For those that aren't familiar, Grainger frequently reworked his own compositions for instrumentations other than the original. Handel in the Strand was originally composed in 1912 as a piano trio (piano, violin & cello with optional viola). It was later reworked in 1930 for piano solo, for orchestra and piano in 1932, and then again in 1947 for 2 pianos 4 hands. The military band version(s) with which many are most familiar was not made by Grainger, but rather either by Goldman or Sousa. There is also a different orchestral version that Grainger may have made for a recording he made with Leopold Stokowski for which he played piano.
I could go on - there is a long list of versions of this piece. I have made two myself; one for quintet, and this one for large brass ensemble. In both instances, I worked from the original 1912 version (shown here).
This arrangement was chosen for illustration purposes for several reasons. First, as with most of Grainger's music, the original score is sometimes very dense and this poses some interesting arrangement problems. Second, it took a really long time and several attempts to arrive at a version for brass ensemble that I felt did some justice to the original. I started working on the brass ensemble version more than 20 years ago, and didn't finish it until this year. Third, the theme & variations nature of many of Grainger's compositions creates a burden on the arranger to use some imagination to keep the variations from getting dull.
Today, I'd like to focus on the first 8 measures. The left hand ostinato is the actual theme of the piece and is derived from a set of variations Grainger himself composed on Handel's Harmonious Blacksmith. The texture is basically 7 voices thinning to as few as 5, and continues on as a 7 voice texture really for about 60 measures. If one were to assign every note in Grainger's setting to an individual member of an 11 piece group, there would only be 4 voices available for color variation. It would also be a dense texture that would be problematic for a solo voice to cut through. And it would be exhausting both to play and to listen.
So my choice was to thin the 7 voice texture to essentially 5 voices. There are some interesting color notes in some of the chords, so in doing this I had to be careful to not miss any, while preserving the voice leading. Since the beginning to measure 40 is one large crescendo, I also chose to 'thin' the texture further at the beginning by using cup mutes (and straight mutes for tuba and horn when they are sparingly used at the beginning).
Because the ostinato comprises a steady eighth note rhythm for eight bars, and brass players need to breathe, AND I'd rather make the decision about where those breaths occur rather than leaving it to the musicians to make the tough choices, thinning to 5 voices created the opportunity to move the ostinato around a bit to create both interest and opportunity to breathe. I was careful to create patterns in the trombones so that voices leaving and entering the texture would not create too much rhythmic 'interest' that would distract from the main melodic material.
I also used the xylophone playing in its second octave (and sounding an octave higher) doubling the top ostinato voice to do 3 things: 1) it adds an interesting color contrast to the trumpets and trombones playing in cup mutes; 2) no matter how the voices shifted around underneath that sound, the xylophone provides a source of continuity for that top voice; and 3) Grainger marks the beginning of his score with the term 'shortish' - the crisp nature of the xylophone helps reinforce the aural illusion of "shortish."
I was mindful of chord voicing, but in this instance perfectly balanced chords took a back seat to voice leading, and making certain I had the interesting (crunchy) notes covered. As is my custom, I first wrote the bottom most line (electing to primarily use bass trombone at the beginning), and then made sure the top voice was covered with a sense of continuity, and then filled in the 'patterns' in between.
Hopefully, all of this effort and thought serves the higher purpose of creating an interesting but subdued canvas for the entrance of the solo horn in measure 5. I could have started with the euphonium which would project better in that register, but wanted to save it for a slightly louder moment 4 bars later (always being mindful of the 40 measure crescendo).
A quick word about mutes (especially cumbersome ones) - understand your exit strategy. You've got to create some time to get the mute out, especially for trombones and tubas. If you don't leave enough time, mute extraction will create a ruckus and you'll get plenty of unpleasant feedback.