| 1 |
\ |
| 2 |
.\" This man page was generated by the Netpbm tool 'makeman' from HTML source. |
| 3 |
.\" Do not hand-hack it! If you have bug fixes or improvements, please find |
| 4 |
.\" the corresponding HTML page on the Netpbm website, generate a patch |
| 5 |
.\" against that, and send it to the Netpbm maintainer. |
| 6 |
.TH "Pnmtofiasco User Manual" 1 "09 April 2025" "netpbm documentation" |
| 7 |
|
| 8 |
.SH NAME |
| 9 |
|
| 10 |
pnmtofiasco - Convert PNM file to FIASCO compressed file |
| 11 |
|
| 12 |
.UN synopsis |
| 13 |
.SH SYNOPSIS |
| 14 |
|
| 15 |
\fBpnmtofiasco\fP |
| 16 |
[\fIoption\fP]... |
| 17 |
[\fIfilename\fP]... |
| 18 |
|
| 19 |
.UN description |
| 20 |
.SH DESCRIPTION |
| 21 |
.PP |
| 22 |
This program is part of |
| 23 |
.BR "Netpbm" (1)\c |
| 24 |
\&. |
| 25 |
|
| 26 |
\fBpnmtofiasco\fP compresses the named pbm, pgm, or ppm image files, |
| 27 |
or Standard Input if no file is named, and produces a FIASCO file on |
| 28 |
Standard Output. |
| 29 |
|
| 30 |
.UN options |
| 31 |
.SH OPTIONS |
| 32 |
.PP |
| 33 |
In addition to the options common to all programs based on libnetpbm |
| 34 |
(most notably \fB-quiet\fP, see |
| 35 |
.UR index.html#commonoptions |
| 36 |
Common Options |
| 37 |
.UE |
| 38 |
\&), \fBpnmtofiasco\fP recognizes the following |
| 39 |
command line options: |
| 40 |
.PP |
| 41 |
All option names may be abbreviated; for example, --optimize may be |
| 42 |
written --optim or --opt. For most options a one letter short option |
| 43 |
is provided. Mandatory or optional arguments to long options are |
| 44 |
mandatory or optional for short options, too. Both short and long |
| 45 |
options are case sensitive. |
| 46 |
|
| 47 |
.UN basicopts |
| 48 |
.SS Basic Options |
| 49 |
|
| 50 |
|
| 51 |
.TP |
| 52 |
\fB--image-name=\fP\fIname\fP, \fB-i\fP \fIname\fP |
| 53 |
Compress the named images, not Standard Input. If \fIname\fP is |
| 54 |
\fB-\fP, read Standard Input. \fIname\fP has to be either an image |
| 55 |
filename or a template of the form: |
| 56 |
|
| 57 |
.nf |
| 58 |
prefix[start-end{+,-}step]suffix |
| 59 |
|
| 60 |
.fi |
| 61 |
.sp |
| 62 |
Templates are useful when compressing video streams: e.g., if you |
| 63 |
specify the template \fBimg0[12-01-2].pgm\fP, then \fBpnmtofiasco\fP |
| 64 |
compresses the images img012.pgm, img010.pgm, ..., img002.pgm. |
| 65 |
.sp |
| 66 |
If \fIname\fP is a relative path, \fBpnmtofiasco \fP searches for |
| 67 |
the image files in the current directory and in the (colon-separated) |
| 68 |
list of directories given by the environment variable |
| 69 |
\fBFIASCO_IMAGES\fP. |
| 70 |
|
| 71 |
.TP |
| 72 |
\fB--output-name=\fP\fIname\fP, \fB-o\fP \fIoutput-file\fP |
| 73 |
Write FIASCO output to the named file, not to Standard Output. |
| 74 |
.sp |
| 75 |
If \fIname\fP is a relative path and the environment variable |
| 76 |
\fBFIASCO_DATA\fP is a (colon-separated) list of directories, then |
| 77 |
\fBpnmtofiasco\fP writes the output file to the first (writable) |
| 78 |
directory of this list. Otherwise, \fBpnmtofiasco\fP write it to the |
| 79 |
current directory. |
| 80 |
|
| 81 |
.TP |
| 82 |
\fB--quality=\fP\fIN\fP, \fB-q\fP \fIN\fP |
| 83 |
Set quality of compression to \fIN\fP. Quality is 1 (worst) to |
| 84 |
100 (best); default is 20. |
| 85 |
|
| 86 |
.TP |
| 87 |
\fB--version\fP, \fB-v\fP |
| 88 |
Print \fBpnmtofiasco\fP version number, then exit. |
| 89 |
|
| 90 |
.TP |
| 91 |
\fB--verbose \fP\fIN\fP, \fB-V\fP \fIN\fP |
| 92 |
Set level of verbosity to \fIN\fP. Level is 0 (no output at |
| 93 |
all), 1 (show progress meter), or 2 (show detailed compression |
| 94 |
statistics); default is 1. |
| 95 |
|
| 96 |
.TP |
| 97 |
\fB--progress-meter \fP\fIN\fP, \fB-B\fP \fIN\fP |
| 98 |
Set type of progress-meter to \fIN\fP. The following types are |
| 99 |
available; default is 1: |
| 100 |
|
| 101 |
|
| 102 |
.TP |
| 103 |
\fB0\fP |
| 104 |
no progress meter |
| 105 |
|
| 106 |
.TP |
| 107 |
\fB1\fP |
| 108 |
RPM style progress bar using 50 hash marks |
| 109 |
|
| 110 |
.TP |
| 111 |
\fB2\fP |
| 112 |
percentage meter |
| 113 |
|
| 114 |
|
| 115 |
|
| 116 |
.TP |
| 117 |
\fB--config=\fP\fIname\fP, \fB-f\fP \fIname\fP |
| 118 |
Load parameter file \fIname \fP to initialize the options of |
| 119 |
\fBpnmtofiasco\fP. See file \fBsystem.fiascorc\fP for an example of |
| 120 |
the syntax. Options of \fBpnmtofiasco\fP are set by any of the |
| 121 |
following methods (in the specified order): |
| 122 |
|
| 123 |
|
| 124 |
.IP \(bu |
| 125 |
Global resource file \fB/etc/system.fiascorc\fP |
| 126 |
.IP \(bu |
| 127 |
$HOME/.fiascorc |
| 128 |
.IP \(bu |
| 129 |
command line |
| 130 |
.IP \(bu |
| 131 |
--config=\fIname\fP |
| 132 |
|
| 133 |
|
| 134 |
.TP |
| 135 |
\fB--info\fP, \fB-h\fP |
| 136 |
Print brief help, then exit. |
| 137 |
|
| 138 |
.TP |
| 139 |
\fB--help\fP, \fB-H\fP |
| 140 |
Print detailed help, then exit. |
| 141 |
|
| 142 |
|
| 143 |
|
| 144 |
.UN advancedopts |
| 145 |
.SS Options for Advanced Users |
| 146 |
|
| 147 |
|
| 148 |
.TP |
| 149 |
\fB--basis-name=\fP\fIname\fP, \fB-b\fP \fIname\fP |
| 150 |
Preload compression basis \fIname\fP into FIASCO. The basis |
| 151 |
\fIname\fP provides the initial compression dictionary. Either use |
| 152 |
one of the files "small.fco", "medium.fco", or |
| 153 |
"large.fco" that come with \fBpnmtofiasco \fP or create a |
| 154 |
new ASCII basis file. |
| 155 |
|
| 156 |
.TP |
| 157 |
\fB--optimize=\fP\fIN\fP, \fB-z\fP \fIN\fP |
| 158 |
Set optimization level to \fIN\fP. Level is 0 (fastest) to 3 |
| 159 |
(slowest). |
| 160 |
|
| 161 |
Be warned: The encoding time dramatically increases when \fIN\fP=\fB2\fP |
| 162 |
or \fIN\fP=\fB3\fP while the compression performance improves only slightly. |
| 163 |
.sp |
| 164 |
The default is 0. |
| 165 |
|
| 166 |
.TP |
| 167 |
\fB--prediction\fP, \fB-P\fP |
| 168 |
Use additional predictive coding. If this optimization is enabled |
| 169 |
then the image is compressed in two steps. In the first step, a coarse |
| 170 |
approximation of the image is computed using large unichrome |
| 171 |
blocks. Finally, the delta image is computed and the prediction error |
| 172 |
is approximated using the standard FIASCO algorithm. |
| 173 |
|
| 174 |
.TP |
| 175 |
\fB--dictionary-size=\fP\fIN\fP, \fB-D\fP \fIN\fP |
| 176 |
Set size of dictionary that is used when coding the luminance band |
| 177 |
to \fIN\fP; default is 10000, i.e., the dictionary is not restricted. |
| 178 |
|
| 179 |
.TP |
| 180 |
\fB--chroma-dictionary=\fP\fIN\fP, \fB-C\fP \fIN\fP |
| 181 |
Set size of dictionary that is used when coding chroma bands to |
| 182 |
\fIN\fP; default is 40. |
| 183 |
|
| 184 |
.TP |
| 185 |
\fB--chroma-qfactor=\fP\fIN\fP, \fB-Q\fP \fIN\fP |
| 186 |
Reduce the quality of chroma band compression \fIN\fP-times with |
| 187 |
respect to the user defined quality \fIq\fP of the luminance band |
| 188 |
compression (\fB--quality\fP=\fIq\fP); default is 2. |
| 189 |
|
| 190 |
.TP |
| 191 |
\fB--tiling-exponent=\fP\fIN\fP, \fB-t\fP \fIN\fP |
| 192 |
Subdivide the image into 2^\fIN\fP tiles prior coding; default is |
| 193 |
4, i.e. the image is subdivided into 16 tiles. The processing order of |
| 194 |
the individual tiles is defined by the option |
| 195 |
|
| 196 |
.TP |
| 197 |
\fB--tiling-method=\fP\fIname\fP, \fB-T\fP \fIname\fP |
| 198 |
Order the individual image tiles (the image is subdivided into; |
| 199 |
see option \fB--tiling-exponent=\fP\fIN\fP) by method \fIname\fP; |
| 200 |
default is \fBdesc-variance\fP. |
| 201 |
|
| 202 |
|
| 203 |
.TP |
| 204 |
\fBdesc-variance\fP |
| 205 |
Tiles with small variances are processed first. |
| 206 |
|
| 207 |
.TP |
| 208 |
\fBasc-variance\fP |
| 209 |
Tiles with large variances are processed first. |
| 210 |
|
| 211 |
.TP |
| 212 |
\fBdesc-spiral\fP |
| 213 |
Tiles are process in spiral order starting in the middle. |
| 214 |
|
| 215 |
.TP |
| 216 |
\fBasc-spiral\fP |
| 217 |
Tiles are process in spiral order starting at the border. |
| 218 |
|
| 219 |
|
| 220 |
|
| 221 |
.TP |
| 222 |
\fB--rpf-mantissa=\fP\fIN\fP |
| 223 |
Use \fIN\fP mantissa bits for quantized coefficients. |
| 224 |
|
| 225 |
.TP |
| 226 |
\fB--dc-rpf-mantissa=\fP\fIN\fP |
| 227 |
Use \fIN\fP mantissa bits for quantized DC coefficients. |
| 228 |
|
| 229 |
.TP |
| 230 |
\fB--rpf-range=\fP\fIN\fP |
| 231 |
Coefficients outside the quantization interval |
| 232 |
[-\fIN\fP,+\fIN\fP] are set to zero. |
| 233 |
|
| 234 |
.TP |
| 235 |
\fB--dc-rpf-range=\fP\fIN\fP |
| 236 |
DC coefficients outside the quantization interval |
| 237 |
[-\fIN\fP,+\fIN\fP] are set to zero. |
| 238 |
|
| 239 |
|
| 240 |
|
| 241 |
|
| 242 |
.UN videoopts |
| 243 |
.SS Additional Options for Video Compression |
| 244 |
|
| 245 |
|
| 246 |
.TP |
| 247 |
\fB--smooth=\fP\fIN\fP, \fB-s\fP \fIN\fP |
| 248 |
Smooth decompressed reference frames along the partitioning |
| 249 |
borders by the given amount \fIN\fP. \fIN\fP is 0 (no smoothing) to |
| 250 |
100; default is 70. This factor is stored in the FIASCO file. |
| 251 |
|
| 252 |
.TP |
| 253 |
\fB--min-level=\fP\fIN\fP, \fB-m\fP \fIN\fP |
| 254 |
Start prediction (motion compensated prediction or additional |
| 255 |
prediction) on block level \fIN\fP; default is level 6. I.e., motion |
| 256 |
compensation is applied to all image blocks of at least 8x8 pixels |
| 257 |
(binary tree level \fIN\fP=6), 16x8 (\fIN\fP=7), 16x16 (\fIN\fP=8), |
| 258 |
etc. |
| 259 |
|
| 260 |
.TP |
| 261 |
\fB--max-level=\fP\fIN\fP, \fB-M\fP \fIN\fP |
| 262 |
Stop prediction (motion compensated prediction or additional |
| 263 |
prediction) on block level \fIN\fP; default is level 10. I.e., |
| 264 |
motion compensation is applied to all image blocks of at most 16x16 |
| 265 |
pixels (\fIN\fP=8), 32x16 (\fIN\fP=9), 32x32 (\fIN\fP=10), etc. |
| 266 |
|
| 267 |
.TP |
| 268 |
\fB--half-pixel\fP, \fB-2\fP |
| 269 |
Use half pixel precise motion compensation. |
| 270 |
|
| 271 |
.TP |
| 272 |
\fB--fps=\fP\fIN\fP, \fB-F\fP \fIN\fP |
| 273 |
Set number of frames per second to \fIN\fP. This value is stored |
| 274 |
in the FIASCO output file and is used in the decoder |
| 275 |
.BR "fiascotopnm" (1)\c |
| 276 |
\& to control the framerate. |
| 277 |
|
| 278 |
.TP |
| 279 |
\fB--pattern=\fP\fItype\fP, \fB-p\fP \fItype\fP |
| 280 |
Defines the type of inter frame compression which should be |
| 281 |
applied to individual frames of a video stream. \fItype\fP is a |
| 282 |
sequence of characters; default is "IPPPPPPPPP". Element |
| 283 |
\fBN\fP defines the type of predicting which should be used for frame |
| 284 |
\fBN\fP; the frame type pattern is periodically extended. Valid |
| 285 |
characters are: |
| 286 |
|
| 287 |
|
| 288 |
.TP |
| 289 |
\fBI\fP |
| 290 |
intra frame, i.e., no motion compensated prediction is used at |
| 291 |
all. |
| 292 |
|
| 293 |
.TP |
| 294 |
\fBP\fP |
| 295 |
predicted frame, i.e., a previously encoded frame is used for |
| 296 |
prediction (forward prediction). |
| 297 |
|
| 298 |
.TP |
| 299 |
\fBB\fP |
| 300 |
bidirectional predicted frame, i.e., not only a previously shown |
| 301 |
frame but also a frame of the future is used for prediction (forward, |
| 302 |
backward or interpolated prediction). |
| 303 |
|
| 304 |
|
| 305 |
|
| 306 |
.TP |
| 307 |
\fB--cross-B-search\fP |
| 308 |
Instead of using exhaustive search the "Cross-B-Search" |
| 309 |
algorithm is used to find the best interpolated prediction of |
| 310 |
B-frames. |
| 311 |
|
| 312 |
.TP |
| 313 |
\fB--B-as-past-ref\fP |
| 314 |
Also use previously encoded B-frames when prediction the current |
| 315 |
frame. If this option is not set, only I- and P-frames are used to |
| 316 |
predict the current frame. |
| 317 |
|
| 318 |
|
| 319 |
|
| 320 |
|
| 321 |
.UN examples |
| 322 |
.SH EXAMPLES |
| 323 |
.PP |
| 324 |
Compress the still image "foo.ppm" to the FIASCO file |
| 325 |
"foo.wfa" using the default options: |
| 326 |
|
| 327 |
.nf |
| 328 |
pnmtofiasco < foo.ppm >foo.wfa |
| 329 |
|
| 330 |
.fi |
| 331 |
.PP |
| 332 |
Compress the video frames "foo0*.ppm" to the FIASCO file |
| 333 |
"video.wfa" using half pixel precise motion compensation at |
| 334 |
a frame rate of 15 frames per second. Intra frame 1 is used to |
| 335 |
predict P-frame 4, frames 1 and 4 are used to predict B-frames 2 and |
| 336 |
3, and so on. Frame 10 is again an intra-frame. |
| 337 |
|
| 338 |
.nf |
| 339 |
pnmtofiasco -2 -p "IBBPBBPBB" -fps 15 -o video.wfa foo0*.ppm |
| 340 |
|
| 341 |
.fi |
| 342 |
|
| 343 |
.UN files |
| 344 |
.SH FILES |
| 345 |
|
| 346 |
|
| 347 |
.TP |
| 348 |
\fB/etc/system.fiascorc\fP |
| 349 |
The systemwide initialization file. |
| 350 |
|
| 351 |
.TP |
| 352 |
$HOME\fB/.fiascorc\fP |
| 353 |
The personal initialization file. |
| 354 |
|
| 355 |
|
| 356 |
|
| 357 |
.UN environment |
| 358 |
.SH ENVIRONMENT |
| 359 |
|
| 360 |
|
| 361 |
|
| 362 |
|
| 363 |
.TP |
| 364 |
\fBFIASCO_IMAGES\fP |
| 365 |
Search path for image files. Default is "./". |
| 366 |
|
| 367 |
.TP |
| 368 |
\fBFIASCO_DATA\fP |
| 369 |
Search and save path for FIASCO files. Default is "./". |
| 370 |
|
| 371 |
|
| 372 |
|
| 373 |
.UN seealso |
| 374 |
.SH SEE ALSO |
| 375 |
.BR "fiascotopnm" (1)\c |
| 376 |
\&, |
| 377 |
.BR "pnmtojpeg" (1)\c |
| 378 |
\&, |
| 379 |
.BR "pnmtojbig" (1)\c |
| 380 |
\&, |
| 381 |
.BR "pamtogif" (1)\c |
| 382 |
\&, |
| 383 |
.BR "pnm" (1)\c |
| 384 |
\& |
| 385 |
.PP |
| 386 |
Ullrich Hafner, Juergen Albert, Stefan Frank, and Michael Unger. |
| 387 |
\fBWeighted Finite Automata for Video Compression\fP, IEEE Journal on |
| 388 |
Selected Areas In Communications, January 1998 |
| 389 |
.PP |
| 390 |
Ullrich Hafner. \fBLow Bit-Rate Image and Video Coding with |
| 391 |
Weighted Finite Automata\fP, Ph.D. thesis, Mensch & Buch Verlag, |
| 392 |
ISBN 3-89820-002-7, October 1999. |
| 393 |
.PP |
| 394 |
FIASCO: An Open-Source Fractal Image and Sequence Codec, Linux Journal, |
| 395 |
January 2001. |
| 396 |
|
| 397 |
.UN author |
| 398 |
.SH AUTHOR |
| 399 |
|
| 400 |
Ullrich Hafner <\[email protected]\fP> |
| 401 |
.SH DOCUMENT SOURCE |
| 402 |
This manual page was generated by the Netpbm tool 'makeman' from HTML |
| 403 |
source. The master documentation is at |
| 404 |
.IP |
| 405 |
.B http://netpbm.sourceforge.net/doc/pnmtofiasco.html |
| 406 |
.PP |