View source for Talk:Gain Structure and Optimizing Levels

I think this is saying two things:

(i) Try to make sure that your initial signal is as strong as possible.  You should keep any gain controls to a minimum.

(ii) Where you must apply gain, try to add a little at each stage in the chain, rather than all at one point.

Personally, I find this a bit confusing.  If my initial source is, say, a CD/tape player without "Line Out", I need to use the headphone/speaker output.  Should I set it low, medium or high?
If I want the strongest possible signal, I can turn the volume to max, but even as a child I knew that this would cause distortion.  So is this where (ii) comes in? 
Should I set the player to the maximum volume where distortion is not discernible to my ear, and then play with the VU meters in Audacity?
Or would you set it to the minimum volume that Audacity can detect, and apply gain within the Audactiy software (on the basis that there might be something different between the volume control in the tape player and what the audio people call 'gain')?