AY-Track: -= X-FILES THEME
=__________________________________________
__________________________________________
UNIVERSAL PT3 PLAYER
-------------------
So, Pro Tracker 3.31 ... In his SETUP'e
have the opportunity to change the table notes: PT or
ST. While you are compiling the music together with the
player'om - everything is in order. But when a couple dozen
Mouzon has only one player - trouble begins ...
The fact that the table is in the music
player (in fact, do not sew
is it to each Mouzon?). Thus, music,
calculated to another plate, to be played so player'om
disgusting (especially if you use Envelope - they have
something to play the old way, regardless of the table notes!).
With such a phenomenon faced by DANIEL
preparation of Deja Vu # 07. In addition obezglyuchennoy
version, country walks version of the journal in which one
Mouzon (moy!) incorrectly played. In addition, the 128-th shell
of this version (in listalke) port arrow.
Fortunately, shortly glitches were seen and
corrected. For the Player had to keep
Two tables of notes (for 192 bytes each) and
for each type Mouzon box, indicating the table type. Moreover,
because of the catastrophic shortage of memory in the 128-th
shell, had one of those plates to throw at
area (257 bytes) interrupt vectors ...
This prompted me to study
Player. It turned out that compiled
module, there is one such ugly Baitik, indicate the type of
table. The player that Baitik is not used, and in the
description of him says nothing ... It is located at offset #
63 from the beginning of the module (without the Player),
immediately after the last byte of information about the artist
music.
Baitik this is zero if Mouzon designed for Pro
Tracker'ovskuyu table notes; unity - at Sound Tracker'ovskuyu.
Just in case, bring yourself a table:
Pro Tracker frequency table:
# C21 # B73 # ACE # A33 # 9A0 # 916 # 893 # 818 # 7A4 # 736
# 6CE # 66D
# 610 # 5B9 # 567 # 519 # 4D0 # 48B # 449 # 40C # 3D2 # 39B
# 367 # 336
# 308 # 2DC # 2B3 # 28C # 268 # 245 # 224 # 206 # 1E9 # 1CD
# 1B3 # 19B
# 184 # 16E # 159 # 146 # 134 # 122 # 112 # 103 # 0F4 # 0E6
# 0D9 # 0CD
# 0C2 # 0B7 # 0AC # 0A3 # 09A # 091 # 089 # 081 # 07A # 073
# 06C # 066
# 061 # 05B # 056 # 051 # 04D # 048 # 044 # 040 # 03D # 039
# 036 # 033
# 030 # 02D # 02B # 028 # 026 # 024 # 022 # 020 # 01E # 01C
# 01B # 019
# 018 # 016 # 015 # 014 # 013 # 012 # 011 # 010 # 00F # 00E
# 00D # 00C
Sound Tracker frequency table:
# EF8 # E10 # D60 # C80 # BD8 # B28 # A88 # 9F0 # 960 # 8E0
# 858 # 7E0
# 77C # 708 # 6B0 # 640 # 5EC # 594 # 544 # 4F8 # 4B0 # 470
# 42C * 3F0
# 3BE # 384 # 358 # 320 # 2F6 # 2CA # 2A2 # 27C # 258 # 238
# 216 # 1F8
# 1DF # 1C2 # 1AC # 190 # 17B # 165 # 151 # 13E # 12C # 11C
* 10B # 0FC
# 0EF # 0E1 # 0D6 # 0C8 # 0BD # 0B2 # 0A8 # 09F # 096 # 08E
# 085 # 07E
# 077 # 070 # 06B # 064 # 05E # 059 # 054 # 04F # 04B # 047
# 042 # 03F
# 03B # 038 # 035 # 032 # 02F # 02C # 02A # 027 # 025 # 023
# 021 # 01F
# 01D # 01C # 01A # 019 # 017 # 016 # 015 # 013 # 012 # 011
# 010 # 00F
Pay attention to the numbers from last
table marked with an asterisk. In PT3 they
were equal and # 3FD # 10A. In the above
table (and in player'e), I replaced them with
value of this ST (apparently, if you copy plates of ST in the
PT3, the author some inaccuracies).
Now about player'e. After completion of his
length has increased by 24 bytes. Such a small increase was
achieved taking into account some features of the frequency
table. The fact that the frequencies of notes two adjacent
octaves are as 1:2. T. is, for example, the frequency of the
notes "to" 1 st octave 2-fold lower than the frequency of the
same notes 2 nd octave. Since the AY'shke for job
sound frequency used something like a period of acoustic
oscillations (or more precisely - the division factor of
frequency), and the period and frequency sound is inversely
proportional to each other, then the ratio will be 2:1.
Thus, knowing the coefficients for
notes one octave, you can get all the rest. Because all the
coefficients - whole numbers, then in order to minimize
computational errors to start with the lowest octave, and
"Frequency" of other music to get by dividing the octave in
half.
All of this is easy to see from the above tables: the number
of 2-th row 2 times manshe numbers first, etc.
The new player is recommended for logs
and various music players.
It removes the problems associated with the sign notes, but
there are still problems with importing music from other
editors: the ST and STP is a parameter HEIGHT (in PT3 it's not);
in ST ornamental looped as well as a sample;
in STP and PT3 different ways lowering the sample volume (at
STP - by subtracting the amplitude of a number of dumping it in
a zero if there was a shift in PT3 - on the table) etc.
Listig the Player is very big, so the
posting it here does not make sense. All
source (in XAS'e) in the appendix.
It uses some of the specific
for XAS'a expression:
label'h - the selection of the high byte.
DEFS 123, # 0000 - 123 Reservation zero.
Also note that the XAS
does not comply with the priority of operations - the
expressions are evaluated from left to right.
Here I quote a small piece of the listing, in charge of
selecting the tables and their "dekryunching:
;----------------------------------------; At the entrance to
the A: 0 or 1 . ;
; Choice of one of two tables:
LD HL, FREQ0 +24
LD DE, FREQ1 +24
PUSH HL
OLD_F CP 0
LD (OLD_F +1), A
JR Z, NO_SWAP
; Swaps 12 notes (24 bytes) so that
; With tags FREQ0 were needed:
SWAP DEC E
DEC L
LD C, (HL)
LD A, (DE)
LD (HL), A
LD A, C
LD (DE), A
JR NZ, SWAP
NO_SWAP POP DE
; Further, from the lowest octave, we obtain
And seven others:
LD L, B
LD B, 12 * 7
FRQ_GEN LD A, (HL)
INC L
LD C, (HL)
INC L
SRL C
RRA
LD (DE), A
INC E
LD A, C
LD (DE), A
INC E
DJNZ FRQ_GEN
; Table "frequencies." Located at the address
; ORG + # 200. Length - 192 bytes.
; In player'e before the first initialization
, This table contains only
, 24 initial bytes (12 notes PT3):
FREQ0 DW # C21, # B73, # ACE, # A33
DW # 9A0, # 916, # 893, # 818
DW # 7A4, # 736, # 6CE, # 66D
; Remaining 192-24 = 168 bytes scored zero
, (Just before the first initialization).
; First 12 notes ST, located at the end
; The Player:
FREQ1 DW # EF8, # E10, # D60, # C80
DW # BD8, # B28, # A88, # 9F0
DW # 960, # 8E0, # 858, # 7E0
;----------------------------------------
In addition, I am a bit (8 bytes) has reduced the
initializer and put EI on the output.
When you call [ORG +0] interrupt is prohibited, followed by
the initialization. At the exit interrupts are allowed (put EI).
When you call [ORG +5] interrupts do not touch. WARNING:
when playing yuzaetsya stack and register IY. The parish at
this time Interrupt highly undesirable! At the exit
in LY entered # 3A (HY player'om does not change), HL 'banging
and not restored!
When you call [ORG +8] does not spoil or register IY, or HL
'. Interrupts are not changed.
Unfortunately, we had to bang all the inscriptions in
player'e. It remains only 3 bytes in which I stuffed "CBX". If
it does not suit kogolibo - replace them with "PT3" or "GDC" ...
- - - -----====***< March 6, 1999
