Skip to content

Optical Music Recognition: A Powerful and Convenient Tool

A Short Musical Background

Before we get into arranging and optical music recognition, I want to take a moment to talk about how I stumbled into the world of music. 

I have been a musician ever since I can remember. From picking up my dad’s mountain dulcimers and guitars to stealing my mother’s Irish whistle, to playing endlessly on my borrowed aunt’s piano keyboard. After enough incessant recorder and flute playing, I ended up with a Western C flute. Fast forward more. I am now a full-time music blogger and artist with a degree in instrumental music education. So how did I come to learn about optical music recognition, (also known as OMR)?

As some of you may know, I am a freelancer on Upwork. I predominately write, edit book covers, photoshop headshots, and edit and produce audio tracks. One day, I got invited to write for ScanScore. While I had heard of the company, I didn’t really know what optical music recognition was. I was offered to try out their software, and I absolutely loved it. (And no, I’m actually not getting paid in affiliate links here.) 

The Time Saver

In my free time, I arrange flute quartets and quintets of popular alternative songs. I create them by referencing pre-existing transcriptions (such as those found on MuseScore), by listening to the song (over and over again), and, reworking the layers of the music until I have something entirely new. But here’s where the drag comes in: I would find a perfectly transcribed, accurate version of the melody to the song that I wanted to arrange. So, what did I do?

Since the melody of the song I was arranging wasn’t in a digitized, editable format, I did what many musicians do: 

I re-transcribed the transcription into Sibelius. Painstakingly. Note by note. Cursing whoever created the PDF. 

Does that sound like a good use of time to you? And, in these moments, we arrangers consider ourselves lucky, because we managed to locate music that we have to double-check and twist around, rather than transcribe ourselves, from the original (Side note: perhaps I should rename my blog “The Lazy Arranger”. Thoughts?).

So. My point is, ScanScore eliminated my need to retype everything inside of my music PDFS back into Sibelius. 

How Optical Music Recognition Works

In a nutshell, optical music recognition software teaches your computer how to read your sheet music. It recognizes the elements, from crescendos to note value to pitch, makes sense of them, and spits them back out in a digitized format. There have been previous programs created like this for text. That being said, these programs were much easier to develop, because music has a lot more information in it than say, a Word document does. 

ScanScore is an Unusual OMR Program

ScanScore is an unusual OMR program, and I mean that in the best way. I stumbled upon the software as part of my professional music blogging career, but, through learning and researching the program, I learned to appreciate it in a totally different manner: from the viewpoint of a musician, composer, and arranger. 

It’s different in the sense that it also acts as a MIDI player, and editing tool. You can change the instruments and pan those instruments using the mixer and can edit misreads inside the document itself. These two functions are hard to find in a program like this. So, in short, it’s not just because I’m a ScanScore insider. It really is better than the other options out there. 

Optical Music Recognition: A Little History

If you know me, you know I love history. So before we wrap it all up, let’s look at how optical music recognition came to be.

The idea to have computers read and digitize music was first spawned by researchers at MIT. Optical music recognition is similar to OCR, which is an optical recognition program type that digitizes text. That being said, OMR is much more complex than OCR, because music has a lot more information to be gleaned than a text document does. 

The first piece of optical music recognition was officially created by Waseda University, and it was called Wabot-2. This OMR software looked a lot more like a person than a computer. And by that I mean, the Wabot played music with its animatronic fingers!

Throughout the years, several problems have arisen in the industry. One of the more recent ones being the recognition and digitization of the music staff. 

Popular optical music recognition programs today are often a sub-piece of notation programs, like Sibelius, etc. But since the main focus of these companies is to produce notation and not to digitize PDFs and images, the result is often lackluster. Anyways…

To close, optical music recognition has come a million miles since Wabot (Want to learn more about the robot that played the organ? Look no further than this link ). I am thankful that it is a tool that we have today, and as a (very) busy musician, am relieved to know that it is still being improved upon and developed even still. 

Arrangements I’ve Made with the Help of OMR

Here are a few arrangements that I was able to make faster thanks to a little help from modern technology: 

Hey There Delilah -The Plain White Tee’s (Featuring the piccolo)

Numb– Linkin Park

What alternative flute cover would you like to hear in the future? Let me know in the comments! 

OMR Wrap-Up

If you’re an arranger like me who spends hours typing in notes that are already transcribed on a page, at least check out the trial version of this OMR program. If you scroll to the bottom of the page, you can pop in your email and get a free trial.

Thanks for stopping by the Fusion blog to learn a little bit more about optical music recognition with me! Be sure to check out our last post, Composer Interview with Kristopher Mariasy. What music/ art topic would you like to learn about next? Let me know in the comments. Until next time! 

Leave a Reply