S  e  c  r  e  t  s     o  f     T  r  a  c  k  i  n  g
            =======================================================

                     written in 2001 by elysis (Gerd Raudenbusch)
 

Content
=======
 

Introducion
    1. What are "frequencies" ?
    2. What are "samples" ?
    3. What is a "Tracker" and is it the right thing for me ?

Chapter   I - Realizing and analyzing your sound-world
    1. Listening conscious
    2. Analyzing sounds
      a) Using resonant tones
      b) Splitting to simple waveforms

Chapter  II - Volume
    1. Envelopes
       a) Spot (Closed decay)
       b) Spot with reverb
       c) Open decay
       d) Soft attack
       e) Soft pad
    2. Accents
    3. Volume crossings
       a) Using the volume column
       b) Using the Mxx-effect-command
       c) Using volume slides
       d) Splitting channels
    4. Volume triggering
    5. Faked echoes
       a) Tracked echoes
       b) Envelope echoes
       c) Sample loop echoes

Chapter III - Panning
    1. Panning methods
       a) Stereo samples
       b) Panning envelopes
       c) Tracked stereo sound
       d) Sample rate shifting
    2. Volume and panning

Chapter  IV - Tempo and Speed
    1. Sample pitching and sample looping
      a) Pitching a sampled rythm-part that it fits to a number of rows
      b) Loop a sample syncronous to a pattern
      c) Pitching two samples to the same frequency
      d) Looping a sample to endless (i.E. to fake reverb)
    2. Syncoped beats
      a) Pattern delay
      b) Altering speed
      c) 6/4-measure

Chapter   V - Effects
    1. Conventional effects
      a) gapper/snipper
    2. Unconventional effects
      a) highspeed-tracking

The ten golden rules of tracking
 

Introducion
===========

First of all I want you to excuse my possible mistakes in the language
because I have german nationality. My English is not bad, but it is still
a barrier sometimes. So many people asked me to write about tracking and
techniques. That made me realizing, how important this informations must
be for some people and that they seem to be honestly interested in tracking
music. I learned tracking by trying ALL the functions of my instruments
(this includes the tracker-software) and by analyzing other modules, that
impressed me. And I want to mention here, that tracking is one of the things
that you especially learn by doing them. This means, if you have read that,
you're not a better musician until you have checked out for yourself, how the
techniques work. It is like learning JAVA, C++ or another programming
language, but here you learn how to program music. And it is not only this
language but the elemenar understanding of what music is. I have got
experience with creating sounds on synthesizers and I want to use this chance
to give you a far-reaching imagination of how it works to create electronic
music with a Tracker and how you can use the different techniques.

I write this because in my opinion you CAN produce professional electronic
music with Impulse-Tracker. Yes, the sound-quality from electronic
instruments is better sometimes but Impulse-Tracker includes all important
possibilities to make electronic music and if you understand to use this
possibilities then you have no problems to use electronic instruments.
Although some people hang on the step-sequencer-concept that many instruments
have I must say that the tracker-concept is a huge expansion of that and I
hope that these people can see this anyway.

I work with the 'Impulse Tracker' because I think it is the best devoloped
Tracker. And so the explained tracker-commands in this tutorial are
Impulse-Tracker-specific. You can download this software
at http://www.noisemusic.org/it. I find that the documentation for the
Impulse-Tracker is very good to understand and I don't want to repeat that
content here. If you have never used a tracker before I strongly suggest you
to read and understand the Impulse-Tracker manual first.
 

1. What are "frequencies" ?

Simple question? Could be, that you are a good tracker but this means not
that you must know anything about frequencies. But it is very useful to know.
Try to understand, what SOUND is. It is the compression and expansion of air.
This is, what every speaker does and what every thing does, if you drop, kick
or hit it. A speaker converts electric impulses into air-compression/expansion
and a microphone does the opposite. Perhaps you know, if you move a magnet
over some kind of electric wire, this is the origin of voltage and reverse.
This means that the moving of the air in the near of a microphone also moves
the membrane and the fixed magnet on it, and this induces voltage into the
wires. This voltage again can move another magnet in a speaker connected with
a membrane, that compresses or expands the air.

Perhaps you have heard about the different kinds of voltage. There exists
AC-voltage and DC-voltage. You can imagine DC-voltage like floating electic
energy. AC-voltage is the same but with a changing flow.
What happens, if you put some millivolts of a battery on a speaker for a
half second ? You will hear a 'click' - it is called 'DC-click' and it is
a change from "NO VOLTAGE" to "VOLTAGE" and again to "NO VOLTAGE".
Now imagine, you could connect and disconnect the battery to the speaker
50 times in one second. You would hear a low tone - this is A FREQUENCY.
The more times you do that, the higher gets the tone (the frequency).
With 'tones' we mean frequencies, that we can hear.
And there exist different elementary types (wave-forms) of frequencies,
SINE, SQUARE, TRIANGLE or SAW and NOISE (random changes). If you are a
tracker, I strongly suggest you to have samples of this waveforms,
because you can create many, many kinds of sound with them.
With 'Sound Forge' you can create this 'simple waveforms'.

      ***                                  *
    *     * SINE                      *     * TRIANGLE
  *         *                         *          *
-*-----------*-----------*-----   --*--------------*-----------*-
              *         *                            *       *
                *     *                                *   *
                  ***                                    *

 *************                      *              *
 *           * SQUARE             * *            * * SAW
 *           *                      *   *          *   *
 *           *                      *     *        *     *
-*-----------*----------*------  ---*-------*------*-------*-----
             *          *           *         *    *         *
             *          *           *           *  *           *
             *          *           *             **
             ************           *              *
 
 

2. What are "Samples" ?

If you want to make music with the PC, you need a soundcard. This soundcard
also contains a digital-analog / analog-digital - Converter. This is a unit
that changes the incoming voltage-changes into ones and zeros and reverse.
And this gives the PC the possibility to record and play EVERY imaginable
sound. A piece of digitalized and recorded sound is called 'SAMPLE'.
This can be a whole song, or just one DC-click. In every case it is a sample.

Now imagine a visible scene and the way how we can record it. We take many,
many pictures in a short time, and if we display these pictures again in the
same speed, we can see the visible scene again. The same principle works on
recording frequencies with the computer. We check many, many times in one
second, what voltage comes from the microphone and the computer records it.
Surely you can imagine, the more times this is done, the better gets the
sound-quality. And the times, how often the voltage (converted to bits) is
recorded in one second is called 'SAMPLE RATE'. To sample music with
CD-quality, a sample-rate of 44.1 kHz is used (44100 times per second).
Why such a strange number ? Because on a computer there are only TWO different
states (0 and 1) and 44100 is TWO multiplied with itself 210 times.

There are many ways, to record and play samples. There are primitive utilities
from the system but also special programs like 'Sound Forge', 'Wavelab' that
give you the possibility to do more except just recording and playing a sound.
Sometimes there are such programs on the CD or disks that came with your
soundcard. What do you think, why could you need a special program ?

Well, here I want to mention an important and old controverse between
sample-rythm and sample-frequency. Imagine, you have recorded your voice,
saying 'HA' four times for example and you needed two seconds for it. If you
want to integrate that in a song with a different speed, you perhaps have to
change the speed of your 'HA'-sample. But what happens, if you let play the
sample faster ? Yes, the rythm is getting faster, but the voice becomes
sounding like 'Mickey Mouse'. This is, why we seperate between 'pitching'
(what means to change the whole frequency) and 'time-stretching' (what means
to change the sample-speed) of a sample. This can both and independent from
each other be done only with these special programs and NOT by a Tracker.

If you are tracking for a time, you don't know it better, but I can imagine
that it would be useful many times, when a Tracker would be able to do
time-stretching and pitching independent from each other.
 

3. What is a "Tracker" and is it the right thing for me ?

Some people just work with programs and never read the documenation because
they are too lazy. This is the reason, why I will shortly explain, what a
Tracker does. With a Tracker, you can build a library of different samples.
And a Tracker gives you a kind of table, where you can decide, when which
samples should be played with which frequency. As I mentioned, you have NOT
the possibility to change the rythm or the pitch of a sample exclusively. So
you can take some drum-samples and create a rythm - these samples are pitch-
sensitive, so you better don't change the sample-speed too much and play the
samples with their original speed. Then you can take some tone-samples and
create a melody. Here you MUST play the samples with different speeds to
create a melody. You can get problems with tone-samples, that also have a
rythm. The simplest are so-called 'Spots'. This are recorded tones, that play
just for a short time. If you play them with a low frequency, the time gets
longer, so you always have a range, where the samples sound good and ranges
where they sound not good.

I saw the first primitive Tracker on a C64 (not with real samples, but with
computer-generated tones). The first Tracker, that really was able to work
with samples was - I beleave - on AMIGA with four channels (This means it was
able to play four different samples at the same time). And the Tracker, where
I began tracking was the 'WHACKER' on PC, also with four tracks. It was
horrible, because the program hang-up all the time. A year later I found the
'Scream-Tracker' written from Future-Crew, a demo-group (S3M-Modules).
I haven't recognized, where and when the 'Fast-Tracker' appeared (XM-Modules).
But in a comprehensive way I can say that many, many songs where written
with these two Trackers, mainly songs for games and for fun.

Many people, also me, tried to make natural-sounding music with the Tracker,
but the controverse between sample-rythm and sample-frequency is the main
reason, why it never sounded really good. Today we have the 'Impulse-Tracker',
in my opinion the best devoloped tracker-program and it is still difficult,
to use a complete-sampled piano for example. You need to sample at least
every sixth tone because the changes are too strong, and the pitch-range
where it sounds good is very small.
Sure, the Impulse-Tracker is able to send and recieve MIDI-messages, so if
you have a good sounding synthesizer, you finally get the natural sound. But
then you have another problem if you want to record your natural playing -
The minimal changes of rythm when you play cause, that the notes are not
always where they should be (no quantisize-option). You better use a
sequencer-program like Cubase then.

All in all you can say that a Tracker is great for producing concept-music
and that is mainly electronic music but not for dynamic-sounding natural
music. So if you wanted to do that, you need a 'Multi-Tracker' - this is a
kind of digital studio for the PC, where you can play-in your instruments
and mix it all together to one song. You need also another tutorial then,
so I say 'Bye' here to all non-electronic musicians, although I make such
music, too.

Well, the good thing therefore is, that you have the complete control over
EVERY SINGLE NOTE with a Tracker and this may be the reason why so many
people use and keep using the tracker-concept.
 
 
 

Chapter I - Realizing and analyzing your sound-world
====================================================

1. Listening conscious
There are different ways to hear. Samples are the base of every tracking
musician so it is important for you to learn to hear consciously and to sample
every good sound. The best thing is to have some portable recorder (This
can be cassette, Mini-Disc, Pocket-Sampler or anything else) so that you can
sample anything anytime. Other people collect stamps - a tracking musician
collects samples.
The more samples you have, the closer you get with the tracked song on your
imagination of it. Some of my best songs have the craziest samples like
my door-keys (as shaker), the pole of my mothers broom (as cymbal), the
coffee-cup of my father (as hihats) - yes, electronic music is experimental,
so experiment with the sounds around you !!

2. Analyzing sounds
There are different ways to analyze frequencies. Mathemathic methods like the
fourier-analysis are good, but cannot be made per ear. But you can use
a similar easier techniques to create sounds.

a) Using resonant tones
The CASIO SK1 Keyboard has this technique included. For example you can
create an organ-sound with a simple sine by overlaying this sine with itself
in different frequencies. Important is, that the differences between the
overlaid frequencies are six halftones (So if you start at C0 for example,
you use G0, C1, G1, C2, G2 and so on with different volumes). The difference
is six halftones because this is the first natural resonance of every tone.
There are more natural resonance tones, but we don't need them here.

b) Splitting to simple waveforms
Another way to analyze a sound is to split it into basic frequencies (square,
sine, triangle, noise). Let's take a bongo-hit and listen to it exactly.
Then you hear two short different tones and a special sound like a 'click'
or a 'tip'. You can rebuild this tone by taking a sine-wave for example and
give it a short volume-envelope. I mean an envelope looking like this :

                 Volume
                    !\
                    ! \
                    !  \
                    !   \
                    !    \
                    -------------------------> time

On many analog synthesizers this is called 'closed decay'. Now play it with
two different frequencies at the same time. Yes, it sounds a bit like a bongo
 - the 808-Bongo has such a sound but there is missing something. Now mix it
all with a DC-click or with a sampled tapping on the microphone or something
like that and you have a pretty synthetic and own-created bongo sound.
Let's take a bass-drum. This is also easy. Take a sine. This sine must be
pitched down very fast. The 'Stomper' uses this technique. So give the sine
a short pitch-envelope (looks like the volume-envelope above). Mix it all
with a DC-click and you have a pretty cool sounding bass-drum. For hihats you
can use high pitched noise, for a snares and claps it is good to use low
pitched noise. I have killed all the presets in my synthesizer and made my
own instruments using exactly this technique.
 

Chapter II - Volume
===================

The most reason that makes modules sounding ugly is missing polyphony and
missing crossings. And the volume is very useful not only to create soft
crossings that let your song sound professional but also to give notes an
accent.
 

1. Envelopes

Why using envelopes ? If you have 'Spot'-sample, and you play it with very
low speed, then the time it plays gets longer. This is the main reason for
using volume-envelopes where the playtime keeps constant. And many times it
is easier to draw such an envelope as tracking it. You can fake a reverb-
effect also. But sometimes the sample is too short and plays not as long as
the envelope. Then you need to loop the samples to endless to use the long
volume-envelopes. I described in Chapter IV how that works. I will shortly
mention the most important envelopes here :

a) Spot (Closed decay)

                 Volume
                    !\
                    ! \
                    !  \
                    !   \
                    !    \
                    -------------------------> time

b) Spot with Reverb

                 Volume
                    !\
                    ! \
                    !  \
                    !   \
                    !    -----____________
                    -------------------------> time

c) Open decay     Volume
                    !\
                    !    \
                    !        \
                    !            \
                    !                \
                    -------------------------> time

d) Soft attack    Volume
                    !            /----------
                    !         /
                    !      /
                    !   /
                    !/
                    -------------------------> time

e) Soft pad       Volume
                    !            /\
                    !         /      \
                    !      /            \
                    !   /                  \
                    !/                        \
                    -----------------------------> time
 

If you can't imagine, how the samples sound with these envelopes, then just
draw them in Impulse-Tracker and try it yourself. The 'note fade' option
in the 'General' menu can be useful if you like to have a polyphonic sound.
Perhaps you know, you can create a 'Spot' not only with envelopes but also
with the damper-effect-command (D0x).
 

2. Accents

Accents give the instruments more life and motion. Just imagine, you use a
closed hihat (Sample #1) like this :
     .
     .
C-4 01 64 .00
C-4 01 64 .00
C-4 01 64 .00
C-4 01 64 .00
C-4 01 64 .00
C-4 01 64 .00
C-4 01 64 .00
C-4 01 64 .00
     .
     .
This is a typical sign in my ears for a tracked song. The people are just
too lazy to invest a bit more time for a better sounding hihat. That could
in the easiest way look like this :
     .
     .
C-4 01 30 .00
C-4 01 30 .00
C-4 01 64 .00
C-4 01 64 .00
C-4 01 30 .00
C-4 01 30 .00
C-4 01 64 .00
C-4 01 64 .00
     .
     .
Or a bit more complex :
     .
     .
C-4 01 20 .00
C-4 01 10 .00
C-4 01 64 .00
C-4 01 30 .00
C-4 01 20 .00
C-4 01 10 .00
C-4 01 64 .00
C-4 01 30 .00
     .
     .
Of course you can use the Mxx-effect-command instead of the volume-column.
This makes the sample sounding so much more groovy but I still miss it in many
modules and the modules keep sounding static and shapeless.
 

3. Volume-Crossings

Often it is not enough just to switch on or off the channels or use a sample
and then just drop it. It is better then to fade-in and fade-out the channel.
There are two possibilities in Impulse Tracker - the first is using the
volume-column (00-64) the other is to use the channel-volume-effect (M00-M40)

a) Using the volume-column b) Using the channel-volume-effect
... .. .. .00                  ... .. .. .00
... .. .. .00                  ... .. .. .00
C-0 01 20 .00                  ... .. .. M00 <---- Channel-Volume-effect
      /!\
       !-------------volume column

c) Using volume slides
You know it is quite easy to do slides in Impulse Tracker with the ALT-K /
ALT-X option. If you use the volume column you just have to mark the area,
insert start- and end-value and press ALT-K.

C-4 01 00 .00 \
C-4 01 .. .00
C-4 01 .. .00
C-4 01 .. .00   Mark Block with Shift+Cursor keys, then press ALT-K
C-4 01 .. .00   to fade in with volume 0 to volume 64
C-4 01 .. .00
C-4 01 .. .00
C-4 01 64 .00 /

It works equal if you use the channel-volume-effect, then you have to press
ALT-X instead. This are really useful functions. I remember with ScreamTracker
you had to do it by hand. If you have used the volume column to use accents
it is better to use the channel-volume-effect for fading-in/out the channel,
if you want the accent-information not to be destroyed by the fading.

d) Splitting channels
What is when you need the volume-column for accents and the effect-column for
effects and you want to do a volume-slide? Well, then you have to split
the source channel like this :

   source           Channel 01     Channel 02     Channel 03     Channel 04

C-4 01 20 D0F    /  C-4 01 20 D0F  ... .. .. .00  ... .. .. .00  ... .. .. .00
C-4 01 10 D0A    !  ... .. .. .00  C-4 01 10 D0A  ... .. .. .00  ... .. .. .00
C-4 01 64 D0A    !  ... .. .. .00  ... .. .. .00  C-4 01 64 D0A  ... .. .. .00
C-4 01 30 D0F --->  ... .. .. .00  ... .. .. .00  ... .. .. .00  C-4 01 30 D0F
C-4 01 20 D0F    !  C-4 01 20 D0F  ... .. .. .00  ... .. .. .00  ... .. .. .00
C-4 01 10 D0A    !  ... .. .. .00  C-4 01 10 D0A  ... .. .. .00  ... .. .. .00
C-4 01 64 D0A    !  ... .. .. .00  ... .. .. .00  C-4 01 64 D0A  ... .. .. .00
C-4 01 30 D0F    \  ... .. .. .00  ... .. .. .00  ... .. .. .00  C-4 01 30 D0F
      .                   .
      .                   .
You take a seperate channel for every diffeerent volume-value, then you can
make your fade without loosing the accent- and effect-information. In our
example, a fade-in for the source would mean to fade channel 01 from 00 to 20,
channel 02 from 00 to 10, channel 03 from 00 to 64, channel 04 from 00 to 30.
 
 

4. Volume triggering

I personally use this technique not so often, because it is used to create
the short-short-long rythm that I personally hate - it is too conventional
for me. But nevertheless it is a useful technique anyway.
You need a long or an endless-looped sample (I described it in Chapter IV how
you loop a sample to endless) then play it one time and change the volumes.
It looks like that :

C-4 01 64 .00
... .. 00 .00
... .. 64 .00
... .. 00 .00
... .. 64 .00
... .. 64 .00
... .. 64 .00
... .. 00 .00
     .
     .

Of course you can also trigger the volume by using the Mxx-effect-command or
by using a square-wave as volume-envelope. Then the zero-point of this square
has to be at a certain point, that is syncronous to a prefered row. I
described the technique how to find this point in the article
'Envelope echoes' in this chapter.
 

5. Faked echoes

You get a richer sound, if you echo some channels. You must fake that echo-
effect, because there is no "echo"-command or something like that. It sounds
good on melodies but also on percussive instruments like 'Claps'.

a) Method #1 : Tracked echoes
You need one possibility to change volumes for that. Copy the channel by
pressing ALT-L and ALT-C to the next channel. Then shift the complete notes
in the new channel three, four or whatever-you-like rows down. If the channel
repeats in one pattern you can mark the last three, four or whatevere-you-like
rows of the PREVIOUS channel and copy these notes into the empty space of the
NEW channel. Now bring this new channel to the half volume-value. Do this
whole procedure at least three times and always take the notes of the
previous channel and copy them to the next channel into the empty space.
It sounds great, if you give the echo-channels another pan.
 

Example :
Channel 01 is the original, channels 02-04 are echoes with a step of 3 rows.
Channel 02 is left, channel 03 is right and channel 04 in the middle again.

  Channel 01        Channel 02        Channel 03      Channel 04
     .
     .
C-4 01 64 X80 -!   ... .. .. X00    /... .. .. XFF   ... .. .. X80
D-4 01 64 ...  !   B-3 01 32 ...  -! E-4 01 16 ...   C-4 01 08 ...
... .. .. ...  !   ... .. .. ...  ! \... .. .. ...   D-4 01 08 ...
E-4 01 64 ...  !-> C-4 01 32 ...  !  ... .. .. ...   ... .. .. ...
... .. .. ...      D-4 01 32 ...  !  B-3 01 16 ...   E-4 01 08 ...
... .. .. ...\     ... .. .. ...\ !  ... .. .. ...   ... .. .. ...
B-3 01 64 ... !    E-4 01 32 ...!-   C-4 01 16 ...   ... .. .. ...
... .. .. .../!    ... .. .. .../    D-4 01 16 ...   B-3 01 08 ...
     .        !
     .        !--This will be shifted out in channel 02 and must be
                 copied then to the begininng in the empty space.
                 This only works as long as the melody repeats every
                 pattern. If the echoed notes repeat over more than one
                 pattern you have to take the last notes of the previous
                 pattern here.

With this technique you can also create pitch-shift-delays (Just mark the
channel and shift the notes up or down with ALT-Q / ALT-A).

b) Method #2 : Envelope echoes
In Impulse-Tracker you can also fake echoes with the volume-envelope. But
this is not easier. First you have to find the certain point, where the
next echo-row is played. In Impulse Tracker this point exists and it is
absolutely syncronous to the row, no matter what tempo and speed you use.
If you like to have an echo with a delay of 3 rows,
then draw something like this :

                 Volume       !----- Shift this, until you hear the
                    !\       /---\   sound when the awaited row is played.
                    ! \
                    !  \      x\
                    !   \     x \
                    !    \    x  \
                    -------------------------> Tick

If you have found out the point, then you can calculate the next. If this
is for example tick 12 then you know the next tick must be tick 24, the next
is tick 36 and so on. And now you are able to draw the whole envelope :

                     ---------------------------------- Syncronous to
                    \!/         !         !         !    row 0, row 3,
                               \!/       \!/       \!/   row 6 and row 9
                    !\
                    ! \
                    !  \      !\
                    !   \     ! \       !\
                    !    \    !  \      ! \       !\
                    -----------------------------------> Tick
                    0         12        24        36

This technique works fine with constant melody-samples like a sine-wave,
human voice, strings or something. If the sample is too short then just loop
it. But if you want to use the volume-envelope on percussive samples to
create echoes there exists another technique.

c) Method #3 : Sample-loop echoes
First of all you must load the sample with a program like 'Sound Forge' or
'Wavelab'. Then insert silence at the end. Save it as new sample. Now you can
loop this new sample in Impulse-Tracker and use a simple 'open-decay'
volume-envelope. I know that many people hate looping or pitching samples by
hand to a special value, but as you see it can be very useful. Here you also
have to find the correct loop-length. If it is too short (repeats too fast)
then you have to raise the loop-end-value, if it is too long (repeats too
slow) then you have to decrement the loop-end-value. I use a system how I
pitch or loop samples correctly, I will explain it later in Chapter IV.

A nice thing is, that you can also echo parts of samples with this technique.
This sounds fine when you use speech-samples (For example 'It will turn
alright') and you loop just the last part ('alright' , every third row)
fitting to the tempo of the song.

By the way - this loop-technique is also good, if you want to use rythms in
your song, that have a length different from the pattern length. Perhaps
you know it from other songs : A rythm is used with 12 repeating notes -
that sounds really great, but if you want to track that with 64 rows, you
need 3 patterns for it and that CAN get complicated, because in songs
with 4/4-measure there belong numbers of patterns together like 2,4,8...
and so on. It is easier to isolate that rythm in an extra pattern and to
create a sample of it. Then you can loop it and just need to play it one
time.
 

Chapter III - Panning
=====================

Yes, the people are often too lazy to programm their songs in stereo.
Using the panning is necessary to create soundscapes that let the listener
beeing able to imagine a 3D-room where the tones or some symetric things fly
around. If you don't use the panning, your song takes no room and it keeps to
be just a flow of tones on one straight line.
 

1. Panning methods

a) Method #1 : Using stereo samples

If you use stereo samples, you need two positions in the library (F3 / F4)
one left, one right and you need two channels in the pattern-mode (F2), when
you track with them. Don't forget to give the channels or the instrument a
prepanning (one left, one right).

b) Method #2 : Using panning-envelopes
The best thing is of course, to use stereo-samples. But what can you do with
monophonic samples? You can do a lot just using the pan-envelopes. It is
often enough just to put in a triangle-wave, so that the sound is panned
to left, right and middle again. But there are no limits for you fantasy.
Why not try something like this :

                 Panning
                    !                         /\
                    !                        /  \
                    !         /\            /    \
                    !/\      /  \          /      \      /\
                    ---\----/----\--------/--------\----/---->
                    !   \  /      \      /          \  /
                    !    \/        \    /            \/
                    !               \  /
                    !                \/

c) Method #3 : Tracked stereo-sound
The easiest way to give some percussive instruments a static stereo-sound is
copying the notes of a channel (containing hihats, claps or something) to
the next channel and then shift the notes in the new channel all a halftone
up by marking the channel with ALT-L and then pressing ALT-Q. Now you just
need to fill in two pan-commands 'X00' in the first and 'XFF' in the second
channel. The pan in the sample-menu (F3) must be switched off. This could
look like this :

C-4 01 30 .00      C-4 01 30 X00   C#4 01 30 XFF
C-4 01 30 .00      C-4 01 30 .00   C#4 01 30 .00
C-4 01 64 .00      C-4 01 64 .00   C#4 01 64 .00
C-4 01 64 .00 -->  C-4 01 64 .00   C#4 01 64 .00
C-4 01 30 .00      C-4 01 30 .00   C#4 01 30 .00
C-4 01 30 .00      C-4 01 30 .00   C#4 01 30 .00
C-4 01 64 .00      C-4 01 64 .00   C#4 01 64 .00
C-4 01 64 .00      C-4 01 64 .00   C#4 01 64 .00
 

But this technique does not work only on percussive samples. If you like to
use it on melodic samples, you cannot shift the notes by a halftone - that
is too much - but you could shift it by a smaller value, using the
FF1-command. You can also use the O01-command, that plays the sample a bit
more from its middle. This is better for long loops running over more
patterns, because they keep syncronous then).

It could look like this :

... .. .. .00      ... .. .. X00   ... .. .. XFF
D-4 01 .. .00      D-4 01 .. .00   D-4 01 .. FF1
E-4 01 .. .00      E-4 01 .. .00   E-4 01 .. FF1
F-4 01 .. .00 -->  F-4 01 .. .00   F-4 01 .. FF1
... .. .. .00      ... .. .. .00   ... .. .. .00
G-4 01 .. .00      G-4 01 .. .00   G-4 01 .. FF1
... .. .. .00      ... .. .. .00   ... .. .. .00
D#4 01 .. .00      D#4 01 .. .00   D#4 01 .. FF1

You see, if you have a tone in the first row, you can not play that tone
stereo, because you need the effect-column to set the panning. But this is
a harmless mistake and if you are a turbo-exact-working musician, there are
other ways to give the channels a pre-panning (for example using the volume-
column, switching to panning by pressing '`' or using the menu pressing F11).

By the way - if you do not pan the channels and all the channels are panned
equal you create a chorus/flanger-effect for the instrument with this
technique. It is interesting then to play with the values of the FFx-command.
 

d) Method #4 : Samplerate shifting
If you also need the effect-colums you can do another thing. Load the
sample two times to the library (F3) and increase the sample-rate of one of
the samples a bit. This detuning is enough to create a stereo-effect by
playing both samples in one row with different panning.
 
 

2. Panning and volume

Volume and panning can play a role using it together to create pseudo-3D
sounds. If you pan a sound to the left or right and decrement its volume
simultanously it appears that the sound 'flies away'.
This could look like this :

               Panning                                        < f(x)=sin(x) >
                 !          /-----\
                 !      /---       ---\
                 !   /--               --\
                 ! /-                    -!
                 ----------------------------------------------------> Tick
                 !                         !-                   -/
                 !                           \--              --/
                 !                              \---       ---/
                 !                                  \-----/
 
 

              Volume                             < = abs(f'(x))=abs(cos(x)) >
                 !---\                   /---\                   /--
                 !    --\             /--     --\             /--
                 !       -\         /-           -\         /-
                 !         \       /               \       /
                 !          !-----!                 !-----!
                 ----------------------------------------------------> Tick
                 !
 
 
 

Chapter IV - Tempo and Speed
============================
 

1. Sample-pitching and Sample-looping

I recognized, that people need a long time until they understand how that
works. And when they understood it, they don't do it because they think it
is too complicated. I often pitch or loop samples and I have devoloped a
system that works fine.

a) Pitching a sampled rythm-part that it fits to a number of rows

Okay, you have a drum-part or stuff and you want to make it fit to your song.
Do the following. Take a new and empty pattern and imagine that the sample
would be right-pitched. Attention : The sample must NOT be looped - switch the
loop-option off!  Then set the sample to the positions, where it should be
played. For example, if you have a sample containing 4 measures then let it
play in row 0,16,32 and 48. Sometimes it is a great help if you programm
yourself a metronome by taking a bassdrum, a stick or anything and set it to
row 0,4,8,12,16,20,24,28,32, and so on in a different channel. Now press F6
to play the pattern. Switch to the sample-library by pressing F3. Select the
songpart-sample and switch over to the sample-speed by pressing TAB. If the
speed is not correct then press (ALT +) until you hear silent spaces between
the loop (It is easier to approximate to the correct speed from a higher
speed).

Here we are. Now listen! Is the loop too fast and can you hear a silent space
between ? Then press (CTRL -) to decrease the sample speed. Wait till it is
played again. Decrease the speed until there is no more silent space (well I
hope, your loop has no silent spaces). Now check, if the loop is played till
the end. If it is interrupted by its new-beginning then it is pitched a bit
too low. Play with the last digits of the sample speed and fill in higher
values until the loop is played correct.
 

b) Loop a sample synconous to a pattern

In Chapter II I mentioned a method how to fake echoes with sample-loops. This
is one reason, why you could want to loop a sample syncronous. Presupposition
is that the sample is long enough, if not then insert silent space at the end
using programs like 'Sound Forge' or 'Wavelab'.

Do the following : Take a percussive instrument like a bassdrum, a closed
hihat or a stick and play it in a new and empty pattern at all the positions
where the sample should play. This help-sample should sound different from the
loop-sample. Now switch to the loop-sample and play in row 00. Start playing
the pattern by pressing F6 and switch to the sample-library by pressing F3.
Select the loop-sample and press TAB to get over to the sample-speed/loop
options. Now turn the loop on. Choose your prefered beginning of the loop if
it isn't the beginning of the sample.

Here we go. Default value for LoopEnd is the sample-end. So decrement the
LoopEnd-value, until the loop is played correct. If you copy the channel with
the help-notes to the next pattern and if you write a short orderlist that
plays the first pattern one time and the second multiple times it is possible
to pitch the sample very exactly. I pitched a sample in one of my songs
(Astralium), so that it plaid syncronous over eight patterns. Well, that's
pretty exact, isn't it?
 

c) Pitching two samples to the same frequency

Take two tone-samples and play them both in a temporary pattern in one row
with the same note. If the samples stop at any time then fill it in more rows
so that you can hear them everytime if you let the pattern play. Now do this
by pressing F6 and switch to the sample-menu by pressing F3. Choose one of the
samples. Use TAB to get over to the section, where filename, speed and so on
is and go to the speed-row. Here we are. Now use your ears! With (CTRL +)
you raise the frequency by a halftone and with (CTRL -) you decrement it by
a halftone. Switch around until the two samples sound ALMOST equal. Now take a
look to the sample-speed. If you have a low sample-speed like 8363 then it
is enough to play with the last or the last two numbers. A greater number
means a higher frequency, a smaller number means a lower frequency than the
actually played. If you have a high sample-speed like 44100 then play with
the last three or four digits until the two samples sound equal.
 

d) Looping a sample to endless (i.E. to fake reverb)

Yes, sometimes you need a sample to be played endless (or until you stop
it) but you perhaps have a 'Spot'-sample or the sample is just too short.
Search a position in the sample from backwards to the beginning, where it
changes its sound not much. This can also be the whole sample sometimes.
Go to the sample library pressing F3 and switch over to the sample speed /
sample loop-section with TAB. Turn on the loop. Go back to the library and
let the sample play. In Impulse-Tracker you always have two lines that mark
LoopStart and LoopEnd. Increment the LoopStart value until the first line is
at the beginning of the  constant area and decrement the LoopEnd value until
the second line marks the end of the area. It can be that you must sometimes
go back to the lib and play the sample again. The looped area must not be to
small. If you have problems with that because you work on a short sample then
try a ping-pong loop, this virtually doubles the length.
Now listen exactly! You can hear two loop-clicks. Play carefully with the
LoopStart value until the volume of one click gets lower. First try the last
two digits and if it don't changes anything then try the last three or four
digits. If you try long enough, you can remove the click almost completely.
Do the same procedure with the LoopEnd value then the clicks are removed.

Such looped samples are great for using long volume-envelopes on them and
especially the 'Spot with Reverb'-envelope is often used to fake Reverb.
 
 

2. Syncoped beats

It is not easy to explain, what syncoped beats are, if you don't know much
about the notation of music and stuff. I'll try to explain it with the
counting-system :

                      How to count rythm       How to count rythm
                      without syncopes         with syncopes

C-4 01 30 .00           1                      1,2
C-4 01 30 .00                                  3
C-4 01 64 .00           2                      4,5
C-4 01 64 .00                                  6
C-4 01 30 .00           3                      1,2
C-4 01 30 .00                                  3
C-4 01 64 .00           4                      4,5
C-4 01 64 .00                                  6
 

a) Method #1 : Pattern delay
A syncoped beat is often used and brings a special kind of groove.
The easiest way is to decrement the initial speed (standard is 6 - must be
decremented to 4) and to use the pattern-delay-effect on every second note :

C-4 01 30 SE1
C-4 01 30 .00
C-4 01 64 SE1
C-4 01 64 .00
C-4 01 30 SE1
C-4 01 30 .00
C-4 01 64 SE1
C-4 01 64 .00
     .
     .

But with this method it is hard (I think impossible) to play also non-syncoped
notes. So if you want to mix syncoped and non-syncoped beat, then better use
another method.

b) Method #2 : Altering speed
Here you change the inital speed from 6 to 4 to 6 and so on in every row per
effect-command. This was often done in older modules, because the older
Trackers had no pattern-delay effect-command.
It looks like this :

C-4 01 30 A06
C-4 01 30 A04
C-4 01 64 A06
C-4 01 64 A04
C-4 01 30 A06
C-4 01 30 A04
C-4 01 64 A06
C-4 01 64 A04
     .
     .

If you want to mix syncoped and non-syncoped beat, you must double the
initial speed (taking 210 as tempo in the F12-menu) and delay the non-
syncoped notes. Because the patterns run through now with the double speed,
you need to convert the old channels by marking the whole pattern
(Press ALT-L for two times) and pressing ALT-F then. Warning : Row 32-64
get lost if you don't double the pattern-length in the CTRL-F2-menu from
64 to 128, too. The result looks like this :

C-4 01 30 A06                   C-4 01 30 .00
... .. .. .00                   C-4 01 30 SD3  <--!
C-4 01 30 A04  <-!              ... .. .. .00     !
... .. .. .00    !              ... .. .. .00     !--- Non-syncoped notes
C-4 01 64 A06    !-- syncoped   C-4 01 30 .00     !
... .. .. .00    !   notes      C-4 01 30 SD3  <--!
C-4 01 64 A04  <-!              ... .. .. .00
... .. .. .00                   ... .. .. .00
     .                               .
     .                               .

c) Method #3 : 6/4-measure
This means to work with 48 or 96 rows as pattern-length. Press CTRL-F2
in the pattern-edit-mode (F2) and change the pattern length from 64 to 48
and - to make the ALT-D-block-mark-command compatible - press F2 and set
'Row hilight minor' to 3 and 'Row hilight major' to 12. Press F12 then and
change the 'Inital speed' from 6 to 4. The tempo is authentic to the bpm then.
Seen from the counting-system that I mentioned at the beginning, this is the
most plausible method, because you work with 6/4-measure now instead of
4/4-measure and the syncoped notes become trioles.

The result looks like this :                      Counting: New     (old)

C-5 01 30 .00                   C-5 01 30 .00              1         (1)
... .. .. .00                   C-5 01 30 SD2  <--!        2
C-5 01 30 .00                   ... .. .. .00     !        3
C-5 01 64 .00                   C-5 01 64 .00     !        4         (2)
... .. .. .00                   C-5 01 64 SD2  <--!        5
C-5 01 64 .00                   ... .. .. .00     !        6
C-5 01 30 .00                   C-5 01 30 .00     !        1         (3)
... .. .. .00                   C-5 01 30 SD2  <--!        2
C-5 01 30 .00                   ... .. .. .00     !        3
C-5 01 64 .00                   C-5 01 64 .00     !        4         (4)
... .. .. .00                   C-5 01 64 SD2  <--!        5
C-5 01 64 .00                   ... .. .. .00    non-      6
     .                                           syncoped
     .
 
 

Chapter V - Effects
===================

Well, I have described how to create the most important effects like reverb,
echo and chorus/flanger with the Tracker. What effects are there remaining
that are not mentioned in the Impulse Tracker manual ?
 

1. Conventinal effects

a)  Gapper/snipper with Oxx-command

You know, with the Oxx-effect-command you have the possibility to play a
sample not only from the beginning but also from anothere position. Every
sample has a specific length, so the first thing is to find out the offset-
length of the sample. Try to play the sample with O10 for example. Raise the
value, until you approximate it to the sample end where you hear nothing.
Now you can use any length of rows over which the sample should be played if
you simply use the offset-effect command and slide from 0 to the end-value.
Let's take a sample and you know its offset-length is 32 for example :

C-4 01 .. O00
C-4 01 .. O02
C-4 01 .. O04
C-4 01 .. O06
C-4 01 .. O08
C-4 01 .. O10
C-4 01 .. O12
C-4 01 .. O14
C-4 01 .. O16
C-4 01 .. O18
C-4 01 .. O20
C-4 01 .. O22
C-4 01 .. O24
C-4 01 .. O26
C-4 01 .. O28
C-4 01 .. O30
C-4 01 .. O32
      .
     .
      .
The sample is played in a shorter time here, than it originally playes, to
it is snipped, but you can also gap it, if you play it over more than 32 rows.
This technique is often used on speak-samples.
 

2. Unconventional effects

a) Highspeed Tracking

Have you ever tried to track with the fastest speed by pressing F12 and
raising the tempo- and inital speed-value as up as possible ?
You can generate tones with percussive samples. Take for example a snare and
let it play 16 times with highspeed. Sounds funny, eh? Now increment the
frequency by a halftone in every row. This sounds like a kind of filtering.
If you have found a good sound, you can make a sample of it. It's hard and
in my opinion senseless to track complete songs with that speed.
 
 
 

The ten golden rules of tracking

01. Sample every sound that you find good and useful. Sample often.
02. Collect samples, sort them and build a big library.
03. Analyze sounds and rebuild them and analyze other modules.
04. Remember that a good module doesn't sound like a module.
05. Use as many functions and techniques as possible.
06. Invest time and work in every song you track and don't hurry with it.
07. Try any style of music until you've found your own.
08. Be open for new influences and become an instrument of inspiration.
09. A programmer keeps the source-codes (The IT-files). Make MP3's.
10. Share the music with other people.