I mean physically cutting those sections from the track. I notice some other mixing engineers' DAWs only consisting of sections of track that actually contain audio, not complete tracks with silence included. I wondered if there was a reason for it, or perhaps they were just where those sections were punched in.
I don't remove every little bit of the "no audio" sections on all tracks, but I will chop most of it out, especially the longer stuff....like not the small mini-pauses between notes or words in a verse...but anything more than a few seconds or more.
I've always done it...to clean up tracks, and to prep them for any editing, comping, or arrangement/production changes...so it's like the first thing I'll do when I get all my tracks in the DAW. I can do it really fast, maybe takes me about 30 minutes to clean up all the tracks on a given song.
Like...if you're playing maybe a few notes on an instrument, only in some parts of the tracks...there's really nothing to be gained from leaving in all the "no audio" sections on a 4-5 minute track...and those sections may not always be silent, especially if you're applying processing on the whole track, some of that processing might pull-up useless noise that doesn't really add any value to the total sound, that wasn't heard before.
Very often when I'm recording a track, unlike the old days of using a gate to keep it clean in-between, I know I can easily chop out stuff in the DAW...so then when I'm playing, I don't really worry about hitting a string accidentally between phrases, or coughing between words, or whatever occurs in-between the actual playing/singing.
It kinda gives me more freedom when recording...rather than worry about any little noises that occur in-between the actual audio.
Another reason I'll end up chopping it out is that I'm already going to cut up some tracks because as I said above, I might be moving things around, maybe changing the arrangement on a section, on the fly in the DAW, or simply because I may comp 2-3 tracks into one...so I'll just go ahead and make cuts where there is relevant content, and by default, remove the silent sections.
Besides that...I track to tape, and while I do record fairly hot, and the playback from the deck is surprisingly clean, there isn't an obvious tape hiss...if I end up boosting some highs on some tracks, there is the chance I will pull up that tape hiss, which in the "no audio" sections could add unwanted noise to the track.
I'm all for getting as clean as possible tracks.
If I was recording a live band, then room ambience, noises and "as it falls" mentality would apply.