I record through a console.  I tweak individual tracks and then tend to mix stems into SONAR buses ITB.  I've been running those stems back though the console and printing the stem to a track back in SONAR.  When I'm happy with all of the stems I once again go back through the console to sum a final mix.  I'm curious how others are accomplishing their analog summing.  Some of the things on my mind are:

1. Is there a best practice for how many times you should run your audio through the console?  In my case it's at the recording, summing an individual stem, and then summing all of the stems for a final mix.  Is there such a thing as "too much" analog goodness?  (I know that this is subjective but wish to hear your thoughts).
2. Is there an argument for recording direct to the box and reducing the number of times that your audio is hitting the console?
3. When printing back to the DAW do you drive the analog levels pretty high at the console and record the stem tracks or mixes at respectable gain staging levels within SONAR?
4. For those of you that are doing analog summing, are you using the console emulators on your tracks or are you letting your summing mixer (or console) add the warmth and character that we are looking for?

I've tried a number of different approaches but haven't yet sorted out my "standard process" (there are always exceptions though!).  Please share your methods, results, and reasoning if you care to participate in this thread.  I look forward to your responses!
 
Best,
 
Dave

I truly believe in recording the best possible sound first. 
All my inputs of course are in the analog world before hitting a track in Sonar. I have bought, instruments, amps, pre amps, compressors and the mikes etc to get the sound I'm happy with. They are 98% done deal before they are recorded.. 
I would see no advantage to run them out and back in as the D/A to A/D would negate any added benefit the console would add if any. 
 
 

As a general rule of thumb the less times you run through converters the better, each time there is a dimension of noise/signal degradation added etc,..really though "best practice" is whatever sounds best...I sometimes run stuff out to an external insert so I can use some real gear...its a pita though...I think what you have described in your own practice is pretty SOP for console driven studios you gotta go through the AD converter a few times no matter what unless you do it all ITB...the real issue here will be the quality of your converters and the resolution you use.. just watched a video with Rupert Neve where he makes a comment about UAD emulating his stuff (he never names them but it is the only plug-in company he works with) in it he makes a comment that at the higher resolutions the plug-ins were very very good and almost imperceptible to his real gear which I found interesting...point being I think if you do your conversion at 96k then you can negate a decent portion of the conversion loss issues...

saw a pro mastering engineer interview once, he prefers music to have no more than 10ms latency.
If you run in/out Sonar twice, your latency will be over those 10ms probably...
I'm still learning but found that the more you process audio, the worse it gets usually (a cliche but true imo).
like above posts describe, it seems to be best to minimize converting and tweak sound before it enters the console or DAW.

Hi Dave,
I do analog summing, in my case I have a 16 channel SSL with their stereo bus compressor and stereo EQ and a bunch of other outboard processors.
re: your questions:
 
1. I don't think there is a best practice for this, other than "until it sounds the way you want". Keep in mind that you're adding distortion with every pass, so at some point it's gonna make things sound bad, but other than that it's a purely subjective thing. Also, if you've got anything resembling decent converters the amount of distortion you're adding with just conversion passes is pretty minimal, so I wouldn't worry to much about it unless you're doing lots of passes. Again, let your ears be your guide, but don't worry about it too much. Lots of folks still use outboard all the time, and if conversion passes were really messing up your sound that much no one would still be buying 1176's and Distressors.
 
2. I don't think so, at least not until it sounds bad.
 
3. Yep, my "print" track is fairly hot, but leave some room for the M.E./Film Editor/whatever if you're sending it out to anyone else. I try not to peak much higher than -6dbfs. Remember that if you're looking for saturation in your console you should be able to saturate your console channels and the internal mix bus and then theoretically turn down your master output to print at reasonable levels back into Sonar.
 
4. I don't use console emulators much. Sometimes I'll drop one on a track because I want to soften it a bit in a specific way, but it's not super common. Keep in mind that you should really mix through your console so you can account for its sound while you're mixing. If you're doing that then you essentially have a "console emulator" on each track anyway, only it's not an emulator at all, it's the actual console. If you want to do something in addition and the console emulator lets you do that, IMHO do what sounds good.
 
As far as my method goes, I route all tracks to more-or-less standard stems (drums, bass git, vox, bgv, fx, and what have you) and then send those out hardware outputs to my SSL. The SSL then sums and sends everything through the bus comp and stereo EQ and whatever other outboard I happen to be using on that mix. From there the SSL has its main out coming into my "print" track which is going to the Master bus in Sonar ("print" tracks are the ONLY things feeding the Master bus in this scenario). I enable live monitoring on my print track so that I'm listening to the signal coming into that track when I mix. The Master bus then has analyzers, A/B tools, metering and ARC?whatever on it so I can do measurements and comparisons. When I'm ready to print I record arm the print track, name it something like "OMG NEXT GIANT POP HIT 9-22-15 A" and hit record. If I make adjustments I create a new print track, change the track title to reflect current date, revision letter and, if necessary, usage ("OMG NEXT GIANT POP HIT 9-22-15 B, TV mix") and record again. This way each of my mix revisions occupies a separate track and I can A/B them or export them individually.
 
Dean

With the antelopes I wouldn't worry too much about conversion loss.  It will be negligible. 
 
The "normal way" (which is not the right way but merely the way many would do it) to use a summer is to take the digital tracks out directly to the board.  Usually you do volume automation in the box before going out, but doesn't the matrix have its own analog automation?  If so, that is where you do it so that any outboard equipment in between (for tracks or busses) gets hit a full volume before it returns to the analog board, where the post-effect automation allows you to take that full volume sound and reduce it.  Most of us don't have flying fader Neves at home, so you automate the track levels before they go to the board.  I would also use digital track/buss effects to fix any obvious problems since digital eq can be very precise.  Once in the board you do whatever you like to the tracks and integrate any hardware outboard you have - comps, eqs as well as bussed hardware reverbs.  At this point in summing I would use eq not to fix problems but as a kind of mojo enhancer.  That is why we spend money on hardware, or some of us.
 
Once you like the mix just print the ssl output back into the antelope.  Voilà.  No need to have multiple re-digitalizing unless you want to save all the analog changes on every track or buss to remix the thing later with that special tube compressor you managed to borrow for the day.
 
That is how I'd do it any.  Part of the alleged plus of analog summing is that the stereo digital output isn't strained by breaking out stems.  That used to be true w/ shallow bit rates, so I'm not sure if that still stands.  I've never felt Cakewalk's digital pipes ever got clogged - it sounds pretty damn pristine.  Which is why I like to output the master buss through colorful hardware and add a little analog saturation hair and shaping.  It is much the same, to my ears, as stemming through an entire summer without nice, extra hardware in line.  If you do both, putting summed stems each through effects, you can really thicken things up.  I don't know how the matrix sounds, but it should be a nice sweetener, and add some nice glue from the buss mixer too.
 
@

If it sounds good it is good.

Thanks to everyone who has responded so far.  A lot of good insight and experience from our great forum users!
 
JohnnyV- I'm with you 100% on getting it as good as you can on the way in.  That does make a lot of sense and I subscribe to that advice as well.
 
Psychobillybob- that's what I was thinking too..... it seems like  less conversion may be better although our gear keeps getting better and better.  I do sample @ 96Khz and I don't detect any loss of quality but I suppose that it's there at some almost imperceptible level.
 
Wouter- ditto.....same as above.  Thanks for your reply!
 
Dean-  I'm happy with the results that I get, but you are correct....every pass adds a little distortion that never goes away.  I think that my converters are acceptable and minimize that problem but I'm always looking for ways to improve.  Using a hot print track seems to work well for me too but as you suggested I'm very careful to keep the levels under control as I print back to SONAR.  It's also good to hear what you suggested about running the channel levels up and keeping the master levels modest.  Again, in my experience that seems to work really well and does indeed give me the "edge" that I'm looking for when I use that method.  Also, thanks for describing your process.....it makes a lot of sense to me.
 
AT-  Yep.  Matrix has automation capability but I'm so used to using the SONAR automation features that I tend to do most of that work ITB.  Your suggestions are interesting and I will give some of them a whirl on the next project.
 
Bitman- sometimes we all worry too much!  You're right......most of the time I am able to get things sounding the way that I want to.  I'm just one of those guys that wonders what the next little tweak is that might make sound just a bit better!  I'm a firm believer though in what you say.  Every situation is different.  Way back in the 60's Abby Hoffman said " If it feels good, do it!"  I think that it's the same thing here.  If it sounds good, then it is good!
 
Thanks to all!
 
 

Possibly the music might not suffer from running through the analog gear (in terms of tone, image and sound it can get better).
But the dynamics will suffer from the DAW importing and exporting (to my knowledge) and not stay fresh like the dry mix (less attack & drive due to latency from the im/exports).
At least, that's what I found A/B-ing a dry mix to one that I processed outside Sonar, the best way for me to hear that is quick A/B-ing  on studio headphones.