Creating Natural Sounding Dynamic Changes with Acoustic Instrument Samples

By Don Muro

Most students use the mixer’s track volume controls to make individual tracks louder and softer. For many acoustic instrument samples, however, lowering the track volume level does not make it sound as if the instruments are being played more softly. This article provides explanations and audio examples to better understand how to create natural sounding dynamic edits with acoustic instrument samples.

 

Digital technology makes it possible to produce radical changes in the sound of a performance. Once recorded, the sounds of acoustic instruments can be pitch-corrected, detuned, transposed, made brighter or darker in timbre, reverberated, and positioned anywhere in a stereo or surround mix. These sound alterations and enhancements can be added so subtly that the listener is unaware of their use. For many acoustic instruments, however, one of the hardest parameters to alter convincingly is the dynamic level of the performance. (For example, there is no way to make a recording of a trumpet part played fortissimo sound as if the part were played pianissimo.) It is possible to edit the dynamic levels of sampled acoustic instrumental sounds in many software synthesizers, but students are usually unfamiliar with the proper technique. 

 

Mixer Track Volume 

Most students use the mixer’s track volume controls to make individual tracks louder and softer (see  Figure 1). For many acoustic instrument samples, however, lowering the track volume level does not make it sound as if the instruments are being played more softly; instead, it sounds as if the instruments are moved farther away from the microphone – a very different psychoacoustic effect.

Figure 1 – a typical software mixer showing the volume level controls for six tracks, two aux returns, and a master fader. 

When acoustic instruments such as the piano, trumpet, and clarinet are played at different dynamic levels, the performers are affecting two properties of sound: loudness and timbre. It is impossible to play these instruments loudly without producing a brighter timbre.  Because of this, lowering or raising the track volume for acoustic instruments will not make it sound as if they are playing softly or loudly because the sound’s timbre is not affected.

Example 1: Lower Volume

Listen to Example 1. This example uses a distorted drum track to help create the illusion of loudness. Chords are played loudly in the upper register of a piano sample in measures 1 through 4. In measures 5 through 8 the same chords are played at a lower track volume in an attempt to create the illusion of a pianist playing more softly. We hear a change in volume but no corresponding change in timbre; instead of sounding softer, it sounds as if the microphone has been moved farther away from the piano.

Example 2: Playing Softer

Now listen to Example 2. Measures 1 through 4 are the same as in Example 1, but in measure 5 a different piano recording is heard. This new recording used the same piano sample (sound) but was played at a soft dynamic level. Our ears and brain immediately respond to the change in timbre and we recognize that the pianist is playing softly in measure 5.

It is very important for you to hear the difference between this sound and the sound of the piano in measure 5 of Example 1, so go back and listen to Example 1 again. The interesting part of this comparison is to note that in Example 2 the volume levels of the loud and soft pianos are almost identical.

Example 3: Piano Comparison

To further understand this concept, listen to Example 3. This example has only the loud and soft piano tracks. If possible, play this example through audio software with display meters so that you can compare the levels of the two pianos. (See Figure 2.)

Figure 2 – when displayed in an audio editing program, the loud piano part (left) and the soft piano part (right) are nearly identical in loudness.

Even though the two pianos are being heard at nearly the same loudness level, our ears and brains tell us that the only way to produce the piano sound heard in measure 5 is by playing softly; therefore, that piano part (in measure 5) is perceived as softer than the piano part heard in measures 1 through 4.

Timbre is the Key

This phenomenon explains how pop singers can whisper and roar at almost the same volume level on a recording and still be perceived as singing softer and louder. Even though the actual loudness level of the voice changes only slightly (by using limiters or compressors to reduce the original dynamic range of the vocal), our ears use changes in timbre as the predominant cue in perceiving a sense of loudness and softness. Again, if you want proof, watch the meters during the song playback.

Editing MIDI Tracks

This is why, when mixing, it is usually best to edit the dynamics of acoustic instrument samples via MIDI velocity levels (see Figure 3).

Figure 3 – this screenshot from Propellerheads Reason shows the note velocity levels displayed as vertical bars at the bottom of the screen. Velocity levels can be redrawn easily using the mouse.

Most software synthesizers use MIDI velocity to affect both volume and timbre simultaneously, so that the sound gets louder and brighter as you strike a key more quickly. In addition, some programs such as Propellerheads Reason and MOTU’s MachFive use velocity switching techniques to trigger additional acoustic samples recorded at louder dynamic levels, producing a more acoustic sounding emulation of dynamic changes.

Therefore, in most cases, MIDI velocity editing should be the first choice for adjusting the dynamics of a sound, and mixer track volume controls should be used for adjusting the balance of sounds in a mix.

©2007 Don Muro

Return to ARTICLES