NOTE: Since this post was written, I have moved my blog to its own domain. This Posterous site is deprecated, so please direct your responses to the new location of this post:

http://blog.mariusmasalar.me/2011/05/daws-vs-notation-software-no-contest/

Thanks!

-----

This rant goes out to the software developers who assemble and maintain the programs that I make a living with, so I hope it’s understood that the critique that follows is in the spirit of improvement and goodwill.

I used to be a die-hard Finale user, and indeed spent 6 years or so writing music using nothing but. This changed in two stages. The first stage was the expected one where someone like me, who writes music for media, suddenly discovers the realm of MIDI sequencers, also known as DAWs (Digital Audio Workstations). The second stage was switching to Sibelius, but that's less relevant — the important thing is that I discovered there was a second category of music production software. This was a whole new world and I admit that at first I really didn’t like it very much, especially since I had always approached music very organically and working in most DAWs seemed rather more mechanical than I liked.

As with any novelty, I soon moved past that point and learned my way around my DAW of choice (Logic) well enough that it no longer felt like a chore or a mechanical process — I was once again simply using the tool to translate music from my brain to a form that I could sell to people.

Today, it’s very rare for a composer in my position to get the opportunity to work with live musicians; budgets for the projects we get simply don’t allow it, and fewer and fewer directors and game designers are considering the huge expenses a worthy investment. To a composer, of course, the value seems self-evident, but in order to understand the market one must step back and put on the horn-rimmed glasses of those on the other side. Doing so reveals that many directors — and, by extension, many eventual audience members — simply cannot tell the difference between a live orchestral recording and a synthesized/sampled one.

This is, first, a testament to the ground that has been covered in the past few years with the quality of sample libraries and music technology in general, but it’s also reflective of the truth that perspective makes a huge difference. At the end of the day, in most cases even those that can tell the difference simply don’t care one way or the other. Especially not in the context of a boisterous film where their attention is 90% on the visuals and only 10% at best on the music.

This shift has meant that for the majority of us, the final product of our efforts is the synthesized recording. The term “mock-up” is still used to describe these products, but in fact the word is no longer applicable. Mock-ups used to be just that — approximations of the final product to give the director an idea of what everything would sound like once it’s recorded. Now that there often is no live recording, the mock-up is the last step. This has led to an increasing number of composers working as I do: directly and exclusively in our sequencers.

There are several reasons for this. First, I believe, is immediacy. For those who are able to write music without relying on notation for input, working straight into your sequencer cuts out the intermediary step of transferring a score from your notation program (or, archaically, manuscript paper) to the DAW. In the fast-paced world of media scoring, where deadlines are constantly nipping at the rear, this is a practical necessity as much as anything else. There simply isn’t time to do both steps, so we pare it down to one.

The second reason is that the notation becomes functionally irrelevant. If you’re not recording with live musicians and you can write just fine without having to work exclusively with notation, then what is the engraved manuscript for? The only valid answer is personal gratification. Maybe you print them out and keep the hard copies as tangible mementos of your work. I know I’ve occasionally done that, so it’s not like it’s unheard of, but the point is that the composer becomes the only remaining appreciator of that manuscript.

So what of it? I have noticed over the past few years that developers of notation software seem to have reacted strongly to this obvious trend. They are, no doubt, threatened by the idea that their product is depreciating in value among the comparatively large sector of composers involved in media work. And this reaction is what I take issue with and what I’m discussing here, because the focus has shifted from making a good piece of notation software to trying to build chimaera programs that try to be the best of both worlds and end up failing, inevitably, because the two paradigms are not meant to be reconciled in a single environment. That’s not to say it isn’t possible, but the result would be an unwieldy monster of a program that (so far) no one has had the resources to develop.

Surely though, trying to simplify the process and close the gap between writing and sequencing is a good thing, right? In principle, I fully agree, but in practice the implementation has so far revealed some serious flaws. It seems to be a path that spirals downward rather than up.

For instance, let’s take the example of notation programs allowing integration with VST plugins for sounds. This is now possible in all major notation packages, and it allows one to load advanced sound libraries and trigger and record their sounds from within the notation program. However, advanced sample libraries are designed to be controlled with complex MIDI messages and precise adjustments in volume, EQ, panning, etc. That whole process is what sequencers were designed to do. They do it very well and provide a level of unparalleled control that is crucial for getting a good sound.

Trying to make use of all that potential within a notation program is akin to buying a Ferarri and then driving it around suburbia at 10km/h. In fact, it’s worse than that. A better analogy would be buying a Ferarri and then putting it on a trailer and pulling that around with an old pick-up truck. Notation programs simply don’t give you the precise MIDI control needed to make proper use of advanced sample libraries.

Thus, developers have tried to resolve this by adding built-in effects, and systems like Sibelius’s SoundSet framework. At this stage, heros like Jonathan Loving and his SoundSet project have done their very best to provide some extra help to those trying to use sample libraries with their notation software. But despite their best efforts, their work is still clouding the fact that it’s a bridge between two countries that don’t get along.

I should say that for many composers who are not at all tech savvy, this rudimentary control is enough — they get better sounds than the built-in ones, and they don’t have to worry about a DAW. But rather than waste all that development effort in trying to get notation software to behave like a DAW and accept plugins for libraries that are designed for sequencers and professionals, wouldn’t it be smarter to keep the program lean and focus that effort on building a customized sound library designed for use within that notation program? Notion got this approach right, and if they built up the actual notation part of their software I’m sure they would take over the game entirely.

At the end of the day, it’s unfair to those non-techy users too: rather than provide them with a simple solution to get better sounds without resorting to a third party, this method forces them into the scary world of advanced samples, which they will then be compelled to purchase without the foggiest idea of the scope of what they’re getting, nor its actual potential. And then when they load it into their notation software and it doesn’t sound anything like the demos they heard on the sound library’s website, they get even more frustrated. How is this a good idea?

On the other end of things, in the sequencing world, it’s worth noting that developers have mostly acknowledged the inferiority of their notation views. Rather than trying to get them to compete with Finale or Sibelius, they simply work on making sure that they’re as seamlessly integrated as possible so that the basic functionality they have is at least perfectly implemented and streamlined.

It seems to me like there is a clear solution to be had. Rather than try to compete with sequencers, notation programs should slim down to become entirely focused on quality engraving, and spend the money and development effort currently directed at DAW-like features on putting together a better integrated soundset for their core users.

Furthermore, for those more advanced users who, like myself, may need a mock-up and and an engraved score for musicians, implement a system that allows for seamless integration between the notation program and the sequencer. Make friends with a sequencer developer and work out some sort of bridge that allows the notation program to serve as an integrated notation view for the DAW, so you can write in the notes and produce a printable score, and then seamlessly flip over into the DAW and exercise the power of that program to create a quality mock-up of the score with all the advanced libraries and synths at your disposal. Sibelius’ ReWire support is a fledgling example of this concept, and I hope to see them continue down that path.

It’s like this: there are 2 basic types of notation software users. The first are those who want to work only in their notation program and just want better sounds: these folks should not be forced to seek a third party or learn a new program just to accomplish this, especially if sequencing is not a skill they need to do their work effectively. For them it makes sense to have a better integrated soundset that gives them quality sounds to work with while they write — ones that are designed to work with the program so they respond appropriately and require little to no additional tweaking.

The second group of users are those who use both a DAW and a notation program. Their biggest problem is getting from one to the other. They do their writing in one and must sequence in the other, so having a method of integrating the two would save them the unnecessary step of exporting/importing a MIDI, or re-entering the notes, or otherwise trying to reconcile the two environments. Until everything can be successfully built into one program (and computers are powerful enough to run the resulting beast), the focus should be on consolidating the separate functions that these two groups offer and making working between the two as easy as possible.

In the end, I would argue that developers on both sides need to come together and stop trying to replace each other. They should work on acknowledging that they are servicing the same market, but in different ways, and their current attempts at being the best for everyone are ultimately leading them down a path that’s detrimental to the users.

Maybe it’s just me.