Oddly, this is how we mixed in the dark ages of 2 ADATs. We had 8 tracks of drums on machine 1, down to stereo on machine 2. Then, say, three more stereo stems of keyboards, guitars and extras (or b-vox). Then the stems got mixed to a master with vocals and bass in machine one again. It could wind up being the equivalent of a 40-channel mix on two machines. Any remix in a stem meant busting out that original tape and rebalancing it.

 

Routing tracks to buses is standard practice.
But exporting these groups as stems has more disadvantages than advantages.
The only real advantage is less workload in the project.

Therefore, I only consider stems useful in these cases:

• To bundle many small audio snippets like SFX or adlibs
• To archive projects as stems to save space
• For stem mastering or delivery as stems

So, for your second question:
Whether or not you should put limiters on buses depends on the material (and what you want to achieve).
I wouldn't recommend adding limiters just for the sake of it.

 

Wow, I've quoted in entirety as this is awesome information, anyone using Logic should watch that video. What a wonderful thing to read/see on New Year Day. Thank you clone.

And before that on two tracks, luckily I came across this quickly. Well worth a read. https://en.wikipedia.org/wiki/Recording_practices_of_the_Beatles

Lot's of opinion and sentiment on bouncing to stems. My 2C is that there isn't a disadvantage if you keep copies of the original audio and mix. If you're using outboard gear you have to bounce to stems.

I'm really interested in trying to learn and understand best practices for outboard gain-staging as I lose confidence when bouncing a sumbus through outboard EQ or Compression... It can be easy to overthink the outboard and end up going round in circles doing realtime bounces and lots of A/B-ing. Once the bounce is made, I suppose it's best to keep it the fader at unity gain and refrain from editing.

 

^^^ Word to the wise.

 

the worst thing is that you can't "unbus" an individual track after it's been bussed. so if you need to make individual preparations before sending it to the bus.

although, in color grading people often send color transform node at the very end of their chain firsthand, then simply making their grading earlier in the chain... so mayube that's not an issue. what do you think?

what I had in mind is that:

1) you have vox bus consisting of "adlibs" "backs" "doubles" and "mains"
2) all of them require some basic common processing like de-essing, leveling and maybe occasional coloration
3) it's handy to do this process in the bus, since it's identical

but then, if you try to apply invididual (!!!) reverb and delay to each of these tracks, you either deal with unprocessed sound, or you... dump the bus and simply copy all your plugin instances to each track? I guess when you have a good PC this is just a matter of two clicks then.

so is there any daw that deals with this problem?

 

The thing with the Beatles that made it really scary was they didn't have any way of synchronizing machines until somewhere around '67 or so and even then it was pretty wonky so when they bounced they were truly committed! At least with ADAT everything was synchronized and it was easy to go back and redo.

 

More files, more storage is not a disadvantage? Okay, storage is cheap these days, so: no problem. But what if you realize that the snare in the chorus is a little too quiet or the kick in region X needs 1dB more at 8kHz? Then mute the stem, revive the entire single-track concert, rework the snare and kick drum, render, kick the old stem into the trash and hope it was the last time? Seriously, no disadvantage? It's not as if we value speed in music production. Who needs creative flow when you can do the stem dance instead?

Do I have to? Not really.
Well, I do that when I'm done, not when I start.

 

See that's the thing... we did it that way because we were forced to but if you don't have to I can't imagine why you would particularly

 

Exactly! It's not as if we still rely on tape machines or xeon foster processors for mixing these days.

 

Happy new year mate
I think the scariest of them all, was the early stages of discography and recording where the artists had one take in the studio which was "printed" on record while they were playing it. No tape no nothing. And only the big names had a second shot. I strongly believe, if we were transferred on a time machine to that era, about 99% of the current pro musicians would fail to cope hahaha.

Hey Happy new year.
I don't see how this is a daw problem. It is a mindset one. Meaning the way you approach mixing vocals. If i have all you said, like main vox with doubles, backing vox and adlibs, to me this is easily 3 discrete sub-groups. With the exception of main vox doubles that can share similar treatment to the main vox (and has happened more than once a considered double to become main), the rest shouldn't share identical treatment if you want your vocals to stand out and be interesting ear candy. After all, vocals are the most important part of a song. All these sub-groups should be then routed to another buss for easy volume comparisons to instrumental mix and/or other instr. groups etc.
For at least a decade i used to mix as you mentioned (one vocals buss) and i would start a song's mix with drums, bass, etc etc. Traditional way. But then somewhere around '99-2K, i realized once you have vocals in a song, everything else's main role is to complement the vox the best way possible. Of course this is common knowledge but how do you make the vocals play the starring role as they should...
Starting the mix from the vocals and sculpting every element around the vox, that's how. At least to me. I will do an old school rough mix starting with drums, bass, then gtrs etc. But when i will get down to the nitty gritty, i will start with the vox, then melodic instruments and generally work my way backwards with the drums & perc last hehe. I found out that for all tracks that mean to "sell" the vocal first and foremost, this approach works better. And no matter if you do it this way or the other or anyway, the discrete vocal busses will help emphasize a unique blend of characters from each type of vocal group. But in my example of starting the mix with the vocals, the different vox busses get +1 point hehe. In the era of unlimited channels and groups/subgroups fx busses and whatnot, i think it would be a shame if we don't put all this power to good and creative use.
All my best

 

Related:

 

He ain't wrong

 

Individual track for correction and tone shaping - so every processing on that track is for that track only, to get what I have in mind, as example, this kick should sound like the kick I have in mind for the track, so I tweak it for that. Some sent to their own parallel channel for FX and processing and will return to the bus of their group. Some needed stereo FX, so the have their own FX channels for that but returning to different group sometimes, depends on the genre. Individual tracks then sent to a (group) bus.

Group Bus processing is for the "mixing" job, how tracks get along with each other, all the processing is more about how the tracks connected to it blending in with others. Send to "STEM" bus. 8 STEMS with 16 total channel is standard practice. I have weird clients that even want to sum it on their own, so it's a good number.

STEM Bus is for mostly for printing and combining them for few type of mixes - Minus One, Acapella, TV etc. STEM bus also for sidechaining the single masterbus compressor so even I'm printing 1 STEM, the MasterBus compressor is reacting correctly and listening to the whole track. That way the sums of single-printed STEM, all other type of mixes sounding as close as possible with the full main mix, and pumping correctly. In a case client want a STEM Mastering, Remix or whatever, they have that option as well because I give them all the possibility for the next step of the production.They get few mixes + stems + pseudo mastered approved mix reference for mastering stage reference.

STEM returning to MasterBus to Print Bus to Listen Bus (studio one). I print the "PRINT BUS". Print Bus has aggressive additive EQ that was like that since beginning (template), I mix into it. "Bro, the mix is too bright", I just modify it on the additive EQ on the PRINT BUS, and print again. Overall dynamic wouldn't be affected much (ME will correct it anyway). 10 revision per mix years ago taught me valuable lesson how to deal with human being.

BUSing is the way to go, if used correctly.

 

Last edited: Jan 2, 2025 at 8:42 AM

IMO his best argument is archiving. The others refer to the mindset, which is not dependent on the used technology.

 

thanks buddy, didn't expcet for such kind answer.

 

This is just what I do: tonal shaping first, mixing second. The problem is: tweaking in solo sometimes gives me a very different sound from what I would get in the context of the other instruments — especially drums and electric bass. So it's usualy a bit of freeze–unfreeze back and forth until I get it right. It's like that 'back to the arrangement' motto. Some things never change, I guess. Composing, arranging, producing and mixing are cleary different stages alright, but they are way more interlocked than textbooks would have you believe.

 

Yeah, but I think that was pretty much his point really. Going analog just helps facilitate that, like being on a certain kind of diet may be easier to do if you don't keep all the things that you don't want to eat around. Like, If you fill your freezer with ice cream but don't want to let yourself eat the ice cream, if you have willpower you can just not eat the ice cream but it's easier if you just don't even have the ice cream.
That said, I'm on your side, I don't want to go through all the expense and hassle of maintaining analog equipment just to emulate the analog mindset and workflow. I've always kind of been analog in a digital world, like, I don't record 5,000 tracks of something and comp them, I just record one track and punch and then call it a day; that sort of thing.