8bitfiles.net/archives/archives.tar/archives/genie-commodore-file-library/Information/SID-CLASS-1

Text File | 2019-04-13 | 12.6 KB | 194 lines

Minutes:
<[Bruce] WHOLE.NOTE> The purpose of these classes will be to provide an
introduction to using a SID Editor to create SID music files.  Required
materials to do this are the book/disk combination "Compute!'s Music System for
the Commodore 128 & 64" by Craig Chamberlain - this is commercia software -
and the Stereo Editor that is available to download here in the Flagship -
#8027 STEREO.ARC.  STEREO.ARC is deliberatly incomplete, requiring a file to
be added to it from the commercial program disk in order to function.
     These classes are intended to supplement the information in the book and
the program documentation.. not to replace them.
     The best way to learn to do SID music is to just dive in!  I suggest you
dive into fairly shallow water for your first plunge or two though!  It is best
to start by doing a 3 voice SID of a fairly simple song that you know quite
well.  And preferably one that you have a recording to listen to as you do it.
     Most SID artists work from sheet music.. almost all in fact.  Try to
select sheet music that is not too complex... you will only be able to play 3
notes at one time in a SID.  Therefore, as a beginner, try to find music such
that not too many will need to be deleted.  Later you will feel comfortable
tackling much more complex music... I have reduced from orchestra scores for
a dozen or more instruments.
     I recommend that you take the time before you even boot the Editor, to
mark the sheet music for the notes you think you may need to delete.  This is
not easy to decide in advance.. you will need to make the final decisions when
you get to the point of entering them.. but there are some guidelines  can
give that might be useful:
      1. First of all, DON'T delete any melody notes! Often (not always) they
         are the highest notes.. or the ones labeled VOCAL or LEAD ;D
      2. A prime candidate is any a note an octave higher or lower than
         another note already being played.
      3. The next consideration requires analysing the chord being played...
         usually I would drop any notes other than the 1st, 3rd or 5th in the
         chord...
      4. ... then maybe the 5th would be the next candidate.. (but not in
         country music, drop the 3rd instead).
      But the reality of it will be that you just have to try it several ways
to see which notes can be cut with the least damage to the music.
      Next you need to select a tempo.  Much modern music will recommend a
tempo (number of beats per minute).  Often you will see M.M.=120 which is
120 beats per minute, as an example.  Classical music frequently does not
stipulate a tempo and this requires judgement and trial and error again.  Just
as a crude guide... 56 is slow, 100 is moderate, and 150 + is rather fast.
       Since the time unit for the length of a note in the SID system is a
jiffie - a 60/th of a second - not all note values are integer values of
jiffies at all tempos. SOooooo.... you need to find the shortest duration
notes you will need to use in the piece and see if they are supported at the
tempo you want to use.  Page 122 of the book has a table that shows the
durations of each note at each tempo.  Put a Stick-Em note on this page..
you'll be in here often!
      As an example.. if you want to do a song at a tempo of 120, and the
sheet music shows that you will be playing sixteenth notes.... you appear to
have a problem because a whole note at 120 (By coincidence) plays for 120
jiffies.. a half for 60, a quarter note for 30, an eigth for 15 ......
no value for sixteenth notes because it would be 7.5 and fractional times
aren't supported... UNLESS... we TRICK the editor!  We can do that with a
command called a JIF. Entering positive values for JIF will make the player
slow down... it alters the clock ;)   Negative JIF values will make it go
faster.  Since we can't enter 16th notes at 120 we can select a tempo of 112
instead and use a negative JIF value to speed it up.  At 112 we are allowed
notes as short as 64ths.  There is another way.. using UTLILITY DURATIONS
that we will see later, but for now, the easiest solution is use JIF.
     FINALLY after all this planning, it is time to boot the Editor program!
You definitely want to use the Stereo Editor.. it is much easier, even for 3
voice SIDs. :)
     Start off by entering the time signature.  This is the 2/4 or 4/4 or 2/3
or 12/8 or some other fraction you see after the treble cleff at the start of
the sheet music score.  This tells how many beats per measure, and allows the
Editor to help you keep track of where you are.  If you see a large "C" instead
of a fraction, that means it is 4/4 time. If there is a slash through the C,
it means it is "cut time" or 2/4.
     Next enter the key.  The editor is designed so you just need to tell it
how many #'s or b's appear at the beginning of the staff.
     All notes that appear on the same line or space as the # or b will need
to be sharp or flat wherever they appear unless they are preceeded by a natual
sign which looks kinda like a sharp with only one arm and one leg left on! ;D
     This symbol means to play the note at it's natural tone..
it overrides the sharp or flat at the begining of the line FOR THE CURRENT
MEASURE ONLY.  The same is true of additional #'s or b's that occur from time
to time.  They revert to their normal value after the end of the current
measure, or until another sign (or accidental) overrides it.
     Having set the key it is time to begin entering commands to the individual
voices.  It is helpful to start by putting a MSR 0 marker at the very front of
each voice.  You often need to come back to the beginning to adjust a voice
and the search command can take you to the MSR 0 marker. :)
     Now comes the hard part when you are first starting.. setting the voice
envelopes.  The book pretty well describes Attack, Decay, Sustain and Release,
so I won't do that here.  The key to getting good voices is trial and error
and looking at the parameters others have used in SIDs you like.  BUT..
parameters RARELY can be copied directly and sound the same.  How they sound
is a funtion of the tempo, the octave played in, other voices chosen and so
forth.  You need to experiment.
    It will help if you stop and think about how the sound is made in the
accoustic instrument you are trying to emulate.  Some examples may be helpful.
    Instruments that are struck give sounds that grow louder very quickly..
attack of 0 or 1 is good for a xylophone or maybe a piano.
    Plucked instruments grow loud at a slower rate.. attack of 2 perhaps for a
guitar or plucked bass.
    Other instruments like a violin, sax, or flute are slower still.. maybe an
attack of 4 or even 7.
     Decay, the falling off of sound after the initial surge seems best about
10 most of the time but work with it and see what you like.
     Sustain is 0 for a piano or harpsichord... maybe 5 or 10 for a flute or
sax... we will see later how we can use sustain values to contol overall volume
of a single voice in the SID.
     Release is often 10 - 12 to give a "persistance" to the sound, but very
short for instruments that are muted or "tongued" as a flute or sax might be.
     PNT is the command that controls when the SUSTAINin' stops and the
RELEASin' starts.  You ought to start out with a value of 3 or 4 and then see
if the note has time to stop playing before the next one starts... if not,
increase the PNT or lower the release.  Higher PNT's give a stacatto sound..
lower PNT's give a legato sound.  PNT = 0 is virtually the same as tying two
notes.
     Next select a waveform. Here are some guides:
           P, or Pulse wave is about the most versatile and the loudest.
              It is good for woodwinds, flute, some guitar... you can do a lot
              with it.
           S, or Sawtooth is good for strings and brass.. it is also rather
              loud.
           T, or Triangle is soft and good for wood flute, Rhodes keyboard,
              bells, etc.
           N, or Noise, is used for percussion sounds and sound effects.
     There are also hybrid voices that are combinations of the above.  The
most common is TP that is used for clavichord, harpsichord, and some guitar.
Others like TS, SP, and TSP are a little risky to deal with.  Experience has
shown that individual computers play these very much differently and many chips
produce only inaudible sounds with them.  You can experiment, but you run the
risk of the music not being portable to other machines.
     If you are using P wave, you need to select a Pulse Width, P-W.  A value
of 2048 will play the loudest, and is the fullest sound.  It sounds kinda
clarinetish.. sorta.  Values higher and lower become tinnier.  Experiment..
don't be restricted by the comment in the book about not going above or below
certain values.. see what YOU like. :)
     After you have selected and entered your vocices, I find it a good idea
to contain them in a DEF - END statement.  Then they can be called at any
point later if you want to change instruments in a voice.  The settings apply
only the voice they are defined or called in. On stereo SIDs, remember that
the first 3 voices and the second 3 form two SEPARATE SIDs that need to have
tempos set and voices defined in then BOTH. You can't call a voice from one of
the SIDs into the other. You CAN call them from one voice to the other within
the SAME SID.
     It is best to give a bit of a delay after the SID comes to the screen and
before it begins to play, by adding a rest right at the start.. before the
first measure starts.  I do this in a special manner to kill another bird with
the same stone.. I put a whole note rest after the envelope parameters in the
FIRST voice, a half rest before and after the params in the SECOND voice, and
a quarter and a dotted half before and after the params in the THIRD.  In each
case it adds up to a whole rest.. but it breaks up all those voice commands
hitting the player at the same instant.. that can cause CLOBBER ERROR..
overload! :)
     Now, if you are in 4/4 time... the Editor will probably have already put
MSR 1 markers in for you, to mark the start of the first measure of the music.
If not...  press F1 to put them in manually.
      Now you can begin entering the notes. At the end of each measure you
should get the markers placed.. if you don't.. go back and look for an error
in durations somewhere in that measure.
      Note entry is pretty straight forward.  One thing that may confuse some
folks is tiny notes that appear before, and are tied to notes that follow them.
These little notes are to be played and their durations borrowed either from
the note before or the note after them.  In baroque music (Bach, Handel and
Scarlatti and them guys) the duration is ALWAYS borrowed fom the note AFTER
the little (grace note) note.  In modern music there is a convention... you
take the duration from the note following ONLY if there is NO line through the
grace note. If there is... take it from the note before.  (This convention is
like computer standards.. not always followed :(   )
     Well.. you may well ask, how much duration and how do I take it from
another note?   The durations is SOMETIMES the value that the grace note
appears to have (and sometimes not!) The key is it needs to be audible, and
short relative to the note that is being reduced. TRY IT AND LISTEN TO IT!
      HOW you do it is easy if the grace note is to be a 1/16 and the reduced
note is a 1/4 note.. reduce the 1/4 note to 1/8 tied to a 1/16 (or a dotted
1/8):)  Life is not always this easy.. you may need to use UTILITY DURATION.
On page 122 of the book it shows the length in Jiffies for each note at each
tempo.  For an example let's say the reduced note starts with a value from the
table of 28 jiffies.  We want the grace note to be short so lets decide to
make it 8 jiffies and then the reduced note will be 20 jiffies. At the point
you want to enter the grace note enter a UTL 8 command followed by the note
with a duration of "u"... utility .. in this case it will play for 8 because we
defined the UTL duration to be 8.  Likewise, for the reduced note enter a
UTL 20 and then the note with a duration "u".. this time it will be 20.  One
word of caution, however... UTL sets a utility value for all three voices...
if you are using UTL in several voices at once, be VERY careful they don't
conflict.. or use UTV instead.. it is specific to the voice it was called in.
     I think this is enough to get you started.  In future clases we will
cover other voice parameters, some pointers for using stereo effectively, and
a bit of discussion on particular instruments, like violin, guitar, drums, and
finally ornaments in classical music.



=END=




Press <RETURN>?