ScotMus.com MIDI Policy

Function:The digital audio realisations of the musical scores published at ScotMus.com are designed to be understood in a similar way to the musical examples that are conventionally used in musicological presentations such as lectures, conference papers, radio talks, TV documentaries, etc. Traditionally, such examples were often played on the piano, but on ScotMus.com, they are almost always played on a range of virtual MIDI instruments. They are designed as representative audio "illustrations" rather than definitive performances. There are two main constraints that apply here:

(1) Current technological possibilities;

(2) Current musicological knowledge.

In addition to these, there is also the more mundane problem of how much labour-time I actually have available for the creation of high-quality sound-files. The primary objective of the ScotMus.com project is the production of high-quality musical score editions according to the principles outlined in my Transcription Policy. As such, there often comes a point in a given production cycle at which spending any further time on finding the best MIDI solution is necessarily sacrificed in order to get a decent score edition online at all. Happily, though, the medium of online publication does allow for the continual re-working of initially-inadequate audio examples. Overall, my sound-files are pragmatic works-in-progress, and I'm more than happy to take suggestions for revisions on board — especially from professional performers with a good working knowledge of historical performance conventions and instrument design (please read the notes below first, though).

One obvious point is that my MIDI realisations generally use equal-temperament tuning at modern "concert"-pitch (a' = 440Hz). There are two main reasons for this: (1) it makes my audio files more compatible with the instruments that most of my musical readers are likely to be using if they want to learn tunes from or just play along with my examples; (2) unfortunately, not a lot of deep research has yet been done into specifically-Scottish historical tuning/pitch conventions. On this latter point, we seem to be lacking significant numbers of modern-day heirs to Alexander Malcolm — any takers? It's particularly odd given that there's a surprising number of MIDI programs available that allow you to set playback tunings to "Malcolm's Monochord". When time allows, I'll give them a proper whirl.

My default transcription editor being Finale, most of my audio files are generated within the limitations of its virtual instruments and MIDI performance settings. Of course, Finale is only really designed to deal with approximately-post-1800 musical scenarios within the common Western traditions. It isn't designed to recreate the sounds of any uniquely-Scottish tradition. As such, I'm always on the look-out for alternative software (for example, the MuseScore open-source project is progressing nicely). But, given that my primary objective is the production of high-quality scores, I just don't have the time to spend on elaborate audio solutions that aren't at least partly compatible with the scores that I (mostly) generate with Finale. Having said that, Finale doesn't cover all of my notational needs anyway. It's a question of working on a case-by-case basis. The following sections elaborate on some of the most important problems:

(1) Instruments versus Voices

MIDI technology is at its crudest when attempting to reproduce the sound of the human voice. And since the entire history of Scottish music is dominated by vocal genres, this is a bit of a problem for me. On the other hand, the vast majority of our vocal repertoire either derived from or spread into the instrumental repertoire at some point in its history anyway (and normally pretty rapidly). Which is very fortunate. So, rather than completely mangling songs by churning them out with MIDI "voices" (which are practically all massed robotic choirs singing only either "ooh" or "aah"), my default position is to find a historically-appropriate instrumental substitute. Thankfully, the sources themselves often provide their own alternative solutions. A good case in point is William Thomson's Orpheus Caledonius which, although being primarily a collection of vocal arrangements, also included an appendix of recorder arrangements in the 1st edition (1725), updated as flute arrangements in the 2nd edition (1733). So, because most MIDI recorders are almost as bad as MIDI voices, I've opted to give my examples on MIDI flute, which is also a universal mainstay of the so-called "Rococo" / "style gallant" that Thomson was one of the first to deploy within Scottish music anyhow. So far so good.

But there are some MIDI arrangement problems that just don't have historically-viable solutions at all. Famously, seventeenth century Presbyterians of the Covenanting tradition actively banned the use of instruments from church music — they were even known to rip the occasional church organ to pieces, rather than just sell it off to the less aesthetically-pious folks south of The Border (I should add, though, that both before and after the seventeenth century, Kirk attitudes towards instrumental music were more diverse). So attempting to create historically-appropriate MIDI versions of seventeenth century Presbyterian Psalms currently has only two bad options: (1) have it sung by a choir of evil Daleks; (2) commit the even worse sacrilege of playing it on an instrument that isn't even controlled by a soul-less wee blobby alien demon lurking inside the tin. And since history clearly proves that not even the most liberal of Calvinists ever got on particularly well with Daleks, there's only one aesthetically un-viable solution worth using — instrumental arrangements that go completely against the Covenanter aesthetic. Either that or I could sing it all myself, but trust me, that's worse than listening to a choir of Daleks. Fortunately, though, Scottish Episcopalians often did use instruments in their church music, which makes my job arranging their stuff a bit easier.

The second major problem with historically-appropriate MIDI arrangements is that of early instruments. Most MIDI sound-fonts just don't cater for the viol, fidil, lute, virginal, hautboy, crumhorn, or anything else that was in widespread use back in the Medieval or Renaissance periods (and yes, all those listed were indeed as common in Scotland as anywhere else). And the few sound-fonts that do try and cater for them don't do it very well at all. However, since the vast majority of older instruments eventually developed into our modern ones, it doesn't stretch the imagination too much to substitute, say, a viola for a viol or even an oboe for a hautboy. On the down-side, in the case of The Skene Manuscript, the Spanish guitar doesn't sound much like a Renaissance mandour (wee lute). But it's still sounds closer to a mandour than a trombone would.

The worst headache of all is the MIDI bagpipe. I'm still working on this.

Again, it's all an eternal work-in-progress, and remarkably few of the problems are actually my fault. Obviously, a good workman shouldn't blame his tools, but he should at least be honest about the fact that his hammer can only be used for the crudest kinds of brain-surgery.

(2) Performance Conventions

I don't know of a single MIDI generator that has a "Scottish" default setting for any of the unique historical performance conventions that I'm trying to deal with here. But another problem is that there simply is no single global performance convention in Scottish music anyway, not even on one instrument. Apart from anything else, all styles of Scottish musical performance have been influenced by not only each other, but also many different foreign styles, on at least several crucial occasions in their history. Nothing is absolutely static, and everything starts looking more and more diverse, the closer you study it. Worst of all, hardly any of it simply pops out of a MIDI robot without vast amounts of techno-tweaking beforehand. Again, it's a question of how much time there is in the day to try and deal with it. And how much you value your sanity.

One thing works in my favour, though — the fact that ScotMus.com is, by definition, dealing primarily with the written traditions of Scottish music, and not the aural traditions. And all traditions of writing in all cultures have always had a useful habit of producing written commentaries on their own writing conventions. Major Scottish sources here are the handy tutorial books published by folks like Robert Bremner or John & Anne Gunn on subjects from guitar and keyboard technique through general theory. In particular, Bremner's undeniable contemporary reputation amongst Scottish musicians working within the written traditions clearly marks his tutor-books out as the key texts for understanding how most "Folk" publications of the eighteenth century were designed to be played. And the answer is, in fact, very "classically" indeed (whether you like the fact or not). But it's also important to understand that this was, like so many other things in these collections, an optional approach, and that other approaches were not only possible, but quite common. So if you prefer using "Folk" grace-notes, ornaments, etc. in your own performances, that's more than fine by me (actually, in many cases, I far prefer it). Nonetheless, since my main work on this website is to present the scores unaltered, I have accordingly adopted a similar policy in my MIDI illustrations of following what's written as closely as possible. Having said that, there just aren't enough hours in the day for me to work out perfectly-Bremner-esque or Gunn-ish solutions for every single grace-note and ornament. In fact, since their's are fairly standard "classical" solutions anyway, I've generally left mine in the "classical" Finale defaults, which at least approximate their general feel (hopefully, only specialist ears will be offended by this). But one day, I might be able to work out a way of doing it all with full accuracy without it totally eating my whole life up.

So, again, provisional compromises are often necessary.

(3) Bass Parts

It's a fairly well-known fact that most of the eighteenth century sources of Scottish music were published with "simple" bass parts. As you surf round my editions, this fact becomes more than obvious (because, unlike most modern editors, I don't edit them out). In some cases, such as ball-room fiddle music, these bass-lines were often just played by a cello. Thankfully, that standard performance convention is incredibly easy to re-create in a MIDI gig — for example, in Niel & Nathaniel Gow's Collection of Strathspey Reels (1784).

However, even in that Gow publication, the title page goes on to state that it's scored "With a Bass for the Violoncello or Harpsichord". Note the "or", though — what instrument you play the bass-line on is often optional (although not always, and generally within a set of given conventions anyway). Bear in mind that although in ball-room gigs, fiddle tunes were often done just with a cello accompaniment, the published ball-room hits were also aimed at the domestic amateur market. And, at least in the well-heeled home, playing both the tune and the bass-line on harpsichord for private fun was very common, too (although personally, I reckon Strathspeys sound like total mince when played as harpsichord solos).

But the other important point about a lot of these bass-lines is that, to some extent, they can be entirely optional to the point of being left out in performance — the tune's the thing, in a lot of cases. Many fiddlers, for example, would have been buying these scores without any intention of playing them with a cellist, harpsichordist, or anyone else on bass — either privately or publically. And, in fact, there are some tunes where the bass-line seems to be added more for the convenience of the home keyboardist than anyone else. With some tunes, it's actually very unlikely that anyone would have played them with an accompanist at all. Personal taste also plays a role here, of course. But, in my view (but not everyone's), some of the longer instrumental variations sets (especially "fiddle pibrochs") just sound like cack if you play them with the published bass part, regardless of how you arrange it. In my view (but not everyone's), there are some things that demand solo performance. The only question is which ones (and that's where opinions differ). So, you'll notice that in some of my multimedia presentations, although my score edition always retains the original bass-part as originally-printed, some of my MIDI realisations of them mute the bass part entirely. But don't take my views as absolutely prescriptive, whether they're based on personal taste or something more substantial — that's one reason I print bass parts, even if I don't always play them.

Equally, some collections that include a bass-line don't give any indication at all about what instrument(s) it's written for — a good example here is William Thomson's Orpheus Caledonius (1725). But in most cases, the conventionally-assumed default minimum of the day was the cello. And, that being so easy to do in MIDI (and in Finale, doesn't actually sound too bad at all), it's the solution I generally use — at least in the early versions of editions that I've only recently got online...

Why the apologetic proviso there? Ah, well, these bass-lines aren't actually as "simple" as they often claim — and they're certainly not as "simple" as a lot of folks are misled into assuming (in fact, the "simplest"-looking ones are often the most complex ones to fully work out in practice). Let's remind ourselves what the Gow collection of 1784 says: it's scored "With a Bass for the Violoncello or Harpsichord". Other tune-books list various other instruments that could be used, such as piano, organ, harp, and pretty-much anything else than can play chords. Now, when these kinds of bass-line were played on such instruments, the performer customarily vamped chords above the bass-lines given in the score, according to a widely-taught and well-understood series of conventional guidelines. In other words, these bass-lines are generally printed in a kind of notational short-hand for full chords that were actually played above them. Looking at it another way, many of these bass-accompanied tunes were, in a sense, published as incomplete scores that the user needs to approach as an arranger, applying compositional skills to flesh them out before they can be performed with any historical "authenticity" at all. The obvious problem here is that it takes time to work out how to complete these parts for a full version — whether you're going to vamp them in real-time (as some clever folks learn to do for live gigs) or scoring them up for a MIDI version. Apart from anything else, there's very often more than one possible solution, even if there are well-documented conventional "rules" to be observed. It all takes time, aside from skill, care, knowledge and (even more problematically) personal "taste" (which has no absolute rules!).

...So, as I've said, at least in the early releases of my MIDI versions, I generally opt to just play the bass-lines on cello alone, even in a few cases where this leads to slight wobbles in historical plausibility (for example, the bass-lines in Thomson's Orpheus Caledonius are far more likely to have been performed in their own day by a harpsichordist, not just a cellist). For adopting such short-cuts (although sometimes they're only temporary), I apologise to any jippy purists whose ears might be offended. :-)

Again, my only excuse is that it's all a vast and endless work-in-progress.

I might also add that you're not paying me a penny for any of it, so, ultimately, you can just like it or lump it. ;-)