Further considerations re: specifying duration
In Global Notation, the general rule is to specify duration for sustainable sounds but not for impulsive sounds (see Unspecified duration). However, the distinction between the two kinds of sound is not always clear, and information about duration may not always be useful for the purposes of the notation even when the sounds are sustainable. In such cases, the decision whether to specify duration or not may call for more thought.
Borderline cases include impulsive sounds of slow decay, such as those of the piano and guitar when the pitch is not too high. Though generated by “impulses” - hammer strokes or plucks - these sounds may be audible for several seconds, and the player can choose to stop the sound before it has died away completely. It will often be helpful to specify duration even for impulsive sounds when, as in these cases, the performer has control over the duration of the sound.
Sometimes it may even make sense to specify duration when the player has no control over it. The highest-pitched strings on the piano have no dampers, and therefore the player cannot control the duration of notes played on them. Yet it would seem odd to notate these notes differently from other notes played on the same instrument.
Since high notes on the piano die away quite rapidly anyway, the transition from high-pitched strings with dampers to higher-pitched strings without dampers is barely perceptible to the ear, so the information as to which strings have dampers is unlikely to be musically significant. For most purposes a sensible approach would be to treat all notes on the piano as “sustainable” and specify their (perceived) duration.
The question might arise how to handle impulsive sounds of slow decay that are not damped, such as church bells and the larger gongs in an Indonesian gamelan orchestra. The long duration of these sounds may seem so prominent a feature of the music that it ought to be reflected in the notation.
Nevertheless, the end of such a sound is not a musical “event” - the sound is simply allowed to fade until it becomes inaudible, which may happen at different times for different listeners - so it would be misleading to specify any definite duration. For most purposes it will be sufficient to specify when each sound begins and to maintain the principle of specifying duration only when it is controlled by the player.
In fact, there may sometimes be reasons for leaving duration unspecified even when it is controlled by the player. If we are only interested in rhythm, then regardless of whether we are dealing with the impulsive sounds of a xylophone or the sustainable sounds of an organ, what matters most is when each sound begins rather than how long it continues. Information about duration might then be an unwanted distraction from the more important distinction between moments where a new sound begins and those where none does.
As always, decisions about what to specify should be based not only on the nature of the sounds, but also on what is most helpful for the purposes of the notation.