Appendix 3

● About the format of binary files

Here, the format of the transmission file is described. Basically, the complete version of the format submitted by YAMAHA with undescribed parts, missing parts, and incorrect parts corrected is recorded. The general structure of the SATURN SCSP format is as follows.

Item	size
Mixer start offset from start	2 bytes
Velocity head offset from head	2 bytes
PEG start offset from start	2 bytes
PLFO head offset from head	2 bytes
Offset of 0th Voice data from the beginning	2 * a bytes
::
Offset of voice data a from the beginning
Mixer Data 0	18 * b bytes
::
Mixer Data b
VelocityData 0	10 * c bytes
::
VelocityData c
PEGData 0	10 * d bytes
::
PEGData d
PLFOData 0	4 * e bytes
::
PLFOData e
VoiceData 0	a {F (x) has 4 * a + Σ (F (i) * 32) for each voice Returns the number of layers i = 0 to be included}
::
VoiceData a
WaveData 0	f ΣW (i) W (x) returns the size of the waveform i = 0
::
WaveData f

● Details

Next, we will analyze each item in detail.

◆ Mixer part format

Mixer is 1 byte per channel, and 7-5 bits are assigned to Effect Send Level and Effect PAN is assigned to the rest.
The size will be 18 bytes.

	7	6	Five	Four	3	2	1	0
Channel 0	EffectSendLev				EffectPAN				1 byte
:	:				:				:
Channel 17	EffectSendLev				EffectPAN				1 byte

◆ Velocity part format

In the Velocity format, in addition to the parameters set on the tone editor, it is necessary to calculate another parameter for each point.
See Sound Tool Guide Vol1 for this parameter.
The size will be 10 bytes.

Slope-coded value 0	1 byte
VelocityPoint 0	1 byte
VelocityLevel 0	1 byte
Slope-coded value 1	1 byte
VelocityPoint 1	1 byte
VelocityLevel 1	1 byte
Slope-coded value 2	1 byte
VelocityPoint 2	1 byte
VelocityLevel 2	1 byte
Slope-coded value 3	1 byte

◆ PEG part format

Since the operability on the tone editor is prioritized, different values are assigned for editing and sending.
The level of PEG to be edited is the table number.

long LevelTable [128] = { 
 	     0, 125, 250, 375, 500, 625, 750, 875,  
 	  1000, 1125, 1250, 1375, 1500, 1625, 1750, 1875,  
 	  2000, 2125, 2250, 2375, 2500, 2625, 2750, 2875,  
 	  3000, 3125, 3250, 3375, 3500, 3625, 3750, 3875,  
 	  4000, 4250, 4500, 4750, 5000, 5250, 5500, 5750,  
 	  6000, 6250, 6500, 6750, 7000, 7250, 7500, 7750,  
 	  8000, 8250, 8500, 8750, 9000, 9250, 9500, 9750,  
 	 10000, 10250, 10500, 10750, 11000, 11250, 11500, 11750,  
 	 12000, 12500, 13000, 13500, 14000, 14500, 15000, 15500,  
 	 16000, 16500, 17000, 17500, 18000, 18500, 19000, 19500,  
 	 20000, 21000, 22000, 23000, 24000, 25000, 26000, 27000,  
 	 28000, 29000, 30000, 31000, 32000, 33000, 34000, 35000,  
 	 36000, 37000, 38000, 39000, 40000, 41000, 42000, 43000,  
 	 44000, 45000, 46000, 47000, 48000, 49000, 50000, 51000,  
 	 52000, 54000, 56000, 58000, 60000, 62000, 64000, 66000,  
 	 68000, 72000, 76000, 80000, 84000, 88000, 92000, 95990};

From this table and PEG Time

	PEGRate = LevelTable / PEGTime;

Is calculated to obtain the PEG Rate. From this, the table number is calculated and used as the transmission data.
The PEG rate table is written in C code as follows.

	long PEGRateTable [128];

	/ * Creation of PEG rate table (100 times the actual value) * /

	for (i = 0L; i <128L; ++ i)
		PEGRateTable [i] = (long) (((float) i * (float) i * (float) 100)	
				/ (float) 16384 * (float) 100); / * Find the slope * /

Delay table value	1 byte
OffsetLevel table value	1 byte
AttackRate table value	1 byte
AttackLevel table value	1 byte
DecayRate table value	1 byte
DecayLevel table value	1 byte
SustainRate table value	1 byte
SustainLevel table value	1 byte
ReleaseRate table value	1 byte
ReleaseLevel table value	1 byte

◆ PLFO part format

This parameter is not used when actually editing in the tone editor, but is edited as a table value. The edited value is written as it is.

Delay table value	1 byte
PLFO frequency table value	1 byte
AMP table value	1 byte
Fade timetable value	1 byte

◆ Voice part format

7	6	5	4	3	2	1	0
FM		Bend range width (0 -13)						1 byte
×		Portamento time (0 -127)						1 byte
×		Number of layers						1 byte
Volume Bias (-127 --127)								1 byte
Layer Data 0								32 bytes
::								::
Layer Data n								32 bytes

FM: For checking if this voice is in FM format

◆ Layer part format

The Layer part is contained in the Voice format.
Part of the format is written directly in the SCSP register.
This part is a little complicated, so I will explain it later.
The data size is 32 bytes.

7	6	5	4	3	2	1	0
×	Beginning of MIDI notes (0 -127)							1 byte
×	End of MIDI note (0 -127)							1 byte
SCSP register part 23 bytes								23 bytes
Base note (0 -127)								1 byte
Fine tune (-63 --63)								1 byte
FM connection 1 (0 -127)								1 byte
For FM connection 2 (0 -127)								1 byte
Velocity number (0 -255)								1 byte
PEG number (0 -255)								1 byte
PLFO number (0 -255)								1 byte

● About SCSP register part

Since this register is written directly in SCSP, one parameter does not reserve 1 byte.

reference

"HARDWARE MANUAL" / SCSP User's Manual / ■ Control register for each slot

15	14	13	12	11	Ten	9	8	7	6	5	4	3	2	1	0
PEG	PLFO	---	(1)	(2)	(3)		(Four)		Loop control		8Bit	StaetAddress (19:16)
Waveform Start Address (15-0)
Loop Start Address (15-0)
Loop End Address (15-0)
DecayRate2 (4-0)					DecayRate1 (4-0)					EG Old	Attack Rate (4-0)
---	(Five)	KeyRateScaling (3-0)				DecayLevel (4-0)					Release Rate (4-0)
LFO M	PEG M	PLFO M	---	---	---	(6)	(7)	TotalLevel (7-0)
FM Modulation (3-0)				FM connection 1 (5-0)						FM connection 2 (5-0)
---	(8)				TMod	(9)
LFOR	LFO frequency (4-0)					PLFO waveform		PLFO depth (2-0)			ALFO waveform		ALFO depth (2-0)
---	---	---	---	---	---	---	---	---	EffectSelect (3-0)				EffectSend (2-0)
DirectSend (2-0)		DirectPAN (4-0)

Parameters written in abbreviated form due to space limitations

Abbreviation	Contents
8Bit	Waveform sample size (stands at 16 Bit)
PEG	PEG switch
PLFO	PLFO switch
LFO M	LFO modulation
PEGM	PEG modulation
PLFOM	PLFOM modulation
TMod	Total level modulation
LFOR	LFO reset

In this table, only the parameters used in SEGA SatToneEd Ver1.00 are listed. The unused parameters are (1) to (9) in the table, and the contents are as follows.

number	Name	Contents
(1)	KEYONEX	Key ON execution (written by driver)
(2)	KEYONB	Key ON, Key OFF registration (written by driver)
(3)	SBCTL	Source bit control (written by driver)
(Four)	SSCTL	Sound source control (written by driver)
(Five)	LSLNK	Loop start link (written by driver)
(6)	STWINH	Stack write protection (write by driver)
(7)	SDIR	Sound Direct (written by driver)
(8)	OCT	Octave (written by driver)
(9)	FNS	Frequency number switch (written by driver)

◆ Waveform part format

The waveform format is saved by arranging the waveforms in the specified waveform order.
There is no header part and it is pure audio data.
Playback information, waveform addresses, etc. are all stored in the layer.

┌─────────────┐
│ Waveform data 1 │
├─────────────┤
│ Waveform data 2 │
├─────────────┤
│ Waveform data 3 │
├─────────────┤
│: │
├─────────────┤
│: │
├─────────────┤
│ Waveform data X │
└─────────────┘