Stem mastering is an intermediate step between mixing and mastering. It involves splitting the mix that is sent for mastering into a set of stems that are appropriate for the situation. A stem is a part of the mix rendered as a stereo file — for example, all drums or all backing vocals. The remaining elements are rendered as one or more additional stereo files. During mastering, these stems are then recombined, with different processing applied to each stem as needed.
As an example, consider a techno track where the mix engineer has created a sidechain relationship between the kick drum and the bass line. In the mix, this works well, but after mastering the kick drum loses too much impact. Raising the kick level in the mix helps, but once the track is mastered to the level the producer wants, the pumping becomes excessive, as all the dynamic processing in mastering reacts to the louder kick drum.
In this case, one possible approach would be to render the kick drum as one stereo file, the bass as another, and the remaining elements as a third file. During mastering, the level of each stem can then be adjusted individually, placing the kick exactly where the compression and limiting no longer dominate, while still preserving the balance and impact the producer is aiming for.
There are countless reasons why stems might be needed, but as a general rule, my experience is that the best results are achieved when the mix is already “right” and is delivered as a single stereo file. That said, I receive a handful of mixes each year where there are valid reasons to use stem mastering. Below are some of the advantages and disadvantages of working with stems.
Advantages
-
It can be cheaper and faster than sending multiple mix revisions back and forth.
-
The mix engineer may have spent too long with the mix and lost objectivity. In those cases, it can be beneficial to hand over decisions about problematic elements to a fresh set of ears.
-
In some cases, stem mastering can result in less overall processing during mastering, as problematic elements can be addressed separately without affecting the rest of the mix. Ultimately, this can lead to a more natural-sounding result.
Disadvantages
-
The mix engineer loses control over the balance of the mix, and the risk of not recognising the mix after mastering increases. Sometimes this doesn’t matter — sometimes it matters a great deal.
-
Mastering becomes more expensive, as stem mastering takes more time. If issues can be resolved in the mix at no additional cost, that is usually the better option.
-
In cases with tight deadlines, the advantages of stems can be reduced or lost entirely. The result may have been better if the time had instead been spent addressing issues in the stereo mix rather than managing stems separately.
-
With some mixes, it can be difficult to render stems that truly represent the original mix when recombined. This can include reverbs that behave differently depending on whether they are fed by vocals and lead elements together or separately, or mixes that rely on sidechain relationships — for example, guitars ducking when vocals enter.
-
The margin for error increases. Human factors such as rendering stems in different formats or with mismatched settings, become more likely.