Now that I am working and earning some income again, I am starting to look at investing in some more audio gear for recording my music. Since most of the gear I am looking at is digital and computer-based, it is interesting comparing the modularity of "music information" systems with more data-oriented information systems.

What is complicated in evaluating these music systems (like those by DigiDesign or Steinberg) is that there are only a couple data interchange standards, and otherwise, most of what these systems do is done using formats and interfaces proprietary from one manufacture to another. So, basically any choice one makes is to use one proprietary system and not others.

At the same time, there are both analog audio signal and digital information formats (e.g., S/PDIF, WAV files, etc.) that can be transferred from one system to another, and also interaction protocols shared by different systems (e.g., VST, HUI, etc.). In general, these proprietary music systems try to use each other's formats at least to support switching tools if not full collaboration.

Altogether, I think there are a lot of similarities between the software architecture philosophy of these music systems and those of systems more regularly discussed on this blog, like ones that use XML, RDF, RSS, etc.

But, what I think is quite different in the digital music information sphere than in the more general information sphere is what I would describe as a drive towards differentiation and diversity rather than a drive towards unified standards and interoperability.

With the drive towards differentiation, there is a lot of development around plug-ins and translators, as well as development of interfaces that allow the chaining of multiple systems in a modular fashion (e.g, SMPTE time code). In other words, interoperability is often realized through supporting redundant proprietary interfaces rather than requiring more singular and standardized ones.

I think this kind of diversity of interfaces has some annoyances, but might be a more healthy model than seeking single standards. At the same time, some key singular standards have been critically important for enabling this diversity to be healthy rather than self-defeating (e.g., MIDI).

As an outlook, I think with music systems, the design is capitalizing on individual's different tastes and styles in the consumption of sound and music. With more data-oriented information systems, I think the design is focused on capitalizing on impersonal business and machine consumption.

Basically, with these digital music systems, there are numerous points of human interaction for which there is essentially no desire to bypass the human. With more general information systems, it is the opposite: there are numerous points of human interaction for which there is essentially a great deal of desire to bypass the human.

Perhaps this is a "law" in the making: the more desire there is for human interaction to be part of an information system, the more system interfaces are required, the more diverse those interfaces must be, and the more redundancy there will be in ways to accomplish things (the more ways there will be to do the same thing).

So, maybe this has already been obvious to everyone for a long time. But, I note all of this because I see the kind of information that interests me as being experienced more like music than like mere "data". The nuances of individual tastes and styles, to me, is an essential thing to be able to "read" in information.