…on singing and playing

In the past, I’ve had a tendency to dismiss or dislike singing and playing, but more recently I’ve come to enjoy it and even start to encourage composers to make use of it. I’m going to slowly rebuild this post to reflect those views, but I think the best example of how well singing and playing can work can be found by listening to/watching Thierry Tidrow’s Clarinetabile, in which I sing and play throughout – what I started to love here, is how radically different the effect of singing and playing different intervals can sound.

I would also advise clarinetists who are starting on this that, in order to avoid pain they should a) not do too much singing and playing at once, b) make sure they have a very open position (a kind of ‘yawn’ feeling in the throat) and c) if there is a lot of singing, do it in a range that’s most comfortable. Tidrow’s piece for example is actually generally written much, much too low for my high singing voice.

I’m going to try to do this in as organised a way as possible in the hopes you can find the information you’re looking for, so let’s consider the voice from three different perspectives: Dynamic, Pitch and in Combination with other effects. I’m also moving back and forth from Eb to Bb to Bass clarinets, and not always being too specific about when I’m playing what. Everything works on all the clarinets, it’s more important in the case of singing and playing to consider the vocal range of whoever you’re writing for!

(If you want to know about singing into the clarinet without mouthpiece, you should visit this post.)


Right, to get us started, a few examples. Here’s a nice long crescendo and decrescendo, singing and playing roughly the same pitch (we’ll get to the specifics of pitch later). First on Bb clarinet on a written middle C:

And then the same again on the bass, on a written middle C and the low C (works quite well when I’m singing an octave or two above!):

Singing and playing at loud dynamics is an easy way to create a lot of distortion. The following example is from Ann Cleare’s eyam i for solo Bb clarinet:


And some more loud noodling, slightly random, but hopefully gives you an idea of some of the possibilities:

It’s best, if you really want to hear the voice, to keep the dynamic relatively low in the clarinet part, and I actually think that low dynamics with singing sounds great and isn’t used often enough. There is an example of this with sung high pitches in the Evan Johnson example in the Pitch section, and one with air sounds in the Combination section, but here’s some noodling to give you a general idea of soft dynamic singing:

I was trying to play around with singing and playing at different dynamic levels, but it doesn’t really seem to be possible. Then again, I’m not sure. Perhaps someone else can do this!


There are a lot of ways that you, as the composer, can define how the performer sings while playing, and some things are of course easier than others. If a composer just writes (+ voice) above a passage, I’ll usually sing something close to the pitch I’m playing (if within range), but not too close, imagining the composer is looking for this interference, especially if the passage is written at a louder dynamic (which it seems usually to be). That will sound something like this (here I’m noodling around quite at random, following the curve of the melodic line on the clarinet with my voice):

You can also ask the performer to (sing a low note)/(sing as low as possible) or to (sing a high note)/(sing as high as possible). Kagel uses the former in his Schattenklänge:

And Evan Johnson uses the second to great effect in his Supplement. The difference with singing as high as possible is that it tends to distort the pitch material a bit, I imagine through the more extreme tightening of the vocal cavity, throat, etc.:

photo 4

Of course one should take note of the vocal range of your clarinetist.

One can, of course, do something more complicated, and write separate lines for the voice and clarinet. This takes some practise to get used to, but as an example, one of the things Ernesto Molinari taught in his masterclasses was to sing in canon with yourself, as in the following example. What keeps this from being overly difficult is that of course I play the pitches before I sing them (so no danger there), and the sung and performed pitches are within a very confined and easy register:

Singing along with fast lines also works relatively well. I can’t sing quite as fast as I can play, but that doesn’t mean the effect isn’t still present, as Aperghis writes in his 280 Mesures:

photo 2

There are a few more effects one can achieve with glissandi. For example, one could start on the same pitch and have the voice and clarinet glissando in different directions (here on Bb):

Or, if one plays a relatively high note, then sings glissandi up and down without bending the clarinet pitch, one can hear the difference tone moving in the opposite direction (again on Bb – if you have a clarinetist who is a man, this might work better on the bass?):

I also think this sounds great on the Eb clarinet:

Something that I do find very difficult to pull off is when the composer asks for a sung pitch and for the clarinet to enter afterwards. The problem is that in order for the clarinet to speak, I have to maintain the same pressure on my vocal chords so that the transition is effective. And it just doesn’t sound very pretty. That said, if this is the sound you’re going for, fantastic. The following is from Chardlie Sdraulig’s piece, never mind, for bass clarinet solo:



There are three techniques I want to demonstrate in combination with singing and playing: flutter tongue, multiphonics and air sounds. (Preparation on the bass clarinet is another one, but I’ve already covered that to some extent)

These techniques are influenced of course by things like dynamic and pitch. For example, there’s a big difference between singing another note to create a chord in a close dyad multiphonic and screaming with something more spectral (as Aperghis writes in his 280 Mesures):

Of course, air sounds and loud dynamics don’t really go together, but I think soft air sounds and sung pitches definitely do:

And finally, one can easily accumulate effects. Here’s a low C (transposed pitch) on the bass clarinet. One of the things I love about these effects is how well they demonstrate the destabilising power of the voice, especially when combined with spectral multiphonics (actually, a great general example from the repertoire of how the voice can do this – and a great example of many other things as well – is the entirety of Evan Johnson’s Supplement, which you can hear here).

1) Low C alone + flutter tongue + singing:

2) Low C alone + spectral multiphonic + singing:

3) Low C alone + flutter tongue + spectral multiphonic + singing:

This entry was posted in air sounds, dyads, flutter tongue, glissandi, multiphonic, singing, spectral multiphonic. Bookmark the permalink.

4 Responses to …on singing and playing

  1. Pingback: Index / Table of Contents (scroll down for recent blogposts) | heather roche

  2. Hugo says:

    Hi Heather,

    This comment is not about any advice I could ask, because I’m not a clarinet player (but I would like to). Actually I’m a student in musical acoustics and I’m currently doing an internship at the acoustics lab in Le Mans (France – Yeah btw I’m French, so I apologize for any English mistake…), where I’m working on digital simulations of multiphonics in the clarinet. This comment is just to thank you, because there really isn’t much informations on the net about singing while playing (which is the basis of my work), but your website is the first one I got on and there are lots of details about it, which kind of helped me in my research work !

    So yeah, thank you :)


  3. Pingback: …on clarinet articulation | heather roche

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.