compression chain question

is it better to use a h/w compressor as a mono send/stereo rtn effect or should it be used on the stereo sum output of the mixing desk?

i want to use it a compressor on my mix, but if it comes after the full summed mix of the mixing console, and i am monitoring from the mixing desk, how do i monitor the mix w/ compression?

i could use it as an aux send in an fx chain, but then (1) the aux sends are mono, and therefore summed, losing the ability to apply slightly different compression to each stereo channel... and (2) the compression would have to br done as a difference... or whatever, meaning that it would be blended with the original audio signal, instead of being the only full audio signal.

does this make any sense?

For what its worth, I perfer to use compressors at the end sum because if I don't I usually make a mistakes and clip.

That being said I also use compressors in specific areas as send returns on individual audio channels before the master compressor.

i agree, but then how do i monitor the signal, if it's applied after the mixing desk (which has ctrl room outs to my monitors)?

I'm not familiar with your studio, but a pretty effective (yet possibly embaressing) method I've used has been to use an old laptop running cooledit's spectrum analyser in real time.

If you did something like this could you just split the signal from the board to the monitors/laptop.