We begin by thinking of orchestration very simply in terms of how to combine elements, both:

"simultaneously" (or "vertically" if you prefer) for voicing chords, doubling melody lines, and handling texture;

"successively" (or "horizontally") for effects like orchestral crescendos.

Simultaneous (Vertical) Combinations

To b(lend) or not to b(lend) … that is the question … for "simultaneous" orchestration. For most tonal orchestrators, the default answer is a very clear "yes, blend!"

Voicing chords

As a rule of thumb for tutti orchestration, blend by treating each section as if it were self-contained. This applies to the large sections (winds, brass, strings), and often also to the smaller sub-sections (e.g., flutes) in the case of larger orchestras where there are greater numbers of each. For each section, follow the principles you learned in four-part writing. That's right, the principles you’ve already learned about chordal writing still apply. Specifically, you'll need to consider:

The inclusion of all (or at least the most important) harmonic pitches.

The number and prominence of each pitch, including avoiding the overdoubling of the third in a triad.

The gaps between voices:

We now see the densest packing in the middle.

Continue to use larger gaps in the bass register (e.g., the octaves given by cello and double bass notated on the same pitch but sounding an octave apart).

Continue to use smaller gaps higher up in the treble clef range.

We tend to see more activity in the very highest registers in orchestral scores than in most other contexts; it's usually best to return to wider spacing in those highest registers.

We have terminology for the different types of chord spacing as summarized in Example 1:

Example 1. Chord spacing types.

Context determines the best choice among these. For a blended sound, we often use integrated "interlocking" voicing, taking care to balance the different strengths of each instrument according to the register; conversely, to achieve a different color for each note, use wider spacing. For more textbook examples, see Adler (2002, 253); Piston (1955, 396).

Doubling lines

The approach to blending in doubling lines again involves finding a compatible balance of (equally strong) instruments and registers. The second part (m.1 4ff.) of Mozart 40/iii is effectively a two-part invention and an intriguing example of blended, balanced lines. Both of the two parts are played in each of the three octaves, and by both winds and strings.

Mozart 40 iii

As always, this default is not a law like that of gravity in physics; composers will diverge from this blended practice for creative and compositional effect in just the same way that they do with any other parameter. Perhaps the most famous example of an odd doubling is provided by Schubert 8/i. The first theme (m. 13) sees the extraordinary doubling of a single oboe and clarinet in unison. While some commentators have seen this as an error, it's probably deliberate on Schubert's part as a way of creating an unsettled feel that entirely befits the mood of the music. It also contrasts with a second theme (reh. A, m. 42): a cello theme accompanied by the optimally blended viola and clarinet in thirds.

Schubert 8 i

Texture

The choice of instruments, register, and more can help support and define the wider texture. For instance:

Polyphonic writing: Tends to see instruments and groups balanced and blended. Examples include fugues and the Mozart invention just discussed.

Melody and accompaniment: Foregrounds the melodic instrument by placing it in its best range, often at the top of the texture (as in other contexts), and with other accompanimental parts contrasting in:

Timbre, accompanying a solo woodwind line by full strings, for instance.

Register (as discussed).

Texture, with less or a different kind of rhythmic activity. This is similar to the kind of complementary rhythmic patterns you often find in polyphonic music: in both cases, we're trying to have more than one element at once, but keep them noticeably distinct.

Successive (Horizontal) Combinations

We’ll look here at forms of succession on two timescales: at the level of individual notes (small scale) and of whole sections (large scale).

Firstly, once again, everything you’ve learned about voice-leading still applies. (Yes, this will be a recurring mantra.) Most of the time, it will be eminently possible to reduce apparently complex orchestral scores down to a two-, three-, or four-part texture with clear and consistent doublings and with "correct" voice-leading.

In handling successive sections, antiphony is a highly favored method for distributing material, creating timbral variety, and articulating the form. This can take the form of whole blocks of material being passed from strings to winds (common in early orchestral scores of the Baroque era), or of more subtle effects. For instance, look at the first page of "In the Hall of the Mountain King" from Grieg’s Peer Gynt suite:

Grieg Mountain King

At the repeat of the tune, the only variation is antiphony (strings-to-winds). The material (both tune and accompaniment) and register are identical. Later, this strings-to-winds principle continues to the upper voices as the register expands upward. Let’s take a closer look at that crescendo.

Crescendos

We have a range of tools at our disposal for achieving an effective orchestral crescendo. Apart from mere dynamic markings, we can:

Add instruments, often moving from the "lightest" to "heaviest" instruments: broadly, strings-wind-brass-percussion (e.g., Grieg rehearsal mark A).

Expand the overall range, often upward from the bass or outward from the center (e.g., Grieg rehearsal marks A and B).

Diversify the note values