I'm building my ultimate template (I use Cubase, but the principle is the same in all daws I guess) and was struggling in routing my STEMS. Do you know how should I dot it properly, in order to have to the same mix result.

Let's take an example, I have 3 kicks which are routed to my Submix named "Kick Sub". On this Submix I may have some plugins (general eq, saturation etc) and some sends (ex. reverb, parallel compression). The Kick Submix is routed to my Mix Buss Master (the buss where all submix are routed to) and also to my Kick STEM buss. The Mix Buss Master (which has the goal to control the overall volume) is routed to my Mix Buss (where I normally have my general plugins effecting the whole mix), which is routed to the final buss PRINT (routed to Stereo out). In the Screenshot I have a buss named "router", it's the place where I've put my general plugins. I know I could only have the mix buss with my plugins inside and directly send it to Stereo Out, but I prefer to have to margin.

Now, how to route my Kick STEM. On the Mix Buss Master there is a volume automation (or could be no automation but some Db down to match -3db of headroom on my mix) and on the Router Buss there are my plugins affecting the whole mix. Should I route my STEM on the Router buss or directly on the Print buss and copy all my plugins of the router on all my SUB Mixes ?

 

I read the whole post three times. If this is normal routing for a typical project I must be behind the times. The Kick submix goes back to your kick stem buss too? As one stem?
Only you can really decide which effects are unique and which can be recreated with another plugin?

 

Last edited by a moderator: May 27, 2017

I had the opportunity to have access to the stems of a hit song, I analysed it and saw that the stems are important. First they summarise all the important parts of the track, it's easier to listen to them from the stems. It permits also to mute some sections for different types of exports of the music, like if I want to create a ringtone version I will mute some stems. And it's easier to have all my stems ready to export for the mastering - I don't need to go back to my sub mix tracks. This helps me to understand a lot and acclimate me to something I should do as a professional learning all by itself.

Yes.Here is a screenshot of my routing of the submix. I can route to multiple places. It goes to the mix buss master and also to the stems. I didn't do a few tests about exporting the stem and comparing it, I'll do that, I thought somebody was familiar with this and could tell me how to do it properly.

 

Why do you need stems for mastering? Aren't stems for others to remix or actually mix the song on a separate DAW?
Maybe there's a slight language barrier on my part.
I would never submix and make a stem with anything important, for instance, three part vocal harmonies would need level tweaking once instruments get added. Same goes for KICK? The attack of the sound might be too loud or soft after adding the bass and synths.

Muting a track or muting a submix....same effort for me at least on ABleton.

Good luck

 

In mastering the best way to master is Stem mastering

 

Interesting topic. I was always insecure about stems, and therefore tried to get the whole mix done without them whenever possible. I will read here carefully, if others contribute to the topic!

 

If you are not certain about your ability to export a 2 channel (assuming you are not mixing in surround X.1) for any reason exporting stems can save you a few reprints from your M.E. or if time is tight and there hard standards you might get a better result from a bad mix than if you sent a 2 channel mix.

There are a few approaches. The upper limit is somewhere between 8 - 12 with one of those being your FULL MIX. If you mixing a standard pop tune with vocals, the old way was the 4gang.

1. Full Mix (reference - this is a 2channel mix with dynamics intact - not an attempt to show M.E. how well you can make sausage)
2. All instruments and background vocals ( with background vocal effects bus and all instruments effects buses)
   
3. Instrumental only (no background vocals - no lead vocals - no vocal effects buses)
4. Lead Vocal only with Lead Vocal effects Bus

From the above you can get a Vox up / Vox Down - if the background vocals are too loud a blend of Instruments with BG and Instrumental only can be used. Instrumental only can be used for placements if you need them. The Instrumental only x Instrumental only can be used with Filters and processes if there is serve problem such as drums being unbalanced or some heavy imbalances in the mix.

The most discreet amount of stems you could ever send an serious M.E. and not be told to go FK yourself is probably 12. Even this is bordering on remix and you have to be really wishy-washy on you ability to put out a 2ch or not really attached to any particular sonic signature. If you get past 2 revision with these discrete stems and you have not been told to go away, then your M.E. is a saint. All these are 2 channel tracks with effects, this means if you plan on exporting this format you will need a Kick Bus, Snare Bus, Percussion Bus, each with their own effects and automation --> your stereo master. If you want interaction between any of those elements for effects they need to be side-chained through the buss not after it. The same goes for every other output bus, it needs to be as close to a finished product as possible and as well balanced as you can get it. no clipping, etc...

1. Kick
2. Snare
3. Rest of the Drum kit
4. Basses
5. Lead group 1 (could be electric guitars or Mid low synths)
   
6. Lead group 2 (different tonal texture to LG1 - Could be sweeps - syncs or a piano)
   
7. Lead group 3 (Interest leads - decorations - sweeps - synth FX)
   
8. Pads and base (Strings - pad brass - etc)
   
9. Lead Vocal
10. BG Vocals
11. Everything else
    
12. !important effects - this is not your entire FX bus, this is for if you have a signature effect or something that you really worked on real hard that you want preserved and you want it to stand out - put it here.

You should send a full mix, but remember to send the best quality you are capable of, your M.E. doesn't want to remix your stuff, if you are your own M.E. and just want stems for archival purposes then a 12 track is also a good format, though the full project will always be better and DVD optical media in proper storage will outlast your lifetime twice over in most cases (100 -200 yrs.) With compression you should be able to get your full resolution projects stored in 4.7 GB.

 

The reason for giving mastering stems is this:

The ME can make only very limited adjustments if u give them a 2ch mixdown. If there's more problems, there may be tradeoffs in fixing them (EQ/phase/imaging, etc). If you give stems, there are no tradeoffs (except for time), cuz they have more flexibility & control.

Good intent, bad implementation. Almost noone has 'proper storage'. Burnt DVDRs won't be fully readable after ~10 years. Use modisc, gold-coated media, data tape, or even analogue audio tape reels. With analogue media you're guaranteed to get at least something after n years. With digital, not so much. You should be able to store a project in 1-2GB. Any more, and it's overkill.

 

Thank you for the explanation. Instruments interacting was my first thought with this method. Seems like proper mixing technique is needed either way.

That track list is reminiscent of a 12 track tape session.

 

I have a question. Sometimes I have effects involved that use the master channel. For example sidechain compressing to give the kick a bit more room. Is this already considered a part of mastering? Or should I render the stems including master channel effects?

 

If it's an effect for mastering and won't mind having it removed, keep it master.

If it's an artistic effect you want to keep, move it to track/bus.

 

Stem mastering is the way to go.

 

In an Episode of Dave Pensado's "Into the Lair" he explains his signal routing. And he had a "Print" Track after the Master Bus to record Stems onto in realtime - like recording it to tape. His whole approach is based around the signal flow of an analog setup: Inputs > Board=DAW > Tape=Stem. But your Busses have to be setup correctly for it to work as expected.

I also think that printing stems this way the sum of all stems will sound different than a stereo mixdown of all tracks hitting the stereo bus and any signal processing that's going on there.

 

Pensado likes to record his master output opposed to rendering a 2 track. Is this a recent episode where he now sends his submix stems to mastering? Seems odd for a engineer known for his mixdown skills.

 

I am not sure I am understanding what you are trying to do here.

If you are ducking your main output with a side-chained compressor are you not also ducking the signal of your bass drum as well?

If you are ducking a submix bus with all the instruments on it except your bass drum and you vocal to get the french SC pumping effect, then you will have to have a gang of compressors on each output you want to attenuate with the signal sent to each one (or from the same key - depends on how you DAW handles this.)

Your ME does not want to make mix decisions for you, so they do not want to decide how much or how little pumping your mix should have, you need to print your stems so that SUM without any effects is what your desired mix sounds like. In an ideal world this also applies to a 2CH mix as well, though there are not rules in art. You have more freedom to simply and easily add effects on your master output when delivering a 2CH mix. Delivering stems can have the same sonic qualities, but far greater complexity in the routing matrix, since you have to duplicate your output effects across a gang of multiple outputs. The routing matrix of effects can get very complicated very quickly when output is large amount of discrete stems.

 

I've been doing this lately, but it's due to running some external hardware /summing on the 2 bus. A production buddy of mine says that he'll render and record and take the one that sounds best. He says it varies for each track.

 

Thanks for your answer. I'm no soundengineer and that led to a description that must have been confusing for you. What I'm really doing is, that I mix until I'm pleased with the "master channel". Then I use this channel as a bus, sidechain compress it slightly with the kick track AND run the dry kick track alongside, and those two tracks (the stereo sum of all and the kick track) go to the final master track (render track, print track, don't know how it's called). The result is not a noticable pumping effect, it just makes the kick stand more prominent in the mix. Here's an example:

I'm not sure how to change this so that the effect is active on individual stems?

 

Thank you for the clarification.

Isn't it strange? ones and zeros should be the same..........

 

I questioned it when he told me that, @dragonhill. He's on a mac using a pro tools setup. Maybe the recording is different because the audio is recorded coming through the converters.

 

Unless you have some nondeterminism like random tone generators or random modulation, a bug, or other issue, he's on crack

(I have some randoms in my track too, but I render once, unless I revise the whole song, unless I have a legit bug)