Why digitize sheet music at all?
Paper tears, disappears and cannot be searched — a digital library fixes all three at once.
Then there is the weight: anyone who rehearses or performs regularly ends up hauling kilos of paper around, and the one piece you need tonight is guaranteed to be at home. Digitally, the whole library lives on your iPad, iPhone or Mac — searchable by title and composer, organized into collections, and available without an internet connection.
Getting started is less dramatic than it sounds: most sheet music already exists as a PDF anyway — bought in online shops, exported from notation software, or scanned yourself. The real question is not whether to go digital, but which format your library should grow in. There are exactly two candidates, and they complement each other well.
PDF: the digital sheet
A PDF is an exact image of the page: perfect for reading, but it knows nothing about the music itself.
That is its strength and its limit in one. The strength: a PDF looks exactly the same everywhere — every system, every fingering, every editorial marking stays precisely where it belongs. On stage that matters: what you see on screen is the same picture you rehearsed from.
The limit: to a computer, a PDF is just an image of dots and lines. It does not know there is a G major chord on the page. Transposing, extracting individual parts or reflowing the layout for a different screen size — a PDF simply cannot do any of that. If all you want is to read, turn pages and annotate with a pen, you also don't need more.
MusicXML: notation as structured data
MusicXML describes every note, bar and voice as data — enabling transposition and reflow.
MusicXML is the open interchange format of the notation world: programs like MuseScore, Sibelius, Finale and Dorico read and write it. The compressed variant uses the .mxl extension — same content, smaller file.
Because the format describes the music itself rather than its printed image, the content stays editable: a MusicXML piece can be transposed, reformatted or split into parts in a notation program and exported again. The trade-off: the visual layout is recalculated on rendering and may look slightly different from the printed original depending on the software.
PDF vs. MusicXML at a glance
Both formats have clear strengths — the choice depends on whether you want to read or edit.
- Appearance: PDF is faithful to the original and layout-stable; MusicXML is re-rendered on display.
- Editability: PDF is frozen; MusicXML stays editable as musical data (transposition, parts, layout).
- Availability: PDFs exist for practically everything; MusicXML mostly where scores were engraved digitally.
- File size: both are small — no deciding factor here.
- Longevity: both are open, widely supported standards.
From scan to MusicXML (OMR with Audiveris)
Optical music recognition turns scanned pages into editable data — with limits worth knowing.
The technology is called OMR — Optical Music Recognition, sheet music's counterpart to OCR. One of the most established open engines is Audiveris. NovaNota has this step built in: the app sends a PDF to the NovaNota server, recognition runs there via Audiveris, and an MXL file comes back; the uploaded file is automatically deleted after 24 hours. The Pro version allows up to three conversions per 24 hours, subject to server load.
Set realistic expectations: cleanly printed, well-scanned sources are recognized well — handwriting, skewed photos and dense engravings remain difficult. The result of an OMR conversion is a starting point that deserves a review, not a finished edition. For "quickly transpose the second part" it is often enough; for re-publishing an edition it is not.
In practice: both formats in one library
The most practical setup keeps PDFs and MusicXML side by side, picking the right format per piece.
That is exactly what NovaNota is built for: PDF and MXL live side by side in one library — sorted by title, composer or custom order, grouped into collections, with covers and fast search. Reading comes with smooth page turns, zoom, portrait and landscape, table of contents and bookmarks; with the Apple Pencil, fingerings, breath marks and highlights go straight onto the page — in both formats, exportable for print.
The library syncs through the cloud of your choice (iCloud, NextCloud or Google Drive) automatically across iPhone, iPad and Mac — and stays fully usable offline when you are on stage.
Conclusion: which format when?
In short: PDF for reading and archiving, MusicXML for editing and transposing.
If you are digitizing a paper collection, a simple rule of three works well: capture everything as PDF first — fast, lossless, playable right away. Convert the pieces you actually want to edit or transpose to MusicXML via OMR. And keep both versions together on the same piece, so rehearsal and editing never drift apart.