using System; //TODO - call on unmanaged code in mednadisc if available to do deinterleaving faster. be sure to benchmark it though.. //a decent little subcode reference //http://www.jbum.com/cdg_revealed.html //NOTES: the 'subchannel Q' stuff here has a lot to do with the q-Mode 1. q-Mode 2 is different, //and q-Mode 1 technically is defined a little differently in the lead-in area, although the fields align so the datastructures can be reused //Q subchannel basic structure: (quick ref: https://en.wikipedia.org/wiki/Compact_Disc_subcode) //Byte 1: (aka `status`) // q-Control: 4 bits (i.e. flags) // q-Mode: 4 bits (aka ADR; WHY is this called ADR?) //q-Data: other stuff depending on q-Mode and type of track //q-CRC: CRC of preceding namespace BizHawk.Emulation.DiscSystem { /// /// Control bit flags for the Q Subchannel. /// [Flags] public enum EControlQ { None = 0, PRE = 1, //Pre-emphasis enabled (audio tracks only) DCP = 2, //Digital copy permitted DATA = 4, //set for data tracks, clear for audio tracks _4CH = 8, //Four channel audio } /// /// Why did I make this a struct? I thought there might be a shitton of these and I was trying to cut down on object creation churn during disc-loading. /// But I ended up mostly just having a shitton of byte[] for each buffer (I could improve that later to possibly reference a blob on top of a memorystream) /// So, I should probably change that. /// public struct SubchannelQ { /// /// ADR and CONTROL /// TODO - make BCD2? PROBABLY NOT. I DONT KNOW. /// public byte q_status; /// /// normal track: BCD indication of the current track number /// leadin track: should be 0 /// public BCD2 q_tno; /// /// normal track: BCD indication of the current index /// leadin track: 'POINT' field used to ID the TOC entry # /// public BCD2 q_index; /// /// These are the initial set of timestamps. Meaning varies: /// check yellowbook 22.3.3 and 22.3.4 /// leadin track: unknown /// user information track: relative timestamp /// leadout: relative timestamp /// TODO - why are these BCD2? having things in BCD2 is freaking annoying, I should only make them BCD2 when serializing into a subchannel Q buffer /// EDIT - elsewhere I rambled "why not BCD2?". geh. need to make a final organized approach /// public BCD2 min, sec, frame; /// /// This is supposed to be zero.. but CCD format stores it, so maybe it's useful for copy protection or something /// public byte zero; /// /// These are the second set of timestamps. Meaning varies: /// check yellowbook 22.3.3 and 22.3.4 /// leadin track q-mode 1: TOC entry, absolute MSF of track /// user information track: absolute timestamp /// leadout: absolute timestamp /// public BCD2 ap_min, ap_sec, ap_frame; /// /// Don't assume this CRC is correct, in the case of some copy protections it is intended to be wrong. /// Furthermore, it is meaningless (and in BizHawk, unpopulated) for a TOC Entry /// (since an invalid CRC on a [theyre redundantly/duplicately stored] toc entry would cause it to get discarded in favor of another one with a correct CRC) /// public ushort q_crc; /// /// Retrieves the initial set of timestamps (min,sec,frac) as a convenient Timestamp /// public Timestamp Timestamp { get { return new Timestamp(min.DecimalValue, sec.DecimalValue, frame.DecimalValue); } set { min.DecimalValue = value.MIN; sec.DecimalValue = value.SEC; frame.DecimalValue = value.FRAC; } } /// /// Retrieves the second set of timestamps (ap_min, ap_sec, ap_frac) as a convenient Timestamp. /// public Timestamp AP_Timestamp { get { return new Timestamp(ap_min.DecimalValue, ap_sec.DecimalValue, ap_frame.DecimalValue); } set { ap_min.DecimalValue = value.MIN; ap_sec.DecimalValue = value.SEC; ap_frame.DecimalValue = value.FRAC; } } /// /// sets the status byte from the provided adr/qmode and control values /// public void SetStatus(byte adr_or_qmode, EControlQ control) { q_status = ComputeStatus(adr_or_qmode, control); } /// /// computes a status byte from the provided adr/qmode and control values /// public static byte ComputeStatus(int adr_or_qmode, EControlQ control) { return (byte)(adr_or_qmode | (((int)control) << 4)); } /// /// Retrives the ADR field of the q_status member (low 4 bits) /// public int ADR { get { return q_status & 0xF; } } /// /// Retrieves the CONTROL field of the q_status member (high 4 bits) /// public EControlQ CONTROL { get { return (EControlQ)((q_status >> 4) & 0xF); } } } //this has been checked against mednafen's and seems to match //there are a few dozen different ways to do CRC16-CCITT //this table is backwards or something. at any rate its tailored to the needs of the Q subchannel internal static class CRC16_CCITT { private static readonly ushort[] table = new ushort[256]; static CRC16_CCITT() { for (ushort i = 0; i < 256; ++i) { ushort value = 0; ushort temp = (ushort)(i << 8); for (byte j = 0; j < 8; ++j) { if (((value ^ temp) & 0x8000) != 0) value = (ushort)((value << 1) ^ 0x1021); else value <<= 1; temp <<= 1; } table[i] = value; } } public static ushort Calculate(byte[] data, int offset, int length) { ushort Result = 0; for(int i=0;i> 8) & 0xFF)); Result = (ushort)((Result << 8) ^ table[index]); } return Result; } } }