A couple of months ago, I was teaching a composition lesson when the subject of freely metered music came up. My student had been listening to Benjamin Britten’s Serenade for tenor, horn, and strings, and while he’d managed to check out a study score from the library, he wanted to know how much information the horn soloist had about what the other parts were doing. The library had the score, but not the parts. Since we were already looking at his music on my iPad, I was able to call up the horn part on the screen and we learned together.
I was able to do this because I was using an early beta of the nkoda music subscription service, which launched earlier this month for iPad on the iOS App Store. With an incredible library of scores and parts, and a promising application, nkoda could quickly become an indispensable resource for me as a composer, teacher, and performer.
When you download the app, you’ll need to set up an nkoda account before you do anything else. This allows them to sync your data across devices, and eventually across platforms (more on that later). You’ll then need a subscription through the App Store at $9.99/mo. or $99.99/yr., both of which start with a 30-day free trial. [Update: As of November 2018, the free trial is 14 days, not 30.] You’re also asked a few on-boarding questions about musical interests to help tailor recommendations.
After that, you’ll be greeted by nkoda’s Discover tab, which will look familiar to anyone who has used a music streaming service or mobile app store in the last few years. There are featured scores and playlists such as “Living Composers”, “Evolution of the Symphony”, “Easy Piano”, and even “Graphic Scores”. Despite the name, the playlists are curated collections of scores. There’s no media playback in the app.
If you have something more specific in mind, you can search the library by composer or title. The search is helpful, but is relatively simple at the moment. You can find works by composer (bizarrely referred to as “artist” throughout the app, a term as out-of-place as “playlist” in this context). However, multiple translations of titles and transliteration of names can throw off searches, and there are no useful filters for things like genre, instrumentation, or time period.
Once you find a work you like, you can save it to your library, download it for offline use, or add it to your own playlists, which are shareable to other nkoda users. This could allow a string quartet to have playlists of prepared repertoire for a program, or teachers to have a set of scores to study for a class.
The library is the single greatest asset of the service, and it alone is worth the subscription price. You will find a huge collection of scores and parts from composers throughout history, including works that have been published in the last few years. Some of the scores here are for works that are only available to rent, and these are exactly the sorts of scores that can be hard to find in libraries and other means. With a large number of publishers, it’s even possible to compare multiple modern editions of many works, such as beautiful Bärenreiter Urtext editions. You’ll even find the new editions of Copland’s Appalachian Spring engraved by Scoring Notes‘s own Philip Rothman!
The nkoda library isn’t perfect nor comprehensive, but it is huge. Approximately 50 publishers have some content on nkoda, including most of the well-known ones; a full list is at nkoda’s web site.
While on the beta, I would regularly spend (waste?) hours digging through scores by Elliott Carter, Thomas Adès, and even piano-vocal collections of popular tunes from musical theater or Radiohead. If you’ve ever listened to a recording and thought to yourself, “I wonder what the score looks like for this”, there’s an excellent chance you can have the answer in just a few minutes with nkoda.
Note: Hamilton is in the library, but here in the U.S., I get a message saying that it is restricted in my region. There may be similar restrictions on certain scores where you live. I would hate to find out on an international tour that I was suddenly locked out of my scores. That said, this seems very rare, and Hamilton is the only example I could find that was restricted from U.S. use.
There are a few notable issues with the structure of the library and the library entries. Some entries have the score but no parts. A few listings have neither scores nor parts, such as Steve Reich’s Tehillim. And several works have two listings, one with the score only and another with the parts only (as in Sofia Gubaidulina’s String Quartet No. 3).
This is another known issue, and it stems from the fact that they are getting data in different formats from different publishers. It’s reminiscent of the metadata problem in classical music recordings.
There is also a wide range of image qualities. All are legible, but some pages look crunchy and pixelated when zoomed in and others look strangely over-smoothed. Unfortunately, nkoda is really at the mercy of its publishing partners to supply them with the highest quality assets, and the quality of printed materials can also vary. Part of nkoda’s processing is to take all pixel-based (raster) PDF images from publishers and auto-trace them to vectors so that they look sharp on any screen. The negative side-effect is that lower-quality source files from publishers can yield some very strange shapes when zooming in. This is mostly fine for normal, full-page reading. Thankfully, it’s only at very high zoom levels that these artifacts present themselves.
nkoda CTO Sundar Venkitachalam told me that in the future, nkoda hopes to add support for other formats, such as MusicXML, which would allow for much sharper and more flexible display options.
Using your own PDFs
The developers of the nkoda app understand that no matter how many publishing deals they make, it’s likely that you’ll need to store and use music that isn’t in their library. To accommodate this, the app allows you to upload your own PDF files. As a self-publishing composer, I was very happy to see this feature included. It allows nkoda users to have all their music in one app.
Unfortunately, there are many limitations to this feature. You can upload from any iOS document provider that shows up in the Files app, like iCloud, Dropbox, Google Drive, or Box. Immediately after that, you’ll hit the first snag. Right when you upload, you’ll get a message that your music is processing, which “normally only takes a few hours” (emphasis mine). So much for a quick score update before rehearsal. In my experience it took about ten or fifteen minutes, but who knows how long it might take as the service grows, and it seems there would be no way to use your own PDFs if the nkoda cloud infrastructure ever went offline. Encouragingly, an nkoda representative told us, “We are speeding up the PDF ingestion time and expect it to have improved by the next release.”
The processing itself is also a bit strange and opaque. The uploaded pages are cropped and processed much like those from publishers in the library. The results were not quite as odd as some of the extremes of what I observed from major publishers, but the results were noticeably worse than simply loading a PDF into Newzik or forScore. And I noticed at least one example where a page was cropped to remove intentional white space in the document. There is no obvious way to prevent or control this processing.
Another frustration with uploaded PDFs: There is no way to do anything with them. You can’t share them with other users (for fear of piracy). You can’t share annotations. You can’t even add them to a playlist. As excited as I was that nkoda allowed me to upload files, I was even more disappointed to learn that those files would always be walled off from the rest of the experience. If PDFs are a big part of your current performance or teaching practice, keep them in whatever app you’re already using. In its current state, nkoda is a very poor place to use your own materials.
Navigation and Score View
When you select a score or part to view, you’re taken to a scrolling navigation view of the PDF document, known as Navigation View. From that view, you can quickly scroll to the first page of music, as most of the documents in the library include the cover and front matter. You can bookmark significant pages, like the start of each movement, and even indicated repeated sections where the reader might need to flip backward a few pages in performance.
Once you enter the Score View by tapping a page, nkoda looks like any other viewer you might use for PDF files. Navigation will take some getting used to for those with experience in other score readers. You can move forward by tapping in either the top or bottom right corners, and backward by tapping in the top or bottom left corners. There are no swipe gestures, as I would have expected. Exiting a score requires tapping on a very small area in the vertical center of the screen along the right edge. It usually takes me more than one try to hit this spot, and I hope the feature gets some refinement. It seems like a bad design decision when the help documents need to include a diagram for such a simple and common action.
Many users familiar with other score readers will be happy to know that nkoda also supports some page turning pedals. It doesn’t support quite as many as other apps, but it will support any pedals the emulate keystrokes (Up/Down Arrow, Left/Right Arrow, or Page Up/Down keys). Unfortunately, for those who own a Bluetooth MIDI pedal, like the iRig Blueboard, you’re out of luck. Dedicated page turning pedals tend to fall in the first category, and multipurpose pedals like the Blueboard and sustain pedals tend to fall in the second. If you rely on a pedal, be sure you test yours during the trial period.
Performing in nkoda is just as practical as performing from any other iPad app. If you’re using something like forScore, you’ll probably be able to pick up nkoda quickly. My one quibble is that the touch targets in nkoda are a bit too narrow for me. In other apps, I either tap nearer the vertical center on the right side, or swipe right-to-left to turn the page. Neither of these is supported by nkoda, but I think I could get used to the nkoda touch targets given time.
When first testing nkoda, I was very disappointed to see that I was not allowed to share my iPad’s screen to a projector, something that I do almost daily in my teaching practice. Currently, if you try to take a screenshot, make a screen recording, or project to another screen, you get a very scary message that “Screen capture is not allowed. You’ll be blocked after three captures”! This is a big, frustrating barrier to using it in an most educational settings, or even just demonstrating the app to a large group. Thankfully, this is another issue that the nkoda team is planning to resolve very soon. My expectation, based on my conversation with Sundar Venkitachalam, is that screenshots will still be prevented, but that AirPlay and other forms of presentation will be allowed. This will be a great help for me, and I hope it will be resolved before classes start in the fall.
Update as of July 11, 2018: This review originally highlighted the fact that presenting scores from an iPad were restricted. As of nkoda 1.0.4 on iOS, released on July 8, 2018, that limitation has been removed. With the latest version of the app, I was able to present both via a wired connection from the iPad’s Lightning port and wirelessly over AirPlay. In both cases, the full screen is mirrored, and the presentation shows the original score as well as annotations. I think this will prove to be a huge benefit to teachers and students using nkoda in schools.
Annotations are simultaneously the most promising and the most frustrating parts of working with nkoda. At first, they look and behave mostly like annotations in other apps. The annotations drawer is permanently attached to the left edge of the screen when looking at any score or part. It includes all the normal markup tools you would expect, like a pencil, highlighter, text entry, and eraser, and you can change the size and color of your annotations. I do my markup with an Apple Pencil, and unlike many other apps, nkoda seems mostly unaware of the Pencil. Because it’s an active stylus, apps like forScore and PDF Expert can distinguish Pencil and finger input and treat them differently. nkoda does not do this.
Writing is also less responsive than drawing input in other apps, and even after the line has started, it actually changes from a jagged to a smooth line a split second after I lift the stylus from the screen. There is also a “magic pencil” input that is supposed to smartly convert handwritten markings into musical symbols. It’s a good idea, but it rarely works in practice. No matter what I draw, it always seems to assume I want a series of tiny slur arcs, rather than an accidental or a hairpin. And even if it got this right, I’m writing in the score because I want information to stand out from the printed markings, not blend in with them. Editing or erasing mistaken markings is even buggier, and I have crashed the app twice while testing annotation editing to write this review (the final release, not a beta). On the whole, nkoda presents a rather poor writing experience.
I mentioned that annotations are also one of the really promising parts of the app, and there are two features in particular that lead me to that determination. First is annotation layers. These work like layers you may be familiar with in most image editors, and just like those layers, they can be combined and hidden however you want. For example, in preparing to perform a work, you might have a set of analytical markings in one layer and performance reminders in another. You may not need to see your indication of a contrapuntal inversion during a concert, and you may not need to see a thrice-circled pair of eyeglasses when doing a harmonic analysis. I could also imagine having different sets of markings for performing the same piece with different ensembles.
The second promising annotation feature is shared annotation layers. If everyone in the choir is reading from nkoda, the conductor could mark up cuts or added fermatas and dynamics in their score, and share that annotation layer via a link with the rest of the choir. Thanks to the layer separation, this shared layer need not include things that only the conductor needs, like reminders for cues and meter changes. Imagine a concertmaster putting bowings in her part, sending it to the librarian who checks it over and makes a couple tempo changes per the conductor’s preference, and then sends the whole layer to the strings with a couple of taps! I would love to mark up score for analysis and send it to my orchestration students. However, it is worth noting that these markings would not transfer between parts of the same piece (say the first violin and second violin parts), or between the score and parts. The nkoda app doesn’t have much semantic understanding of the score’s contents, only the image. Also note that this shares a copy of the annotation layer, so any changes made in the original won’t update those copies.
If the basics of annotation improve, the layer and sharing features could be a major win for lots of different kinds of musicians. One thing to keep in mind, though: your annotations are locked to nkoda. There is no way to export them, so if you ever decide to part ways with the service, you lose access to that data until you resubscribe. If nkoda goes away or changes their business model, you could lose your annotations entirely. forScore, by comparison, allows exporting in its own 4SC format, clean PDF with no annotations, and annotated PDF.
nkoda is available now for iPad from the iOS App Store. After a free trial period, the subscription fee is $9.99/mo. or $99.99/yr in the US; pricing may vary in other countries. It’s available in many countries around the globe, and supports English, French, German, Spanish, Portuguese, Italian, Russian, Polish, Japanese, Korean, and Chinese (simplified).
This is just the beginning for nkoda, and the team has great ambitions to launch improved search, reading, annotation tools, and more, on top of their planned applications for desktop and mobile platforms. The library is continuing to expand on a regular basis to include new publishers and more works from existing publishers. There is a lot of data to go through, and processing each new score will take time, but knowing that there are already deals in place for most major publishers makes me optimistic that more of my favorites (Karel Husa, anyone?) will be showing up soon.
The iOS beta lasted several months, and during that time, I corresponded sporadically with a person whose title is “Musicologist in Residence”. To me, a technology startup like nkoda creating subject-specialist positions like that is a promising sign for how seriously they take their responsibility to musician-customers. For all its faults — and I have outlined some very serious faults in this review — I have been impressed with the refinements that have been made over the beta period, and I am optimistic about the future of the service and the app. The software team is currently working on nkoda apps for Android, macOS, Windows tablets & desktop, and iPhone.
For now, though, if you’re a regular iPad user for sheet music, it’s worth taking the free trial for a spin to see how much of your repertoire needs can be covered by nkoda. For me, when I have the option of using paper, forScore, or nkoda, I’ll pick forScore almost every time. However, as frustrating as nkoda’s many limitations are, I find the breadth of its library to make it worth considering. In fact, I usually don’t have the option of using paper or forScore, because some of the scores I want to study aren’t as easily available as they are in nkoda. That’s why I plan to start paying for it when my trial expires next month.
All photos and screen captures in this review were taken by David MacDonald for Scoring Notes, and not furnished by nkoda.
This sounds very interesting and enormously useful.
Is anything known about how the publishers/authors are compensated? Presumably, some sort of royalties-agreement with each individual publisher must be in place, and publishers will then adhere to own contracts with authors. Though, I’ve heard that Music Sales has sold off printing rights to Hal Leonard, and many publishers already provide scores on Issuu, so this is certainly the future. The “problem” could arise with hire material. If hire parts are so readily available, I find it dificult to imagine that the publishers have any possibility of actually enforcing the copyright. Percentages of hiring fees can be a significant source of income for a composer.
Thanks for commenting, Martin.
Like Spotify, I’m guessing that the terms of the content deals are carefully protected. But I think it’s safe to assume that the publishers are compensated based on some measure of usage of their scores. Of course, as you point out, the extent to which that would redound to the composer isn’t up to nkoda; it’s up to the terms of the publisher-composer contract. Issuu scores are explicity for perusal, and the Issuu platform doesn’t have the tools for performance, annotation, organization, and sharing that nkoda has.
Regarding rental materials, some newer works on the platform show a banner above them which state that users must hire materials from the publisher to perform them, along with a link to where to do that. You can see an example of this in the first screenshot under the Library heading above.
I don’t think it’s any harder for publishers to enforce rental terms here than from photocopies. If anything, the fact that hiring parts can be done virtually (without shipping paper across the globe) makes it easier and more appealing to do correctly and legally. I think this is yet another example of convenience bringing greater value than free or cheap.
Hereby I cancel my free trail NKODA!