Over the last few weeks, composer-pundits have been have been throwing a collective tantrum over a new feature announced in an upcoming release of MakeMusic’s Finale. The feature allows users to scan a score and import that into Finale. Furthermore, users can import a PDF that has been previously scanned. This is the part that has everyone talking like MakeMusic has introduced a feature which will retroactively destroy all music you’ve ever created and possibly eat your cat. The outrage is primarily coming from John Mackey and Jennifer Higdon. I should say that I have immense respect for both of these individuals: they do fantastic work of making their scores readily available for perusal and purchase. I have often used this openness to study their works. However, they are doing so by distributing restricted PDFs that cannot be printed. Finale’s PDF import feature will be able to import these PDFs just like any other. Stated like that, the torches and pitchforks seem justified.
However, Philip Rothman offers a crucial, “turns out” rebuttal:
[Finale’s PDF import feature is] really a re-packaging of existing music OCR technology, which has existed for more than two decades. Just like any other document on the planet, if you can see it, it can be read and interpreted by OCR. It doesn’t matter if the document is a piece of paper, a print-restricted PDF that’s viewable on a computer screen, or an image on your phone.
I’ve been trying to come up with these words for a week now, and I’m glad Philip got there before me. The issue here is not the technology. I scan things from my phone all the time, sometimes from other screens. Streaming audio and video works the same way! Any time you’ve got encrypted or DRM-protected media, there’s necessarily a point where it is unencrypted for consumers’ eyes and ears. At that point, all bets are off, and there is literally no technological way to prevent it.
The worst possible outcome of this kerfuffle is that composers like Mackey and Higdon–not to mention major publishers like Boosey & Hawkes–remove or limit access to perusal scores on the Internet. This would be a completely understandable reaction, but would make us poorer as a music community to lose this resource.
At some point, you need to trust people to do the right thing. I can walk into a library, pull a score off the shelf, and do exactly the same thing that people are fretting about. Even without scanning, I could just input the notes into Finale and have a clean copy just like that. This would be possible with any perusal score, digital or paper.
Earlier this year, Microsoft announced that they had created a machine intelligence called Tay that would chat with users and learn from them. Within hours, Tay started spouting off some pretty vile racist remarks. This was not because Microsoft engineers are Neo-Nazis (they aren’t). Rather, Tay was simply learning from the horrible racist remarks people were saying to her. Just like the PDF-scanning feature in Finale, the problem with Tay isn’t the technology. It’s people.
So to composers and publishers: keep your stuff right where it is. Put more stuff online. Don’t kill such a valuable resource because someone might abuse it. After all, we’re not talking about a Nirvana record on Napster. The market for these works are musicians and educators. We “get” the economics of the whole situation. We want you to keep writing more great stuff and distributing it online. Besides, I’m much more likely to plop down $100 for a piano/vocal score of Higdon’s new opera Cold Mountain if I can take a look at it first. I see this as a virtuous cycle between publishers/self-publishers and performers. Note that the cycle doesn’t include software developers. I would hate to see a technological development in Finale break the wonderful, direct composer-to-fan relationship the Internet has fostered.
- Yes, Sibelius has been doing this for a decade. If there was ever any sturm und drang over it, I never saw it. People are weird, right? ↩
- Also, while this is probably only a temporary limitation, I’ve never had a score scan completely correctly the first time in Sibelius, which is using nearly identical technology. Usually, I find that unless it’s an exceptionally clean engraving with no mixed meters, tuplet rhythms, or other features of music written after 1860, it’s more trouble cleaning up than just entering notes the first time. ↩