Logic Pro 9 TV Stem and Full Mix Delivery

I am currently working on a weekly television series. I need to deliver my mixes in six individual stems as well as a full stereo mix. I have a Logic template set up where I route the outputs of all my instruments to individual stem buses, which then all go to a master bus.

When I am mixing, I realize that if I put any bus plugins on the master bus, the full mix will then not be equal to the sum of all the buses.

I'm curious how others, who deliver stems and a full mix for film and TV, use bus compressors on both the individual stems and the mix.

I'm always looking for better ways to get this work done quickly...

Currently, I don't do any Stem bussing until after the mix is approved. Stems are the very last thing I do. After performing a Save as I change the outputs of the tracks to the correct buss. Often times I need to create additional Delay and Reverb auxes specific to each stem. So if the orchestra has been going to the same reverb, but I have 4 orchestra stems, I'll recreate the same reverb 4 times, one for each stem. This way each Stem has the reverb or delay appropriate to that stem.

The Aux Channel for each stem might need to get the same Master Channel treatment. So if I was using EQ or Compression to color the mix this needs to go onto each stems Aux. Obviously this means turning off the processing on the master channel. Which is fine because now you can see what you stems summed will be doing to the output level. I like to avoid any overs once the stems are summed.

Lastly. I'll send the output of each of the stem auxes to an audio channel and print all the stems at once in the session. This way I can check for any problems or errors and if needed reprint the one or two stems that need adjustments.

It's a PIA to be sure, but I haven't found an easier way to print them all at once AND check the work as I go. The only way this gets easier... Is by assigning the task to your engineer or assistant :p

Would love to hear if anyone has better ways of getting this done. All of this was a little easier in Pro Tools, but Logic gets it all done in essentially the same way.

Hi Bob,
I work similarly in that I send a stereo mix for approval and then do my stem assignments and tweaking once I've received the 'okay'. I've never had to deliver so many stems and was balking at the thought of replicating onto the stem auxes the plugins that I would normally put onto the Master Bus.

I just had a great 'work flow' session with Eli Krantzberg and am now quite pleased with my setup. I'm not balking:) He'll probably write to you about the specifics of my setup.

We have my stems and full mix routed out of my Logic soundcard into a ProTools 003 since I have to deliver my work in a ProTools session. I can record all of the stems and full mix at the same time for each cue. I'm going to try monitoring through ProTools at this stage of the bouncing process because I've found that my mixes always sound different in PT and I found myself remixing tracks once I'd listened to them in PT.

So I think we're actually doing things in a similar fashion.

Yep, that is a great way to do it...

We have an HD system sitting Idle in Studio A... Maybe I'll print stems that way tomorrow? Depends on if I feel like dealing with 2 computers... : )


Hi Bob,

Sounds like you have a nice workflow. When you say you assign your stem auxes to audio channels, I'm not sure exactly what you mean. Do you mean you assign them to unique physical outputs and then re-record them through unique physical inputs? If so, there's an easier way you can do it. Assign each stem aux to an unused bus. Then create several new audio tracks, and assign the unique busses as the inputs. You can then record enable all of those new audio tracks at once and effectively "re-record" all the stems at once to new audio tracks. All without having to do any outboard physical patching!

Here's a rough overview of how I have Judith's system set up for her. She has a MOTU 828 mk 3 and a Digi 003.

• Each stem arrives at an aux with the unique stem bus as the input. All set up in a row. Let's call this Row A

• Each of those auxes outputs to another unique bus each which I use as kind of a "stem router".

• I create two additional rows of aux tracks. Each row contains an aux track for each stem.Let's call these Row B and Row C.

• The stem router bus is the input of each corresponding aux in Row B and Row C.

• So now we have the stems arriving at auxes, and then effectively split off into two duplicate rows.

• On Row B, I output each aux to another unique bus I call MasterBus. This is for the full mix.

• An aux is set up with the MasterBus input. The whole mix arrives here. This is then split off with another router bus to go to an additional Aux in Row B and Row C.

• The Output of this MasterBus in Row B goes to her monitoring for playback.

• Each of the stem router tracks in Row C outputs to a unique pair of physical Outputs on her MOTO 828 mk 3.

• There is another row of Output Faders, for each of these physical outputs.

• The physical outputs are physically cabled to inputs on her Digi 003, including litepipe.

•Â She then sets Pro Tools to generate MTC, and Logic to auto detect incoming MTC.

• We then set up a group of tracks in PT with each input set and named to correspond with the signals they are receiving from the 828 mk3.

• She can then record in PT and all streams are discreetly captured in PT, including a full mix.

This is probably complicated to follow like this. It's a brain twister following the signal flow throughout the various rows! But after lots of tweaking and trouble shooting; it works!!
Hey Eli,

Regarding your first question, no. I mean that My Stem Auxes get get bussed internally to new audio tracks (whose inputs match the buss). This way I can just hit record and everything is recorded / ready to be verified. From there I export the audio to the desired format. Usually 48 /24 which is what the session is at. I find It can be tricky to keep track of the auxes and channels but practice and color coding make it easy enough. Sometimes there can be a problem with a really large session that might be tapping out the CPU, but I haven't had a session yet where creative freezing couldn't get me through.

I had to reread through your setup several times, but I think I understand the signal flow. I may try this later and ask you if I run into any questions.



Hi Bob,

Yeah bussing your stems out to new audio tracks and re-recording them; that's the way to go. You're already on top of it :)

The setup for Judith was fairly complex to set up. But now that it's done, it does seem to work well (fingers crossed!). It allows for recording all stems plus a full mix out to PT simultaneously.
This is a bit fussy, but seemed to work. What you can do is print your mix as a stereo file, without plugins, compression etc, then bring the stereo file into a track in Logic but don't send it to any outputs. Setup a sidechain on your buss effects that looks at the printed stereo file. The as you print your stems they will respond to the overall mix as hoped for. This worked, but I'm not totally sure that it was worth the hassle.
Michael Brook
Eli's methodology is roughly the same way I set it up for Steve Porcaro on his present TV series and he has been quite happy with it also.

That said, I hate giving a picture mixer the level of control over my music that giving him/her six stems provides. He/she can f$*k it up royally. I try to contain it to 3 at most.
Hey Jay,
I too really dislike giving up my mix as stems but I understand why the mixer wants the flexibility. So far, the mixer has reported to me that he hasn't used the stems. So despite the fact that I spend a lot of time creating these darns splits, it's nice to hear that they're mostly not necessary........... ugh.
This is fascinating to see how others work. But Judith's question was about effects on the stereo master, and how to apply those - or not - to the stems.
I think we all agree that delivering stems, and the unpredictability of the mix engineer's use of them is a PITA, and is always going to lead to some sort of compromise. My 2¢ worth - Make the stereo master as good as you want it to be, then just leave the shared effects compression off the stems. Let the mix engineer deal with the sweetening in exchange for you giving him added flexibility.

While you can duplicate reverbs and keep them within the stem busses, Judith's correct concern is that compression is going to react differently when your stems are mixed off separately. I like MBESQ's suggestion of sidechaining, but as he says, it's fussy. Would setting the sidechain to the stereo master bus have the same effect? Alternatively, take the pain and bounce off your stems one at a time.

Roger Jackson
Film Music
Oxford UK


Make the stereo master as good as you want it to be, then just leave the shared effects compression off the stems. Let the mix engineer deal with the sweetening in exchange for you giving him added flexibility.
This is exactly what you DON'T want to do. You want to make as certain as possible that if they have to go in to the stems, that there is as little chance as possible of it not sounding the same as the full mix! You want to limit there ability to change it sonically, not enhance it. If they need to go into the stems is likely not for reasons having to do with the mix of the music, but probably for editing reasons - maybe the director wants to shorten a cue or alter some of it's component elements somehow.

Alternatively, take the pain and bounce off your stems one at a time.
This is just not possible in Judith's case. Imagine the situation - you have to deliver an episode each week and each episode may contain between a dozen to eighteen separate cues. It's already a lot work re-recording each cue into a Pro Tools session. Imagine multiplying that by the number of stems for each cue. She is delivering six stems plus a full mix. She would then need to do seven passes for each cue instead of one.
I respectfully side with ELI.

First, not only is it not desirable to leave the final processing to the post engineer / Sound Stage, but it isn't professional. They often have a full plate as it is and it is our responsibility to provide them stems that in the best case can be put up to zero and never touched. You'll make good friends with the supervisor and the Re-Recording engineer if they don't have to think about the composers music mix.

This means coming up with a workflow that allows you to hear the stems as they print, allows you to check the summed level, and sounds as close to if not the same as the Stereo mix.

I set up all the stem busses with the exact same processing as I used on the master. It does sound a little different, and maybe it needs a small mix adjustment, but it is an acceptable difference to me. it's also the last time I'll get a chance to make minor tweaks before delivery, and there are always minor tweaks.

One thing I could do in Pro Tools HD that I can't do in Logic. I would sometimes determine my stems at the start of the mix and create all my busses to begin with. Everything was bussed to the correct destination as I went through the mix, that way the final stem print always sounded exactly the same as the mix. With Logic we are writing as we are mixing to save time so I'm not able to be this creative with the bussing for time sake, and for latency issues when recording.... I'm looking into the apogee Symphony to maybe help address some of that.

The Stereo Master processing is usually, in my experience, a quick way to clean up the mix for the client approval. The actual mixing of the score has been happening with nothing or very minimal processing on the master. That's planning ahead so I'm not road blocked into something that becomes a nightmare. : ) Keep in mind too that this is commercial, TV, or Film score mixing. The Stereo Master processing is not going to be the same kind of thing it would be for a pop or rock mix. It physically can't be.
The stem routing template that Eli set up for me has worked perfectly! He created a monitoring switch in the environment so that I can monitor my mix through either my Logic or ProTools soundcards. My 7 stems are recorded directly into ProTools because I have to deliver PT sessions, thus saving me the time of bouncing and then having to manually place them into the PT timeline.

When I'd begun gearing up for this project, I'd feared having to get a TDM system but am pleased that the 003's analog and ADAT inputs have been sufficient and that switching between the 2 apps has been seamless.

It took me a while to get used to mixing with the stems but it's worked out fine and I'm sure that what the film mixer is hearing, be it my stereo mix or the stems, is pretty much 'my mix'. Now what he does with the stems is another story:)
You are all right of course, and my slightly provocative suggestions were only positing a solution for where for whatever reason that ideal isn't possible - lack of time (and/or budget) being the usual bugbear in our business.

Roger Jackson
Film Music
Oxford UK
They make you deliver a PT session? That is a PIA... At least on the films I've done there is a Music Supervisor who's handled that. I just give him audio files. Same for the commercial work.
That whole process can sometimes be enough work for another person to handle, I'd hate to have to deal with that on top of everything else... Too many things to check and verify and keep running for one person as it is...
but we alway do what we need to don't we : )
Most of the time I deliver PT sessions and now that I can record directly into PT easily, it's no big deal but you're right, it's another process and one that has to be checked carefully. I'm trying to delegate some of the work and am certainly eyeing the whole mixing / bouncing stage as one that might be eligible.

That said, I find that when I mix I also 'tweak' the music. I add tracks and a bit of musical polish because i usually mix after I've not listened to the tracks for a few days as I wait for the director to approve the rough stereo mix. Being away from the music is great because I then go back to it with fresh ears and ideas and I enjoy the 'tweaking' as I mix stage.

Unfortunately this process becomes quite time consuming and a rather personal task, not one that I can delegate.. at least not with my current work flow of 'tweaking and mixing'. As my schedule tightens and I get busier, I'll have to revise how I work.... a happy dilemma:)
They make you deliver a PT session? That is a PIA... At least on the films I've done there is a Music Supervisor who's handled that. I just give him audio files. Same for the commercial work.
That whole process can sometimes be enough work for another person to handle, I'd hate to have to deal with that on top of everything else... Too many things to check and verify and keep running for one person as it is...
but we alway do what we need to don't we : )

Assuming you are bouncing .wav files, which are actually timestamped broadcast wave files in Logic, you can bring them into a PT session and just use their spot mode to place it at the appropriate SMPTE position.
I do work with .wav files but prefer to record the stems directly into PT which eliminates the added step of placing the bounced files into the session. Even with the spot mode, it's still another persnickety thing to do. But most importantly for me, I like monitoring my mix through the PT sound card. I used to have to remix certain cues once I'd imported them into PT and listened to them because they simply didn't sound the same as when I'd mixed them in Logic.