



                  Copyright 1996 S.D.Rodrian


Neither this text or the contents of these MIDI music files are in the
Public Domain. You may copy and/or distribute only the zipped
GOLDBERG.EXE file with 'gold-a.mid,' 'gold-b.mid,' and gold.txt' in it.
You may write for other permissions to me at sdrodrian@aol.com. This is
especially directed to those who like to 'correct' 'wrong notes' and
other 'errors' that such 'self-geniuses' happen on in works like mine.

J.S.Bach's Art of The Variation (commonly known as The Goldberg
Variations) is a set of 30 pieces using an Aria movement both as point
of departure and as its ultimate destination. (The basic figured bass
has additional significance, but not as far as this little note is
concerned.) But the character of this Aria, although its notes are
identical in both cases, should be quite different in each of the two
situations. Which highlights the entire approach of this interpretation:
Here the 'set' of individual pieces is at every point considered an a
complete unity.

Have you ever heard a complete performance of The Goldberg Variations?
Or, have you heard only the individual variations sequentially played...
with a long enough pause between them to render each piece logical in
and of itself?  Well, this is the only 'performance' (to my knowledge)
of the Goldberg Variations in which The Work is rendered as a single,
unified piece, played without interruptions throughout except after
Variation 15 (Variation 16 being, or course, an Overture).

This is not as easy an accomplishment as might be imagined--Played
individually, each variation needs to takes off literally from silence
and then lead up to yet still another anticipated silence. This means
that silences themselves frame the nature and character of each piece.

Conceived as an unbroken whole, and played throughout without
interruption (the way it's done here)... each variation is framed by the
preceding piece, and anticipates the variation that follows. This method
not only gives unity to the work, but also enriches the interpretation
every step of the way as it moves along to its culmination. (Of course,
with a midi recording, the experiment to perform is to listen to a given
variation in the context in which it is placed here, and then listen to
it again, this time just by itself... the computer does not 'know' it's
playing the same piece in two different contexts, so it will not make
the adjustments a human player performing the same experiment will
invariably find it impossible NOT to make.)

This will immediately show you why to understand the unity of this music
requires that listening to it in two uninterrupted segments: The only
interruption comes between the 15th and 16th variations (and just as the
opining Aria is executed like a right and proper introduction and its
reiteration at the end of the world is played in the manner of a
finale... every note of the 15th variation must cry out that it is the
place to which all the previous music had been leading to all along; and
every note of the 16th variation must announce its role as an overture).

The pieces can always be presented consistently, of course, and thought
of as a whole (except for the few places in which I omitted the final
chords). Thus the variations can all be played in the same spirit and
style as one would play a sonata movement, rather than as the collection
of an arbitrary number of disparate pieces.

But always listen to the music, not the instrument. Very few people
really understand the basic fact about music: It is THE most basic
language of all (most animals use it, practically all mammals), and like
all languages its logic--the way it is understood--is based on a
standardized grammar, usage... sentences, paragraphs, and phrases most
of all (remember how trying listening to President Carter as he placed
pauses in the middle of his phrases): The difference between the good
musician and all the rest of'em is that the good musician understands
and interprets the logical phrases within the music.

The bad news, of course, is that this piece was not originally written
for the pianoforte and in order to approximate a clavier-playing style
the fortes and pianos must always be held in strictest control, within
as narrow a range as possible (but this something which is tailor-made
for the computer, of course): A humming piano would muddle a
performance, but the steely brilliance of the Midi piano is about as
close as one can get to a marriage of sprawling piano sound and that
dry, unmessy purity that is the hallmark of the clavier.

This is also just about the ONLY uncut 'performance' of this singularly
Great Work you're ever likely to run across: All the repeats are played.
Also: In the beginning I added subtitles to each variation to help me
keep in mind the character I was after for each of them--and I have
retained these subtitles inside these files even when in most cases my
current approach to them has evolved from what the subtitles are saying
I tried to get at originally.

It remains my own individual interpretation, of course: You will hear
unique chord & trill modifications, including omissions and inclusions
which, I'm sure, will outrage all purists and traditionalists (for, if
nothing else, THIS is the most untraditional interpretation of The
Goldberg Variations you are ever likely to hear). But I hope that a
conscientious hearing will eventually lead all listeners to the
conclusion that it is the instrument that dictates the performance (and
this particular 'performance' is wholeheartedly wed to the Midi piano).

But rather than computer-played it is a computer-aided performance:
Instead of recording it into a sound-only recorder and then re-recording
(piercing together) corrections, this 'performance' is encoded from the
start into a MIDI transport (Master Tracks) so that corrections were
easily made until the end result was as seamless a performance as I
consider it to be finished.

Because of its existence as a MIDI file, the performance can always be
recreated in just about any MIDI-equipped instrument (including any
standard/conventional acoustic piano connected to a MIDI interpreter).
Just keep in mind that you can't just take a piece for the violin and
play it on a saxophone. The character of a piece of music, as it is
interpreted, is ALWAYS dictated by the instrument through which it is
being played; and this particular interpretation is specifically adapted
to the Sound Blaster AWE32 piano program #1.

This is the music that plays in my head, not through my fingers; so you
are in no way listening to the limits of anybody's playing skill:
Because it is the computer that is performing the music instead of human
fingers every last note plays exactly where I intended it to play with
such a degree of accuracy bordering almost on the absolute: What you
hear is what the Goldberg Variations sound to me from the music on the
paper. (Which brings up the worst part about all computer-generated
music: the clock ticks are so accurate that the hardest part of all is
probably to prevent the notes from sounding too close to the tempo:
There are still a few 'rough' passages in this version in which this is
the case, but I will not tell you which ones they are because this will
only emphasize this defect.)

Nevertheless, no work of art (and all works of art are human) can ever
be considered perfect. And, of course, this one as well: There are
passages that perhaps should be faster or slower, there are phrases
which should be a little tighter. But 'fixing' them is a future endeavor
(when I have more time, and wisdom). For now, what remains is as near
perfection as I would have it be--whether anybody else agrees with me or
not; this, after all, is where every maker of art parts company with
every critic of it.

You have to realize that I've had the benefit of sitting in front of the
music for days at a time and consider, and reconsider, exactly how a
particular phrase or ornamentation would be played without such intense
focus having any effect at all on any part of the music other than just
that note or group of notes I was concentrating on... long enough for
reality itself to quite literally become a wash and dissolve into utter
sameness.

The interpretation is always guided by a double purpose--to introduce
absolute unity (the interpretation of each variation depending entirely
upon the manner in which the previous variation was interpreted... an
approach to the music more for its intrinsic meaning than for its
historical contexts); and to treat the music (not its historical
position) on a personal lever: I regard the music throughout as
contemporary as if Bach had handed me the manuscript personally, rather
than historically (so there is no need for me to resort to any
historical traditions outside the most obvious and basic). The only
purpose I serve is the music's, exactly as it is set down in its
notation--always to be interpreted by the individual musician who may be
playing it at the time.

I add a few ornamentations, which the nature of this, my very unique
'instrument,' demand, and which I have the freedom to include (not
having to fear the wrath of first-row 'purists' and the other intense
maniacs one runs across all over the place... I might not fully
understand Glenn Gould's paranoias, but I can almost sympathize with
them), no matter how unorthodox or downright outrageous they my be to
anybody else. I also exclude any and all ornamentations which in any
way, shape, or form, distract from or arrest the flow of my own
particular musical reading, or simply do not jive with the sampled midi-
piano for which I intended this 'performance.'

I have seen too much contrapuntal music played as if the different
voices had absolutely no relation to each other and just only happened
to be being played at the same time. My view is that at every point...
if a given voice is not the focus of the music then it must support and
advance the one voice which IS the focus--Ignore this dictum and you run
the risk of making baroque music itself alien and out of favor (and then
any whiz kid with a trunkful of one-fingered melodies may become all the
rage).

There is usually a huge amount of ornamentation in baroque music for the
keyboard, and the Goldberg Variations have more of it than you can
discover from just the look of the printed notation or a casual
listening: This is because baroque ornamentation seldom becomes
distractive. Unlike Classical ornamentation, which is much too often
simply an out-and-out aside, baroque ornamentation (and most especially
of all in the music of Bach, who is both the focal point of the baroque
as well as its culmination), baroque ornamentation is almost always
integrated into its content in the same manner that gargoyles and little
angels are an integral part of baroque architecture.

In this sense, for example, the trills of the 7th variation are
particularly telling: Fully developed they sound so sweet in a plucked
instrument; but, obviously, the hammerklavier does not work by plucking:
In this sense the piano is strictly a percussion instrument. And those
same flowing trills almost sound like they're being played on a snare
drum with chop sticks. The alternative is to change the entire character
of the whole interpretation from a seriously, intensely pianistic one
throughout (which is what I strive for) to one which bounces from the
monumental to the lovable. Glenn Gould does this with some success (but,
then again, he never tackled my ideal of an interpretation whose
hallmark throughout is this uncompromising an unity).

Two files are included: "gold-a.mid" holds the opening aria, followed by
variations one through fifteen, while "gold-b.mid" holds variations
sixteen through thirty and the reprise of the Aria.

.MID files can be loaded into almost any music sequencer program; and
there are dozens of them. I use Master Tracks for the IBM.

To approximate the way in which I configure my own equipment to
listen/record the piece as if it were being done in stereo (and remember
that the most crucial aspect of a good reproduction will be the quality
of the sampled piano):

I use an AWE32 Sound Blaster (with sampled instruments). To simulate
stereo I have distributed the notes among four tracks. once you load the
MID file(s) into you own sequencer (preferably MASTER TRACKS) you can
then either copy the separated tracks back to track one, or break them
up among even more tracks for an even more realistic stereo effect.

In Master Tracks, I then 'pan' each tracks (with its own Channel, or
Voice) so that the higher notes are played mostly to the right and the
lower notes mostly to the left. In Master Track Controller #10 is the
PAN controller (but I am not familiar with other sequencers): The values
for each of the four tracks channels are (1) 84, (2) 76, (3) 65, (4) 51.

I also connect my Sound Blaster card to a dolby radio/amplifier/
receiver (Pioneer) set to 'simulated' stereo mode, delay 15 (if your
receiver gives you a choice between Phantom/Wide/Normal for its 'center'
mode choose Phantom first, and then experiment, of course, and always).
This will enhance the bass/treble and stereo effects.

My own AWE32 Sound Blaster's master controls are set as follows: I raise
the bass and treble as high as they'll go. And so too with the overall
master volume (also as high as it'll go). While the specifically Midi
volume is set to 80% or so.

Remember that these files hold a recording, not a computer's rendering.
So if you wish to convert it into a multi-instrumental piece you will
have to either manually separate the individual voices into individual
channels... or acquire a straightforward note-by-note input version of
the manuscript. The only computer-generated parts of this 'performance'
are the tempo corrections (you can see these in the Tempo Map screen of
your sequencer), and when the computer, like any other recording
machine, plays the recording. Thus, the tempo of the music will seldom
agree with the tempo/meter markings whatever sequencer you use to play
your midi files through may show.

The notes in these files will transmit exact ON-Velocity values to a
synthesizer, or a PC card, which can recognize ON-velocities the
interpretation will be exactly as I intended it: ON-velocity values
allow you to specify the volume of each note. (So if your synthesizer
and/or PC midi-card cannot recognize On-velocity values, again, you may
not get the exact results intended by me.)

I record (ultimately to cassettes and VHS HiFi cassettes) from a Pentium
120 running Master Tracks Pro with the following settings: Sync
Source=Internal; SMPTE Format=30 Frames/Sec.; Timer Resolution=Extra
High.

My PC midi card is a Sound Blaster AWE32 with a bank of sampled
instruments. I cannot guarantee how it will turn out on another card.


(C) 1996                                  Mr. S. D. Rodrian
                                          1121 Deutz Avenue
                                          Trenton N.J. 08611
                                          ( sdrodrian@aol.com )
