gc-forever - Gamecube/Wii Forums

pin  Signal
1    AISLR (audio bus)
2    5V
3    AISD (audio bus)
4    5V
5    AISCLK (audio bus)
6    5V
7    DIHSTRB
8    5V
9    DIERRB
10   Ground
11   DIBRK
12   DICOVER
13   DIDSTRB
14   DIRSTB
15   DIDIR
16   Ground
17   DID7
18   Ground
19   DID6
20   Ground
21   DID5
22   Ground
23   DID4
24   Ground
25   DID3
26   Ground
27   DID2
28   MONI
29   DID1
30   MONOUT
31   DID0
32   Ground

Pin: Signal
--- Streaming Audio ---
1: AISLR ("Audio: Stereo L/R")
3: AISD ("Audio: Data")
5: AISCLK ("Audio: Clock")

--- Data Communication ---
7: DIHSTRB ("Host Strobe")
13: DIDSTRB ("Device Strobe")
15: DIDIR ("Direction Line": indicates who drives the Data Bus)
17, 19, 21, 23, 25, 27, 29, 31: DID0-7 (8-bit bidirectional Data Bus)

--- Interrupt ---
14: DIRSTB ("Reset": Cube may instruct the DVD drive to "reset")
9: DIERRB ("Error State": indicates drive error to mainboard)
12: DICOVER ("Door/Cover State": indicates door/cover is open, to mainboard)
11 DIBRK

--- Other ---
28 MONI
30 MONOUT

--- Power/Ground ---
2, 4, 6, 8: +5V
10, 16, 18, 20, 22, 24, 26, 32: Ground

Upon Power up or reset , 2 commands must be issued for proper use of audio streaming:

DVDReadDiskID A8000040,00000000,00000020
DVDLowAudioBufferConfig E4xx00yy,00000000,00000020

xx=byte 8 [0 or 1] from the disk header retrieved from DVDReadDiskID
yy=0 (if xx=0) or 0xA (if xx=1)


All immediate commands [Ex] respond with 4 bytes.

DVDLowRequestAudioStatus E2xx0000
xx=0 is stream playing?
response: 0,0,0,y
y=1 yes;0=no

xx=1 what is the current address
response: y,y,y,y
y=address of current audio buffer that is burning on disk with 0x2000 granularity
remember address is actaul byte offset>>2
i.e. it has a granularity of 0x8000 bytes which is 0x2000

xx=2 what was the start address of currently playing audio stream
response: starting address of current audio stream >>2

xx=3 what was the length of currently playing audio stream
response: length of current audio stream


DVDCancelStreamAsync E1010000,xxxxxxxx,yyyyyyyy
response: 0,0,0,0
cancel current audio stream

DVDStopStreamAtEndAsync E1000000,xxxxxxxx,yyyyyyyy
response: 0,0,0,0
start audio stream at x with length y
if x and y are zero then will stop audio stream at end of buffer.
if not zero then will add another stream to queue, when current one finishes
then next one will be played.
Queue is only 1 level deep, and is replaced by last command issued.

Some other commands:
DVDLowInquiry 12000000,00000000,00000020
response: 32 bytes with what appears to Date/Version info
00,00,00,00
20,02,04,02 Date code
61,00,00,00 Version
rest are zero

DVDLowSeek AB000000
seek to offset, reponse 4 bytes

DVDLowStopMotor E3000000
stop motor


To properly handle Door open and DIERRB:
When door is opened and command is issued [sometimes software doesnt check status of door open]
assert DIERRB until E0000000 command is issued.
Deassert DIERRB and send 4 byte response:

Error Responses:
high byte:
00 Ready.
01 Cover is opened.
02 Disk change.
03 No Disk.
04 Motor stop.
05 Disk ID not read.

low bytes:
000000 No error.
030200 No Seek complete.
031100 UnRecoverd read error.
052000 Invalid command operation code.
052001 Audio Buffer not set.
052100 Logical block address out of range.
052400 Invalid Field in command packet.
052401 Invalid audio command.
052402 Configuration out of permitted period.
062800 Medium may have changed.
023A00 Medium not present / Cover opened.
0B5A01 Operator medium removal request.
056300 End of user area encountered on this track.
020401 Disk ID not read.
020400 Motor stopped.
040800 Transfer protocol error.

Unseen wrote:AISLR should be driven by the drive in either case.

--GC
DID: inout std_logic_vector(7 downto 0);
DIHSTRB : in std_logic;
DIDIR : in std_logic;
DIDSTRB : out std_logic;
DIERRB : out std_logic;
DIBRK : inout std_logic:='Z';
DICOVER : out std_logic;
DIRSTB : in std_logic;

AISCLK : in std_logic;
AISD : out std_logic;
AISLR : in std_logic;

hornpipe2 wrote: I'm also planning to start one of these (based around some fast ~100mhz ARM microcontroller and an SD card) so I'll add my own notes as I discover issues.

// GC DVD ADPCM decoding algorithm 
// NOTE: the ADPCM format used by the dvd drive is *NOT* the same as the ADPCM format used by the DSP
// they are similiar but the dvd adpcm format requires less multiplies
// previous public decoder by Shinji Chiba is innacurate
// -Crazynation

#include <stdio.h>

int predL1,predL2,predR1,predR2;
int new_pred;

init_predictors()
{
	predL1=0;
	predL2=0;
	predR1=0;
	predR2=0;
}
int decode(int predictor1,int predictor2,int nibble,int index)
{
	int i,j,k;
	short  scalex,scaley;

	int data1,data2;


	switch(index>>4)
	{
		case 0:
			scalex=0;
			scaley=0;
			break;
		case 1:
			scalex=0x3C;
			scaley=0;
			break;
		case 2:
			scalex=0x73;
			scaley=0xFFCC;
			break;
		case 3:
			scalex=0x62;
			scaley=0xFFC9;
			break;
		default:
			break;
	}
	i=predictor1*scalex;
	j=predictor2*scaley;
	i=i+j+0x20;
	i=i>>6;
	if(i<(int)0xFFE00000)
		i=0xFFE00000;
	else if(i>0x1FFFFF)
		i=0x1FFFFF;

	data1=i;

	j=(nibble&0xF)<<12;
	j=(short)j>>(index&0xF);

	i=j<<6;
	i=i+data1;
	
	data2=i;
	new_pred=data2;

	j=data2>>6;
	if(j<(int)0xFFFF8000)
		j=0x8000;
	else if(j>0x7FFF)
		j=0x7FFF;

	return j;



	

}
char wav_header[]={0x52,0x49,0x46,0x46,0x26,0x40,0x21,0x00,0x57,0x41,0x56,0x45,0x66,0x6D,0x74,0x20,
0x10,0x00,0x00,0x00,0x01,0x00,0x02,0x00,0x80,0xBB,0x00,0x00,0x00,0xEE,0x02,0x00,
0x04,0x00,0x10,0x00,0x64,0x61,0x74,0x61,0x00,0x40,0x21,0x00};

int main(int argc, char *argv[])
{
	int i,j,k;
	int count,length,offset;
	int nibble_high,nibble_low;
	FILE *fin,*fout;
	unsigned char buffer[32];
	short wav[128];
	

	if(argc<3)
	{
		printf("decode <adpcm in file> <output wav file>\n");
		exit(1);
	}
	fin=fopen(argv[1],"rb");
	if(fin==0)
	{
		printf("cant open input file %s\n",argv[1]);
		exit(1);
	}

	fout=fopen(argv[2],"wb");
	if(fout==0)
	{
		printf("cant open output file %s\n",argv[2]);
		exit(1);
	}
	fseek(fin,0,SEEK_END);
	length=ftell(fin);
	fseek(fin,0,SEEK_SET);

	init_predictors();
	offset=0;
	fwrite(wav_header,44,1,fout);
//-----------------------------------------------------------------------------
decode:
	i=fread(buffer,32,1,fin);
	if(i==0)
		exit(1);

	if(buffer[0]!=buffer[2])
	{
		printf("invalid ADPCM file\n");
		exit(1);
	}

	if(buffer[1]!=buffer[3])
	{
		printf("invalid ADPCM file\n");
		exit(1);
	}
	count=0;
	do{

		nibble_high=buffer[count+4]>>4;
		nibble_low=buffer[count+4]&0xF;

		k=decode(predL1,predL2,nibble_low,buffer[0]);
		wav[count*2]=k;

		predL2=predL1;
		predL1=new_pred;

		k=decode(predR1,predR2,nibble_high,buffer[1]);
		wav[(count*2)+1]=k;

		predR2=predR1;
		predR1=new_pred;

	count++;
	}while(count<28);

	fwrite(wav,28*4,1,fout);
	
	offset+=32;
	if(offset<length)
		goto decode;

	length=ftell(fout);
	fseek(fout,4,SEEK_SET);
	i=length-0x6;
	fwrite(&i,4,1,fout);
	fseek(fout,0x28,SEEK_SET);
	i=length-0x2C;
	fwrite(&i,4,1,fout);

	


}

meneerbeer wrote:What kind of chip is on the DVD drive anyways? Also just a standard microprocessor?

hornpipe2 wrote:

Unseen wrote:AISLR should be driven by the drive in either case.

Actually... AISLR may be Cube controlled too. Check this VHDL for the CrazyNation version?

Unseen wrote:

hornpipe2 wrote:

Unseen wrote:AISLR should be driven by the drive in either case.

Actually... AISLR may be Cube controlled too. Check this VHDL for the CrazyNation version?

Indeed, I misremembered that - when an I2S receiver is the master device, the only signal driven by the sender is the data line.

meneerbeer wrote:

hornpipe2 wrote: I'm also planning to start one of these (based around some fast ~100mhz ARM microcontroller and an SD card) so I'll add my own notes as I discover issues.

Interesting. I have been thinking about that as well, but I am much more a hardware designer, so I usually settle with an FPGA. What kind of chip is on the DVD drive anyways? Also just a standard microprocessor?

    Matsushita MN103S13BGA Optical Disk Controller
    Matsushita MN102H60GFA MicroComputer clocked at 34mhz