--- amv.pl	2007/07/19 21:16:30	7
+++ amv.pl	2007/07/21 17:03:56	23
@@ -7,14 +7,38 @@
 # Various useful links used to produce this:
 # http://www.moviecodec.com/topics/15431p1.html
 # http://en.wikipedia.org/wiki/RIFF_(File_format)
+# http://www.obrador.com/essentialjpeg/HeaderInfo.htm
+# http://lists.helixcommunity.org/pipermail/datatype-dev/2005-January/001886.html
 
 use strict;
 
 use Data::Dump qw/dump/;
 use Carp qw/confess/;
+use File::Path;
+use Getopt::Long;
+
+my $dump = 0;
+my $debug = 0;
+my $dump_dir = '/tmp/dump/';
+my $dump_avi = "dump.avi";
+my $no_jpeg_header = 0;
+my $jpeg_q = 100;
+my $jpegtran = '-flip vertical';
+
+GetOptions(
+	"dump!"			=> \$dump,
+	"debug!"		=> \$debug,
+	"dump-dir=s"	=> \$dump_dir,
+	"no-jpeg-headers!" => \$no_jpeg_header,
+	"jpegtran=s"	=> \$jpegtran
+);
 
 my $path = shift @ARGV || die "usage: $0 movie.amv\n";
 
+
+rmtree $dump_dir if -e $dump_dir;
+mkpath $dump_dir || die "can't create $dump_dir: $!";
+
 open(my $fh, '<', $path) || die "can't open $path: $!";
 
 # offset in file
@@ -24,11 +48,20 @@
 my $d;
 
 sub hex_dump {
-	my $bytes = shift || return;
+	return unless $dump;
+
+	my ( $bytes, $offset ) = @_;
+	return unless $bytes;
+
+	my $old_o;
+	if (defined($offset)) {
+		$old_o = $o;
+		$o = $offset;
+	}
 
 	my $ascii = $bytes;
 	$ascii =~ s/\W/./gs;
-	my $hex = unpack('h*', $bytes);
+	my $hex = uc( unpack('h*', $bytes) );
 	$hex =~ s/(..)/$1 /g;
 	# calculate number of characters for offset
 	#my $d = length( sprintf("%x",length($bytes)) );
@@ -45,6 +78,8 @@
 			last;
 		}
 	}
+
+	$o = $old_o if $old_o;
 }
 
 sub x {
@@ -59,14 +94,14 @@
 	hex_dump( $bytes );
 
 	if ( $bytes eq 'AMV_END_' ) {
-		warn "> end of file marker AMV_END_\n";
+		print "> end of file marker AMV_END_\n" if $dump;
 		$d->{eof}++;
 		return;
 	}
 
 	if ( $format ) {
 		my @data = unpack($format, $bytes);
-		warn "## unpacked = ",dump(@data),"\n";
+		print "## unpacked = ",dump(@data),"\n" if $debug;
 		return @data;
 	} else {
 		return $bytes;
@@ -81,19 +116,209 @@
 	if ( $expected_len ) {
 		confess "expected $expected_len bytes for $part got $len" if $len != $expected_len;
 	}
-	printf ">> %s - %d 0x%x bytes\n", $part, $len, $len;
+	printf "<< %s - %d 0x%x bytes\n", $part, $len, $len;
 	x($len) if $skip;
 	return $len;
 }
 
-my ( $riff, $amv ) = x(12, 'Z8Z4');
-die "not RIFF but $riff" if $riff ne 'RIFF';
-die "not AMV but $amv" if $amv ne 'AMV ';
+sub quality {
+	my @table = @_;
+	die "quantization matrice needs to have 64 bytes!" if $#table != 63;
+
+	my $in = join('', map { chr($_) } @table );
+	my $out;
+
+	foreach my $t ( @table ) {
+		$t = int( ( $t * $jpeg_q ) / 100 );
+		$t = 255 if $t > 255;
+		$out .= chr($t);
+	}
+
+	if ( $dump ) {
+		print "## quantization table original\n";
+		hex_dump( $in );
+		print "## quantization table for $jpeg_q %\n";
+		hex_dump( $out );
+	}
+
+	return $out;
+}
+
+my @subframes;
+my $frame_nr = 1;
+
+# how many subframes to join into single frame?
+my $join_subframes = 0;
+
+sub mkjpg {
+	my ($data) = @_;
+
+	confess "no SOI marker in data" if substr($data,0,2) ne "\xFF\xD8";
+	confess "no EOI marker in data" if substr($data,-2,2) ne "\xFF\xD9";
+	$data = substr($data,2,-2);
+
+	if ( $#subframes < ( $join_subframes - 1 ) ) {
+		push @subframes, $data;
+		print "## saved $frame_nr/", $#subframes + 1, " subframe of ", length($data), " bytes\n";
+		return;
+	}
+
+	my $w = $d->{amvh}->{width} || die "no width?";
+	my $h = $d->{amvh}->{height} || confess "no height?";
+
+	my $header = 
+	# Start of Image (SOI) marker
+	"\xFF\xD8".
+	# JFIF marker
+	"\xFF\xE0".
+	pack("nZ5CCCnnCC",
+		16,			# length
+		'JFIF',		# identifier (JFIF)
+		1,1,		# version
+		0,			# units (none)
+		1,1,		# X,Y density
+		0,0,		# X,Y thumbnail
+	).
+	"\xFF\xFE".
+	"\x00\x3CCREATOR: amv dumper (compat. IJG JPEG v62), quality = 100\n".
+	# quantization table (quaility=100%)
+	"\xFF\xDB".
+	"\x00\x43".
+	# 8 bit values, table 1
+	"\x00".
+	quality(
+    0x10, 0x0B, 0x0C, 0x0E, 0x0C, 0x0A, 0x10, 0x0E,
+    0x0D, 0x0E, 0x12, 0x11, 0x10, 0x13, 0x18, 0x28,
+    0x1A, 0x18, 0x16, 0x16, 0x18, 0x31, 0x23, 0x25,
+    0x1D, 0x28, 0x3A, 0x33, 0x3D, 0x3C, 0x39, 0x33,
+    0x38, 0x37, 0x40, 0x48, 0x5C, 0x4E, 0x40, 0x44,
+    0x57, 0x45, 0x37, 0x38, 0x50, 0x6D, 0x51, 0x57,
+    0x5F, 0x62, 0x67, 0x68, 0x67, 0x3E, 0x4D, 0x71,
+    0x79, 0x70, 0x64, 0x78, 0x5C, 0x65, 0x67, 0x63,
+	).
+	"\xFF\xDB".
+	"\x00\x43".
+	# 8 bit values, table 1
+	"\x01".
+	quality(
+    0x11, 0x12, 0x12, 0x18, 0x15, 0x18, 0x2F, 0x1A,
+    0x1A, 0x2F, 0x63, 0x42, 0x38, 0x42, 0x63, 0x63,
+    0x63, 0x63, 0x63, 0x63, 0x63, 0x63, 0x63, 0x63,
+    0x63, 0x63, 0x63, 0x63, 0x63, 0x63, 0x63, 0x63,
+    0x63, 0x63, 0x63, 0x63, 0x63, 0x63, 0x63, 0x63,
+    0x63, 0x63, 0x63, 0x63, 0x63, 0x63, 0x63, 0x63,
+    0x63, 0x63, 0x63, 0x63, 0x63, 0x63, 0x63, 0x63,
+    0x63, 0x63, 0x63, 0x63, 0x63, 0x63, 0x63, 0x63,
+	).
+	# start of frame
+	"\xFF\xC0".
+	pack("ncnncc9",
+		17,			# len
+		8,			# sample precision in bits
+		$h,$w,		# X,Y size
+		3,			# number of components
+		1,0x22,0,	# Component ID, H+V sampling factors, Quantization table number
+		2,0x11,1,
+		3,0x11,1,
+	).
+	# Define huffman table (section B.2.4.1)
+	"\xFF\xC4".	# Marker
+	"\x00\x1F".	# Length (31 bytes)
+	"\x00".		# DC luminance, table 0
+	"\x00\x01\x05\x01\x01\x01\x01\x01\x01\x00\x00\x00\x00\x00\x00\x00".
+	"\x00\x01\x02\x03\x04\x05\x06\x07\x08\x09\x0A\x0B".
+	# Define huffman table (section B.2.4.1)
+	"\xFF\xC4".	# Marker
+	"\x00\xB5".	# Length (181 bytes)
+	"\x10".		# AC luminance, table 0
+	"\x00\x02\x01\x03\x03\x02\x04\x03\x05\x05\x04\x04\x00\x00\x01\x7D".
+	"\x01\x02\x03\x00\x04\x11\x05\x12".
+	"\x21\x31\x41\x06\x13\x51\x61\x07\x22\x71\x14\x32".
+	"\x81\x91\xA1\x08\x23\x42\xB1\xC1\x15\x52\xD1\xF0".
+	"\x24\x33\x62\x72\x82\x09\x0A\x16\x17\x18\x19\x1A".
+	"\x25\x26\x27\x28\x29\x2A\x34\x35\x36\x37\x38\x39".
+	"\x3A\x43\x44\x45\x46\x47\x48\x49\x4A\x53\x54\x55".
+	"\x56\x57\x58\x59\x5A\x63\x64\x65\x66\x67\x68\x69".
+	"\x6A\x73\x74\x75\x76\x77\x78\x79\x7A\x83\x84\x85".
+	"\x86\x87\x88\x89\x8A\x92\x93\x94\x95\x96\x97\x98".
+	"\x99\x9A\xA2\xA3\xA4\xA5\xA6\xA7\xA8\xA9\xAA\xB2".
+	"\xB3\xB4\xB5\xB6\xB7\xB8\xB9\xBA\xC2\xC3\xC4\xC5".
+	"\xC6\xC7\xC8\xC9\xCA\xD2\xD3\xD4\xD5\xD6\xD7\xD8".
+	"\xD9\xDA\xE1\xE2\xE3\xE4\xE5\xE6\xE7\xE8\xE9\xEA".
+	"\xF1\xF2\xF3\xF4\xF5\xF6\xF7\xF8\xF9\xFA".
+	# Define huffman table (section B.2.4.1)
+	"\xFF\xC4".	# Marker
+	"\x00\x1F".	# Length (31 bytes)
+	"\x01".		# DC chrominance, table 1
+	"\x00\x03\x01\x01\x01\x01\x01\x01\x01\x01\x01\x00".
+	"\x00\x00\x00\x00".
+	"\x00\x01\x02\x03\x04\x05\x06\x07\x08\x09\x0A\x0B".
+	#/* Define huffman table (section B.2.4.1) */
+	"\xFF\xC4".	# Marker
+	"\x00\xB5".	# Length (181 bytes)
+	"\x11".		# AC chrominance, table 1
+	"\x00\x02\x01\x02\x04\x04\x03\x04\x07\x05\x04\x04".
+	"\x00\x01\x02\x77".
+	"\x00\x01\x02\x03\x11\x04\x05\x21".
+	"\x31\x06\x12\x41\x51\x07\x61\x71\x13\x22\x32\x81".
+	"\x08\x14\x42\x91\xA1\xB1\xC1\x09\x23\x33\x52\xF0".
+	"\x15\x62\x72\xD1\x0A\x16\x24\x34\xE1\x25\xF1\x17".
+	"\x18\x19\x1A\x26\x27\x28\x29\x2A\x35\x36\x37\x38".
+	"\x39\x3A\x43\x44\x45\x46\x47\x48\x49\x4A\x53\x54".
+	"\x55\x56\x57\x58\x59\x5A\x63\x64\x65\x66\x67\x68".
+	"\x69\x6A\x73\x74\x75\x76\x77\x78\x79\x7A\x82\x83".
+	"\x84\x85\x86\x87\x88\x89\x8A\x92\x93\x94\x95\x96".
+	"\x97\x98\x99\x9A\xA2\xA3\xA4\xA5\xA6\xA7\xA8\xA9".
+	"\xAA\xB2\xB3\xB4\xB5\xB6\xB7\xB8\xB9\xBA\xC2\xC3".
+	"\xC4\xC5\xC6\xC7\xC8\xC9\xCA\xD2\xD3\xD4\xD5\xD6".
+	"\xD7\xD8\xD9\xDA\xE2\xE3\xE4\xE5\xE6\xE7\xE8\xE9".
+	"\xEA\xF2\xF3\xF4\xF5\xF6\xF7\xF8\xF9\xFA".
+	# Start of Scan marker
+	"\xFF\xDA".
+	pack("nC10",
+		12,			# length
+		3,			# number of components
+		1,0x00,		# Scan 1: use DC/AC huff tables 0/0
+		2,0x11,		# Scan 2: use DC/AC huff tables 1/1
+		3,0x11,		# Scan 3: use DC/AC huff tables 1/1
+		0,0x3f,		# Ss, Se
+		0,			# Ah, Ai (not used)
+	);
+
+	if ( $dump ) {
+		print "## created JPEG header...\n";
+		hex_dump( $header, 0 );
+	}
+
+	my $frame = join('', @subframes ) . $data;
+	@subframes = ();
+
+	my $path = sprintf("$dump_dir/%04d.jpg", $frame_nr++ );
+
+	my $fh;
+	if ( $jpegtran ) {
+		open($fh, '|-', "jpegtran $jpegtran > $path") || die "can't create $path: $!";
+	} else {
+		open($fh, '>', $path) || die "can't create $path: $!";
+	}
+
+	if ( ! $no_jpeg_header ) {
+		print $fh $header . $frame . "\xFF\xD9" || die "can't write jpeg $path: $!";
+	} else {
+		print $fh $frame || die "can't write raw jpeg $path: $!";
+	}
+	close $fh || die "can't close $path: $!";
+	print ">> created $frame_nr ", $no_jpeg_header ? 'raw' : '', " jpeg $path ", -s $path, " bytes\n";
+}
+
+my ( $riff, $amv ) = x(12, 'Z4x4Z4');
+die "$path not RIFF but $riff" if $riff ne 'RIFF';
+die "$path not AMV but $amv" if $amv ne 'AMV ';
 
 while ( ! defined($d->{eof}) ) {
 	my ( $list, $name ) = x(12,'A4x4A4');
 	die "not LIST but $list" if $list ne 'LIST';
-	print "> $list .. $name\n";
+	print "< $list * $name\n";
 
 	if ( $name eq 'hdrl' ) {
 
@@ -108,7 +333,7 @@
 		} x($len, 'Vx28VVVx8CCv');
 
 		printf "## %s %d*%d %s fps (%d ms/frame) %02d:%02d:%02d\n",
-			$h->{path},
+			$path,
 			$h->{width}, $h->{height}, $h->{fps}, $h->{ms_per_frame},
 			$h->{hh}, $h->{mm}, $h->{ss};
 
@@ -124,15 +349,20 @@
 		while (1) {
 			my $frame = $d->{movi}++;
 		
-			my $len = next_part( '00dc', 0, 1 );
+			my $len = next_part( '00dc' );
 			last unless $len;
-			printf ">> %s 00dc - frame %d jpeg %d 0x%x bytes\n", $name, $frame, $len, $len;
+			printf "<< %s 00dc - frame %d jpeg %d 0x%x bytes\n", $name, $frame, $len, $len;
+			mkjpg( x($len) );
 
-			my $len = next_part( '01wb', 0, 1 );
-			printf ">> %s 01wb - frame %d audio %d 0x%x bytes\n", $name, $frame, $len, $len;
+			$len = next_part( '01wb', 0, 1 );
+			printf "<< %s 01wb - frame %d audio %d 0x%x bytes\n", $name, $frame, $len, $len;
 		};
 
 	} else {
 		die "unknown $list $name";
 	}
 }
+
+my $cmd = "ffmpeg -i $dump_dir/%04d.jpg -r 16 -y $dump_avi";
+system($cmd) == 0 || die "can't convert frames to avi using $cmd: $!";
+