ReasonTalk.com

In the past I have used the normalize option to clean up merged audio clips so that they actually become one long clip instead of being treated as one in the sequencer, but this changes the levels in the clips. Is bounce to new recording a better method for retaining original levels while also cleaning up the audio to remove embedded versions of the old clips?

Yes. You have to delete the unused recordings afterwards but  yes Bounce to new recordings keeps the same levels. 

Isn't there a clean up unused recordings option somewhere?

It's under the right-click menu of the clip, but only shows up if there are indeed unused recordings.

To be honest, I have never really understood what makes something qualify as used or unused. Bouncing things is a bit of a head scratcher for me at times.

A clip is unused if no parts of it are active in the comp editor. If you have an audio track that's the entire length of the song, and cut it in half, and then record a new half. The original clip is still half used, while you can delete it from the new recording it will still take up the full space in the file. But if you cut the clip in half, and then bounce both halves, they will be stored as two individual recordings, and the original will be completely unused, and can be deleted from both halves, and then when the file is saved and optimized, the data will be discarded. Then you can replace one half with a new recording, and the half what was replaced will also be unused. If you join clips together and then bounce, they will be represented by a single recording.

Yeah, that probably didn't clear anything up. (But I think I just found the topic for my next instructional video.)

The more time spent in the comp editor, the more obvious it becomes about working with a sliced single recording, verses multiple recordings.