Orchestral Voicings Part 16: Dynamic Envelopes

A big factor that helps us to recognize instrumental colours is in fact not only the mixture of the harmonics of the sustaining sound but at least to an equal part its dynamic envelope, particularly the attack of the sound. As attack time, we usually mean he time it takes until the air in the tube or the string string etc. has reached a continous vibration. But with acoustic instruments that phase itself has such a distinct sound that it is part of the individual quality of an instrument. 

Today we will focus on how these properties influence how we write orchestral voicings. To read the previous parts of this ongoing tutorial series click here.

In acoustics, the dynamic envelope of a tone is usually split into 4 phases which are called ADSR. A= Attack, D= Decay, S = Sustain, R = Release.

These curves are very prominent in all sorts of synthesizers where the individual lengths of these phases usually can be adjusted. As you can see in the graphic above, the y-axis is labeled with "amplitude" which is the basic understanding of such a curve, however with acoustic instruments, all these phases have specific properties that go way beyond only amplitude.

Let's for example imagine a bowed note played on a string instrument. Starting to drag the bow across the string creates a small moment of "scratch" sound before the string itself starts to vibrate. This is the attack phase. The decay phase is depending on how accented the player plays that note and can be quite noticeable in a dip of intensity after the attack or practically non existant in a softer played note. The sustain phase is the actual note that we're hearing and even in that the sound is far from stable waveforms that keep looping. After the bow is lifted from the string, it will vibrate on for a while until it stops vibrating. That is the release phase.

Why am I telling this? Because these properties have a considerable impact on how you can mix colours and how you can create voicings.

One of the most prominent instances where you are really confronted with these properties is when you try to voice a very soft chord in heterogenous instrumental combination. For some instruments, it is incredibly tricky (also depending on the register) to start a note super softly. Most prominently here are brass instruments. It simply needs a certain amount of air pressure behind the lips to make them start to vibrate. So you will always get a noticeable attack on brass notes, no matter how low the dynamic marking. If you make the unfortunate choice to for instance combine clarinets and trumpets in a soft voicing, you will not get the beginning of the voicing to really feel like it is lining up with each other. The clarinets are capable to sneak in practically inaudibly while the trumpet no matter how hard they try will not start the note out of nowhere. So the result will be a disbalance in the beginning of the voicing even if you somehow manage to balance the voicing after all the notes reach their sustain phase (which will be tricky in this combination anyway).

But there are also combinations of different attacks that are rather unsuccessful in lower dynamics. Bassoons for instance suffer from a similar problems as trumpets as (unless it is an excellent player). You need a certain amount of air pressure to start the double reeds to vibrate. So again, a combination of clarinets and bassoons in a very soft setting will create some uneven attacks even if the combination of these two textures in sustain works very well.

However, be aware that this problem only is noticeable if the voicing is very exposed.

Also on the other end of the dynamic spectrum, there can be problems. Practically every wind and string instrument is capable of creating a forced attack - more or less an accent which starts the note rather explosively. With woodwind/brass instruments this is created by building up air pressure behind the tongue that is blocking the air stream and suddenly releasing that air. On string instruments, this effect is achieved by starting the note with higher bow pressure and a more rigid start of the bow movement (or hitting the string while the bow is already in motion). The accented attacks work particulary well on brass instruments that due to their rich harmonic spectrum in high dynamics are capable to create a noticeable rhythmic event on their attack phase. While in isolation, woodwind instruments can create nice punchy attacks, too, they are not as effective as the brass attacks and it is not really possible to create a really sharp and percussive attack in most cases. So again even if the voicing might be balanced during the sustain phase, it might not neccessarily be effective on its attack.

A special case here are of course percussion instruments that move right from attack to release in most cases. And while it is completely plausible to use percussion instruments that have a specific pitch like timpani or glockenspiel to highlight the attack phase in a voicing, it is also worth seeing percussion instruments of indefinite pitch as part of a voicing. Doubling the start of a voicing with a hit on a snare drum or crash cymbal will subjectively highlight the attack. A snare drum basically adds a brief burst of "white noise" to the beginning of the voicing which is quite similar to a high energy attack phase of brass instruments. It is useful in such cases to not only look at individual pitches of the voicing that you are writing but looking more at how you occupy the frequncy range. Adding a swell on a suspended cymbal to a crescendoing chord does nothing else than adding high frequencies which generally we feel like an increased intensity.

So far we only spoke about these "standard" ADSR phases that presume a rather constant volume during the sustain phase but of course, there are also instances where we write a crescendo or decescendo over a voicing. In these cases we should be cautious as well when voicing chords. While of course, some instruments are capable of creating a much higher volume at their higher dynamics than others, it should also be noted that the textures in different dynamics change differently. A crescendoing brass instruments adds more and more high harmonics to its sound while a crescendoing woodwind does that only mildly. So if you voice a chord that has crucial tones in one or several instruments that don't get as prominent in the higher dynamics, it might create problems. Even though they might balance great in softer dynamics, the voicing might fall apart when they all crescendo.

Of course, this effect can also be used creatively where for instance a decrescendoing voicing reveals a more complex harmonic structure with the different parts of the voicing reaching a more even balance towards softer dynamics.

It should also be noted that it is of course possible and can create intriguing effects to "synthesize" or orchestrate your own ADSR curve. A fairly common effect is to double a woodwind chord in the strings playing pizzicato. So while the woodwinds sustain the chord, the pizzicato only adds a stronger attack to the sound and then dies away. But of course, practically anything goes in this regard. Creating a heavy attack on a string chord by having the brass play a loud staccato at its beginning etc. But also the other way around works well. A common practice to orchestrate a loud sustaining (possible climax) chord that decrescendos is to stagger the end points of different instruments or sections. For instance in a loud tutti chord you let the brass come off first, then woodwinds and last strings, you can create a "written out" decrescendo but also a "timbre decrescendo" which can have a very attractive quality.

As I said, the possibilites playing with these things are endless. Having a tutti crescendo to a big fortissimo and suddenly stopping all but a soft triad in the flutes or whatever instrument that keeps sustaining can feel like a very attractive and exotic release phase. Some composers go as far as for instance indicating the harp or piano player to not play but also not actively dampen their strings in such a loud "rip off"-chord and just let the resonance that made the strings vibrate on their own ring out which feels like a weird and amost artficial reverb tail.

So the bottom line here is: Voicings don't neccessarily mean to start and end everybody at the same time holding the same dynamic and playing with the same articulation but you can be way more creative with these things. As usual, if you understand the individual properties of the instruments and have experience using them, you can create incredibly interesting and attractive soundscapes, just by playing around with these things. But of course, inexperienced orchestrators could also end up writing something that unintentionally sounds rather strange.

So as usual, it helps being aware about these things and using them to your advantage.