- This topic has 28 replies, 9 voices, and was last updated 3 years, 6 months ago by .
-
Posts
-
Good afternoon,
I’m starting to make stems (yes, I should have been doing that all along!:-) in anticipation of plugins no longer working due to future DAW updates.
My question is: do you bounce each stem by simply soloing the appropriate tracks in your stereo mix (allowing for fader automations to keep the mix elements) OR do you disengage automation and bounce each stem at unity? Which is the best practice?
Thank you so much.
PaoloSounds like you’re mixing terminology – printed tracks w/ FX vs. Stems. Stems are a group of like tracks such as Orchestra stem, Rhythm Tracks stem, Synths stem, etc.. To me it sounds like you’re talking about printing tracks with FX which are technically not stems.
Once you figure out which it is you’re after, the next question would be “why”? And at this point I can’t discern that. If it’s for future remix, I’d print everything dry, and possibly without automation. If it’s so you can come back and tweak an instrument slightly, then print auto and FX. If it’s for REAL stems, then absolutely wet and affected. When you define your intent, you can make a decision as to whether or not to print FX and/or automation.
Personally, I don’t do stems unless asked for them. I do alt mixes instead. When I do stems (real stems, not individual tracks) ALL fx, verb, EQ, etc are printed so that the combined total of the stems equals exactly the 2 mix – making sure that no FX or verb cross polinate between stems.
In my world, Stems are not for recall (that’s the DAW’s job), they are for re-mixiing engineers who want to rebalance the mix, or remove elements, etc..
No, stems are not dry without automation. They are singled out categories of the mix, and as such need have their fx, eq, etc. printed with them. They are grouped elements of the mix and should sound like they would if you solo’d “percussion” or “guitars” or “keys” during a mix. They would have their FX.
If you’re remixing and just doing a few mix tweaks in the future, go ahead and print FX and automation.
If you’re starting the mix over from scratch, changing the arrangement, adding or deleting parts, re-writing, etc.., at that point I’d print without automation or FX.
That is the only time AFAIC. Otherwise I’d be printing FX.
OK. I finally think I see where you’re coming from.
When I have soft synths in “midi-land” I ALWAYS print them as audio tracks before I start mixing. The only time I use midi automation is if I’m affecting filters or other synth / sampler parameters, or for the very rare volume move. I always automate levels, etc. and mix in PTHD. Never in midi. That may answer your question.
IME – always print midi tracks as individual audio tracks dry or dry-ish (if a sound has verb on it and I like it, I’ll print it with the verb) for mixing, and for later remixes. For stems, FX and automation always get printed.
Sure, you’re welcome. thumbsup
Printing audio tracks and automating those instead of midi renders a lot of future-proof sessions and the ability to transfer sessions to someone else to work on them seamlessly. Plus working with audio is so much smoother and more enjoyable than automating midi.
When I have soft synths in “midi-land” I ALWAYS print them as audio tracks before I start mixing. The only time I use midi automation is if I’m affecting filters or other synth / sampler parameters, or for the very rare volume move. I always automate levels, etc. and mix in PTHD. Never in midi. That may answer your question.
Thanks for that tip LAwriter. I am guilty of keeping my tracks in MIDI for mixing because I am always finding little things to tweak in the MIDI during mixdown. In other words, I tend to mix before my track is finished. I need to work on my self discipline and that little tidbit helped.
Regarding stems, I was asked for stems for a specific placement one time. I bounced out what they wanted from the mix. Then had to decide if I was going to deliver as is or run the stems through my mastering session. Is there a standard practice?
I “assumed” they should ideally be able to combine the stems equally to recreate the original mix. If I run each stem through my mastering session as is, the combined stems would clip. I think I ran the stems through the mastering session at about -3dB to -4dB but I cant remember for sure. What would you have done?
Funny end to the story, the stems were used on “The Marvelous Mrs Maisel” for a whopping 10 seconds, very lo-fi, coming out of a black and white TV viewed from outside the window, under dialog. I would think a simple band pass filter would have been a lot easier and faster. But I guess that’s why they earned those Emmys, ha.
Alan – wthout getting too detailed, IMO, this is the best way to mix stems if you KNOW you are going to mix stems. IMO of course…
For the random sake of argument, let’s say the stems we need are : 1.) Orchestra (input bus 1/2), 2.) Solo instruments (input bus 3/4), 3.) Rhythm section (input bus 5/6), 4.) Synths (input bus 7/8), 5.) Vocals (input bus 9/10). Then a stereo two mix – (input bus 11/12). Adjust whatever stems you decide on to suit your particular piece of music’s requirements and/or client requests. Put all tracks into “input” and “solo safe” them.
So….you will end up printing 6 Stereo tracks when you print your mix. 5 Stems tracks that ultimately feed the 6th track which is your 2 mix bus. Send the output of all Orchestra tracks and associated reverb and FX to internal bus 1/2, the Solo Instruments outputs and their reverb and FX to internal bus 3/4, and etc. This should have all your individual track outputs feeding their respective stem inputs properly. Then, feed ALL the stem track outputs to bus 11/12 which will end up being your 2 Bus stereo mix. Go ahead and put whatever 2 bus processing you use (I use analog gear on hardware inserts) on your mix bus. Monitor your stems tracks and 2 bus mix track on input while you mix, and your stems and your 2 bus will be perfect level wise. Or at least they will be where you want them. 🙂
This type of mixing can get quite elaborate, but saves a HUGE amount of time, and creates stems that are properly designed and mixed.
And to answer your question – yes, the stem outputs combined should recreate your mix bus – sans processing on the mix bus.
I had a similar issue regarding the dilemna of keeping things MIDI or bouncing them to audio. before i would mix and master with MIDI tracks unbounced, but because of resource constraints i had to resort to having 2 session files per song (MIDI unbounced and bounced). I would have my mastering plugins on the bounced version though (again due to resource constraints)…then i came across a situation where a library asked me to change a string sound cause they didnt like the one i used, so i had to go back to the unbounced session, change the vsti, made some level and articulation tweaks, bounce to audio, then import that to the bounced session and remixed/remastered the track. a lot of work…if it was just timing or level issue i could have edited the bounced session, but if its an instrument or arrangement change i have no choice to go back to the MIDI track…
i havent had an issue with my stems clipping through the mastering chain though…i’d assume the sum of all the stems is greater than the stems separately…
I always render (Cubase terminology) my midi tracks and then have a Macro key command to hide and disable the midi track. This saves resources and also places the hidden midi track just above the audio version if I need to go back. Regarding stems I just solo the relevant tracks and bounce out. There can be certain situations where the the send fx (reverb etc) need to be separate which requires a different setup from the outset. Jason Graves has a YT channel where he describes this in detail.
This is how I do it:
Stems for a client:
– everything printed running through the master FX, but with a little less compression.
– I then (if there’s time) load these stems into a fresh new session and reduce the volume overall until there’s no clipping. Re-bounce / exportStems (stripes) for a mix:
– no master FX, except for any gain automation. Still print all the track automation, or else the mixer will have a rough time.
– Depending on the mixer, and their process, I will / won’t print reverb into each track.does anyone know if the editors mix stems with compression or just basic level adjustments?
- You must be logged in to reply to this topic.