[16] | 1 | <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> |
---|
| 2 | <html> |
---|
| 3 | <head> |
---|
| 4 | |
---|
| 5 | <meta http-equiv="Content-Type" content="text/html; charset=iso-8859-15"/> |
---|
| 6 | <title>Ogg Vorbis Documentation</title> |
---|
| 7 | |
---|
| 8 | <style type="text/css"> |
---|
| 9 | body { |
---|
| 10 | margin: 0 18px 0 18px; |
---|
| 11 | padding-bottom: 30px; |
---|
| 12 | font-family: Verdana, Arial, Helvetica, sans-serif; |
---|
| 13 | color: #333333; |
---|
| 14 | font-size: .8em; |
---|
| 15 | } |
---|
| 16 | |
---|
| 17 | a { |
---|
| 18 | color: #3366cc; |
---|
| 19 | } |
---|
| 20 | |
---|
| 21 | img { |
---|
| 22 | border: 0; |
---|
| 23 | } |
---|
| 24 | |
---|
| 25 | #xiphlogo { |
---|
| 26 | margin: 30px 0 16px 0; |
---|
| 27 | } |
---|
| 28 | |
---|
| 29 | #content p { |
---|
| 30 | line-height: 1.4; |
---|
| 31 | } |
---|
| 32 | |
---|
| 33 | h1, h1 a, h2, h2 a, h3, h3 a { |
---|
| 34 | font-weight: bold; |
---|
| 35 | color: #ff9900; |
---|
| 36 | margin: 1.3em 0 8px 0; |
---|
| 37 | } |
---|
| 38 | |
---|
| 39 | h1 { |
---|
| 40 | font-size: 1.3em; |
---|
| 41 | } |
---|
| 42 | |
---|
| 43 | h2 { |
---|
| 44 | font-size: 1.2em; |
---|
| 45 | } |
---|
| 46 | |
---|
| 47 | h3 { |
---|
| 48 | font-size: 1.1em; |
---|
| 49 | } |
---|
| 50 | |
---|
| 51 | li { |
---|
| 52 | line-height: 1.4; |
---|
| 53 | } |
---|
| 54 | |
---|
| 55 | #copyright { |
---|
| 56 | margin-top: 30px; |
---|
| 57 | line-height: 1.5em; |
---|
| 58 | text-align: center; |
---|
| 59 | font-size: .8em; |
---|
| 60 | color: #888888; |
---|
| 61 | clear: both; |
---|
| 62 | } |
---|
| 63 | </style> |
---|
| 64 | |
---|
| 65 | </head> |
---|
| 66 | |
---|
| 67 | <body> |
---|
| 68 | |
---|
| 69 | <div id="xiphlogo"> |
---|
| 70 | <a href="http://www.xiph.org/"><img src="fish_xiph_org.png" alt="Fish Logo and Xiph.org"/></a> |
---|
| 71 | </div> |
---|
| 72 | |
---|
| 73 | <h1>Ogg Vorbis I format specification: helper equations</h1> |
---|
| 74 | |
---|
| 75 | <h1>Overview</h1> |
---|
| 76 | |
---|
| 77 | <p>The equations below are used in multiple places by the Vorbis codec |
---|
| 78 | specification. Rather than cluttering up the main specification |
---|
| 79 | documents, they are defined here and linked in the main documents |
---|
| 80 | where appropriate.</p> |
---|
| 81 | |
---|
| 82 | <h2><a name="log">ilog</a></h2> |
---|
| 83 | |
---|
| 84 | <p>The "ilog(x)" function returns the position number (1 through n) of the |
---|
| 85 | highest set bit in the two's complement integer value |
---|
| 86 | <tt>[x]</tt>. Values of <tt>[x]</tt> less than zero are defined to return zero.</p> |
---|
| 87 | |
---|
| 88 | <pre> |
---|
| 89 | 1) [return_value] = 0; |
---|
| 90 | 2) if ( [x] is greater than zero ){ |
---|
| 91 | |
---|
| 92 | 3) increment [return_value]; |
---|
| 93 | 4) logical shift [x] one bit to the right, padding the MSb with zero |
---|
| 94 | 5) repeat at step 2) |
---|
| 95 | |
---|
| 96 | } |
---|
| 97 | |
---|
| 98 | 6) done |
---|
| 99 | </pre> |
---|
| 100 | |
---|
| 101 | <p>Examples:</p> |
---|
| 102 | |
---|
| 103 | <ul> |
---|
| 104 | <li>ilog(0) = 0;</li> |
---|
| 105 | <li>ilog(1) = 1;</li> |
---|
| 106 | <li>ilog(2) = 2;</li> |
---|
| 107 | <li>ilog(3) = 2;</li> |
---|
| 108 | <li>ilog(4) = 3;</li> |
---|
| 109 | <li>ilog(7) = 3;</li> |
---|
| 110 | <li>ilog(negative number) = 0;</li> |
---|
| 111 | </ul> |
---|
| 112 | |
---|
| 113 | <h2><a name="float32_unpack">float32_unpack</a></h2> |
---|
| 114 | |
---|
| 115 | <p>"float32_unpack(x)" is intended to translate the packed binary |
---|
| 116 | representation of a Vorbis codebook float value into the |
---|
| 117 | representation used by the decoder for floating point numbers. For |
---|
| 118 | purposes of this example, we will unpack a Vorbis float32 into a |
---|
| 119 | host-native floating point number.</p> |
---|
| 120 | |
---|
| 121 | <pre> |
---|
| 122 | 1) [mantissa] = [x] bitwise AND 0x1fffff (unsigned result) |
---|
| 123 | 2) [sign] = [x] bitwise AND 0x80000000 (unsigned result) |
---|
| 124 | 3) [exponent] = ( [x] bitwise AND 0x7fe00000) shifted right 21 bits (unsigned result) |
---|
| 125 | 4) if ( [sign] is nonzero ) then negate [mantissa] |
---|
| 126 | 5) return [mantissa] * ( 2 ^ ( [exponent] - 788 ) ) |
---|
| 127 | </pre> |
---|
| 128 | |
---|
| 129 | <h2><a name="lookup1_values">lookup1_values</a></h2> |
---|
| 130 | |
---|
| 131 | <p>"lookup1_values(codebook_entries,codebook_dimensions)" is used to |
---|
| 132 | compute the correct length of the value index for a codebook VQ lookup |
---|
| 133 | table of lookup type 1. The values on this list are permuted to |
---|
| 134 | construct the VQ vector lookup table of size |
---|
| 135 | <tt>[codebook_entries]</tt>.</p> |
---|
| 136 | |
---|
| 137 | <p>The return value for this function is defined to be 'the greatest |
---|
| 138 | integer value for which <tt>[return_value] to the power of |
---|
| 139 | [codebook_dimensions] is less than or equal to |
---|
| 140 | [codebook_entries]</tt>'.</p> |
---|
| 141 | |
---|
| 142 | <h2><a name="low_neighbor">low_neighbor</a></h2> |
---|
| 143 | |
---|
| 144 | <p>"low_neighbor(v,x)" finds the position <i>n</i> in vector [v] of |
---|
| 145 | the greatest value scalar element for which <i>n</i> is less than |
---|
| 146 | <tt>[x]</tt> and <tt>vector [v] element <i>n</i> is less |
---|
| 147 | than vector [v] element [x]</tt>.</p> |
---|
| 148 | |
---|
| 149 | <h2><a name="high_neighbor">high_neighbor</a></h2> |
---|
| 150 | |
---|
| 151 | <p>"high_neighbor(v,x)" finds the position <i>n</i> in vector [v] of |
---|
| 152 | the lowest value scalar element for which <i>n</i> is less than |
---|
| 153 | <tt>[x]</tt> and <tt>vector [v] element <i>n</i> is greater |
---|
| 154 | than vector [v] element [x]</tt>.</p> |
---|
| 155 | |
---|
| 156 | <h2><a name="render_point">render_point</a></h2> |
---|
| 157 | |
---|
| 158 | <p>"render_point(x0,y0,x1,y1,X)" is used to find the Y value at point X |
---|
| 159 | along the line specified by x0, x1, y0 and y1. This function uses an |
---|
| 160 | integer algorithm to solve for the point directly without calculating |
---|
| 161 | intervening values along the line.</p> |
---|
| 162 | |
---|
| 163 | <pre> |
---|
| 164 | 1) [dy] = [y1] - [y0] |
---|
| 165 | 2) [adx] = [x1] - [x0] |
---|
| 166 | 3) [ady] = absolute value of [dy] |
---|
| 167 | 4) [err] = [ady] * ([X] - [x0]) |
---|
| 168 | 5) [off] = [err] / [adx] using integer division |
---|
| 169 | 6) if ( [dy] is less than zero ) { |
---|
| 170 | |
---|
| 171 | 7) [Y] = [y0] - [off] |
---|
| 172 | |
---|
| 173 | } else { |
---|
| 174 | |
---|
| 175 | 8) [Y] = [y0] + [off] |
---|
| 176 | |
---|
| 177 | } |
---|
| 178 | |
---|
| 179 | 9) done |
---|
| 180 | </pre> |
---|
| 181 | |
---|
| 182 | <h2><a name="render_line">render_line</a></h2> |
---|
| 183 | |
---|
| 184 | <p>Floor decode type one uses the integer line drawing algorithm of |
---|
| 185 | "render_line(x0, y0, x1, y1, v)" to construct an integer floor |
---|
| 186 | curve for contiguous piecewise line segments. Note that it has not |
---|
| 187 | been relevant elsewhere, but here we must define integer division as |
---|
| 188 | rounding division of both positive and negative numbers toward zero.</p> |
---|
| 189 | |
---|
| 190 | <pre> |
---|
| 191 | 1) [dy] = [y1] - [y0] |
---|
| 192 | 2) [adx] = [x1] - [x0] |
---|
| 193 | 3) [ady] = absolute value of [dy] |
---|
| 194 | 4) [base] = [dy] / [adx] using integer division |
---|
| 195 | 5) [x] = [x0] |
---|
| 196 | 6) [y] = [y0] |
---|
| 197 | 7) [err] = 0 |
---|
| 198 | |
---|
| 199 | 8) if ( [dy] is less than 0 ) { |
---|
| 200 | |
---|
| 201 | 9) [sy] = [base] - 1 |
---|
| 202 | |
---|
| 203 | } else { |
---|
| 204 | |
---|
| 205 | 10) [sy] = [base] + 1 |
---|
| 206 | |
---|
| 207 | } |
---|
| 208 | |
---|
| 209 | 11) [ady] = [ady] - (absolute value of [base]) * [adx] |
---|
| 210 | 12) vector [v] element [x] = [y] |
---|
| 211 | |
---|
| 212 | 13) iterate [x] over the range [x0]+1 ... [x1]-1 { |
---|
| 213 | |
---|
| 214 | 14) [err] = [err] + [ady]; |
---|
| 215 | 15) if ( [err] >= [adx] ) { |
---|
| 216 | |
---|
| 217 | 15) [err] = [err] - [adx] |
---|
| 218 | 16) [y] = [y] + [sy] |
---|
| 219 | |
---|
| 220 | } else { |
---|
| 221 | |
---|
| 222 | 17) [y] = [y] + [base] |
---|
| 223 | |
---|
| 224 | } |
---|
| 225 | |
---|
| 226 | 18) vector [v] element [x] = [y] |
---|
| 227 | |
---|
| 228 | } |
---|
| 229 | </pre> |
---|
| 230 | |
---|
| 231 | <div id="copyright"> |
---|
| 232 | The Xiph Fish Logo is a |
---|
| 233 | trademark (™) of Xiph.Org.<br/> |
---|
| 234 | |
---|
| 235 | These pages © 1994 - 2005 Xiph.Org. All rights reserved. |
---|
| 236 | </div> |
---|
| 237 | |
---|
| 238 | </body> |
---|
| 239 | </html> |
---|