8bitfiles.net/archives/archives.tar/archives/genie-commodore-file-library/Information/SID-CLASS-3

Text File | 2019-04-13 | 12.0 KB | 169 lines

Date: 10/23/91   Time: 23:18EDT
Minutes:
<WHOLE.NOTE> Before we get into balancing SID sounds and some specific instru-
ments, I'd like to cover something called POLYCON.  This is a neat little
technique that can use a program by that name included on the Enhanced
SIDPlayer disk.  Some instruments are only capable of playing one note at a
time... like a clarinet, for example.  Others, like a keyboard or string instr-
ument can play more than one.  Whenever a clarinet plays a second note, the
first one is pretty much ended before the second one starts.  This is the case
also when a single voice in a SID plays a series of notes.  However, often
keyboard notes continue to ring even when the next note is played... maybe just
briefly, but they do.  The way to capture this on a SID is to alternate the
notes on two voices.  Then each one can continue their release while the next
note starts on the other voice.  POLYCON is a program that will take the notes
on one voice and spread them over two, placing rests where required to keep
the two voices in time.  This is a VERY nice effect, BUT...there is a bit of a
bug in that program.  It gets confused on HED or DEF/CALL commands, I forget
which exactly - so I don't use it. I do something else instead, even though it
is a bit more work.  I merely take the voice or passage from one voice and copy
it into another voice using the Cut and Paste feature.  Then I align the voices
so that they are at the same starting point, be sure the space bar is pressed
to lock them to move together, and begin manually changing them.  I start with
one voice and change every other note to a rest. (Consider tied notes as ONE
note for this purpose.)  Then I reset the voices to the beginning of the
passage again and change the notes in the second voice to rests if they are
still being played in the first voice.  It sounds much more complicated than it
really is.  If you try it you will find that it can be done quite quickly. When
you play it back you will be startled at how much fuller it sounds.  You may
need to change the envelope a bit from what you liked without POLYCON.  You
might want a longer RELease in particular.
     Let's talk abaout a few instruments in some more detail.  Violin, electric
guitar, drums and harpsichord are the ones I have chosen since they are pretty
good examples for illustration.
     Let's start with a violin.  There are several ways to do a violin, and I
have my favorite. I use ATK 4, DCY 12, SUS 6-10, REL 3, PNT 3, WAV S,
VDP 9-12-15, and VDP 3.  The attack is slow, but if notes are played very
quickly and more staccato, I will shorten it.  I figger the violinist is
perhaps striking the strings a bit harder and faster in those passages. Sustain
is a range because you can control the VOLUME of the violin sound with this
parameter. THIS IS VERY IMPORTANT, and we will talk about that in a bit. REL is
made smaller for staccatto effects or pizzicato, when you want the note to end
quickly.  You should also increase PNT for those notes.  I use an S waveform.
Some folks prefer to use a P. If you have a voice free to do it.. if you
repeat the violin parts exactly again in that free voice, and use a T wave..
and this plays together with the S wave voice, you get a really great rich
sound.  SylviaL did this with great effect on some of her cello SIDs.
The various numbers I mentioned for VDP refer to the discussion in the earlier
class about deepening vibrato on longer notes.  These settings seem to work
rather well at any tempo.
     Electric guitar is rather a general term since there are lots of different
sounding guitars from the Les Paul country type to the distortion guitars of
METAL.  The one I'll discuss here is the one I used in DREAMING.MSW, the Yngwie
Malmsteen song.  Malmsteen plays with incredible speed.. I used ATK 0. Then
DKY 11, SUS 6 (could be most anything to change note volume), REL 11, PNT 3,
WAV P, PW 485 (low, to get it a bit sharper, more metallic), PS 20 (this sweep
gives it a bit of a dirty, distorted sound) VDP 2, VRT 2, TPS -12 (you usually
need to drop guitars an octave from the sheet music) and I used DTN 1 (just for
the fun of it!)  BUT.. I did something we haven't talked about yet.  I copied
all the guitar parts from voice 1 into voice 4.  This made them twice as loud,
relative to the other instruments.  It also made the guitar seem to be in the
middle, instead of on either side.  And more importantly, I had voice 4 play 1
jiffy later by inserting a UTL 1 and UTL REST in voice 4.  This delay,
admittedly short here.. ya can go to 3 or 4 sometimes, causes a reverb effect..
or a digital delay effect, that truly makes the instrument stand out.  Solo
guitar leads, trumpets, most any lead instrument improves immeasurable when
this technique is used. On the delayed part in voice 4 I lowered the DCY to 10
to tone it down just a bit and I think that improved it.
    Bends or slurs on the guitar are done with POR, as we discussed last time.
I used POR 40 in that song, but it really depends on tempo.  For hammer-on's or
for hammer-off's I just tie the notes.  You may develop your own techniques,
but this sounded close enough I thought.
     Drums are tough.  There are different ways to do them... look at SIDs in
the Editor and if you hear some you like, see what they did.  They are also
often more illusion than exact synthesis.  I'll give ya an example.. and it
will illustrate the basic techniques.  If you download the SID GIRLS.SDA that I
uploaded and play it on StereoPlayer V10.3, the drums sound, I think, fairly
adequate.  If you then load the song into the SID Editor, you will notice that
they sound a bit different.  Most voices do.. it's a complicating factor, but
they just sound somewhat different.  Now play voice 6 alone.. that is the drum
track.  They sure sound pretty bad, huh?  BUT.. they did the job in the song.
This is where the illusion seems to play a part.  I tinkered with them until
the SONG sounded okay.. even though the drums could be better individually.
     No instrument is more sensitive to tempo than the drums.  The example of
the drum in GIRLS is at tempo 150.  At this tempo I used ATK 0, DCY 14, SUS 11,
PNT 1, RLS 2 WAV P, and POR 25.  Basicly drum sounds are made my having a note
sweep downward very quickly.  Set the starting point using an A or Absolute
instead of a duration, and then the terminating point with a note about 2
octaves lower with the desired duration.  Since there are a lot of commands
used for each note, I usually use a DEF statement for each of the bass, snare,
and toms... then call the DEF's as needed.  Drum sounds are quick, so I usually have a rest imbedded in the
DEF also.  Here are some examples from GIRLS.MSW:
           The Bass drum was defined as:
               DEF 4, RLS 4, PNT 2, E2A, E0 sixteenth, RLS 1, sixteenth rest,
               PNT 1, END
  (WOW!  Lots more in there than I said ya needed!)  The thing is, the snare is
   gonna use different values for RLS and PNT so they need to be reset!
           The Snare was: DEF 5, DCY 6 E3A, E1 sixteenth, DCY 14 sixteenth
               rest, END
    I use ATK at 0 and SUS at 11 it this case... always attack 0... sometimes I
need to lower the sustain from 15 (best value for loud drums).. in order to get
the sound to end fast enough.  OR sometimes I actually have to make them
quieter, but not often! :D
    Drums are so sensitive to tempo that they are really tough... you have to
balance the POR, and durtrations and DCY to get a crisp sound.  And when you
like a set of parameters in one SID, they just never seem to translate to the
next one....and you have to start all over again. :(  But, as I mentioned
above, the drums in GIRLS that I used for an example don't sound very good at
all alone.. but they WORK in the song.  The illusion is often sufficient.
    After you have established DEF's for the drums, the rest of the voice may
only be CALL's of those DEF's.  Or maybe you need to make some TOMs for
occaisional accents.  Just do it analogously to those, at the pitches that
sound appropriate to you.
    The next instrument to be discussed is the harpsichord.  For harpsichord I
use ATK 1, DCY 12, SUS 0, REL 9, PNT 3, WAV TP, PW 850-1300, PS -8.  The
harpsichord is a plucked instrument.. not a hammered instrument like a piano.
That's why the slightly slower attack.  The sound is very soft and doesn't
persist.. in fact it is muted when the key is released.. so SUS 0 is correct.
PNT of course depends on tempo.  TP is a voice that gets quieter as the PW
increases.  I use higher values to tone down the lower notes.  Values of 800 or
so give it a metallic sound that is appropriate.  Pulse Swep can be fairly high
-8 and it seems to enhance the effect.  These settings seem to be pretty good
most of the time.. until you get to 32nd notes or less... In baroque music you
run into that a lot too!  SOOooo.. I use DEF's to set up modified parameters
for the short notes, and another to restore it.  Usually, right up front I'll
set up a DEF 1, SUS 4, REL 5, PNT 0, END.  This allows the very short notes to
stand out.  SUS may have to be tinkered with to get the right volume. PNT at 0
is virtually the same as making the notes tied.. and I often tie them also in
ornaments.  In order to restore the envelopes after the short notes are past,
I have a DEF 2, SUS 0, REL 9, PNT 3, END.  You could call the original envelope
again, but since you only changed 3 parameters, its less strain on the ole CPU
to just restore those three. :)
     I'll cover ornaments and other things specific to classical music in the
next session.
     Now lets talk about ways to make the voices work together in a SID.  I'll
talk about stereo now, since it has more opportunities and more problems
because of the 3 additional voices.  For most music, I guess it is safe to say
that the melody is the most important part.  That may sound rather obvious, but
it is amazing how many SIDs seem to have the melody get lost.  You want to be
sure that the melody is done in a voice that can stand out.  Sax, Electric
Guitar, trumpet - these are good choices...They all are also sustained voices..
and you can use SUS to control their volume.  Flute is tough if you use a
Triangle wave... that wave is rather quiet.  You CAN double it up, using the
reverb effect we discussed earlier.  Or.. you can be sure the other voices are
softened.
     The three elements in the music are usually the lead, the accompaniment
and the base.  Let me use another of my SIDs to illustrate how the instruments
were balanced to provide appropriate volumes.  HORNPIPE.SLR is from Handel's
Water Music.  Voices 1 and 2 are oboes.  They are frequently carrying the
melody.  I used suutain 6.  this kept them loud enough to be heard plainly.
Voice 3 was the bassoons... they were too loud... so I used sustain 1...a
bassoon needs a filter and the filter was set very harsh... AUT 1 and RES 2..
this doesn't pass a lot of the sound.
     The second SID contains the strings.  Now the strings OVERALL were playing
too loud at first, but since I grouped them in their OWN SID I can use the VOL
control to drop them all at the same time.  This SID has VOL 5, whereas the
first SID has the default value of VOL 8.  The Violins are at SUS 8.. since the
VOL was dropped, they could be set higher and I liked them at that sustain.
    The Bass was too loud even at Volume 5 so again a strong filter was used...
AUT -1 and RES 0 and SUS was dropped to 0 also.  In measure 11, horns are added
and again the parameters are used to control their relative volumes.
    The balancing of the sounds is the key to making the instruments play
together rather than clashing or overpowering each other.  From passage to
passage, you can change the SUStain, use the BUMP command to raise or lower the
whole SID, or any of the other efects we described to change the emphasis from
instrument to instrument as the lead is perhaps traded off.  Developing this
variety, and introducing dynamics to the SID keeps it interesting and prevents
folks from running to the space-bar before its ended! :)
     Next time we'll cover some topics specifically related to SIDding classical
music. :)