I’m fairly new to Cubase Artist. I understand the use of ‘Snap to Zero’, but it seems like something you would want to use all the time although the fact that is has an on-off button suggests that there are times when you would not.
When would you not want to use ‘Snap to Zero’ ?
You mean zero crossing? You don’t want that if you have to allign the part you just cut to the grid or another part. The cut you make at the zero crossing is not exactly where you placed the cut, so the part will be slightly too short or too long, and thus out of sync when you move it to the grid.
Yes, but isn’t the movement caused by the snap at such a microscopic level that it would never be noticed? I thought that was the point, you could use snap to zero crossing because it made such a small adjustment it would not interfere with audible timing.
You wouldn’t notice it generally, but when for instance doing this to live recorded drums, captured from multiple microphones, the zero crossings won’t be identical. The tiny shift in time can cause phase issues between the different recorded signals.
I hardly ever use “snap to zero crossing”. Only for scratch track edits. I want to have total control (not to let the software decide what to do) and then use crosfades (not automatic) at every cut point.
One time you can nearly always use it is when you are selecting a portion of audio to then apply effects in place. It can help prevent clicks caused by whatever processing is taking place on the selected audio. No guarantees, but usually helpful. Or you can do a crossfade.
Don’t ever use it when:
If it’s a track in isolation, then you should be okay - I use it a lot to top and tail vocals. As a rule, though, I leave it off because using in error can be tragic.
Now , comping lanes is when I thought it would be the most useful 
In my limited experience I have always used it when comping, and have never noticed a problem, althought I can see that it might leave small spaces or overlaps.
Crochety, when comping , what do you do? Use crossfades at each cut? Also, what do you mean by a track in ‘isolation’ and ‘top and tail’ vocals?
Those overlaps are what can cause the problem. If you have several events one above the other but not the same length, you may leave exposed the tail of one of the longer ones, which may contain/cause a glitch.
Yes. I make as many cuts as I need (right across all lanes) then glue the ones that didn’t need it back together. It might be a good idea at this point to select all the events in the final comp and put them in a Part. You loose some functionality by doing this but by now that shouldn’t matter too much and it protects the comp by isolating it from the remaining events. JHP did a video about this (if you can keep up!): Computer für Audioanwendungen - WaveLab - Steinberg Forums
If you don’t want to keep the events you’re not using, select all the events in that section and do “Delete Overlaps” and it will put just the chosen events on the top lane and delete the rest.
Now I x-fade (just the used events) but check/adjust them, especially if there are strong beats at the cuts. The x-fade always straddles a cut equally so it can blur an attack. You can avoid this by overlapping events manually before x-fading.
A track where I’m not using Lanes or Group Editing, so doesn’t need to synch with anything. “Top and tail” means trimming the start and finish either side of a lengthy break.
Hope that’s clear cos I’m off to bed now. Nighty-night. 
Thank you for these explanation! I was noticing phase issues, now everything makes sense!!