There are two overarching concepts that determine how modern music is mixed. The first is the room workflow, which is divided between ITB and OTB mixing. ITB (in the box) mixing means that you are working totally within the confines of a DAW (digital audio workstation) application, like Logic Pro, Cubase, or Pro Tools. OTB (out of the box) mixing means that you are working in a studio with a physical console and a myriad of outboard audio processing gear. This is how music was always mixed before the invention of the DAW. Modern studios and mix engineers use both techniques. Often a hybrid approach is employed, such as mixing in a DAW, but routing some of the signals out to external hardware and back into the DAW.
The second concept is whether you are mixing bottom-up or top-down, which is the subject of this post. In a typical bottom-up mix nearly all of the processing is applied to the individual tracks or instruments. Those channels are mixed together though the stereo output or 2-bus. Only basic compression and limiting is applied to the 2-bus signal to “glue” the mix together and tame signal peaks.
In the opposite approach – top-down mixing – a lot of processing is applied to the 2-bus to shape and control the signal (EQ, expanders, exciters, compression, limiting, etc), but very little is applied to each individual channel/track/instrument. The idea here is that you are “driving” the signal into the effects chain on the 2-bus, where the total mix is being shaped.
Analog console evolution
Bottom-up versus top-down workflows stem from the evolution of analog console design. Mixing consoles are built around a series of input channel strips, whose signals are combined into a mono, stereo, or multichannel output. The earliest consoles only had channel strips with preamp and volume controls. As these evolved, most console channel strips also gained high/low-pass filters and equalization. Any other processing (noise gates, de-essers, exciters, reverbs, compression, limiting, etc) had to be handled by external hardware. Naturally, such equipment was and still is expensive, so most studios could only afford a limited amount of this gear. The result was often to only apply this processing to the output of the mix.
Solid State Logic changed this with the introduction of its consoles, notably the SSL 4000 series. Their key innovation was that the circuits for each channel strip integrated a full dynamics control section (compression, gating, expansion) in addition to the preamp, equalization, and filtering. This design made it possible for a recording engineer/mixer using an SSL console to employ a wider degree of control for each individual instrument without the need for external gear. Mixers also liked the tonal decisions made by SSL’s design engineers, so SSL consoles became popular in studios around the world. Many of the most beloved rock records of the late 20th century were recorded and/or mixed using SSL consoles.
Applying analog console design to digital software
Fast forward to the present and you’ll see that the way people design studios and mixes – whether ITB or OTB – is based on concepts from the analog days. Modern DAW software mimics the layout of tracks and channel strips. Depending on the application, some have channel strips with built-in processing effects, some rely only on effects inserts (built-in and third-party plug-ins), but many use a combination of the two.
For example, Fairlight (Blackmagic Design DaVinci Resolve) offers a selection of built-in channel strip effects. Click on a section and adjust the controls. In Logic Pro, if you click the EQ panel at the top of any channel strip, you activate the default digital parameter EQ on that channel. Click on the gain reduction panel and the default Platinum compressor is applied. Of course, you can swap these out for other built-in or third-party tools.
The downside of this approach is that each separate effect has its own GUI, so opening just an EQ and a compressor for multiple channels quickly covers your screen. Many users prefer various third-party channel strip emulations, such as those that mimic SSL or Neve hardware. The advantage to these is that all of the different processing tools for a channel strip open up within a single, cohesive GUI.
There are plenty of SSL clones, but the Waves CLA Mixhub takes this up a notch by turning the channel strip plug-in into a virtual console. Apply a Mixhub plug-in to a series of channels (up to 64 total) and assign each to one of eight “buckets.” Then in Mixhub’s bucket view you can see up to eight channels of EQ or dynamics side-by-side within a single window. It’s like having a virtual SSL console on your computer screen.
Some analog console manufacturers offered a set of integrated submix buses. This architecture has been carried over into DAWs. You can combine and route a group of similar instruments over to one or more buses. The individual channel levels are set relative to the rest of the channels in that group and then the bus level is set relative to the other buses as part of the mix.
Typically, buses are either VCA or summing. In the simplest of terms, a VCA bus is a glorified remote control where a single bus fader applies relative volume changes to the channels within the group. In most cases the VCA bus isn’t actually working with a combined signal (although it looks like that in the GUI). This affects the gain of the signal and whether or not effects (plug-ins) can be applied to that bus. A summing bus actually combines the individual signals into one and then allows for absolute volume changes to that group, along with the addition of processing. This design requires more attention to gain-staging, but makes the hybrid top-down mixing solution possible.
Pros and cons to analog emulation
There is no right or wrong way to mix. There are plenty of award-winning mixers who fall into any of these camps. You have to develop a methodology that works best for your needs and style. I mix music on a casual basis, mainly using Logic Pro. I own a number of channel strip plug-ins that were inspired by or emulate noted analog consoles – SSL, Neve, Focusrite, etc. I’ve created mixes using each of these, as well as just with stock Logic Pro plug-ins.
Regardless of how many demos you’ve seen and heard or how many YouTube influencers have touted a product, you might not hear much of a difference between these different tools. The truth is that all compressors, all EQs, all reverbs, and so on do pretty much the same thing. Some have different coloration to the sound. Some are “character free” – i.e. clean. But once you get everything into a mix and are no longer fixating on isolated channels, you’ll realize that the differences are pretty slight. This is especially true when listening to your mix on headphones or small/medium near-field desktop speakers as opposed to A-level recording studio monitors.
If you’ve applied a plug-in like an analog-style channel strip onto each channel, then this becomes the virtual equivalent of an analog console and represents a starting point for a bottom-up mix. Likewise, you can insert a stack of processing plug-ins onto the 2-bus and push the mix into these as a classic example of top-down mixing.
The hybrid mixing method
Here’s the approach I’ve settled on. The first step is to bring everything in with flat faders. Not everyone agrees, but I do worry about gain build-up in the process. Logic Pro runs in 32-bit float, but I’m not sure if that’s true for every third-party plug-in that I use. I have heard distortion with some when the input level was too hot. So, I’ll typically drop the gain of each track/channel by -4dB to -6dB. Next, I’ll group common instruments into a set of Summing Track Stacks, which is Logic Pro’s way of automatically creating a summed group bus.
At a minimum on a rock or pop track, I’ll have buses for drums, bass, keys, guitars, and vocals. These in turn are routed to a new bus, which becomes my Submix bus. It feeds the Stereo Output (2-bus). I will then apply my channel strip emulations (or comparable plug-ins) to each instrument group (Track Stack). Usually some light compression and limiting is applied to the 2-bus.
The bulk of my processing to shape and color the sound of similar instruments is happening at the buses, which is a classic top-down mixing method. However, it’s a hybrid, because I’m doing that to instrument groups instead of the full mix. After all, if I’m going to apply the same basic effects to each instrument, it’s more straightforward to apply it once to the group. In addition, I can still control the level and processing of one bus relative to another, rather than have to tweak individual channels. For instance, once I get the channels of the drum mix right, then I only need to deal with the whole drum kit as a single unit.
Since I’m applying processing at the bus level, I can then “push” the mix of those instruments within the group into that processing. Using this method means that there are fewer effects that I need to apply to individual channels. Sure, I might EQ an individual vocal track or apply an guitar amp emulation to a guitar DI track. But, that’s less involved and often better sounding than monkeying with each individual track.
Let’s boil this down to some simple steps: 1) Import tracks, create buses, and get a nice overall balance of volume and panning. 2) Apply processing to each group/bus and adjust. 2) Tweak individual tracks to enhance the balance, clarity, and depth within that group and the mix. 3) Apply light 2-bus processing to “glue” the mix.
Which emulation or plug-in chain should be used on the group buses? I’m primarily looking for EQ and filtering to shape/color the sound and compression to solidify the group and tame extreme level changes. Picking a single chain is something I’m still experimenting with. I recently mixed five songs – each with a different set of plug-ins applied to the groups. This wasn’t intended to be a “scientific” comparison. Rather, I’m looking for the best workflow for me. My choices included stock Logic plug-ins, Waves CLA MixHub, Waves Scheps Omni Channel 2, Sonimus SatsonCS, and Kiive Audio Filkchannel Strip | MK2. A different approach for each mix.
I’ve worked with a range of options for 2-bus processing. Usually that means a little extra EQ followed by compression/limiting. I like the stock Logic compressor (picking one of its seven variations), TDR Nova, FabFilter Pro-MB, Sonible smart:comp 2, and/or others. I’ll bounce out that mix and do a separate mastering pass, which I previously explained. Of course, depending on the style of the music, I will use additional and/or alternative plug-in options. But this is a quick description of how I apply the hybrid top-down mix concept.
Not all third-party plug-ins work equally well in all audio and video applications. Most of mine are fine in Logic Pro and Audition, as well as in Premiere Pro and Resolve/Fairlight. However, Final Cut Pro has definite issues with many of these third-party plug-ins. Logic Pro crashes when I use the Brainworx emulations. Therefore, when selecting plug-ins test a trial version first if it’s available.
For a more in-depth look at these workflow ideas, here’s a good tutorial by Nashville mixer Joe Carrell: Part 1 and Part 2. Another good listen is this recent interview with Grammy-winning mixer, Andrew Scheps, who discusses his approach to mixing, as well as the collaboration with Waves on the recently-updated Scheps Omni Channel 2 plug-in.
©2024 Oliver Peters
digitalfilms 