An interview with Urs Liska

Urs Liska is the one who suggested “Scores of Beauty” as the name of our blog, so he has a special place in our hearts 🙂

Janek Warchoł: Hello Urs, I’m glad to have this opportunity to speak with you. We’ve already seen one post of yours (I hope that more will come :)), and we’d like to know something more about you.  Could you please introduce yourself and tell us how you had learned about LilyPond?  What was your reaction when you used it for the first time?

Urs Liska: Hello Janek, I’m honored to have ‘won’ this special place through that contest. Not because of being a central figure in the LilyPond universe (which I’m not), but maybe my background is somewhat representative for one approach to LilyPond and therefore qualifies for this interview 😉 

It’s always difficult to describe ‘who I am’ or what my profession actually is (because there are several quite disparate ‘fields’ to mention). Originally (and most professionally) I’m a classical pianist with a strong focus on two seemingly unrelated areas: classical ‘lied’ interpretation and contemporary (and electronic) music. Probably my most worthwhile achievement so far is the first complete recording of Arnold Schoenberg’s songs that was released last year ( and On the other hand I’ve studied music theory and musicology (unfortunately two separate academic disciplines in Germany) and gradually drifted towards philological realms, e.g. ‘creeping’ into manuscripts of, say, Schubert or Schoenberg and trying to get as much information out of them as possible.

Original performance material Click to show as PDF

Original performance material
Click to show as PDF

I have to admit that I don’t know anymore how I learned about LilyPond, but I remember quite well my first experiences with it. My first ‘real’ LilyPond score that still exists was created when rehearsing Tunjuk by Dieter Mack with the Stuttgart Radio Symphony Orchestra. There was a quite intricate rhythmic section that had to be played synchronously with harp and crotales (IIRC), and It was somewhat confusing to distinguish the 32rd and 16th notes and rests. 

My LilyPond version Click to enlarge

My LilyPond version
Click to show as PDF

I had been quite appealed by the introductory text on LilyPond’s web site and decided to give it a try. So I typeset one page of it with LilyPond, and the result was totally convincing: I practically could play the music prima vista now, which is due to the better spacing and the relation of the thicknesses of the elements. I suppose that if we all had LilyPond scores we actually would have played synchronously in that concert 😉  (You probably can’t judge that from the two images within this text. You should at least compare the full PDFs, possibly placing a printout in realistic distance on a music stand.)

The next score I recall is a fragment of a piano part from an ensemble piece by Brian Ferneyhough I had to learn. The – handwritten – material from the publisher was rhythmically somewhat unclear, and I wanted to have an uncorruptible model (and maybe a MIDI file…) to practice it. It was quite obvious from the start that I wouldn’t want to tackle that task with Finale 2001 (which I had used up to that point…) so I tried it with LilyPond – and once again the results were very promising. While being far from publication quality the out-of-the-box results were (and I still consider this a crucial point) perfectly readable and playable, something I had never experienced with Finale scores of considerable complexity.

Janek Warchoł: Why do you think that out-of-the-box output quality is so important?  After all, publishers are supposed to fine-tune the scores they sell… Does this mean that performing from “self-published materials” is more common than I thought?

Urs Liska: Well, I don’t really know what is common usage of LilyPond and other notation programs and can only speak for myself, but I think that preparing publications is only a rather small share of actual use. Composers preparing performance material, pianists doing transpositions, arrangements for performance, teaching material – all these are scores that do not necessarily need to be polished to perfection. They have to be usable. LilyPond creates outstanding out-of-the-box results because it is non-realtime and can really consider layout choices thoroughly. Which results in scores that can be played from right away after entering the content, with rather rare cases needing manual intervention for that purpose. This is also an important point when preparing scholarly editions: I (as the editor) can concentrate on the content until this is musically perfect, I’m not tempted or even forced to move items around when I should rather think about the musical text. When everything’s finished I or someone else can work on the typographical part. Bringing LilyPond to the attention of the scholarly community is one of my bigger projects BTW.

Janek Warchoł: I see – you make a good point.  And incidentally this brings us to a question that I had intended to ask: how would you describe working with LilyPond – in particular, what is your opinion on its text input? Some pros and cons?

Urs Liska: You don’t seriously expect a balanced answer to that question, do you? Of course I’m quite biased, but I’ll do my best to be fair 😉

Let’s start with the cons. Using text input for writing scores implies a learning curve that can’t be denied. Even with quite some experience you’ll often have to look up things in the documentation instead of just skimming through menus. This is even more true for people without any programming background. But (in my opinion) this issue is definitely outweighed by the advantages of having all the content written explicitly in plain text. Everything is there ‘in black and white‘ and your software doesn’t have the opportunity to make fun of you by arbitrarily reverting your manual changes. And there are editors like Frescobaldi that assist you in a way that is heading towards the comfort of a WYSIWYG editor.
Text input allows to profit from experiences and techniques of software development, once one starts to see it that way. It endorses code reuse, development of libraries, generating/editing code with programs, and – above all – versioning. It would be too much to elaborate on this now, but I’m about to finish an essay on this topic, which will of course be announced here 😉 In the meantime the readers could look for your posts about crowd engraving
To sum it up: I wouldn’t ever want to switch back to WYSIWYG tools, and I have mostly abandoned word processors too (writing most of my text documents with LaTeX nowadays).

Janek Warchoł: Indeed, i feel the same way – i wouldn’t want to go back to opaque file formats such as those used by Finale or MSWord.  However, I suppose that for some of our readers (who may not have similar technical background), your arguments about libraries, code reuse and versioning might not be very tangible – it may seem that text input is something useful only for serious, big projects.  What is it that LilyPond can offer to ordinary musicians in everyday use?

Urs Liska: You have to distinguish between LilyPond and text input here, and as we were talking about that I will comment on the latter. Probably I’m overemphasizing just because I’m so fascinated about these potentials for sophisticated set-ups. But the text input is also a good thing at the other end of the spectrum. If you are going to sketch just a short phrase, prepare an exam paper or music examples in a text, you just have to write down these few items instead of opening a whole document. For example you can enter LilyPond code in OpenOffice/LibreOffice documents (with the help of the OOoLilypond extension), so just typing
{ time 3/4 c8 d e f g a b4( c') }
will already give you a complete music example.
Another handy possibility is provided by the editor independence of text files. You can sketch music in any text editor, for example on a friend’s computer which doesn’t provide LilyPond, or on your smartphone or in an email client. Then just paste the code in a file at home and compile your score from it. Or use it on Wikipedia.
Finally you should also consider the (discussed) quality of LilyPond’s default output in this context. I think I’ll demonstrate this in a separate post soon.
[Edit: Here it is]

Janek Warchoł: Indeed, there are situations when LilyPond’s text format is hardly replaceable – when I have to send a tune to someone in a text message, notating it in LilyPond syntax is an obvious choice.

Keeping all that in mind, in what directions and “market areas” do you think LilyPond will (or should) develop?

Urs Liska: Once again I can only speak for myself, that is from my perspectives and interests. My greatest wish would be to get LilyPond a wider acceptance in publishing houses. Of course one can’t really blame them for that but most of the bigger publishers insist on their established workflows, i.e. on getting scores delivered as Finale or Sibelius files.
Another – partially related – field is to increase LilyPond’s exposition to scholarly work. This is something I’m actively working on, also by actually developing tools for and in LilyPond. Currently I’m planning to support annotations in the text files that can be used for a variety of purposes. The idea is to have all the work that is necessary to prepare critical reports within the score file, that is directly beside the music it is referring to. (This and other projects are assembled in the project openLilyLib. Feel free to join and assist …)
And finally I’m watching with interest the efforts to increase the usability of LilyPond through editing tools like Denemo, laborejo or Frescobaldi. They are heading in a promising direction, continuously minimizing the usability gap to editing scores visually, at the same time offering choice in the selecting your tools and style of working. LilyPond should definitely lose the image of being a nerdy thing …

Janek Warchoł: Absolutely!  I hope that the day when the first major publisher will accept LilyPond format is near (smaller publishers already use it, see here).

Thanks for the interview, Urs!  I’m sure we’ll soon see you again on this blog 🙂

3 thoughts on “An interview with Urs Liska

  1. Redly

    Personally I dislike his pretentious unfounded alien-styled inhumane chaotic pseudo-random material, which exists only for it’s own sake and creates sensory responses that are not of the composer’s intention, but just happen to occur.
    Make no mistake: Ferneyhough is no real composer; and the fact that this has never been accordingly stated or criticized shows the times in which we live: Feed the people any rubbish, with just a hint of added intellectual implication and they’ll believe it and worship your ‘message’.

    … Ferneyhough… the charlatan king of pretentious wishful implication

    1. Urs Liska

      Personally I dislike pretentious unfounded anonymous pseudo-random rants which only exist for their own sake. But as it’s our declared policy to only mark bots’ comments as spam we just have to acknowledge that this comment has some relation to the blog post.

      1) You are randomly ripping a tiny snippet from this post and “comment” on it completely out of context.

      2) Opinions are not stated but expressed.

      3) You’re free to like or dislike the so-called “new complexity” and you’re free to announce that publicly (even here). But it’s just more convincing doing so from a position of poised supremacy. Denying someone to be a real composer without giving the slightest reasoning is just plain bullshit.

      I suggest you piss off and come back when you can give at least the impression of having a sufficient level of understanding to judge these matters.


Leave a Reply

Your email address will not be published. Required fields are marked *