typical -db setting for normalize ?

I am using the Audacity 2.1.1 with Mac Snow Leopard 10.6.8 and when using the normalize effect with the default minus 1 db setting for amplitude.
On playback I hear no distortion but peaks are appearing as red, although obviously nothing is going over minus 1 db .
I just wondered whether it is a matter of personal preference where you set your -db… or does it depend on the material ?
On playback, mine sounded a tad “hot” and I was thinking about trying -2 or 3 db’s to contrast .
I just wondered what other forumites have done with their recordings

If you have only one track in the project, normalizing to -1 (minus 1) dB will not allow clipping.
The “orange zone” in the meter is the recommended level for recording (so as to allow a little headroom). After Normalizing it is absolutely OK for the level to go into the “red zone” provided that it does not go so high as to cause the clip indicator to go red.

The “clip indicator” is the little red vertical line on the extreme right end of the meter. This image shows the clip indicator red, which indicates that clipping has occurred (the level is too high):

When exporting as MP3 (which is an inexact “lossy” format), the peak level in the exported file may be a little higher (or lower) than the original audio in Audacity. This is unlikely to cause noticeable distortion, but if (like me) you prefer to keep the peaks below 0 dB, you may want to Normalize a little lower than -1 dB before exporting as MP3. -2 dB (minus 2) should be a sufficient amount of headroom in all real-life cases.

When exporting in a lossless format (such as WAV or FLAC), -1 dB (minus 1) should be plenty, though some people prefer to go a little lower (or even as high as 0 dB).