Even if you could, the effective width that sounds best on headphones is generally not what will sound best when played back over speakers - usually what sounds best on headphones is too narrow over speakers.
Correct, right now since I don't have any speakers anymore, I'm limited to headphones, so I try and get a mix that sounds great over headphones when I merge the channels by 10-15%. Another way is to get a good mix, then add about 2-3db of side to it (almost sounds too much) to spread it out a bit. That way it's a compromise between headphones and speakers. The downside is I spend a lot of time tinkering with the mix as a result to get it where I want it in post.
[m/s n00b]how do you guys mix the mid-side in post? i have Soundforge 6 and Adobe Audition 1.5...i assume i can use one (either?) of these programs to do it?
my lack of knowledge on this subject is the only reason i haven't run this config [/m/s n00b]
Well, there are (in general) two ways. One is to use a plugin and tweak with the fake nobs and stuff till it sounds like you want. The second is to take the two tracks, duplicate both, pan the originals to a full left, and the duplicates to the right, then invert the duplicate side track and play with the decibel level of the resulting mix. The later is how I do it in Audacity since I can't use most VST plugins and I havn't found a good AU plugin. The former is easy to do, just takes some time to listen for what you want the mix to be. The later is a little tougher, but you get to play with the channels individually so if you want to rotate the mix to "look" left or right, it's easier to do.*#*
*#* Presuming that the VST plugins link the channels, if they don't then this is moot.
Also, if you don't pad the side with extra gain (or remove the same amount afterward), the stock 1:1 ratio yields an image which is reminiscent of a hyper pattern with an angle of approximately 135.
If you are really interested in midside recording, take a look at Trevor Owen de Clercq's Master's thesis paper on M/S imaging. I don't recommend you read it in one sitting though, read 4 pages and put it down for digestion. Just remember that results (in theory) will be better for identical capsules, and not a mishmash as he uses in one of the experiments. It's a shame he didn't use something like the LSD2. (
http://www.midside.com/pdf/nyu/masters_thesis.pdf)