view DOCS/tech/mncf.txt @ 15171:c047b2d7f7ce

releaseclean should now work as expected, crate of Coke going in my direction.

	NUT Open Container Format DRAFT 20041005 (Michael's experimental fork)
		----------------------------------------


                
			Intro:

Features / goals: 
	(supported by the format, not necessary by a specific implementation)

Simple
	use the same encoding for nearly all fields
	simple decoding, so slow cpus (and embedded systems) can handle it
Extendible
	no limit for the possible values for all fields (using universal vlc)
	allow adding of new headers in the future
	allow adding more fields at the end of headers
Compact
	~0.2% overhead, for normal bitrates
	index is <10kb per hour (1 keyframe every 3sec)
	a usual header for a file is about 100bytes (audio + video headers together)
	a packet header is about ~1-8 bytes
Error resistant
	seeking / playback without an index
	headers & index can be repeated
	damaged files can be played back with minimal data lost and fast
	resyncing times


        
			Definitions:

MUST	the specific part must be done to conform to this standard
SHOULD	its recommanded to be done that way but its not strictly required



			Syntax:

		Type definitions:

f(x)	n fixed bits in big-endian order
u(x)	unsigned number encoded in x bits in MSB first order

v
	value=0
	do{
		more_data			u(1)
		data				u(7)
		value= 128*value + data
	}while(more_data)
        
s
	temp					v
	temp++
	if(temp&1) value= -(temp>>1)
	else       value=  (temp>>1)

b (binary data or string)
	for(i=0; i<length; i++){
		data[i]				u(8)
	}
	Note: strings MUST be encoded in utf8

vb
	length					v
	value					b


		Bitstream syntax:
packet header
	forward ptr				v

align_byte
	while(not byte aligned)
		one				f(1)

reserved_bytes
	for(i=0; i<forward_ptr - length_of_non_reserved; i++)
		reserved			u(8)
	a demuxer MUST ignore any reserved bytes
	a muxer MUST NOT write any reserved bytes, as this would make it
	inpossible to add new fields at the end of packets in the future in
	a compatible way

main header:
	main_startcode				f(64)
	packet header
	branch					v
	version					v
	stream_count				v
	max_distance				v
	max_index_distance			v
	global_time_base_nom			v
	global_time_base_denom			v
	for(i=0; i<256; ){
		tmp_flag			v
		tmp_fields			v
		if(tmp_fields>0) tmp_timestamp	s
		if(tmp_fields>1) tmp_mul	v
		if(tmp_fields>2) tmp_stream	v
		if(tmp_fields>3) tmp_size	v
		else tmp_size=0
		if(tmp_fields>4) tmp_res	v
		else tmp_res=0
		if(tmp_fields>5) count		v
		else count= tmp_mul - tmp_size
		for(j=6; j<tmp_fields; j++){
			tmp_reserved[i]		v
		}
		for(j=0; j<count && i<256; j++, i++){
			flags[i]= tmp_flag;
			stream_id_plus1[i]= tmp_stream;
			data_size_mul[i]= tmp_mul;
			data_size_lsb[i]= tmp_size + j;
			timestamp_delta[i]= tmp_timestamp;
			reserved_count[i]= tmp_res;
		}
	}
	reserved_bytes
	checksum				u(32)

stream_header:
	stream_startcode			f(64)
	packet_header
	stream_id				v
	stream_class				v
	fourcc					vb
	average_bitrate				v
	time_base_nom				v
	time_base_denom				v
	msb_timestamp_shift			v
	decode_delay				v
	fixed_fps				u(1)
	reserved				u(6)
	codec_specific_data			vb

video_stream_header:
	stream_header
	width					v
	height					v
	sample_width				v
	sample_height				v
	colorspace_type				v
	reserved_bytes
	checksum				u(32)

audio_stream_header:
	stream_header
	samplerate_nom				v
	samplerate_denom			v
	channel_count				v
	reserved_bytes
	checksum				u(32)

 
frame
	frame_code				f(8)
	if(stream_id_plus1[frame_code]==0){
		stream_id			v
	}
	if(timestamp_delta[frame_code]==0){
		coded_timestamp			v
	}
	if(flags[frame_code]&1){
		data_size_msb			v
	}
	for(i=0; i<reserved_count[frame_code]; i++)
		reserved			v
	data
                
Index:
	index_startcode				f(64)
	packet header
	stream_id				v
	index_length				v
	for(i=0; i<index_length; i++){
		index_timestamp			v
		index_position			v
	}
	reserved_bytes
	checksum				u(32)

info_packet: (optional)
	info_startcode				f(64)
	packet header
	for(;;){
		id				v
		if(id==0) break
		name= info_table[id][0]
		type= info_table[id][1]
		if(type==NULL)
			type			vb
		if(name==NULL)
			name			vb
		if(type=="v")
			value			v
		else
			value			vb
	}
	reserved_bytes
        checksum				u(32)
        
sync_point
	frame_startcode				f(64)
		global_timestamp		v

file
	file_id_string
	while(!eof && next_code != index_startcode){
		main_header
		for(i=0; i<stream_count; i++){
			if(next_packet==video_stream_header)
				video_stream_header
			else
				audio_stream_header
		}
		while(next_code != main_startcode){
			if(next_code == info_startcode)
				info_packet
			else{
				if(next_code == frame_startcode)                
					sync_point
				frame
			}
		}
	}
	index
                
forward_ptr
	size of the packet (exactly the distance from the first byte of the
	startcode of the current packet to the first byte of the following packet

file_id_string
	"nut/multimedia container\0"

*_startcode
        all startcodes start with 'N'

main_startcode
	0x7A561F5F04ADULL + (((uint64_t)('N'<<8) + 'M')<<48)
stream_starcode
	0x11405BF2F9DBULL + (((uint64_t)('N'<<8) + 'S')<<48)
frame_startcode
	0xE4ADEECA4569ULL + (((uint64_t)('N'<<8) + 'K')<<48)
	frame_startcodes SHOULD be placed immedeatly before a keyframe if the
	previous frame of the same stream was a non-keyframe, unless such
	non-keyframe - keyframe tansitions are very frequent

index_startcode
	0xDD672F23E64EULL + (((uint64_t)('N'<<8) + 'X')<<48)
info_startcode
	0xAB68B596BA78ULL + (((uint64_t)('N'<<8) + 'I')<<48)

branch
	1 for my fork, 0 for the official nut container

version
	2 for now

max_distance
	max distance of frame_startcodes, the distance may only be larger if
	there is only a single frame between the 2 frame_startcodes
	this can be used by the demuxer to detect damaged frame headers if the
	damage results in a too long chain
	SHOULD be set to <=32768 or at least <=65536 unless there is a very good
	reason to set it higher otherwise reasonable error recovery will be
	impossible

max_index_distance
	max distance of keyframes which are represented in the index, the
	distance between consecutive entries A and B may only be larger if
	there are no keyframes within this stream between A and B
	SHOULD be set to <=32768 or at least <=65536 unless there is a very good
	reason to set it higher
        
stream_id
	Note: streams with a lower relative class MUST have a lower relative id
	so a stream with class 0 MUST allways have a id which is lower then any
	stream with class > 0
	stream_id MUST be < stream_count

stream_class
	0	video
	32	audio
	64	subtiles
	Note the remaining values are reserved and MUST NOT be used
	     a demuxer MUST ignore streams with reserved classes

fourcc
	identification for the codec
	example: "H264"
	MUST contain 2 or 4 bytes, note, this might be increased in the future
	if needed

time_base_nom / time_base_denom = time_base
	the number of timer ticks per second, this MUST be equal to the fps
	if the fixed_fps is 1
	time_base_denom MUST not be 0
	time_base_nom and time_base_denom MUST be relative prime
	time_base_nom MUST be < 2^31
	examples:
        	fps	time_base_nom	time_base_denom
		30	30		1
		29.97	30000		1001
		23.976	24000		1001
		sample_rate	sample_rate_mul	time_base_nom	time_base_denom
		44100		1		44100		1
		44100		64		11025		16 
		48000		1024		375		8  
		Note: the advantage to using a large sample_rate_mul is that the
		      timestamps need fewer bits

global_time_base_nom / global_time_base_denom = global_time_base
	the number of timer ticks per second
	global_time_base_denom MUST not be 0
	global_time_base_nom and global_time_base_denom MUST be relative prime
	global_time_base_nom MUST be < 2^31

global_timestamp
	timestamp in global_time_base units
	when a global_timestamp is encountered the last_timestamp of all streams
	is set to the following:
	ln= global_time_base_denom*time_base_nom
	sn= global_timestamp
	d1= global_time_base_nom
	d2= time_base_denom
	last_timestamp= (ln/d1*sn + ln%d1*sn/d1)/d2
	Note, this calculation MUST be done with unsigned 64 bit integers, and 
	is equivalent to (ln*sn)/(d1*d2) but this would require a 96bit integer

msb_timestamp_shift
	amount of bits in lsb_timestamp
	MUST be <16

decode_delay
	maximum time between input and output for a codec, used to generate dts
	from pts
	is 0 for streams without b frames, and 1 for streams with b frames, may
	be larger for future codecs
        
fixed_fps
	1 indicates that the fps is fixed

codec_specific_data
	private global data for a codec (could be huffman tables or ...)

frame_code
	the meaning of this byte is stored in the main header
	the value 78 ('N') is forbidden to ensure that the byte is always
	different from the first byte of any startcode

flags[frame_code]
	the bits of the flags from MSB to LSB are KD
	if D is 1 then data_size_msb is coded, otherwise data_size_msb is 0
	K is the keyframe_type 
		0-> no keyframe, 
		1-> keyframe, 
	flags=4 can be used to mark illegal frame_code bytes
	frame_code=78 must have flags=4
	* frames MUST not depend(1) upon frames prior to the last
	  frame_startcode
	depend(1) means dependancy on the container level (NUT) not dependancy
	on the codec level

stream_id_plus1[frame_code]
	must be <250
	if its 0 then the stream_id is coded in the frame

data_size_mul[frame_code]
	must be <16384

data_size_lsb[frame_code]
	must be <16384

timestamp_delta[frame_code]
	must be <16384 and >-16384

data_size       
	data_size= data_size_lsb + data_size_msb*data_size_mul;

coded_timestamp
	if coded_timestamp < (1<<msb_timestamp_shift) then its a
	lsb timestamp, otherwise its a full timestamp + (1<<msb_timestamp_shift)
	lsb timestamps are converted to full timesamps by:
	mask = (1<<msb_timestamp_shift)-1;
	delta= last_timestamp - mask/2
	timestamp= ((timestamp_lsb-delta)&mask) + delta
	a full timestamp must be used if there is no reference timestamp
	available after the last frame_startcode with the current stream_id
        
lsb_timestamp
	least significant bits of the timestamp in time_base precission
        Example: IBBP display order
		keyframe timestamp=0                     -> timestamp=0
		frame                    lsb_timestamp=3 -> timestamp=3
		frame                    lsb_timestamp=1 -> timestamp=1
		frame                    lsb_timestamp=2 -> timestamp=2
		...
		keyframe msb_timestamp=257               -> timestamp=257
		frame                    lsb_timestamp=255->timestamp=255
		frame                    lsb_timestamp=0 -> timestamp=256
		frame                    lsb_timestamp=4 -> timestamp=260
		frame                    lsb_timestamp=2 -> timestamp=258
		frame                    lsb_timestamp=3 -> timestamp=259
	all timestamps of keyframes of a single stream MUST be monotone

dts
	dts are calculated by using a decode_delay+1 sized buffer for each 
	stream, into which the current pts is inserted and the element with
	the smallest value is removed, this is then the current dts
	this buffer is initalized with decode_delay -1 elements
	all frames with dts == timestamp must be monotone, that means a frame
	which occures later in the stream must have a larger or equal dts
	then an earlier frame
	FIXME rename timestamp* to pts* ?

width/height
	MUST be set to the coded width/height

sample_width/sample_height (aspect ratio)
	sample_width is the horizontal distance between samples
	sample_width and sample_height MUST be relative prime if not zero
	MUST be 0 if unknown
        
colorspace_type
	0	unknown
	1	ITU Rec 624 / ITU Rec 601 Y range: 16..235 Cb/Cr range: 16..240
	2	ITU Rec 709               Y range: 16..235 Cb/Cr range: 16..240
	17	ITU Rec 624 / ITU Rec 601 Y range:  0..255 Cb/Cr range:  0..255
	18	ITU Rec 709               Y range:  0..255 Cb/Cr range:  0..255

samplerate_nom / samplerate_denom = samplerate
	the number of samples per second

checksum
	adler32 checksum

index_timestamp
	value of the timetamp of a keyframe relative to the last keyframe
	stored in this index

index_position
	position in bytes of the first byte of a keyframe, relative to the
	last keyframe stored in this index
	there MUST be no keyframe with the same stream_id as this index between
	2 consecutive index entries if they are more then max_index_distance
	appart

id
	the id of the type/name pair, so its more compact
	0 means end
                
type
	for example: "UTF8" -> String or "JPEG" -> jpeg image
	Note: nonstandard fields should be prefixed by "X-"
	Note: MUST be less than 6 byte long (might be increased to 64 later)

name
	the name of the info entry, valid names are
	"TotalTime"	total length of the stream in msecs
	"StreamId"	the stream(s) to which the info packet applies
	"StartTimestamp" 
	"EndTimestamp" 	the time range in msecs to which the info applies
	"SegmentId"	a unique id for the streams + time specified
	"Author"
	"Description"
	"Copyright"
	"Encoder"	the name & version of the software used for encoding
	"Title"
	"Cover"		an image of the (cd,dvd,vhs,..) cover (preferable PNG or JPEG)
	"Source"	"DVD", "VCD", "CD", "MD", "FM radio", "VHS", "TV", 
			"LD"
			Optional: appended PAL,NTSC,SECAM, ... in parentheses
	"CaptureDevice"	"BT878", "BT848", "webcam", ... (more exact names are fine too)
	"CreationTime"	"2003-01-20 20:13:15Z", ... 
			(ISO 8601 format, see http://www.cl.cam.ac.uk/~mgk25/iso-time.html)
			Note: dont forget the timezone
	"ReplayGain"
	"Keywords"
	"Language"	ISO 639 and ISO 3166 for language/country code
			something like "usen" (US english), can be 0 if unknown
			and "multi" if several languages
			see http://www.loc.gov/standards/iso639-2/englangn.html
			and http://www.din.de/gremien/nas/nabd/iso3166ma/codlstp1/en_listp1.htmlthe language code
	"Disposition"	"original", "dub" (translated), "comment", "lyrics", "karaoke"
	Note: if someone needs some others, please tell us about them, so we can
	      add them to the official standard (if they are sane)
	Note: nonstandard fields should be prefixed by "X-"
	Note: MUST be less than 64 bytes long

value
	value of this name/type pair
	
stuffing
        0x80 can be placed infront of any type v entry for stuffing
	      purposes

info_table[][2]={
	{NULL			,  NULL }, // end
	{NULL			,  NULL },
	{NULL			, "UTF8"},
	{NULL			, "v"},
	{NULL			, "s"},
	{"StreamId"		, "v"},
	{"SegmentId"		, "v"},
	{"StartTimestamp"	, "v"},
	{"EndTimestamp"		, "v"},
	{"Author"		, "UTF8"},
	{"Titel"		, "UTF8"},
	{"Language"		, "UTF8"},
	{"Description"		, "UTF8"},
	{"Copyright"		, "UTF8"},
	{"Encoder"		, "UTF8"},
	{"Keyword"		, "UTF8"},
	{"Cover"		, "JPEG"},
	{"Cover"		, "PNG"},
	{"Disposition"		, "UTF8"},
};

			Structure:

the headers MUST be in exactly the following order (to simplify demuxer design)
main header
stream_header (id=0)
stream_header (id=1)
...
stream_header (id=n)

headers may be repated, but if they are then they MUST all be repeated together
and repeated headers MUST be identical
headers MAY only repeated at the closest possible positions after 2^x where x is 
an integer and the file end, so the headers may be repeated at 4102 if thats the
closest possition after 2^12=4096 at which the headers can be placed

headers MUST be placed at least at the begin of the file and immedeatly before
the index or at the file end if there is no index
headers MUST be repeated at least twice (so they exist 3 times in a file)

a demuxer MUST not demux a stream which contains more than one stream, or which
is wrapped in a structure to facilitate more than one stream or otherwise 
duplicate the role of a container. any such file is to be considered invalid

info packets which describe the whole file or individual streams/tracks must be
placed before any video/audio/... frames

		Index
Note: in case of realtime streaming there is no end, so no index there either

		Info packets
the info_packet can be repeated, it can also contain different names & values
each time but only if allso the time is different
Info packets can be used to describe the file or some part of it (chapters)

info packets, SHOULD be placed at the begin of the file at least
for realtime streaming info packets will normally be transmitted when they apply
for example, the current song title & artist of the currently shown music video

		Unknown packets
MUST be ignored by the demuxer

			demuxer (non-normative)

in the absence of valid header at beginning, players SHOULD search for backup
headers starting at offset 2^x for each x players SHOULD end their search from a
particular offset when any startcode is found (including syncpoint)


			Sample code (GPL, & untested)

typedef BufferContext{
	uint8_t *buf;
	uint8_t *buf_ptr;
}BufferContext;

static inline uint64_t get_bytes(BufferContext *bc, int count){
	uint64_t val=0;

	assert(count>0 && count<9)
        
	for(i=0; i<count; i++){
		val <<=8;
		val += *(bc->buf_ptr++);
	}
        
	return val;
}

static inline void put_bytes(BufferContext *bc, int count, uint64_t val){
	uint64_t val=0;

	assert(count>0 && count<9)

	for(i=count-1; i>=0; i--){
		*(bc->buf_ptr++)= val >> (8*i);
	}
        
	return val;
}

static inline uint64_t get_v(BufferContext *bc){
	uint64_t val= 0;

	for(; space_left(bc) > 0; ){
		int tmp= *(bc->buf_ptr++);
		if(tmp&0x80)
			val= (val<<7) + tmp - 0x80;
		else
			return (val<<7) + tmp;
	}
        
	return -1;
}

static inline int put_v(BufferContext *bc, uint64_t val){
	int i;
        
	if(space_left(bc) < 9) return -1;

	val &= 0x7FFFFFFFFFFFFFFFULL; // FIXME can only encode upto 63 bits currently
	for(i=7; ; i+=7){
		if(val>>i == 0) break;
	}

	for(i-=7; i>0; i-=7){
		*(bc->buf_ptr++)= 0x80 | (val>>i);
	}
	*(bc->buf_ptr++)= val&0x7F;
        
	return 0;
}

static int64_t get_dts(int64_t pts, int64_t *pts_cache, int delay, int reset){
	if(reset) memset(pts_cache, -1, delay*sizeof(int64_t));

	while(delay--){
		int64_t t= pts_cache[delay];
		if(t < pts){
			pts_cache[delay]= pts;
			pts= t;
		}
	}

	return pts;
}

		        Authors

Folks from MPlayer Developers Mailinglist (http://www.mplayehrq.hu/).
Authors in ABC-order: (FIXME! Tell us if we left you out)
    Beregszaszi, Alex ([e-mail])
    Bunkus, Moritz ([e-mail])
    Diedrich, Tobias ([e-mail])
    Felker, Rich ([e-mail])
    Franz, Fabian ([e-mail])
    Gereoffy, Arpad ([e-mail])
    Hess, Andreas ([e-mail])
    Niedermayer, Michael ([e-mail])