Michael Good, the inventor of MusicXML [encore]

Scoring Notes
Scoring Notes
Michael Good, the inventor of MusicXML [encore]

If you’ve ever needed to open a Finale file in Sibelius, or a scan music in SmartScore and bring it into Dorico, or write something in StaffPad and upload it to MuseScore.com, you’ve relied on MusicXML to make it happen. MusicXML is the standard interchange format for music notation applications, and it’s what makes such robust interoperability possible between hundreds of programs. We might take it for granted now, but that doesn’t mean we should.

MusicXML was invented by Michael Good, and on this episode, Philip Rothman and David MacDonald talk with Michael about how he came to create such important technology more than two decades ago. He tells us the inside story of what it took to get it off the ground, and puts us in the room at pivotal moments in its development, from its humble origins all the way through to its current incarnation as part of a thriving worldwide community. We also learn what to expect in the next version, MusicXML 4.0, and how this now-ubiquitous established format is keeping pace with the newest technological changes.

This episode is an encore of the episode that first aired on March 13, 2021, with a few updates at the beginning about the release of MusicXML 4.0 and Dolet 8 for Sibelius.

Read the transcript of this episode

More on Scoring Notes:


  1. Philip Benjamin

    I believe that MusicXML should become a key part of a combined notation, MIDI, audio file format (with separate formats continuing to exist for each). For this format to work people like Michael Good – and other music notation experts – would need to be instrumental in its creation.

    My reason for suggesting this isn’t only because there are times when one might want to move the three kinds of information as one – it’s also because I believe that there is a need to preserve relationships between audio, MIDI and notation data – and only with some kind of combined format would this preservation be possible.

    What kind of relationships do I mean? I mean that for example it should be possible for a file format to associate a specific notated part with an audio recording of it. At the moment in music apps these associations must be remembered by the user/composer/arranger. This should not be the case – it is not for example possible for someone to look at audio waves and know what they sound like – instead it should be possible for programs to “link” tracks of different types within an element (just as DAWs do now only with tracks of the same type). We need apps to have some kind of entity (they could be called ‘elements’) which include one or more audio tracks, one or more MIDI tracks, and one or more notation tracks (staves). Once this kind of structure exists (once apps are collections of elements instead of collections of tracks) it will enable apps to provide clever features which to this point don’t exist. For example a monophonic audio track within an element could be set to automatically be copied to an MIDI track in the same element, or we could have the pitches of a MIDI track – when changed – automatically change the pitches on the notation stave but not the rhythm – and vice versa. Or we could have the velocities for the MIDI track determined by the audio recording. These kinds of associations would ensure that anything that can be had for free is had for free while people are then not limited by the associations. For example if the user wishes to they could have a separate playback track for a notated stave which has a series of trills – instead of relying on the app to play the trills the way they want them played.

    If anyone from Scoring Notes likes these ideas and can ensure that people that matter are given this suggestion I would appreciate it. If we want music notation to be a first class participant in DAW apps – and if we want audio to be implemented effectively in notation apps – I believe my suggestions – or some kind of features which achieve similar – need to exist.

    1. Philip Rothman

      Hi Philip. These are all excellent suggestions, of course. Although I haven’t kept up with all of the details, I believe that at least some of these ideas are under active consideration, if not development, in the next-generation music interchange format, which is known as MNX.

      If you’d like to contribute to that discussion, you could join the W3C Music Notation Community Group and also read more about all of their work so far in this area.

      1. Philip Benjamin

        Thanks for the suggestion Philip – and for your instant reply. Let me too express my appreciation for your constant work – and the work of those who work with you at Scoring Notes. I know it’s only part of that work but I regularly listen to and enjoy your podcast.

        I will send my ideas through to Adrian Holovaty.

        There is of course perfect compatibility between my suggested concepts and how apps work at present. We need only think of the existing audio tracks, MIDI tracks – and if they are separate – the notation data – within any app as each being “an element’ – until at some point individual apps enable them to be formally grouped in future versions – hopefully preceded by some kind of file format which gives guidance about how to best make those associations. I believe that any DAW app which begins to make these associations will make their notation features exponentially more useful (both by the presence of DAW features within the app and due to the power of making associations between different track types). For these reasons I see the battle between the existing music notation providers – who with the exception of MakeMusic also make their own DAW app (Steinberg, Presonus, Avid) as a battle between companies – not a battle between apps (this leading me to wonder if MakeMusic’s future might lie in cooperation or merger with a company that makes a DAW).

        I would love to see open discussion about these issues in a podcast episode. I believe that there is a very locked in attitude in music notation specialists which if it continues will only delay vendors providing solutions which will end up being of most help to all types of users.

        Is anyone going to tell me this is not the future? Is anyone going to tell me that ten years from now the most capable music notation environment will not be either:
        – a music notation app which due to the continued progress of dedicated DAWs is a substandard DAW?
        – the result of notation features having been integrated into a powerful existing DAW app?
        – an app which does all these things which will have been built from scratch (who can bankroll that – and where does such talent exist in a group of developers)?

        Now is the time to talk about these things – before developers forge paths which will prevent any kind of universal interconnectivity from existing. For example one iOS developer has just added audio tracks to their score app – the audio tracks turn on and show in the middle of the score – or not. That seemed to me an odd way to handle things (instead of making audio tracks directly linked to staves – it’s the kind of thing we don’t want to happen (if it remains as is) to assets precious to music notation people.

        I believe that those reviewing music notation apps whose ‘new features’ relate to connectivity between their app and DAWs – or their implementing features which have been in DAWs for years – need to see and speak about these features in an appropriate light. This is NOT a criticism of Dorico 4 – for example its excellent MIDI import features would be needed for any music notation capable app – DAW or not – to handle incoming MIDI data. I’m saying it’s something to keep in mind when one sees companies beginning to build MIDI editors and go to huge efforts to create playback capabilities (recognising that such capabilities are in part special because they recognise notation symbols). I don’t know if the reasoning that sees apps currently developed in separate silos is that more money can be made by selling two apps not one (I suspect that the bigger reasons are either trust – people may not have faith in other teams in their company – or technical incompatibility) – but I can’t see why money should be a reason for building separate apps. For example Steinberg could make Nuendo, Nuendo Notate, and Notate – it’s not as if companies like Steinberg are pricing their DAW apps now according to what Apple charges for Logic. And looking from the DAW end DAWs currently exist in no man’s land for music notation – they provide more than notation editing and yet what is more than that is too weak to help those serious about scoring – while not being used by those who don’t care – the result being that they an odd approach – they provide scoring capabilities which work for almost no-one.

        Computer processing power is no longer a reason standing against these kinds of mergers.

        If we want traditional music notation to remain a prominent part of the future it needs to exist in the apps where the young are making their music.

Leave a Comment

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