Stems and producing

SuprSi

A friend of mine asked me to send him the stems for my last track. Not knowing what he meant, I did a bit of digging and found out that they are the individually recorded audio tracks from your sequencer.

The thing is, I didn't record any stems. All of my mastering is done on the live mix, so when I record my final mix it's one big audio file based on the sequencer recording. In other words, I didn't record each instrument separately into an audio file.

My question is, is it normal practise to record each instrument or track individually and use them to create the final mix?

holystungun9

I'm only getting into production so someone with a better answer should come into this soon but here is my take on this.
When some one is going to remix your track or you are going to send it to a professional to mix and master it they will want the individual tracks. Don't know what software you are using but you can surely bounce the track stems out individually.
The person will probably want all EQ, compression, reverb, etc. turned off as they will need to apply these in their own way to fit each of the tracks in their mix so it sounds well.
The only general exception is if there are specific effects which actually really characterise a track so they don't have to figure out how you did it.
A simple example is if your track has a vocal with a vocoder or something and this is a signature part. They will probably want that as it is.
Anyone else?

mordeith

I'd echo everything said above. I'm more curious as to how you arrange your tracks if you are unable to output stems. Surely you have seperate tracks for the various instruments in your DAW? You would need to in order to be able to tweak your levels and apply fx to individual tracks. I'm using a fairly outdated version of FL Studio so I would have to export each individual track spererately, but I imagine more up to date daws have the ability to export each track as a seperate audio file in one operation.

SuprSi

Thanks, that's a great explanation. I have my tracks separated but only in midi format, not wav. I'm using Cubase 5 and have just done a test to export the tracks individually and it worked! I never knew I could do that! I have to say I'm having great fun getting back into this stuff

Thanks both!

mordeith

SuprSi wrote: »

I have my tracks separated but only in midi format, not wav.

Am I correct in assuming that you're using the MIDI channels in Cubase to control external synths?

SuprSi

mordeith wrote: »

Am I correct in assuming that you're using the MIDI channels in Cubase to control external synths?

Almost - midi channels to control soft synths. I got rid of my hardware 10 years ago unfortunately.

mordeith

SuprSi wrote: »

Almost - midi channels to control soft synths. I got rid of my hardware 10 years ago unfortunately.

Ah right. I've never had any external hardware except for two midi keyboards. I never see the step sequencer as being just midi notes. Midi always seemed like something whereby you had to some some kind of physical cable connection. My brain isn't clued in to the fact I use midi every time I create something

SuprSi

Haha yeah I get ya. I miss having a keyboard and such easy access to sounds, I find it's not as fluid writing stuff solely on software but you have so many more options, tools and configurations so it balances out.

TroutMask

Subgroups

Stab*City

these days alot of the fx processing is done during the writing stage in the mix so its kinda hard to remove all fx from single channels. The track wouldn't sound anything like what you made if you wrote a track and then removed all fx from each channel. I always bounce my tracks down to stems/audio tracks. Easier to manage the end product and then when all is said and done i just remove fx from the master this seemed to suit the guy who mastered my tunes recently.

Rob Humanoid

I'd be the same... I'd usually have two files. One for working on, and the other that would be for finishing up (all bounced down to stems).

Audio is a lot easier to deal with plus it's cpu efficient.

SuprSi

Why is audio easier to deal with? If you notice an issue with any part of the arrangement you have to re-record it so surely working on the live instrument (if possible) is better? I see what you're saying about cpu efficiency, which was a major problem on my last PC but not so much on my current laptop.

Rob Humanoid

Well, I've a fairly new mac with 8gigs of ram. And as soon as you start stacking up the soft synths, with a load of FX, things start to go down to a crawl. It really depends on your plugins too - Some 3rd party plugins use way more than a DAW's native ones.

Complex sounds in synths like Diva or Spire with eat into you memory. And the more complex your music is the more synths you use, etc. You can whittle this down by choosing a more cpu efficient set up (ussually the DAW's own). But then you might be loosing out in terms of quality. I've yet to see a soft synth that can match Diva's quality in the analog emulation realm.

Also you might want a consistent sound through out, it's a lot harder to predict things like effect tails, delay or reverbs when the synths are left free running.

Plus, there is also the fact that you can easily split up audio, and reuse snippets you like, even little reversed phases or whatever. You can't do that in the same way using midi.

For example, you might like a bit of a certain synth bass, but as the tune progresses it looses its character, as the LFO moves out of phase, or the key tracking is slightly off (it can be hard to get it spot on sometimes). That's why samplers are key. Take the bit you like, bang it in the sampler and you got a patch that works on every note.

Certain plugins, like a sampler will need audio. A decent sampler will have a shed load more parameters than any soft synth .

Mixing is easier too - you might have 15 audio stems to mess around with all nicely bounced down. Now, rearranging your track is a lot easier then having to worry about redoing the automation or altering any of the FX.

Using S1 affords me the luxury of bouncing down, but then reverting back to an instrument channel if I decide that I need to chance something. But obviously that's an S1 advantage...

mordeith

Rob Humanoid wrote: »

. . .
Also you might want a consistent sound through out, it's a lot harder to predict things like effect tails, delay or reverbs when the synths are left free running.

Plus, there is also the fact that you can easily split up audio, and reuse snippets you like, even little reversed phases or whatever. You can't do that in the same way using midi.

For example, you might like a bit of a certain synth bass, but as the tune progresses it looses its character, as the LFO moves out of phase, or the key tracking is slightly off (it can be hard to get it spot on sometimes). That's why samplers are key. Take the bit you like, bang it in the sampler and you got a patch that works on every note.

Certain plugins, like a sampler will need audio. A decent sampler will have a shed load more parameters than any soft synth . . .

Some great tips here. Every now and then I run into issues with a synth line misbehaving at some point. Sampling the relevant section is something I'll have to bear in mind the next time I'm going mad trying to sort out some annoying niggle.

Stab*City

SuprSi wrote: »

Why is audio easier to deal with? If you notice an issue with any part of the arrangement you have to re-record it so surely working on the live instrument (if possible) is better? I see what you're saying about cpu efficiency, which was a major problem on my last PC but not so much on my current laptop.

Audio is way easier to deal with when it comes to the cpu/hdd load i have an i7 MBP with 8gb and trust me when you get up to the 30 channel mark all loaded with vst's, fx and the rest your machine will buckle no matter what it is! I dont do re-recording as i dont bounce to audio until im 99% happy. I dont delete the midi channel i group the midi and the audio most of the time and then just disable the midi channel and use the audio for arrangement That way i can go back and edit whenever i like..

SuprSi

Ah ok, 30 tracks with fx and everything else and I can see the reasoning behind it. I've only managed about 12 - 15 max at the moment and my laptop is cruising along (until I pull the power cable out!). I always wondered whether how I wrote a track was the same as everyone else and it's interesting and reassuring to find out that there are plenty of similarities.

I'm curious to know if you had a PC that could handle 30+ tracks with fx & vsts would you use audio at all? As in, is it how you work or do you do it because your laptop inevitably runs out of steam?

paulo6891

It all depends, I have around 70 tracks in my latest project containing mainly zebra, omnisphere and East/West Hollywood Strings, and not enough bussing - I still have plenty of room left for more without the project crashing! Most channels have effects. I've never tried mixing audio stems, but I may next time - if nothing else it will stop you from fiddling with your track non-stop!

Stab*City

Another handy feature in the program i use Ableton is the freeze track function. Using this i can freeze a track and ableton will temporarily convert the output to audio this saves big time on proc power too during the creation/idea phase. Then when im happy later down the road ill unfreeze convert to audio and then disable the midi/vst original channel. So you end up with two one editable but disabled and one just audio.

Also when grouping tracks you may have 20 groups and some may contain a couple or more channels.

Here is a little sum up of my process lately:

Make beats with Maschine group it and record each drum sound to a separate channel in the same group. Sometimes then i may remove Maschine from the group but not delete it and add compression/fx to the group to gel or add fx to separate drum parts to scuplt the sound.

I pretty much only use Ableton add on's or Native Instruments stuff for my vst's bar the odd time. I just really like the quality of sound with them and the presets that come with the NI stuff are great starting points. I use these pretty much for bass/atmosphere i also have a couple other synths like the minimoog that was free a while back. I think all my software is legit and licensed or free as far as i remember. this really helps in the crash department i think.

dont want to bore ya anymore!

SuprSi

Trust me, you're not boring me at all! I'm very curious to know your process - you've got stuff released on a label so you're obviously doing something right!! That's a very interesting feature of Ableton and I've just checked - Cubase does it too I must check it out this evening.

I mainly use Alchemy for synth stuff, though have started playing with NI Massive and find it really intuitive. But what I've been trying to focus on is percussion and fx, which is leading me to all sorts of mad sounds! It's great fun and it's a shame I wasn't as interested in it 10 years ago as I am now.

Stab*City

SuprSi wrote: »

Trust me, you're not boring me at all! I'm very curious to know your process - you've got stuff released on a label so you're obviously doing something right!! That's a very interesting feature of Ableton and I've just checked - Cubase does it too I must check it out this evening.

I mainly use Alchemy for synth stuff, though have started playing with NI Massive and find it really intuitive. But what I've been trying to focus on is percussion and fx, which is leading me to all sorts of mad sounds! It's great fun and it's a shame I wasn't as interested in it 10 years ago as I am now.

Thanks. Do you use the pc version of alchemy? I have the ipad version been meaning to fire it up and check it out. How is Cubase these days i tried it a few times but i found it too rigid..

SuprSi

Stab*City wrote: »

Thanks. Do you use the pc version of alchemy? I have the ipad version been meaning to fire it up and check it out. How is Cubase these days i tried it a few times but i found it too rigid..

Yep, it's the PC version. It's excellent, though I find myself just modifying existing patches rather than creating my own, which is creating in a round-about way I guess! It sounds amazing though and as you can use samples in it along with the standard synth sounds you can do some amazing stuff with it. Highly recommended.

As for Cubase, I have v5 so I'm a bit behind the times, but haven't bothered upgrading as tbh it does what I need. I have always used it in some shape or form, though a few years back tried Fruity Loops Studio as a recommendation from a friend who produces some really good progressive stuff, but I could never get into it. I think once you're familiar with a DAW it's best to stick with it, unless a different one has features you badly need.

It's maybe not the best DAW for dance music but again I think it's more a familiarity and ability thing - if you're good or dedicated enough you'll make anything work!