Gene Gajewski

Personal Blog

Close this search box.

Transposing Instruments

This should really be titled “And another thing…” because you just know I’m just going to go off on another rant. This little snippet of music made me pause. It’s a mid-part instrument change; not an unexpected situation in a symphonic score – players handle multiple instruments of the same basic type all the time.

Here’s the example of an English Horn player subbing in a short bass oboe part, alongside of an already existing bass oboe player.

Bass Oboe Solo

You see there are extra rests indicating the English Horn is silent while player #2 plays a few bars with a bass oboe. Player #3 is also playing a bass oboe. But think a moment – players #2 and #3 aren’t playing some funky chord together, they are playing the exact identical notes! What unholy abomination is this?

First off, although the bass oboe isn’t a transposing instrument, its range starts with C3 instead of a regular oboe’s C4. That’s not our problem, however.

The English horn is a transposing instrument, and it sounds a perfect fifth lower than it is written (seven semitones). Player #2 is playing the same bass oboe part as player #3, but it is notated on an English horn’s transposed scale. Things like this will hurt your brain. They’re playing identical notes.

UPDATE: I had assumed that these were instrument changes – but I am wrong here. These are just “cues” placed in the score for the player to orient themselves to another (different) orchestral player. So, this explains my curiosity with the transposition. Even though it is a different instrument, it is displayed in the key and clef of the player being “cued”, not the player who is the one being used as a cue. Lesson learned!

Ok – I’m just going to keep adding to this particular post and use it as a running commentary of my conversion of Gustav Holst’s The Planets

Clef changes (mid-bar) and enharmonic shifts

There’s a ton of this kind of fun (not) going on all over the place in Holst’s score. I can only imagine that the reasoning for this is some sort of British fastidiousness to use the least amount vertical space in a score. Here’s the bassoons.

The signature coming into this section is 6/4. In the second measure here, the composer uses the C-clef to indicate a tenor clef, before going back to a F (bass) clef. This means the indicated notes of the ledger lines will have shifted. This may be the “technically” correct thing to do – but I’ll bet a wooden nickel that a modern version of this score would dispense with this fuddy-duddy, clef shifting trick and just notate up vertically. But we’re not done with the tricks here yet.

On the first bassoon, just past the tenor clef follows the notes D#, C (natural), D. Why does the C have a (courtesy?) natural attached? It’s not because of an earlier prior accidental of the same note, but an accidental on the same ledger line (despite the new clef rearranging things). We wouldn’t want to confuse C (natural) with that F# from the previous bar, now would we? (Sarcastically)

I guess the point I’m making here is that musical notation is more geared towards the convenience of the composer than the performer. There’s a lot of convention in musical notation that, for a performer, you must simply commit to automatic, short-term memory, these conventions to be able to successfully perform. Therefore, instrumentalists practice thousands and thousands of hours to be successful. But, from a MIDI perspective, the notion of clefs has no meaning, except for display purposes.

By display purposes I mean that pitches are always absolute in MIDI. If you view music as a piano-roll, as every DAW will do, there’s no mumbo-jumbo with accidentals other than as names for some of the notes. It’s only when you switch to a score (ledger) notation view does the clef assumes importance.

Since MIDI data has no notion of a clef, when looking at MIDI data in score form, the DAW applies what it thinks is the correct clef to use depending on the instrument in use. But it’s a guess and not always right. And the DAW has no way of knowing that the composer switched to a tenor clef in the middle of a bar. The midi score notes the DAW displays are technically correct, they just don’t look the same as the score. If you go and manually enter that tenor C clef, notes then display in the same positions as the original score.

It really helps to be able to read music to ensure that a conversion to midi from ledger notation sticks to the original intent. It’s not that most DAWs don’t do this correctly, but there’s always the chance of a bug occurring, and you won’t be able to spot it otherwise.

For a lot of the work I do, I can simply follow the rhythmic landmarks in the score to identify which notes will need some extra help and identification for the DAW to properly articulate. But even then, you must be careful. A staccato eighth note will appear in a MIDI-ized score view to be a 16th note, because time is absolute as well, it has no notion of “staccato” – there’s not much you can do about it.

Also, a DAW will usually stick to proper beaming of notes across beat divisions, where the composer doesn’t, usually to eliminate having to print extra rests and ties. Fortunately, the DAW I use, Cubase, has options to control some of these things so that that its score view can hew closer to what the original score looks like, but it’s not perfect. My other DAW, Cakewalk, simply isn’t as sophisticated and provides no such features – and that makes score comparison impracticable.

In short – my view is that (modern) musical notation was designed more for the brevity of the composer than for the obviousness of the reader. Manually copying musical manuscripts must’ve have been a real pain in the ass, back in the day.

I gripe – but there really isn’t any alternative. Besides, I thrive on obtuseness – I love a good puzzle. You need to have a foot in two worlds here – one in musical notation, and one in MIDI representation. You don’t need to be a sight reader – you have the leisure to read at your own pace, but you do need to be able to read and understand. Today’s internet makes it possible to learn things on your own that would have been difficult without an instructor in the past.

Oh yes, I mentioned enharmonic shifts, didn’t I?

Yeah, I’m catching them. That’s fine, but if the first trombone is using D flat, the second one shouldn’t be using C sharp.

In all fairness, this is an exceedingly difficult score, and there’s bound to be a lot of little trip ups here and there. Most of them I can ignore. This one is just annoying because it makes visual comparison difficult.

Also, you can see a lot of hidden notes in the score. These don’t sound, but they sure make things ugly!

Those clefs again!

This is another gotcha where Cubase was confused. The reason for this is the trouble treble clef for the tenor tuba shown here. Tubas are non-transposing instruments when using a bass clef. But for some reason, a tenor tuba, also called a euphonium, is treated differently. According to Wikipedia, a euphonium is treated as a transposing instrument and used with a treble clef in British bands. I would have thought to use a tenor clef as it is tenor voiced, but no, those blasted limeys must be different.

Anyway, thank goodness for Wikipedia clearing things up.

Another unholy thing

This thing haunted me for a bit. What the hell are tremolo symbols doing between notes? For that matter, how can there be two half notes in the same bar when the meter (not shown) is 2/4? How is that possible? Well, it seems there be a tremolo articulation in here somewhere! My guess is that the two half notes are for telling the violins to split divisi, one note for each half, and putting the tremolo slashes between them means to treat them as a singleton for time purposes? If that’s the case, then why not just put them both on the same stem with a proper tremolo mark and note divisi in the score? Actually, you could forego a divisi directive altogether since splitting divsi is implicit whenever there’s more than one voice on a staff. Who knows? Anyways, another scoring peccadillo solved.

Leave a Reply

Share the Post: