Another tool that gets used in mixing just like aux sends and subgroups is parallel processing. Running a track into two buses which can then be processed separately is a multi-tool which can be very useful. The first and most obvious benefit is the increase in volume that another gain stage can provide. It also allows flexibility for mixing. Being able to A/B the changes you make with another version or having a dry version to send straight on through, bypassing the effect chain, are some examples.
The real power is the ability to use both versions at the same time. There are a variety of effects that can be dialled in at harder settings than usual and then backed off to get just the right amount. Some processors sound different, however, depending on how much you crank their settings. Vintage style compression is a common one to be used this way. The settings can be adjusted to be very compressed and then a dry signal run through a parallel track so the sweet spot on the compressor can be used without compressing it too hard.
This type of processing is now being included with many of the new processors and plugins but at its core, this concept is very basic and universal to the workflow of audio processing and is used very much so during the mastering stage. Very often, it’s used on drums as well and it gets used on most lead vocals.
Other more specialized techniques stem from this method. Vocals can really cut through a mix when a parallel version is distorted or taken slightly out of pitch. Pan pitching is one such technique where several versions of the voice track are panned to the left and right and pitched up and down. This technique is not dissimilar to the way sound designers tune the oscillators on a synthesizer to thicken it up.
Side-chaining is another trick where the parallel of the track is sent to a processor where it is used to control that processor. This allows a compressor, for instance, to trigger on a kick drum instead of reacting to the signal being compressed. This technique is popular in EDM but has been used on more practical engineering long before that. Using an EQ on the side-chain signal offers the ability to remove low frequencies that may be triggering the compressor when they get too loud. It works well with noise gates too. The signal can be used to tune out only the room noise or microphone hum. When using this technique, the processor has to support it by including a side-chain input. Plugins are designed to allow the host software to route an extra input to the side-chain. In either case, the signal sent to the side-chain input is usually never heard and is only used to control the way that processor reacts.