Stem Mastering - Part 2: Selecting and Exporting Stems(Letto 62 volte)

NOTE: This is the second part of the article: Stem Mastering – Definition and Operating Principles

---

Which stems

Gang of Four

One possibility, suggested by sound engineer Bob Katz and applicable to most pop-rock, funk, dance and similar genres, is to simply divide the track into four stems, which he has nicknamed The Gang of Four (Gang of Four):

• Stem 1 – Full mix
• Stem 2 – Instruments 
• Stem 3 – Instruments + Backing Vocals
• Stem 4 – Lead Vocals or Lead Instrument of the song 

Considering the possibility of using stems in an additive or subtractive manner, it is easy to imagine the simple operations that can be achieved.

Example

Let us suppose, for example, that in the mix the voice is too loud, lacks body and is too compressed.

By using stem no. 4 with reverse polarity, we can reduce the volume of the voice considerably, even to the point of making it disappear.

With another track having the same stem 4, but used in straight polarity, we can re-enter the vocal track into the audio programme at the desired volume, taking care to give the track more body and a touch of expansion to counteract excessive compression.

Magnificent Seven

 You will forgive me for the “conceptual game” with the “Gang of Four” in defining the choice that follows with the name “The Magnificent Seven".

Personally, I believe that in some cases it would be better to operate with a few more chances, obtaining a greater number of stems, and in particular:

• Stem 1 – Full mix
• Stem 2 – All instruments
• Stem 3 – Instruments + Backing Vocals
• Stem 4 – Vocals or Soloist of the song
• Stem 5 – Bass
• Stem 6 – Bass drum
• Stem 7 – The instrument (or group of instruments, or instrument session) that best defines the harmonic rhythmic feel of the song (often an acoustic or electric guitar or a piano, but sometimes it can also be a synth, a string quartet or something else).

In this way, in addition to the stems suggested by Bob Katz, you will have other elements at your disposal, including those that are particularly critical at low frequencies, such as the bass and the drum bass drum.

By using additive and subtractive processes, we will be able to effectively resolve many of the imbalances that may be found in the mix.

NB: In special cases, it is possible that the mastering engineer may request a single source or group of sources not included in the above lists. This may occur if the source or group in question exhibits tonal or volume imbalance with the context.

Advantages of Stem Mastering

This allows the mastering engineer a wide range of options for intervention, in order to best resolve any distortions introduced by the mixing engineer. In many cases where single track mastering was planned, stem mastering had to be used in order to control and resolve certain flaws in the mix.

Risks of Stem Mastering

The risk is that the overall “flavour” of the original mix may be slightly distorted, especially if many tracks are involved. For this reason, it is advisable that this be done in the presence of the production manager.

How to export stems correctly

A forward-thinking mixing engineer should take care to export a sufficient number of stems when bouncing. This would guard against the risk of having to make revisions to the mix or even simply having to export the stems requested by the mastering engineer at a later stage.

It is remarkable to note that not all sound engineers know how to export stems correctly if they have no previous experience of this type.

In summary, the correct export of a stem produces a file whose audio programme is perfectly identical to that of the same source immersed in the mix, including the reverb fields, without any variations in volume, tone or dynamics.

NB: It is also possible to proceed with a stem purged of all reverberation, which will allow full control over the source but not over the reverberation fields. In this case, for example, a subtractive action will not eliminate the reverberation of the source. Using a stereo stem within the M/S matrix will ultimately allow us to significantly reduce the impact of the stem's reverb by simply lowering the volume of the S component of the matrix.

The export procedure is very strict but simple. However, it is easy to lose control of some details, rendering the stem virtually unusable in many applications.

The basic rules

• The tracks, whatever they may be, must be exported once the mix is perfectly defined. 
• Synchronised stems are obtained by starting and ending the export at the same point on the timeline, similar to the complete mix. 
• You will simply need to mute any tracks that you do not wish to export.
• This operating mode forces any reverb or other process that has been applied to the entire mix to be reflected in the stems. In certain cases, however, it may be necessary to export a stem of the reverb only.
• In order to maintain the balance of the original mix, it is absolutely necessary that all tracks routed to the same bus are exported as a single stereo element (this recommendation is particularly important if a dynamics processor is in use in the group and if there is an effects return in it).

In the field of additive correction, but even more so in subtractive correction, the mix between the complete track and a single stem will only work satisfactorily if the stem corresponds “perfectly” to the same source inserted into the mix.

To function effectively, compared to the corresponding track immersed in the mix, the stem:

• it must contain the same stereo reverb (so the stem will also be stereo)
• shall not interact by means of sidechain compressor control
• it must have undergone the same equalisation, compression and other processes, using the same plugins
• must have the same volume
• must be positioned in the same place in the stereo image, using the pan pot.
• It must also be identical with regard to the variations imposed by the automatic functions programmed by the sound engineer, both for the volume function and for all other functions.
• It must not be part of a group or bus in which dynamic control processes of multiple sources have taken place, as such processes would create dynamic interaction with the other sources contained therein, making the stem different from the corresponding source contained in the mix. For the same reason, any dynamic intervention on the master track will also be prohibited.

Reduced effectiveness of stems, especially in subtractive contexts

If one or more of the above elements are not complied with, the additive techniques may be only partially effective and it may be completely impossible to use the subtractive techniques.

Please note that many “coloured” filters used in a stem, particularly analogue emulation filters (even a simple analogue emulation equaliser), will often be unable to guarantee a waveform that is perfectly identical to the previous ones at each export cycle, which could reduce the effectiveness of the process, especially in subtractive use (but not only).

Finally, the use of any type of automation on the track, especially volume automation, could produce stems that do not correspond 100% to the characteristics of the same source within the mix, due to the reactive latency differences that sometimes occur in different exports.

However, even partial but high efficacy (e.g. 90%) can be used successfully in most applications.

To evaluate effectiveness, simply add the stem to the mix after reversing its phase polarity, bringing it to a value of 0dB Unit Gain. If the sound of the stem source disappears completely, then the effectiveness will be total. If, on the other hand, a trace of the source remains, its volume will give us a measure of effectiveness.

---

NOTE: This is the second part of the article: Stem Mastering – Definition and Operating Principles

---