For many readers of this blog, Daniel Spreadbury surely needs no introduction. For the unacquainted, Daniel started working at Sibelius in 1999 and, when he left Avid in 2012, was its senior product manager. During his time there he independently started the very blog you’re reading now, for which he wrote hundreds of posts. So it was a special privilege to invite him to talk about what he’s been doing for the last two-plus years as product marketing manager for Steinberg.
For the better part of two hours we spoke via Skype about a variety of topics related to his current work on a new music notation program. An edited version of his remarks appears below.
Progress on the new product
Q: How are things going?
Things are going really well. It has been a learning experience. We came into this thinking, “well, we know how to do this, we’ve been doing it for years.” But when you’re faced with a blank canvas to paint on, it can actually be very difficult to make decisions. You start to realize how much you take for granted the foundation that you were building on top of.
As far as our progress is concerned, you can look at it in two ways. On one hand, you could say, well, it’s taking longer than we thought it would, when we first started talking with Steinberg about working together. On the other hand, we’re committed to doing it right the first time. With a relatively small team, and in a relatively short space of time, for the things that we’ve done so far, we’re achieving results that are superior to the current top-end scoring programs. The things that we’re actually going to have in our first version are going to be much better than the things that are already in these mature products, and we’ll have achieved it — by some definition — in a fraction of the time.
Goals and details
Q: You said you’re committed to “doing it right.” What does it mean to “do it right” when it comes to music notation, with its wide variety of styles and permutations?
Well, there isn’t one “right” answer, but there is a fairly narrow range of good answers. We want the program to get as close as possible to the desired end result automatically. All the things that we’re talking about in this regard are engraving details. We’re not happy with getting it right, say, half the time or three-quarters of the time. We want to be getting it right nine times out of ten, at the very least. There’s already a huge amount of sophistication in the existing notation programs. I’m talking not only about Finale and Sibelius, but also, and especially, LilyPond. LilyPond is the one that lives and dies on the automatic results that it gets because it’s so time-consuming and, for the average user, awkward to tweak the results that it gets. So LilyPond gets closer to what you might call traditional engraving by default than the other two do, but the other two make it much easier to tweak it.
Interestingly enough, I’ve been spending some time looking at beaming at the moment, comparing the default beam angles and stem lengths that are produced by various programs. We’re to the point that we’ve now honed in on tiny details, and I would say that already our beam angles are at least the equal of the defaults you get in any other program, but we’re still not quite satisfied. I recently picked up the Alfred Essential Dictionary of Music Notation to read it again, and one of the things that I found very interesting that hadn’t struck me before is that it says that it’s very time-consuming to adjust beams in computer programs, and so it essentially cops out on giving very specific guidelines!
That’s the exact kind of cop-out that we’re trying to avoid. We’re trying to make it such that when you type in the music, it just looks good. If you want to change it, you can — you’re provided with a number of tools to do so — but the goal is that music that is entered without any tweaking could be printed and put in front of a musician without the musician having any problems understanding it.
That’s a pretty tall order, because it encompasses not only fairly obvious things like page layout, staff and note collisions, and so on, but also how the music is actually notated: how the rhythmic durations are broken up; how beat groupings are represented; where you have ties and where you don’t; how you show the metrical emphases in the music. And then it comes down to the really minute details, like making sure that a tie doesn’t clash into a sharp, or that a beam angle isn’t misleadingly steep for a relatively small interval. All of these are things that professional music preparers spend an awful lot of time fixing up, because it makes such a difference when you put the music on a stand in front of a player for the first time. Most musicians can’t articulate why music looks right or wrong; it’s more subliminal.
So even if, as a user, you don’t like the very specific solution we’ve come up with to a given notational problem, we want you to at least find it acceptable. Depending on the amount of time and effort you’re then willing to expend to go beyond that, the default bar that we’ve set is that much higher, and the distance between there and exactly what you want is much shorter. We’re trying to make the program behave so that making local changes to a given tie or a given beam is not necessary; instead, you’ll be able to set an engraving rule that communicates your intentions not only in that individual instance, but also in other cases like it.
The user base
Q: Who do you envision using your program?
We’re unashamedly targeting pro users. When I say “pro users,” though, I don’t necessarily mean the technical term of somebody who actually makes the majority of their living from using notation software; those people are pretty few and far between. What I mean is that we are trying to build something that will satisfy those people, on the theory that if we can satisfy those people that can distinguish quality engraving, efficient workflow, and configurable playback, then we can more easily satisfy the broader universe of people who are currently less demanding, but who are on their way up that ladder.
We’re not saying that we aren’t interested in amateur users, but I would say that the least experienced user we’re targeting would be the university student — not the high school student, not the elementary school student — the university student, the music major, who is serious about doing music prep, music production, composition, or arranging. This is the program that is going to provide the kind of tools that those people need to do their job as efficiently as possible, and to produce results that stand up to the level that they would expect.
Existing programs like Finale and Sibelius surely target those users, but I don’t think they do as good a job of meeting the needs of those people as they might. That’s where we see the opportunity — to be the professional choice.