio clips to the same volume level, which is what I think you want.
If I can make a workflow suggestion based on your statement "My ultimate goal is to have my voice have roughly the same volume across the entire video".
1. Select the voice clips you want to have set to the the same level. Tag them as Dialog in the Essential Sound panel.
2. Turn on Loudness>Auto-match. This will scan through each file and adjust it to a LUFS level of -23dB. Each clip will be adjusted by a different amount, but the total perceived volume between all the clips should be the roughly the same.
3. "get my voice volume anywhere close to 0dB or even 3dB" What needs to happen next is to add an equal amount of make up gain to raise all the clips by the same amount. So, with all the clips still selected, turn up the "Clip Volume>Level..." section of the Essential Sound panel. Play the audio and watch the meter. Adjust the Level slider until you reach the desired output.