Sunday, April 2, 2017

Ostinatos; making a lot from a little (2. Rite of Spring)

Further to my previous blog entry
If the 339 consecutive repetitions of a 2-bar rhythmic ostinato pattern in Ravel's Boléro (1928) don't constitute overuse of an idea, then what does?
Or, to rephrase the question in a less acerbic way, Ravel's Boléro is his most popular work (according to Wikipedia), yet from start to finish it repeats a 2-bar rhythmic pattern without change or interruption; how does Boléro maintain our interest? Why are most listeners not bored, troubled, or driven mad by the 339 incessant repetitions of this ostinato rhythm?

The explanation is that there are other musical aspects that change and evolve continuously throughout the work, and these are what sustain our interest, such as: 
  1. It begins almost inaudibly (so quietly that, when I worked as a record department sales clerk as a student, many people who purchased the album tried to return it, thinking there was something wrong with the audio), and grows steadily and inexorably over its 15-16 minute length to become as loud as possible at the end; it is a study in how to write an extraordinarily-long crescendo for orchestra
  2. As such, it is a masterpiece of orchestration, filled from start to finish with colour and texture changes that reflect Ravel's brilliance as an orchestrator. The two-part theme is repeated many times, but each presentation uses a different orchestration (and hence a different colour), and the orchestration also changes within thematic presentations as well.
  3. Its form is essentially a theme and variations, and just as in the best examples of this form, our interest is sustained by hearing many permutations of the main theme, instead of becoming annoyed that a given theme is played over and over again. The unusual aspect in Boléro is that the pitch content of the theme is never varied (except for a modulation in the final 40 seconds), just the colour (orchestration and texture) and dynamics. 
  • This idea is not original to Ravel, however; another mono-thematic work that begins quietly and, over the course of multiple thematic repetitions, eventually becomes very loud, is Grieg's "In the Hall of the Mountain King," from Peer Gynt (1875), composed 53 years earlier.
If Boléro consisted only of 339 repetitions of a short idea, and nothing else, I'll go out on a limb and speculate that the composition would be less popular.  😄

Or perhaps it would have become celebrated as the grand-daddy of minimalism… 😴

Igor Stravinsky: The Rite of Spring
Igor Stravinsky's The Rite of Spring (1913) makes extensive use of ostinati, and, more generally, repetitive elements, so today we will look at some of the ways Stravinsky used one particular ostinato, the repeated four-16th-note idea in the excerpt below.

It is introduced at the end of the first section (Introduction) of Part 1, played by pizzicato violins, while a clarinet sustains a lengthy trill (this is at about the 3-minute mark in the YouTube video at the end of this post; all other timings indicated below refer to this video):

It is a pretty simple idea by itself, but its simplicity makes it useful, because it works well with many other ideas. Therein lies one of the keys to writing a good, multi-purpose, ostinato: While good music can be written with longer and more complex ostinati, a short, simple idea is probably more flexible, because it can support a variety of other ideas.

The above excerpt introduces the ostinato somewhat hesitantly, instead of introducing the idea and immediately continuing with almost incessant repetitions, as we found (see previous post) in Boléro. Stravinsky is saving more regular repetitions of this idea for the next section. He is "planting a seed," giving us a taste of an idea that will become increasingly prominent and repetitive.

Sure enough, about 10 seconds later, he gives us a somewhat longer version of the ostinato, which leads to the strongly-rhythmic bitonal (E (enh.) in the lower strings, Eb7 in the upper strings) chord repetitions that are one of this work's most memorable features, which mark the start of the next section, Augers of Spring/Dances of the Young Girls. This starts at 3:20 in the video below:

Incidentally, for those interested in "golden mean" ratios, note that in its first three presentations, the ostinato is heard twice, once, (thus three times so far), and Five times, adding up to eight times, all of which are numbers in the Fibonacci series (1, 2, 3, 5, 8…). Sadly, the next time we hear the idea (see excerpt below) it is only presented four times, which is not a Fibonacci number. 😢 It is used eighteen times consecutively the time after that (part of which is two excerpts below), and this too is not a Fibonacci number. 😭 This is a work clearly in need of a revision! 😎 [Okay, I promise to insert no more emoticons in today's post. The are very useful, though…]
Very soon after the previous excerpt, at about 3:35 in the video below, the heavily-accented repeated chords are interrupted, and we hear the first presentation of the ostinato figure with counterpoint, consisting of an arpeggiated chord in the bassoons, which is itself repeated. Notice that the bitonality continues, this time between the ostinato (Eb7) and the counterpoint below it (E enh.):

This next excerpt, which starts at about 3:45 of the video, gives us another melody, once again with repetitive elements, this time above the ostinato. Note the frequent colour changes in Stravinsky's orchestration of the melody; the reduction below does not accurately reflect this, but there are 6 colour changes to the melody in 6 bars. When you listen to the full orchestra version (in the video below), note as well that the ostinato at this point is buried in the texture, almost inaudible within the heavily accented chord repetitions (not shown in this example, but the chords are the same as in the last 2 bars of the previous example):

New melodic fragments are superimposed on the 4-note ostinato over the two minutes that follow the previous excerpt, one of which is this one, which is again repetitive, and occurs at about 5:25 of the video:

Please suggest other works that make prominent use of ostinati – one such piece is Stravinsky's, L'Histoire du Soldat – and I will possibly (see explanation at the end of this post) discuss them in future blogs… At the very least, I can compile a list with your suggestions. If you can, try to be specific about where the ostinati occur within the work you are citing. In the mean-time, I already have an idea for a third post in this series.

Below is a recording of The Rite of Spring on a video that shows the score. Try to find other uses of the ostinato discussed above. As well, try to listen to it more than once, in order to find other ostinati, and the degree to which repetition of musical ideas is used. If you don't have enough time to hear the entire piece, listen to at least the first 6 minutes; all of the above examples occur within that time.

Explanation of my use of the word "possibly" with regards to doing more blogs on this topic:

My only hesitation is that the amount of time involved doing a post such as this one is daunting. It involved making score reductions in Finale of the sections I wanted to use as examples, saving them as GIFs and importing them into this blog, recording the musical examples, importing them into Audacity and splicing them into sections that corresponded with the notated examples, saving them in two different audio formats (MP3 and OGG) because not all web browsers read MP3s, uploading the audio files to my website, and then inserting the code that lets the Blogger website read and play audio files on all major web browsers. Plus the time spent fixing things that didn't work along the way. Not complaining, mind you! I enjoy doing this, but it is time consuming.

That said, and to repeat what I wrote above, I do have another post in the works on this topic, so there will be at least three in the series, and possibly more if people make suggestions regarding other works with prominent use of ostinati.

Thursday, March 30, 2017

Ostinatos; making a lot from a little (1. Boléro)

An ostinato is a musical idea that repeats immediately (as opposed to returning later in the composition) and persistently (it usually is repeated more than once). It can be melodic or rhythmic, and is usually fairly short – one to four bars – but it can be longer. Kids love 'em.

And not just kids; it is widely used in many musical styles and periods.

The attractiveness of ostinati for composers is easy to understand; you can generate a lot of material from a relatively short musical idea, and, if you do it well, audiences may respond well to the music.

With the advent of computer-notation software, and, more specifically, the "copy" and "paste" commands in those programmes, it has become extremely easy to use ostinati in compositions. And, with programmes like GarageBand, which comes bundled with every Mac computer, you don't even need any musical knowledge to write loop-based music; in this context, "loop" and ostinato mean the same thing.

The downside of repetition, however, is that too much can make a composition overly predictable, unless the composer finds ways of varying, interrupting, growing, evolving, or otherwise adding interest to repeated patterns; music that is overly predictable can lose the listener's interest.

An example of how to successfully repeat an idea to an almost absurd degree is Ravel's Boléro. It uses the two-bar rhythmic ostinato figure below throughout the work; this two-bar rhythmic unit never stops repeating until the work's (very loud) conclusion, about sixteen minutes later:

There is even further repetition within this two-bar ostinato: The rhythm on beat one is used on the first two beats of each bar; four of the ostinato's six beats are identical. This is repetitiveness ad absurdum, and I won't stand for it!!! [Just kidding, of course; the piece is awesome.]

The above pattern is repeated 339 consecutive times in Boléro (yup, I counted), which means that the rhythm on beats one and two of each bar is heard 1,356 times.

That's a lot of repetition!

One can argue that the uninterrupted repetition of the same short rhythm for sixteen minutes in a composition is a bit much – or a lot much – but Boléro is Ravel's most popular piece, so clearly, millions of people have no issue with it. Indeed, its popularity may in part be due to this rhythmic ostinato!

So, the question I have for you is this: What makes it work? What does Ravel do to keep our interest despite the 339 ostinato repetitions? Why do audiences cheer enthusiastically following the conclusion of the work, rather like sports fans cheering an exciting overtime win by their favorite team, instead of standing up to boo the repetitiveness?

I once listened to a radio documentary on Ravel's Boléro in which orchestral musicians were asked to give their thoughts on the work. Many said that they don't look forward to performing it because they perform it so often, there is such a high degree of repetitiveness, and, in some cases, once the piece starts they have to wait an extremely long time before they get any notes to play. However, once  rehearsals begin, they gradually feel their resistance melting and become ensnared by the hypnotic power and beauty of the work, to the point where they feel like standing up and cheering along with the audience after reaching the triumphant final chord.

Please share your thoughts in the comments section below.

Here's a performance of Boléro, conducted by an unshaven man with a toothpick instead of a baton,  if you wish to have a listen:

Tuesday, January 24, 2017

Music Notation Software – Pros and Cons for Composers

There are several music notation programmes available for computers and tablets, but the brands that I suspect most composers use are Finale and Sibelius. Another excellent notation product is Notability Pro (for Mac only), which some composers I know swear is the best product out there, and is now free.

In today’s post, I will explore how the use of notation software can affect the composition process, in ways we may not realize, both positive and negative. The first seven points below list many of the unambiguously-positive aspects of using notation software, and the remaining points concern some of the potential challenges that can arise from its use, some of which we may be unaware of.

If any readers can think of pros and cons not listed below, please let me know via the "comments" area, and I'll add them to the list if merited.

Music Notation Software
Cons (or commentary on pros)
1. Can produce polished, publication-ready scores. 1. This is indeed true. However, it takes considerable skill, and the knowledge of all the minutiae of notation conventions, to produce publication-ready scores, whether one uses notation software or not.

Because a score produced with notation software generally looks far better than a hand-copied score (although some highly-skilled hand-copyists can also produce beautiful scores), we may be seduced into thinking our score is as good as it needs to be, when in fact it may need a lot more detailed work to reach a point of being truly publication ready.
2. Scores look better than hand-copied scores. 2. This is generally true. I doubt that many students are trained in the art of hand-copying music any more (I was, which reflects the period in which I was trained (cretaceous)) – but it isn’t always true; a sloppy computer-notated score looks far worse than a beautiful and meticulous hand-copied score.

Again, this is more a product of the user's limitations than of the software, however.
3. Pitches and accidentals are notated clearly and correctly.

3. No disadvantages! Here are some of the reasons this is such a valuable advantage for notation software:

• Sometimes, in hand-copied scores, pitches are less than 100% clear because they are notated in such as way as to "spill" into the territory of an adjacent pitch.

• Notation software also notates 2nds correctly. Sometimes, in a hand-copied score, students attempt to vertically align notes that are a second apart, which looks very messy.

• Notation software also aligns accidentals correctly (again, students sometimes try to fit them on top of each other, causing collisions and all manner of visual mayhem).
4. Other score information, such as text and articulations, is clear (hand-written text can be somewhat challenging to read if a composer has poor calligraphy skills). 4. This again is generally true, unless the composer uses a font or font-size that is difficult to read.
5. You can remove or add bars without recopying entire pages.

• It is also easier to change/add/remove notes and any other score information (such as dynamics, slurs, articulations, text, etc.

• Software also lets you do A/B comparisons, listening to a version with bars added, and then comparing with a version with those bars removed.
5. This is a huge advantage of notation software; having to recopy an entire page by hand in order to add or subtract a few bars is such a hassle that it can become a disincentive to make such changes. Anything that gets in the way of making even small improvements in your compositions is a significant problem.

•And yes, the possibility of doing numerous A/B listening comparisons is a tremendous advantage in using notation software.
6. Parts can be generated automatically.

• This is a huge advantage in using notation software.
6. Generating parts can still involve some work, however, because you may need to adjust the layout, number of bars per system, fix any new collisions that may have shown up, plan page turns, do any necessary last-minute edits (you sometimes notice problems in parts that you didn't notice in the score), etc. But there's no question that generating parts is a much faster process with notation software.
7. Transpose, Invert, Retrograde, and other commands, as well as plug-ins.

• Did you know that Finale has commands for melodic inversion, and retrograde? These (particularly inversion) can be useful when considering possibilities of how to grow/extend/transform a melodic idea.

• There are also third-party plugins available, such as Patterson Plug-Ins for Finale,  which are designed to speed up and generally improve workflow.
7. Composers can obviously do these things without a computer, but the computer does them much faster. Plus, having these options so readily available makes it easier to try them in order to see if they can be used in your composition.
8. Dynamics look as they should, and are usually well positioned. 8. Notation software does indeed produce dynamics that are beautiful.

• They are not always well positioned however; in Finale, you have probably found many cases where a dynamic collides with something else, such as an accidental, note, or slur, which requires the user to re-position the dynamic, or the other objects with which it collides; I'm not sure this happens as frequently in hand copied scores.

• One potential issue to be aware of is that in some software programmes, a dynamic intended for one instrument (e.g., below the flute staff) in an orchestral score can show up in an adjacent instrument's part (e.g., above the clarinet) when parts are generated. When positioning a dynamic, Finale uses a temporary dashed line to indicate the note to which the dynamic is attached, which reduces the likelihood of misplaced dynamics.

• There are potential playback issues in the use of dynamics, described in section 9.
9. You can hear what you write as you write it, performed at the indicated tempo, or at a slower tempo if you prefer, which allows you to listen repeatedly, carefully, and critically.

• You can also hear and evaluate any indicated tempo changes (including rit. and accel.), and dynamic levels (including cresc. and dim.).

• You can also listen to the composition, or a section thereof, repeatedly, tweaking it until it sounds as good as you can make it, no matter what time of day you play it, and no matter what your mood is.
9. Being able to hear an approximation of what you write in real time is a huge benefit of notation software.

• There are, however, significant issues or limitations in relying too heavily on MIDI playback as a realistic indicator of what your music will sound like; these include:
  • Unwittingly writing parts that are either extremely difficult or even unplayable, because the computer plays them without any problem whatsoever (!). A computer plays unidiomatic lines flawlessly, while a performer might struggle in attempting to play them, or even refuse to play the piece. The computer can lull the user into thinking that the line is perfectly idiomatic, when in fact it is extremely difficult or even impossible. I am not sure how much different this would be in a hand-copied score, but in producing a hand-copied score, a composer usually spends hours playing each line, usually on a piano, which might flag any such issue;
  • Balance problems: The balance in a MIDI ensemble is often not very realistic; 
  • Further to  balance problems, sometimes, in an attempt to bring out a line that is insufficiently prominent, we may temporarily give it an extreme dynamic boost, such as marking it fff instead of f, so we can hear it better in the MIDI playback, but then forget to change the dynamic to its correct value (fff back to f) before giving the parts to the performers, resulting in performers blasting the heck out of that line in the first rehearsal, when all we intended was for it to be more prominent than the lines around it. Or sometimes, an inexperienced composer may use an extreme dynamic boost (e.g., f to fff) intentionally, thinking it necessary to bring out the line to the desired level, perhaps not realizing that if a line is marked  f, while the other instruments are marked mf, the performers and/or conductor will make sure that the f line is heard more prominently than the others.
  • It is also possible that the previous example (extreme dynamic boost) might be the result of poor orchestration; if a musical line is insufficiently prominent in MIDI playback, perhaps it needs to be reinforced in some way (e.g., octave doublings, or the addition of other instruments to that line), or perhaps the material around it is too busy and needs to be thinned out in some way.
  • Unrealistic representation of the nuanced colour and dynamic changes in different registers of an instrument or voice; 
  • MIDI playback is only as good as the quality of the samples  in your computer's sound-bank. 
  • Glissandi, heard through MIDI playback, usually elicits a chuckle from class members, presumably because it often sounds so unrealistic or even ridiculous.
  • The computer will play any glissando, even impossible ones, which may entice composers into writing impossible glissandi. We need to be aware of the possible glissandi for different instruments; always show your work to a performer of that instrument to be sure.
10. Copy and Paste.

• Musical material, from the smallest ideas to entire sections, is often repeated, either immediately or brought back later; the Copy & Paste functions let you do this with great ease.
10. Again, a very useful tool. I recommend exercising some restraint in its use, however.

• One of the most wonderful attributes of great classical compositions is that ideas are often altered in some way when repeating or recapitulating them. This provides both the comfort of familiarity, since we recognize the ideas, but also an element of surprise, if we recognize that some aspects have been changed.

• You can make such modifications when repeating ideas in notation programmes, of course, but, at least in student work, it seems as though the ease with which the paste command can be executed often leads to not making modifications.

• My advice to students is to explore modification possibilities when re-using (pasting) an earlier idea into a later section.
11. Other limitations and challenges. 11. Using different metres in different staves simultaneously, and having bar lines that don't necessarily line up with each other (vertically).
12. Other limitations and challenges. 12. Using a time grid at the top of your score (e.g., a grid in 5 second increments), with no bar lines.

• You can hide bar lines, of course, and create a graphic to represent the time grid, but this involves more work than it would if done by hand.
13. Other limitations and challenges. 13. Graphic notation can be difficult or even impossible.

• Again, you can create graphics on a computer, but it takes some skill to do this well, and doing it by hand is often faster.
14. Other limitations and challenges. 14. Oversize metres in orchestral scores (e.g., a large 4/4 that spans the height of the entire woodwind section) are either impossible or very tricky. Oversize metres are generally much appreciated by conductors, because they can be easily read at a glance. When my orchestral music has been played, I often get the score back with oversize metre changes written in by the conductor.
15. Other limitations and challenges. 15. Unless you invest in an expensive sample library that includes extended techniques in all instrument families, your MIDI playback will probably not be able to reproduce such sounds. This is not necessarily an impediment to using extended techniques, but I suspect they would be used more if we could hear a reasonably-accurate reproduction of these techniques during playback of our scores.

These performance techniques include: col legno, col legno battute, sul pont., sul tasto, different mute types for brass instruments, hand-stopped notes (for horn), play with bells in the air, multiphonics, flutter-tongue, harmonics, harmonic gliss. ("seagull effect") for strings (particularly for cello), senza vibrato, scraping sound created by heavy bow pressure and slow bow speed, a myriad of sounds available by slapping, scraping, muting, picking (with a guitar pick) strings inside of a piano, prepared piano sounds, etc.

Have I missed any significant advantages or disadvantages in my list? Are there times when you feel the notation software is pushing you to notate an idea in the way that it wants, as opposed to the way that you want? Please let me know in the comments section below, and thanks for reading!

Thursday, January 12, 2017

Recommendation Letters – How Students are Evaluated

At the end of last semester, from about mid-November to mid-December, it seemed to me that I was writing more student recommendation/reference letters than usual — which is fine, of course — but, as I was doing so, it occurred to me (as it does every year) that students requesting such letters fall into different categories. There are students who are good at everything, students who are good at some things, pretty good at others (e.g., good at composition, okay in theory), students who are pretty good at 1-2 things, rather weak in others, students are are really good at one thing (e.g., composition) but really weak at others (e.g., theory), and every permutation of these and other variables you can imagine.

If I am asked to write a letter, the academic areas on which I comment are the ones that I teach, namely music theory  and composition. But, as you can see in the list below, professors are often asked to rate students in a surprisingly-large number of areas that go far beyond specific academic disciplines, and so I thought it might be useful to list these areas, and, more generally, provide information on the evaluation process for anyone that may wish to request such a letter at any point in their future.

If you are in a Bachelor of Music programme and there is even a remote chance that you might apply to graduate school upon completing your degree, today's post may have some useful information for you. And, if you have no interest in this topic, I promise to get back to composition-related topics in my next post!

Asking for a reference letter
 • Don’t be shy about asking a professor to write reference letters; it is a part of our job, like correcting/grading, office hours, extra help for students that seek it, committee work, research/composition/performing, and of course, teaching.

• Be aware, however, that it takes time – for me, typically about 1-2 hours – to write a thorough reference letter.  

Why does it take so long?
Part of that time is spent going through a student’s transcript carefully (which I request from the student), reviewing how the student did in my courses (which involves opening the spreadsheet files for every course taken with me in order to see the student’s results in all parts of the course evaluation scheme), and ranking that student’s final mark in relation to the class (e.g., 3rd out of 34 students), because some graduate schools ask for that information, and if they don’t, I usually provide it anyway.

Time is also spent listing all of the positive attributes I can think of for that student, particularly those that I believe would be relevant to studying composition or theory at the graduate level.

I also list any weaknesses or concerns I have about the student, because admissions committees want a balanced and honest assessment; if I describe a student in only positive terms, but the transcript reveals that the student has a 75% average over the most recent 1-2 years, then it seems unlikely that an admissions committee would take my letter seriously; a B+ average is nothing to be ashamed of, but it’s not an A average, and, in my opinion, a professor's recommendation letter should expand on the evidence found in a student's transcript, without making claims that are unsubstantiated in the transcript.

What Student Attributes are Evaluated?
Professors are often asked to rate, and comment on, specific personal attributes or character traits, such as:
  • Ability – All around
  • Ability to work under pressure
  • Ability to work independently/Willingness to take responsibility for learning
  • Academic potential
  • Academic preparation/Demonstrated academic ability
  • Achievement
  • Communication skills – Oral
  • Communication skills – Written
  • Curiosity – Willingness to explore new approaches, ideas, particularly those that may fall outside the student’s comfort zone
  • Industriousness
  • Initiative
  • Intellectual capacity
  • Leadership
  • Likelihood that my institution would accept the student, if we had a comparable programme
  • Likelihood that the student will complete the degree
  • Organization - Planning
  • Originality/Willingness to “think outside the box”
  • Creativity/Talent
  • Persistence 
  • Punctuality
  • Research ability
  • Response to constructive criticism
  • Teaching ability (if known) 
  • Teamwork (work well in a group)
The rating system varies, but it often consists of a series of clickable boxes that correspond to percentile values, such as top 5%, top 10%, top 20%, top 50%, lower than top 50%, and "no basis for judgment." The professor clicks on the appropriate box, in their assessment, for each category.

Universities typically ask for a letter, in addition to asking the professor to rate the student in a number of categories, as above, but no university with which I am familiar asks for the student to be assessed in every  category listed above; the above list is compiled from a variety of different universities' lists. Also, some universities have comment boxes in their on-line recommendation forms that allow professors to add a sentence or two explaining our rating in that particular category.

So What?
I have no idea how much weight is placed on this aspect of the evaluation process – I suspect that it matters, but perhaps not as much as the student’s transcript, composition portfolio, and the recommendation letter, but that’s just my guess. At the very least, however, these categories tell us something about what graduate schools value in prospective students.

How would you assess your level in each of the above categories? Most people have at least a few weak areas; if you can identify your own, I encourage you to work at them so that they become strengths.

How to request a letter
When requesting a letter from a professor, give them lots of lead time, and provide all necessary information. More specifically:
  • Make your request at least two weeks in advance, because of the time it takes to write these letters. If this is not possible for some reason, then at least one week in advance, but be aware that professors are generally pretty maxed-out in terms of demands on their time, and if you don’t give us enough time, we may not be able to write the letter you request.
  • Make your request in writing, as well as in person (if feasible). The reason for this is that I don’t necessarily remember every request, but if you put it in an E-mail, then I have something to remind me. Confirming this request in person is mostly a social nicety, but it may also give you a sense of the degree to which the professor is willing to write you a positive letter.
  • If you are asked to include a composition portfolio, ask your composition teachers for their opinions on what to include, and make as many of the changes/improvements suggested by your instructor as possible. If applying for a theory master's and they require the submission of a theory essay, ask your theory teachers for suggestions and corrections on any essay you plan to submit.
  • Provide the professor with an up-to-date transcript if you can, for reasons described above. The transcript can be a series of screen shots taken on your computer from your “university self-service” portal, or a PDF of the actual transcript.
  • Provide a list of every course you took with the professor, and the semester in which you took that course.
  • List any relevant achievements that might strengthen your case, e.g., Senior Rose Bowl Winner, Kiwanis Festival of (what region or city), Gower Band Terra Nova Competition Winner, etc., with the date(s) of the awards. You could include non-musical achievements as well, e.g., cycled from St. John’s, NL, to Victoria, BC, April-August, 2014... If I think they reflect  strengths in your character, I might mention them. And yes, one of our students did this!
  • Indicate in writing every school for which you’d like a letter.
  • Include the submission deadlines for each school.
  • Indicate in writing the specific graduate programmes to which you are applying (e.g., MA in music theory, MMus in composition, etc.).
Electronic Submission?
Most schools allow professors to submit their letters electronically; I much prefer this, so if that’s an option, please choose it. I would guess most professors find E-submissions most convenient, but double check with the professor to see if they prefer to submit the old-fashioned way (i.e., a hard copy via "snail mail").

On the pros and cons of “Gentle Reminders”
To the best of my knowledge, universities automatically send applicants an electronic confirmation message every time they receive a recommendation letter for that student (UPDATE: Maybe it's a glitch in their systems, but they occasionally don't, according to a student who just contacted me… However, I think they are all supposed to do this). If the deadline is near, and you have not received such a notice, here’s what I suggest:

Send the professor a “friendly reminder,” or “gentle nudge” a few days before the submission deadline . For professors, work life tends to be pretty hectic all semester, but particularly so towards the end of a semester. Also, if you have, say, eight students who each want three reference letters, that’s twenty-four references; that's a lot of references, and it is possible to miss some. Not only that, but E-mail is a somewhat inefficient way to communicate, in that every day our inboxes are bombarded with SPAM and “noise” (messages that are not SPAM, but clutter up our in-boxes), so it is easy for the occasional legit E-mail (one that requires a response) to get buried among the debris.

I know that not everyone appreciates reminders, however – I remember as an undergraduate student seeing a professor in the hallway one day, and reminding him as politely as I could that the deadline for that letter was that day, only to have the professor blow up at me and proclaim, in an outdoors (not indoors) voice, “Look! If I say I’ll write you a reference letter, I’ll write you that g**d*** letter!”

It occurred to me that I might have destroyed any hope I had of going to graduate school, which I found rather daunting at the time. In today’s world, however, this kind of situation can be mostly avoided, because a thoughtfully-worded gentle reminder via E-mail is unlikely to elicit such a rude response. Speaking only for myself, I don’t mind them at all.

Trust the Process; How do you know the professor will write a good letter?
Speaking of daunting, it can feel a bit scary to request a letter and then trust that the professor has written a good letter. But, once your request has been made and the professor has agreed to recommend you, I would suggest trusting the process. In my opinion, a conscientious professor should tell you if they can’t write you a good letter, rather than agree to write a letter, and then write only negative things about you.

I have turned down requests to write recommendation letters on very few occasions — when I do, I always suggest that the students find other professors to write recommendation letters —  for one of three reasons:

(i) The student's request came too close to the submission deadline. This is just a practical matter, not personal — requests often come at the busiest time of a semester, and there are times when I cannot set aside all the other work I have to spend two hours writing a letter that is due the day after I got the request. On the other hand, if I have recently written a letter for you, and you ask me to send it to more universities a day or two before a deadline, this is less problematic. It still can take a little more time than you might expect, particularly when the original letter needs to be modified in any way (such as if the original letter needs to be updated due to new information), but if I can do it, I will. But I'd still prefer 1-2 weeks notice, of course;

(ii) I don’t know the student well enough to say anything really positive about her/him; e.g., they only took one course with me, and it was a few years ago, and the student did not get a very good result in that course;

(iii) I know the student pretty well, but what I know isn't particularly positive… Perhaps they took several courses with me, but did not do very well in any of them, or they were rude/disrespectful on multiple occasions, or they skipped or were late for classes on multiple occasions, or they missed assignments or quizzes several times, or they committed academic dishonesty (the most common example of this is students collaborating with each other on assignments), or they generally disregarded my compositional advice and suggestions, etc.

To be clear, I’m not talking about an occasional disagreement with a student; these are not unusual in teaching, and I don't mind them, as long as the discussion is respectful. I’m referring more to an on-going and troublesome pattern of behaviour, which is extraordinarily rare; I can think of only about three such cases in thirty years of teaching.

The only time I recall declining a recommendation letter request for reason (iii) above involved a student who was extraordinarily sullen and rude. Don't get me wrong — I have encountered occasional instances of rudeness over the years, but this particular student really stood out from the rest (and not in a good way!), so much so that I actually had other students complain to me about the student’s behaviour. I was therefore surprised when the student asked for a recommendation — I'm not sure what sort of karmic good-will credits they thought they were accumulating while engaging in disrespectful behaviour — and so I suggested they would be better off asking someone else for a recommendation.

I don't hold grudges, or try to undermine students in any way. Had I decided to write a letter for this person, I would have found positive things to say about her/him (no matter who you are, I can usually find positive things to say about you, and this student was definitely smart and talented), but I would have also had to mention the problematic behaviours, which I felt would have sabotaged their chances of getting into a graduate programme, which is why I suggested they ask someone else.

Not sure about applying?
If you want to go to graduate school but are thinking that perhaps you ought not to apply because your marks aren't great, or your composition portfolio is not strong enough, I would suggest applying anyway, because you never know what the outcome will be. Sometimes, average students (IMO) have been accepted into master's programmes, while stronger students have been rejected. You don't know the circumstance or criteria used in the evaluation of your application; in some years, and in in some universities, you might be in competition with sixty other students; other years, you might only be in competition with eight other students for six available spots.

If graduate schools reject you, and are not willing to give up on your dream, find a way to turn this into a positive experience: Identify your weaknesses, work hard on improving in those areas, and then apply again. For composition, this might entail writing more pieces, perhaps longer and more substantial, perhaps for larger ensembles, or written more idiomatically for the instruments involved, or exploring new techniques, or providing more score detail, or better recordings, or just writing better pieces; most of us acknowledge that our early compositions are not as good as later compositions. If you had some weak results in music theory classes, explore the possibility of taking some of them again, this time working harder than you did previously, in order to gain a better understanding of the material.

Types of applicants
It is easiest students to write strongly-positive recommendations for students with these qualities:
  1. At or near the top of every composition course they took with me;
  2. Developing into very good composers — this is slightly different than #1, because a students course grade can be near the top of the composition class by getting full marks in all the extra-compositional assessment areas — excellent preparation, excellent contributions to class discussions,  making the required number of thoughtful composition blog comments — while producing compositions that are pretty good (B+, A-), but not extremely good (A, A+);
  3. At or near the top of every theory course they took with me;
  4. Actively seek out new scores to study and new music to hear, and are excited by new discoveries;
  5. Willing to try new things without pre-judging them, and to embrace at least some of them;
  6. Do all their work on time with a minimum of excuses;
  7. Do more work than most other students;
  8. Respond well to compositional suggestions,
  9. Have a great attitude, and
  10. Participate regularly in class, and, in particular, they are good at giving constructive criticism to fellow students in our composition seminars.
If I were assigning a score of 1 (= terrible!) to 10 (= whatever is better than excellent!) to each of the above qualities, I feel confident in saying that I have never had a student who I would have rated as a 10 in every category.

However, I have had occasional composition students who are strong in all of these areas, meaning I would perhaps rate them as an 8 or higher in every category. Such students tend to be accepted into graduate programmes, provided they didn't only apply to the most selective universities, like Indiana or Yale (but our students have gone to Indiana University and Yale, for example, so why not you?).

What do you do if you are pretty good at composition, but pretty weak in, say, theory?
 Most students are strong in some of the areas, but not all, and in many cases such students have been accepted into graduate programmes. Some students are very good at composition, but very weak in music theory; this can be a deal-breaker for some universities, since a graduate programme in composition usually involves music theory courses as well. I still recommend that such students apply to graduate schools, and, if not accepted, then consider doing extra work in music theory (like re-taking music theory courses to try to obtain better results), and then applying again in the following year. Realistically, however, this does not give students much of an opportunity to improve their standing in theory courses, since applications are often due by the end of the fall semester, which would give a graduating student only one semester in which to demonstrate an improvement.

If you are reading this in the early years of your academic studies, the clear conclusion to draw from this is to become as strong as possible in every possible area. It starts with taking every course seriously, and not falling into the trap of making excuses for poor results in, say, theory courses, such as, "this is nothing but a set of rules for OLD music; I want to make NEW music, and break any rules I wish!"

I will confess to speaking from experience on this matter. When I started taking music theory, I could not relate to it, and thought it had nothing to do with the music I wanted to write. Consequently, I failed grades 3 and 5 harmony at the Royal Conservatory of Music (RCM), and didn't do very well in grade 4 harmony either. After about two years of poor results at the RCM, I started studying privately with an excellent composition teacher, Dr. Sam Dolin, who made me redo music theory from the beginning all over again, and, while I resented it a bit at first, he made it interesting, fascinating, and relevant for me, and my love for music theory – which is, basically, learning how music works – has never abated. And yes, I got better at it.

You can get off to a bad start in something, and end up becoming very good at that thing; in my case I just needed a combination of a good teacher and an attitude readjustment.

This has become much longer than I anticipated, but if you have questions on anything relating to recommendation letters that I did not address in this post, please ask them in the "comments" section, and I'll try to respond.