Bouncing stems – same or different volume from stereo mix?

[LAwriter]

think about it…. Re-Mix engineers usually like less compression not more. It makes their job easier and sounds better. So for most feature applications, I’d say no compression is added, but we’re definitely in a transition phase as those coming into the power positions have never heard music that isn’t smashed to ****

[LAwriter]

If you want your music smashed, then mix through the stems mixes into your stereo mix as I described above. Put your processing on the stems instead of the 2 mix. You’ll have more flexibility, and the sum total of your stems WILL equal a smashed 2 bus mastered mix with no processing required on the 2 bus.

for those requiring heavy compression / limiting to their music, this is the best of both worlds. If you’re using hardware to accomplish your sound though, this is pretty difficult to do without a LOT of outboard. There can be no cross pollenization of verbs, delays, etc., and you need enough hardware for multiple stems, so I’d say that in all but the most simple of stem mixes, this is unrealistic. (Hardware processing that is….)

[Art Munson]

Check this thread also https://musiclibraryreport.com/forums/topic/new-mike-patti-video/

[Tim]

In the production music world, stems are meant to give greater control to the editor, etc. You should be able to compile your stems and they should play back just like your original mix. So, FX etc should be printed. Sending stems dry or with no automation defeats the purpose of this. What of your snare is compressed with a short reverb but your kick is dry? They are grouped together so
If someone wants to start with a fresh mix and add compression and reverb, they’d be affecting/effecting (whatever) the entire kit, right? Not the goal. If you want to print individual tracks
without FX/auto then you are providing a multi track, not stems. You can do that too for your own backup or if someone truly wants to do a fresh mix. One thing we do is take stems with FX etc but unmastered so in case we want to get a cleaner, punchier, etc remastering that can be done. The end goal is to make things as easy as possible for your client.

[Tim]

Apologies for misspellings above. Typing on my phone.

Re: smashing/bricking Your mix I would highly advise not to. Get it loud but keep it dynamic. Most editors can see the wave form and if it’s bricked may actually see it as a dated track because smashing/bricking was the thing to do years ago. Give them every reason to use your music. Cheers!

[Logika]

If the music is for a library which later may or may not need stems, then how important is it that the stems combined exactly equal what was given to them as the stereo master?

Point being that generally the full stereo mix may have been “pseudo-mastered” (a.k.a home mastering with plugins on the 2-bus). If so, then the stems will not equal the full mix.

So do you master the full mix, with the stems and/or alts not equalling the mix? Or, do you instead put plugins (i.e. saturation, tape, compressor, whatever) on each stem bus—with nothing on the 2 bus—so that the “mastering” is done via the stem bus processing? (that way the stems would equal the mix). I don’t mean smashed mixes…I’m talking about maintaining as much of a dynamic mix as possible with appropriate and expected loudness.

Basically I’m asking what is the best way to go.

And on this topic, do you incorporate any tape / saturation plugins, without issue? I’ve heard that it can cause problems in broadcast use (causing harmonics that may alias), but don’t have a definitive answer on that.

[PAMMusic]

If the music is for a library which later may or may not need stems, then how important is it that the stems combined exactly equal what was given to them as the stereo master?

Point being that generally the full stereo mix may have been “pseudo-mastered” (a.k.a home mastering with plugins on the 2-bus). If so, then the stems will not equal the full mix.

So do you master the full mix, with the stems and/or alts not equalling the mix? Or, do you instead put plugins (i.e. saturation, tape, compressor, whatever) on each stem bus—with nothing on the 2 bus—so that the “mastering” is done via the stem bus processing?

I am currently creating stems using this “pseudo-mastering” method for a library and would be really interested on hearing what the best practice is on bouncing them for a library.

[Alan]

I’m currently taking the Berklee Online course “Audio/Music Production for Visual Media” and stems was a recent topic. Here is the take from a top industry film score mixer (not final dubbing mixer). All stems MUST combine at fader unity to be equal to the final mix. The point being the dubbing mixer has the option to quickly and easily edit parts of the mix that may be competing with dialog, rather than heavy handedly turning down the entire mix. Every stem must also be autonomous, with no effects bleed etc.
This of course means there can be no two bus processing on the final mix, it is pre master.
Based on this “new information,” I consider stems and alt mixes as separate beasts. Moving forward I will continue to master my alternate mixes the same as the main track. If stems are requested, Stem01 will be the full, unmastered mix and the rest of the stems will sum at unity to exactly match Stem01.

[PAMMusic]

So in that case when submitting say an acoustic pop track to a library requesting cutdowns and stems, good practice might be to submit:

– Alts with “pseudo-mastering” stereo bus processing – Full mix, 60secs, 30secs, light mix, no drums etc
– STEMS without stereo bus processing – full mix, lead guitar, rhythm guitar, bass, drums etc.

Am i interpreting this correctly?

[Alan]

So in that case when submitting say an acoustic pop track to a library requesting cutdowns and stems, good practice might be to submit:

I interpret “cutdowns and stems” as cutdowns (:30, :15, Sting) and alternate mixes (No Lead, Bed, Drums and Bass). If it were me, I would simply ask the library to tell me exactly what they want.
“Dear A&R,
The definition of a stem in open to interpretation across this industry. Would you like a pre-master mix plus stems that combine to equal it, or would you prefer mastered stems treated as alternate mixes?”

PAMMusic, I think there is a very good chance the A&R will not know the answer to that question without asking around the office or they may even need to call a few clients to find the answer.

[Alan]

Here is a direct quote regarding stems from a film industry professional:

Beware of the distinction between stems in this context—premixes made at the dub by the dubbing mixer—and music stems. You can think of music stems as premixes made by the score mixer rather than on the dubbing stage. Sometimes the word “stems” overlaps and there is momentary confusion about the meaning.

All the more reason to ask the library what they want.

[abellboy]

I think it means different things to different people as I have had stems mean slightly different things depending on the context.

I just delivered a soundtrack to a director and there was a “theremin stem” (don’t ask)….
This allowed the mixer to used it or not depending on decisions down the road.
In this case I gave him a “full mix” (Theremin included) and a “full mix minus Theremin” with everything except Theremin, and then a “Theremin stem”. When the sound person puts the “full mix minus Theremin” and “Theremin stem” faders to 0, it sounds EXACTLY like the “full mix”.

I have worked with libraries that want “stems”. The stems were meant to recreate the track if all faders were set to 0 much like the above example.

A few other libraries have asked for “alts” and are usually versions that stand alone and have no intention of being used to reconstruct the original mix. IE. “Orchestra Only Mix”, “No Choir Mix”, “No Drums Mix”.

As for mastering. When I’m asked for alts I just apply the same mastering chain and adjust to taste. I’m not sure this is best practice, but some of those cues have been placed, and none got rejected for sounding measurably different.

[PAMMusic]

Thanks for the info.
I reached out to the library about it and they were fairly laid back about stems not sounding exactly like the full mix, and said they weren’t that important. I have also been reading through the agreements for some libraries I am working with soon, and they explicitly ask that no master bus processing be used.

i guess the main thing to take away from this is that one of the first things to do when writing music for a library that requests stems is to make sure you ask exactly what they want from their stems before you even open your DAW to avoid any messy processing later.
My next question would be – how do you get the Main mix to sound glued together without the use of master bus processing such as compression?

[Advice]

As been said, it totally depends on what the library wants. Many want “alt mixes”, not stems. The terms get mixed up all the time. With alt mixes, each one is a stand alone for the music sup to pick from. Most of the time, your cable TV reality TV show oriented libraries want these. It’s not just volume, mastering, etc. it’s also whether to leave space in the front if those instruments don’t even come in until, let’s say, 6 seconds in. Stems are truly the component tracks that would allow the sup/editor to remix. These should line up and combine as well as possible. *Generally*, these are for higher end libraries/placements, though it can vary.

Bottom line: ALWAYS ask the library for clarification. Never assume anything.

[Author]

Posts