A question that I’m frequently asked is “what is the best music notation software for Arabic music?” Since it seems to be a perennial question, I thought I’d make a post to collect my thoughts on the question. Though not an exhaustive evaluation, this will be kind of a deep dive. If you just want my conclusions here they are:
If you’re a professional and creating scores for performance purposes: Dorico
If you’re only doing Maqam/makam music and relatively basic scores: Mus2
If you’re pretty computer savvy and don’t mind a text/code-based system: LilyPond
If you are on a tight budget, or are a more casual user: MuseScore
Okay, now if you want a bit more info on the programs and how I came to those conclusions, read on.
First, I’ll explain my criteria for evaluation. There are a few things a notation program has to do well in order to handle maqam-based music. These are:
• Microtonal accidentals and key signatures
• Complex meters and time signatures
• Free/Open meter (no time signature or bars, for notating taqasim, mawwal, layali, and similar metrically free passages)
• “Nonstandard” key signatures (anything other than the traditional 14 major/minor key signatures)
There are three “major” notation programs currently being used by most professionals: Finale (by MakeMusic), Sibelius (by Avid), and Dorico (by Steinberg). There is also a relative newcomer that has made big improvements in the past few years, MuseScore. The “Big Three” are all relatively expensive: Finale lists for $600 (though right now it appears to be on sale for $299), Sibelius uses a subscription model priced at $199 per year, and Dorico is $580. MuseScore is open-source and can be used for free. I’m only comparing the “full” versions of the programs here—they all offer lower-priced and even free options with fewer features but it’s not really worth considering those, partly because many of the features that aren’t available in those versions are ones you will need for notating maqam-based music.
So right off the bat, MuseScore has the advantage of being essentially free to use. If you are a casual user that may really be the only thing you need to know: MuseScore is fully-featured, can manage almost anything you can imagine and is free to use.
Before I get into a more detailed comparison of those four options, I’ll note two other choices to consider: LilyPond and Mus2.
LilyPond is a full-featured program that allows you to produce sheet music by writing code in a text-based format. If you are comfortable with any kind of computer coding, the language used is pretty simple and easy to grasp; I would put it on par with basic HTML in terms of complexity (it’s Unix-based and is similar to LaTeX). There are some slightly tricky aspects of getting the actual program to run if you are on a Mac, as it requires access to Terminal and another program called MacPorts. The actual notation produced is generally excellent and it can do microtonal music, open meter and complex meters easily. Since it’s text-based, copy and paste functions are really straightforward. The downside is there’s no user interface to speak of, you can’t see the music visually as you are writing it, and there’s no playback option so it’s not really useful for trying things out. If you are copying a composition that is already written out, these aren’t really significant problems, but if you are composing or arranging, it is more challenging—you’d probably want to write everything out on paper first and then copy it into LilyPond.
Mus2 is a Turkish program that is expressly made for notating makam music. It’s pretty straightforward and does a really good job with classical Ottoman music and Turkish/Armenian/Greek/Macedonian folk music, and is well-suited to Arabic music as well. It’s priced as a subscription model, but relatively inexpensive (about $25/year). It claims to be able to do other styles of music, but in my opinion the results for non-makam music are supbar when compared to the other options. If all you are doing is straightforward charts for things like Saz Semai or other traditional music (mainly a single melody line) then this could be a great option as it is the only software that is really specifically designed for this music. The interface is fairly straightforward. When I first tried it, almost all the documentation was in Turkish but it seems that they have provided a lot more English resources now so it’s more accessible to non-Turkish speakers. Documentation is still pretty limited though.
If you want to be able to do larger arrangements or music other than maqam/makam music, and want a program with a graphical interface, then the main programs mentioned above are the ones to really consider. So let’s get into the comparison.
Finale, originally introduced in 1988, has been around the longest. This has some pluses (large user base, lots of plug-ins, lots of resources for learning, tons of features) and some minuses (bloated codebase, outdated design, lots of strange quirks). The biggest strength of Finale in my opinion is its flexibility: pretty much anything you can imagine in notation can be accomplished one way or another. It’s also possible to produce very high-quality scores in any style. The biggest drawback is that its ‘default’ output looks pretty terrible and is nearly unusable for professional purposes without tons of manual tweaking. The second biggest weakness is that although nearly anything can be accomplished, many things require convoluted workarounds because the basic functionality of the program doesn’t support it.
I was a Finale user for over 20 years, so I got pretty used to a lot of the workarounds necessary for Arabic notation (and even invented some of my own) but they remained time-consuming and frustrating (for example, the workarounds for non-standard key signatures do not allow for transposing). Playback of microtonal music is possible to some extent but prohibitively complex and time-consuming to achieve. I like to do initial “proofreading” of scores by ear, so even though I wasn’t using the playback function for any ultimate purpose most of the time, it was still annoying that microtonal scores would always sound wrong in playback. It technically can produce nonstandard key signatures (including microtonal ones) but they cannot transpose and the process is so complicated and unintuitive that it’s usually better to “fake” it using workarounds. Complex time signatures, mixed meters and other complicated metric choices (including dashed barlines) are relatively easy, though they require a weird familiarity with the idiosyncratic MIDI time values of notes (e.g., dotted quarter is 1536 for some reason). Free/open meter is not an option but can be faked. The process is relatively simple, but rather laborious for extended or complex passages like a taqasim or mawwal.
In the end, it’s a fine program overall and can produce absolutely great notation in the right hands, and is still possibly the best program for producing 20th-century style modernist “graphical” scores, but I’m hesitant to recommend it at this point for new users. The developers are still reasonably invested in continued upgrades and support, so it may be that some of the drawbacks of the program will be addressed, however most of the issues that affect users interested in maqam music are deeply rooted in the fundamental design architecture and are unlikely to ever be a priority to fix. The priorities seem to primarily be around improving playback results and serving the needs of educators and students.
Sibelius, I confess, I have much less experience with. It was introduced in 1993, at which point I was already using Finale. Being a fluent Finale user, there was never much real need for me to work with Sibelius: though it has improvements over Finale in many respects, it lacks some of the flexibility of Finale. The improvements, while generally worthwhile for new users, were not significant enough to justify switching. The main improvements are that the workflow is generally more intuitive and the learning curve is lessened and the default output is much improved (recall that Finale’s defaults were nearly unusable). Sibelius does a good job of preventing users from creating “wrong” notation.
Quarter-tone support is much improved, but key signatures still require a lot of workarounds, as do free/open meters. Complex meters and mixed/changing meters are straightforward and well-supported. If the choice was only between Finale and Sibelius, I would (until recently) generally have recommend Sibelius for most people; it’s flexible enough and is much easier to use. Having been around for a long time, it has many features and robust plug-in support. However, the developers (Avid) seem to have deprioritized improvements in recent years, which is not encouraging for the long-term viability of the program. The shift to a relatively expensive subscription-based pricing model may lead to more upgrades, but is probably a deal-breaker for all but the most prolific users.
Dorico is the newest of the programs, first released in 2016. Built by many of the people who were behind the design of Sibelius, it reflects a much more coherent, consistent and modern design philosophy than Finale or Sibelius. This comes at the expense of having a pretty steep learning curve: the design is aimed at making the program very fast and easy to use once you know how. To do a lot of tasks, you need to learn key commands and shortcuts. There are tons of settings and options for default behavior; this is ultimately a faster workflow because you set something once and it will be the way you want it everywhere, but in the beginning it can be hard to find the relevant settings and changing things manually is somewhat trickier than in the other programs. Once you learn the principles and details of how to work in Dorico, it’s a breeze and very fast. Many things that take complicated “faking” in the other programs are very very easy to do in Dorico, but getting to that point can be frustrating and requires dedicated learning.
Quarter-tones or any other microtonal system can be easily created in Dorico, and saved for later use. It also includes a system for creating custom key signatures. The tool for doing so is intuitive and easy to use, and the resulting set of accidentals and key signatures can be saved as a “Tonality System” that can be re-used in any future projects, so you really only have to do this once. Both nonstandard and microtonal key signatures can be created using any accidental that is defined in your tonality system. Complex meters are easy and can be entered simply using key commands and text. Certain things, like a series of changing meters can be a bit more work but are doable within the basic functionality of the tools. Free/open meter is actually the default in Dorico, and notating taqasim etc. are a breeze with no action needed to accommodate the lack of metrical structure. Line breaks are simply determined by how “full” the line gets, but this can be overridden if desired (for instance you really wish to reflect the phrase structure). You can insert barlines anywhere you like.
Dorico’s default output is easily the best of any of the currently available major programs, though LilyPond arguably is comparable. I think this is definitely the best, most professional option, especially If you also are creating sheet music in different styles or arranging for modern ensembles. The developers are just getting started and each new release has major upgrades and new features, so support is robust and looks to be that way for some time. The learning curve and price may not be tenable for more casual users, though, in which case MuseScore may be something to consider.
MuseScore has been around a while, as an open-source program it was initially extremely limited and the development has be an ongoing public project. The first widely usable version was released in 2009. Initial versions had a lot of problems, was buggy, lacked important features and provided sheet music that was not remotely professional-quality. However, in 2017 the company was bought by Ultimate Guitar and a renewed effort was made to bring the software up to professional standards. The most recent version, released in 2022, was a major overhaul that brings it impressively close to the other major programs in functionality, stability, and output quality. Many of the improvements seems to be drawn from the work of the Dorico developers, providing an element of healthy competition.
Because of the open-source nature of the program, some of the trickier possibilities are implemented via plugins that are developed by contributors. The basic implementation of microtonal accidentals is built-in, but playback requires specific plugins. I haven’t tried this specifically, but reports seem to be that how well this works depends on the specific plugin and tuning system you are trying to use. It’s possible though, the program appears to be flexible in this regard and it is something that will likely continue to improve in future versions. Nonstandard and microtonal key signatures are supported, and relatively simple to implement, however they cannot be transposed. Complex meters are mostly relatively easy to create, but some aspects of mixed or changing meters can be difficult to implement and may not display consistently. Free/open meter is not supported, but like Finale and Sibelius, it can be faked with some effort.
MuseScore is in the middle of major improvements and the developers seem committed to continuing to add new features. The open-source plugin community helps fill gaps in functionality, and I expect the program to continue to get better, though the financial prospects of the developers seems unclear since the program is provided for free. Which is of course the major selling point of MuseScore—it’s free and anyone can use it. Price aside, it’s no competition for Dorico, but for most people a program that does most of what you need reasonably well and is free probably makes a lot more sense.
I made a comparison chart for evaluation:
|Program Name||Microtonal Accidentals||Microtonal Key Signatures||NonStandard Key Signatures||Complex/
|Free/Open meter||General Notation Quality|
but very difficult.
Can be done via
with lots of extra work.
Don't transpose. Can be faked with workaround instead
Don't transpose. Can be faked with workaround instead
|Yes, reasonably straightforward||No, but cumbersome workarounds
can fake it
|Can be excellent, but defaults are poor|
|Sibelius||Yes. Playback is possible via a
|Not supported, can be faked
|Not supported, can be faked with workarounds||Not supported, can be faked
|No, but cumbersome workarounds
can fake it
|Very good, defaults
|Yes. Playback is possible via plugins to
|Yes, but they cannot be transposed||Yes, but they cannot be transposed||Mostly yes, but some limitations||No, but cumbersome workarounds
can fake it
very good, defaults
are mostly acceptable
I can’t think of any reason in 2023 that someone who wasn’t already using Finale or Sibelius would choose to start there. If you want a professional, full-featured program, Dorico is the clear choice. If you want something specialized for makam music, Mus2 makes the most sense and is inexpensive. If you want a free option that is powerful, has good results and is reasonably flexible, MuseScore or LilyPond are the logical options.