how do i mixed bitwig

[minortom]

Feels like a stupid question, but it's never too late and I can't get into Bitwig Nerd Camp without it.

What process do people use and recommend for streamlined and efficient mixing in Bitwig? Do you bounce tracks to audio? In place or to separate tracks (or a separate project)? With volume automation or without? Do you automate track levels or the Tool? Do you automate the Volume or the Gain knob, and what is the difference?

I looked and couldn't find a comprehensive answer on my own, a link to something I've missed will also be appreciated.

[antic604]

Do you bounce tracks to audio?

Not if I don't have to, i.e. my CPU can cope with the load. I like to tweak the stuff to the very last minute

In place or to separate tracks (or a separate project)?

To separate tracks and then I disable (Alt+A) and hide the source for potential future editing. Turns out I don't bounce-in-place stuff as much as I imagined I would. Ideally I'd love to end up with 2 Groups in a project, but I haven't done that fully yet:
1) MIDI - with all the MIDI originals in disabled tracks, which means it doesn't take any CPU and RAM
2) Audio - bounces for mixing

This way it's all neatly encapsulated in single file, that's still light.

With volume automation or without?

I try to avoid volume (meaning "track fader") automation, really. If for some reason I still want to do that e.g. for a fade in/out, then I do multiplicative modulation instead, so then I can move the fader anywhere and it still works.

Do you automate track levels or the Tool? Do you automate the Volume or the Gain knob, and what is the difference?

99% of the time with Tool. Which knob depends on need - Volume if I need to go to complete silence of -inf dB, Gain if I'm doing adjustements within its range of I believe +/-36dB

[UltraByte]

Feels like a stupid question, but it's never too late and I can't get into Bitwig Nerd Camp without it.

What process do people use and recommend for streamlined and efficient mixing in Bitwig? Do you bounce tracks to audio? In place or to separate tracks (or a separate project)? With volume automation or without? Do you automate track levels or the Tool? Do you automate the Volume or the Gain knob, and what is the difference?

I looked and couldn't find a comprehensive answer on my own, a link to something I've missed will also be appreciated.

antic already covered it , just want to add some points.

large projects i will mix in reaper, or sometimes harrison mixbus. however you can do it in bitwig too and its only sometimes a little bit more hassle.

before bouncing tracks i create a group for the track first and have different iterations in it. it seems clumsy first but can get very handy. bouncing can get especially useful when you work at higher samplerate. 96k etc (dont forget TDR Ultrasonic to take care of IMD in that case)

knowing you may need to bounce, its handy to group a track from the beginning and have level/pan automation on your "bounce group". that way you can bounce as many times without affecting it (like sends etc)

tool is awesome but for automated gain changes i use airwindows puregain (which features a great fade too) or bitshiftgain (but that one is not relevant for automation, just to get the level up before the end chain when you mix before at 0VU.

[minortom]

Thank you both for the answers!

Not if I don't have to, i.e. my CPU can cope with the load. I like to tweak the stuff to the very last minute

I guess that's a personal preference depending on whether or not you're in the "commit everything to audio" camp. I'm still undecided.

To separate tracks and then I disable (Alt+A) and hide the source for potential future editing. Turns out I don't bounce-in-place stuff as much as I imagined I would. Ideally I'd love to end up with 2 Groups in a project, but I haven't done that fully yet:
1) MIDI - with all the MIDI originals in disabled tracks, which means it doesn't take any CPU and RAM
2) Audio - bounces for mixing

This way it's all neatly encapsulated in single file, that's still light.

Why not bounce-in-place, though?

I can also see duplicating track groups for mixdown (and multiple copies for different mixes) without bouncing.

I try to avoid volume (meaning "track fader") automation, really. If for some reason I still want to do that e.g. for a fade in/out, then I do multiplicative modulation instead, so then I can move the fader anywhere and it still works. [...] 99% of the time with Tool.

What I meant was, when you bounce does the bounced track include printed volume fades or not. Pre or post fader, in other words. Multiplicative modulation is an interesting idea!

Does that mean that you do all your volume curves with the Tool and not touch mixer volume/track faders at all? Or still use them for broad balancing?

Which knob depends on need - Volume if I need to go to complete silence of -inf dB, Gain if I'm doing adjustements within its range of I believe +/-36dB

I'm guessing the purpose of Gain is also broad overall static cut/boost, and Volume for automated curves?

Just trying to summarize which role each of the three level control plays in the mixing process...
* Get track material levels in the same ballpark with Tool Gain
* Automate coarse volume changes with Tool Volume
* Mix with track levels set in place and make changes to Tool Volume automation? Or add automation to track level?

before bouncing tracks i create a group for the track first and have different iterations in it. it seems clumsy first but can get very handy. bouncing can get especially useful when you work at higher samplerate. 96k etc (dont forget TDR Ultrasonic to take care of IMD in that case)

Nice! Didn't know about TDR Ultrasonic (then again I'm still happy at 48k). Do you mean a group for different versions of the same track? Or a group _of_ different versions, to be bounced out as a single track?

[UltraByte]

Nice! Didn't know about TDR Ultrasonic (then again I'm still happy at 48k). Do you mean a group for different versions of the same track? Or a group _of_ different versions, to be bounced out as a single track?

at least before the final limiter but usually you would want to do it before every plugin that introduces THD.

its being discussed in these threads (also by TDR)
https://www.gearslutz.com/board/masteri ... -post.html
(Ultrasonic was created as conclusion of this thread)

and more recent here
https://www.gearslutz.com/board/music-c ... st15091789

(and further posts in that thread)

[antic604]

Thank you both for the answers!

I probably shouldn't have replied, because I'm not very deliberate and principled about mixing. I mix as I go, meaning I'll have an intro and I'll get it 95% ready, I'll move to say "verse" and I get that 95% ready, and so on. Normally "mixing" is when someone has their recordings and arrangement done. But I don't separate the process, so I do what I think works in given moment and won't constrain me in the future (so fader automation is a no-go, 99% of the time).

Anyway, to reply to your questions:

1) I'm not bouncing-in-place, because that way I lose MIDI. I know I can move it to the side, or to Clip Launcher or drag to Browser to save it as a clip; but I don't find any of this "elegant". Although I can make use of it, e.g. when I have the same sequence repeating twice: once by itself and 2nd in some densepart of the track, so I'll bounce-in-place the 2nd repeat so that it doesn't tax CPU as much, because I still have the 1st version

2) No, I wouldn't print fade in/outs to audio, because I lose the control over duration, strength and curve of it. Automating a Tool or multiplicative fader modulation is what I do then.

3) I try to use faders for static mix, i.e. the overall balance of sounds, whereas Tool is for gain-staging within the device chain - Gain knob - and also for FX, like rhythmic gating, ducking or indeed fades - Volume knob.

But those aren't rules, just personal choices

[Hez]

I've ended up with a workflow where heavily processed tracks have their own group, which separates processing I want to leave until the mixing stage onto the group track (aggressive compression/EQ/reverb), and leaves stuff that I know I want to print on the actual track itself (the source synthesiser, more CPU intensive advanced effects etc.).

This lets me easily bounce + deactivate and hide the heavy CPU stuff, but still have control over final mixing touches, without needing to duplicate devices or do any crazy audio receiver routing.

Simpler tracks (one shots, background noise etc.) I'll usually just bounce to a new track and then deactivate the source.

I normally leave bounced + deactivated tracks next to their original track, as I find it easier to navigate the project this way.

I have show/hide deactivated tracks mapped to a keyboard shortcut so it's very quick for me to have a look at them in case I've changed my mind and want to rebounce something with some tweaks to the source device.

I usually end up with projects a 3 deep group hierarchy, which I find quite organised and easy to navigate. For some extreme sounds I'll go to another layer of depth, but this is usually when I'm messing about with complex routing and tracks that affect one another (such as vocoders etc.).

By working mainly with deactivated tracks hidden, and carefully separating processing across group tracks vs. individual tracks, I can work pretty quickly this way without ever needing to move tracks or change volumes/compressor settings etc. after a bounce. I would struggle to go back to the 'freeze' workflow in e.g. Ableton - the Bitwig paradigm of gradually bouncing each stage of the processing once I'm happy with it feels much more elegant to me.

[C-quence]

Thanks for sharing @Hez, sounds like a good workflow, will try it out!

[Taika-Kim]

The way I mixed my new album (https://kadonneetmaat.bandcamp.com/releases), apparently successfully since I have received very good feedback on the sound:

When the tracks were about ready, I rendered everything to stems.
Then I put everything in a group called "Project" and disabled it, freeing all CPU resources.
Then I imported all stems and put then in a group called "Mix".
I had to redo send amounts and automation, which was a bit time consuming. But this was the only way for the send fx (global reverbs only in the way I work) to reflect any changes I make within Mix.
Now it was very comfortable to only focus on the sound balance, especially using volume automation, which can get very surgical and surpasses any compressors etc automatic processors.
I also added some elements and reworked the tracks a bit if I felt like it, and it was easy to make new bounces from the Project group if I noticed some sound needed a change.
I then made mixes where the overall balance between different tracks elements was good, but the overall tone and timbre of the album was not yet set. I rendered these out to stems.
Then I imported all of these again back into Bitwig, this time a completely new project called Master, where I made everything almost final, and matched the timbre of the different tracks to each other, decided how the dynamics will be, etc.
I rendered everything out at -10dB and recorded the tracks onto tape with Mix:Analog to smooth off some of the digital sheen and give the sound just a slight bump. On a few tracks I also used the Pulteq and Fairlight processors, but again only made smallish changes.
I then again imported these into the Master project, and made the final very slight adjustments mostly to recover some of the high end the analog processing ate. At this point I used Zynaptiq Intensity on very slight settings to give the tracks a tiny but in my opinion important final polishing of the details.

Besides the analog equipment and Intensity, I used only Bitwig native devices for mastering. I don't find they lack anything. Especially useful is the ability to sidechain device parameters from another track.
I used a lot of dynamic EQs, filter ducking etc to make the various sounds sit comfortably with each other.

[antic604]

I had to redo send amounts and automation, which was a bit time consuming. But this was the only way for the send fx (global reverbs only in the way I work) to reflect any changes I make within Mix.

We should really lobby the devs to add option for rendering of FX Track processing to stems.

Alternatively - or in addition to the above - regular bounce should have an option to copy all devices past the bounce point + their automation, as well as track's settings (fader, panorama, sends) & automation to the newly created audio track. That would make the process you describe - to which I'm aspiring, too - much faster and safer.

[Taika-Kim]

Rendering sends to separate stems wouldn't help much, because then you could not change either anyway without them becoming decorrelated from each other.
I'm not even sure what would be a good workflow here in an ideal situation.

One thing is, that if there are fx on group level, they need to be copied too to the mixing project, along with any automation. In hindsight I think it would be best never to put any fx on the group level if the goal was this kind of workflow. But that would obviously be limiting in a quite big way, since the a major point of groups is to be able to collectively process tracks.