The Creative Art of Submixing
Bouncing Down to Greatness...
First I will mention the advantage of using a recording system with a large number of available audio tracks.
When working with the number of tracks that most commercial productions have available nowadays there is no need to bounce down any tracks to stereo. You are likely to always have tracks available to overdub your 30th guitar track or have 20 tracks for the drumkit (yikes!) and you can feel safe in the knowledge that even right up to the last second before you declare your mix complete, you can go back and change any setting on any one of those tracks in the mix. The obvious advantage here is flexibility and having a single session where all your mix variables are situated.
However I believe there is one aspect to this method that is hugely restricting. Your creative decision-making becomes impeded. What do I mean by that? Well, it's a bit like having a fear of commitment. In fact it becomes exactly like that. There is a danger of you feeling like your mix is never quite finished and the thought of committing a particular sound or setting without being able to change it later can become too much to bear. Maybe it's the "I've got 128 tracks, so I have to use them" syndrome or the "If I commit the sound now then it might not be right later on" mentality.
The result can be a mix that, whilst sonically good is a little lacking in soul. I believe that a large part result of a good mix is the creative energy put in by the mix engineer. I just think that mixes sound better when certain commitments to the sound are made along the way.
The submix process is a fine art and one which is fundamental to getting all the tracks to "gel" together nicely, making the overall sound carry more punch and weight. It's like glue that holds together all the instruments as one big sound rather than loads of little sounds all competing for the same space. Also in a software-based Digital Audio Workstation (DAW) I believe that the more stressed the CPU of the computer (i.e. the more tracks it has to process during mixdown), the more detrimental effect there is to the sound of the final result. This may not apply to a powerful system like Protools HD, which has it's own dedicated processing power for just that and indeed may not even be true on any system, but I have quite strong opinions on this based on my own experiences.