Baroque-friendly music software?

Andrew wrote in the Kellner suites thread: '"Speaking as a person that nowadays spends most of my time engraving, with Lilypond and Dorico . . . " (I’m starting a new topic in an attempt to be a good digital citizen.)

I am curious about which music software is most helpful when engraving Baroque music. I’ve been using MuseScore, which is generally very good. For instance, it includes the spiky box-shaped C clefs used by French Baroque engravers. The only thing I haven’t been able to do with it (so far) is reproduce the diagonal bracket shapes used by Couperin to indicate legato. And I do like to support open-source software, which MuseScore is.

I looked at Finale but it didn’t seem as good as MuseScore, based on my initial examination. Are Sibelius, Dorico or any others better for Baroque music? I’ve never used Lilypond, but I have used LaTeX and XeLaTeX to typeset books. They are very powerful but also a pain in some ways, which led me to move to other software. I suspect the same is true of Lilypond, which is a member of the TeX family;

The answers to your questions are long and complicated.

First, for Baroque I think we can discount Finale and Sibelius for a host of reasons. I bought Dorico - it’s quite expensive - and while it is a really good engraving program and has a lively and responsive user forum (done in Discourse the same as what I am using for Jackrail, by the way) it is entirely focused on contemporary cinematic scoring, jazz, Common Era music and so on. Using it for work outside these major customer areas gets difficult if not impossible, and one of my critiques of Dorico as a product/company is that as somebody who does both Baroque and modernist New Complexity School engraving, when you step outside Common Era music engraving rules you get long lectures on why you should not write in this that or the other way and why they will never change or update Dorico to add badly needed features because they are for a tiny customer set. I wish they would come down off their high horse someday. For example, putting notes in the centre of a bar - common Baroque practice, and, amusingly also common in modernist scores - is well nigh impossible. Dorico is heavily rule bound and that helps students and scorers in a hurry, by not allowing you to make mistakes, but you can’t override the rules to suit different forms of music. If it’s not regulated by Elaine Gould in Behind Bars, forget about it. Hopefully this will change in the future. Dorico is only five years old.

Now I used Lilypond for ten years and it is the most powerful engraving program by a long shot. But being text based data entry with a particular relatively complex syntax it has a steep and long learning curve. Lilypond was developed by two Dutch programmers who were also musicians. In the very earliest days 25 years ago there was some influence from MusicTeX but that is long gone by now and it would not really be right to say it is part of the TeX family. There’s a superficial syntactical similarity, and it is likewise text based, but that’s where it ends. Lilypond output is supremely excellent without tweaking, but every detail can be tweaked and adjusted to as fine a degree as you want. Lilypond is open source. It is highly extensible by writing code in Scheme, a very nice member of the Lisp family of programming languages, but again complex with a long learning curve. I have a library of Scheme functions I wrote over years to extend Lilypond to do all sorts of ‘impossible’ things.

One person has written and extensive library of French Baroque articulations which are very nice, but it is quite technical to get it going.

The only reason I switched from Lilypond to Dorico is that Lilypond is a compiled language and even on my fast processor the 300 page (not a typo) string quartet I was engraving was taking twenty minutes to compile and if you have to change one note as a correction you are in for a twenty minute compile again. Lilypond is by design and musical necessity single threaded, so having more cores does not speed it up/ After a while this became untenable for me. But for shorter pieces it’s superb.

As an aside Lilypond has first class support for Gregorian chant and medieval music and so on.

While MuseScore has many virtues, compared to Lilypond the output always looks inferior. You can spot a MuseScore engraving instantly. It’s not to my taste. Last time I looked at MuseScore it could not do triple dotting - essential for Handel for example. While I suspect they have addressed that I think you cant do more dots - and so you cant use it for modernist scores etc.

Lilypond has very strong support for arbitrary lines and curves and you can even write in Postscript if you like. The main virtue of Lilypond is that it is unlimited and can be made to do anything you can throw at it. It’s the only program that can be said of.

There’s a graphical developement environment called Frescobaldi which is nice but collapses on very large scores. I use vi with some clever integration for the work - it’s really great. [No need to mention the best environmen for Lilypond is Linux. Way better that the Windows and Mac versions.]

Just coming back to Dorico, there is a commercial font called November 2 that does pretty nice Couperin. Here’s the sample page from the font developer:

But I don’t think Dorico can do the angled slurs!

Enough for now. This is a complex topic.

Le 06/12/2021 14:30, Andrew Bernard via The Jackrail écrit :

First, for Baroque I think we can discount Finale and Sibelius for a host of reasons.

Well, I’ve done quite a bit of engraving of Baroque music with Finale,
and I’m happy to hear I’m discounted, but would also enyoy hearing some
of these reasons. Several well-known publishers of Baroque music use
either Finale or Sibelius. And I know of none that use either Lilypond
or Musescore.

Oh dear, so sorry - yes that was quite unintentional but could come across as offensive I see. Sincere apologies.

It’s true that Lilypond has no presence in commercial print publishing, and probably never will have.

Lilypond can do the flat slurs Couperin-style:

As for Sibelius, it can’t do Couperin slurs. Of course it is possible to mimick them, as I did in the picture attached. But they aren’t “slurs” for Sibelius, they are just three lines at an angle. You can alter the angle, the length, the thickness. But if you move the spacing, or redo the layout or the note size or whatever, the three lines detach and you haven’t a slur anymore. It’s something you do as the last thing after the definitive layout is done.

I must say I agree with Dennis. A lot of Baroque music is done with Sibelius, a lot I did myself. I didn’t miss anything, any symbol, excepted the Couperin slurs (which aren’t in a font but are drawn by Sibelius itself). Even the beautiful seagull-like mordants you find in Couperin and other French editions can be done easily, provided you have a suitable font as November 2. I agree with Andrew on November 2, I own it and it’s a great font with a lot of symbols.

And if you haven’t a suitable font you are not lost as you can import graphics, or draw yourself, or even make a font with just one symbol you need. This last thing doesn’t call for some particular skill, I have done it many times and I am no font expert.

However, Dennis, even Musescore has been used for a noted Baroque music edition: It’s the edition of Gustav Leonhardt harpsichord transcriptions of Bach’s works for ensembles or other instruments, and was done in Musescore.


PS one thing I miss from Sibelius is an automatic handling of everything about a critical edition: square brackets, cut slurs, footnotes. I am able to get everything I need but it’s tiresome.

Sorry, forgot the picture. Here it is.

Don’t look at the spacing, the font etc. It’s just a sample of the Couperin slurs.


One of the things Lilypond can do by virtue of being extensible is that I could define say a Couperin angled bracket slur as a proper musical object that automatically adjusts its size/angle/parameters when notes are changed or moved or the layout shifts and so on. This relieves you from having to have special workflows where you adjust all the things that are not ‘fully aware’ or emulated by gluing bits together in the other programs.

I am not saying Lilypond is the best or that everybody should use it - after all, I have had to abandon it for Dorico. But it is immensely powerful in so many ways.

I was also about to mention that Siebe Henstra did the Leonhardt edition in Musescore and I believe the publisher printed it directly off his engraving without resetting it. I may be wrong but in correspondence with him this is the impression I got.

It is not a goal of the Lilypond open source development to establish the program as a major force in commercial publishing. Apart from being hard to learn, there is no commercial support for it, one of the standard reasons some companies shy away from open source software still today.

As a Sibelius user ever since it was only available for Acorn computers (pre 1990, if I recall) and a user of Dorico from its more recent beginning, I agree with all Andrew’s comments about Dorico. However, it must be noted that the people who designed and develop Dorico were in charge for many years at Sibelius. The same push back was evident there when suggestions were made that did not find approval. I have so far done nothing significant in Dorico and am greatly frustrated by the narrow-minded interests of those who work there. I do not have a copy of Elaine Gould’s book, but it sounds like the Chicago Manual of Style all over again. Like Domenico, I have been able to make use of the slightly greater flexibility of Sibelius, but I have largely lost interest in that program, in the, perhaps vain, hope that Dorico will improve.

The new Bärenreiter Couperin edition has the flat slurs, but I don’t
know what software they used.

Speaking of Couperin, one of the things Finale can’t (or couldn’t - I’m
still using a rather old version) do without a workaround is 258th notes.

I like the idea of source content with macros in plain text much more than a musicXML standard, which is not really human readable, and which is not suited for version control systems.

The relation between TeX-like source material and its compiled form has been characterized as how a blueprint relates to the building it represents. TeX-like macro languages probably allow you to focus on the content more than you would in WYSIWYG programs.

Interesting. Sorry for asking the obvious: can’t you split up the work? And maybe run concurrent compilation jobs as needed? I wouldn’t know, I’ve never typeset that much music. Is the entry below helpful?

Thanks to everyone for the replies, and particularly to Andrew for his detailed explanations. I ‎have learned a great deal from this thread. A few things in no particular order:‎

The current version of MuseScore (3.6) is a major upgrade in terms of engraving quality. If you ‎checked out this program a while back, it might be worth a second look. (And it has done triple ‎and quadruple dots for some time.) MuseScore also has many features for early music that I ‎don’t use myself — things like mensural notation, lute tablature, etc. — so the developers are ‎used to people wanting arcane stuff. That doesn’t mean they will add every requested feature, ‎since they have to balance programming time with potential utility, but I have never felt an ‎arrogant “our way or the highway” attitude on their forum. MuseScore is not perfect for critical ‎editions, since things like footnotes have to done manually by adding a text frame.‎

I will definitely check out the November 2 font — thanks for mentioning this! I also have a ‎music font of my own that I have created that I would be willing to share. It’s an ad hoc kind of ‎thing; when I need a symbol for a particular project, I add it to the font. But it has grown to ‎include quite a large number of symbols. I can send a PDF that shows the contents as of now.‎

I am intrigued by the Frescobaldi front end for LilyPond. The comments about the power and ‎extensibility of LilyPond do not surprise me, given my experience using TeX for book ‎typesetting. It is excellent to know that LilyPond can do Couperin-style slurs. Maybe I will give it ‎a try. But I did not stick with TeX long-term, despite its power, for several reasons and suspect I ‎will end up feeling the same about LilyPond.‎

I sense that within this group, people are mainly looking for easy-to-use graphical tools; but as long as we’re talking about Linux and text-based, programming-language-like music tools, let me also mention the “abc” notation language, and the many free tools out there for manipulating it and typesetting it, such as “abcm2ps”. I can’t comment on the more-sophisticated things that people are discussing, like straight-line slurs, but I have used abcm2ps to set a number of pieces by J. S. Bach, and it has always had the necessary features and capabilities. Like Lilypond, it is a developer-community-supported open-source kind of effort. But I prefer the “abc” language for music setting; it does have a significant learning curve, but it’s nice and compact (unlike my impressions of Lilypond and other tools that derive their syntax from TeX and/or XML), and e.g. the pitch syntax is reminiscent of “Helmholtz” notation.

There’s a “homepage” for abc notation, with a most uninventive name:
It was originally developed for folk and roots type music (traditional fiddle tunes, etc.), and the tone of the webpage reflects that; but the open-source aspect has allowed it to be greatly extended into Baroque and many other musical styles.


How well do the the various products handle French unmeasured preludes? From Andrew’s description it appears that Dorico can’t even get started.

There’s a useful site dedicated to engraving news and tutorials and so on:

Much less biased than me!

The general answer is not at all well. Advice given is usually that you need to do all those graceful and curvaceous multiple overlapping long slurs in an illustration program like Inkscape or Adobe Illustrator after the engraving step. Not at all satisfactory.

I gues those pieces are super specialised, but there’s actually some modernist music that uses long spans of complex curves as slurs and phrasing and breathing and other marks, equally difficuly to do in current software.

Indeed, Sibelius too just can’t do the curved lines resembling to long slurs from a note to nothing, typical in the unmeasured preludes, even if it has a “lines” features. The program awaits a line to start from something and end to something else.
The remedy is the same for Sibelius as well: export the pdf and modify it adding the curved lines in inkscape or illustrator. Or the opposite way: draw the lines in one of the other programs and import in Sibelius.