Stem Strategies

I rarely do them, but on the occasional project I'll be asked for stems. because I use so much outboard, I'll ask the artist how they want them, with or without effects, and I'll explain why they'll sound different than a finished mix. If they are still OK with it, I'll just run each group through the mixer like I'm doing a mixdown. If they don't want that but still want the stems, I'll just do rough levels and export them.

I'll generally refuse stems if someone tries to give them to me, mainly because I have so much stuff on single tracks and rarely do bus effects other than compression.

If it's a project I'm producing or overseeing, then it's all cut live to the Radar, and mixed from Radar thru console.
At mix time, I do these versions, printed usually to a Masterlink 24/96 using a 2192 or a Hedd Convertor.
#1 Main stereo Mix
#2 Vox Up +1db
#3 Vox Dn -1 db
#4 Instrumental #1 No lead Vox or FX, but has Bk Vox and Harmonies
#5 Instrumantal #2 No Vox at all !
#6 Lead Vox w fx
#7 BK Vox with FX

Now if I'm mixing (and NOT in charge) this is what I do.
Since I don't use PT....First I put all the tracks into DP- 8
At this point I use my portable rig, which means from the computer it goes out to a Apollo 16, into 16 channels of Helios Type 69's into a Neve 8816 summing mixer.
The Summing mixer is set up so that all the direct outs / Mutes are pre fader. SO the fader on the 8816 are down!

I mix into the Helios/Neve using any and all the outboard I want...
I then take the 16 direct outs back into the Apollo 16 and record stems as such:
#1 Kick
#2 Snare
#3 Drum L
#4 Drum R
#5 Bass
#6 Solo
#7 GTRS-L
#8 GTRS-R
#9 LD VOX
#10 Vox-2
#11 BKVOX-L
#12 BKVOX-R
#13 KEYS-L
#14 KEYS-R
#15 FX / SPARE-L
#16 FX / SPARE-R

All of these 16 channels are assigned to one stereo output via the Apollo's digital out.
Using a DAC-1 I take the digital output convert it back to Analog and insert that into the Neve's BUSS INPUT on the 8816 mixer.
So this new stereo mix is now getting the Neve's transformer out sonics as well, as everything before it, and I also have an original SSL 384 Mix Buss Compressor in one of the 8816's patch points, (this one has a blend control!)
and a Summit Stereo EQ (in MS) on the other mix insert.

At this point I take the stereo mix out of the Neve, and into either a Hedd, a 2192, or my UNL-8 and record the 2 mix as well as all the stems in one pass.
This gives the artist / label what they want, and if asked I usually can make any changes required just using the 16 stems.
Now even though the stems have all the processing imbedded in them, they will NOT sound like the STEREO MIX~!
Simply because I ran them back into the NEVE thru the SSL and the SUMMIT, so once I change the stem levels in a remix, I just run everything thru the Neve as I originally did.
(job security)
Done...

But by the same token a bad can change levels of the stems on a laptop and send me the files and I'll just run everything back thru...
Simple and easy.

Now this is some great info.  Thanks!