1Pamtotiff User Manual(0) Pamtotiff User Manual(0)
2
3
4
6 pamtotiff - convert a Netpbm image to a TIFF file
7
8
10 pamtotiff
11
12 [-none | -packbits | -lzw | -g3 | -g4 | -flate | -adobeflate]
13
14 [-2d]
15
16 [-fill]
17
18 [-predictor=n]
19
20 [-msb2lsb|-lsb2msb]
21
22 [-rowsperstrip=n]
23
24 [-minisblack|-miniswhite|mb|mw]
25
26 [-truecolor]
27
28 [-color]
29
30 [-indexbits=bitwidthlist] [-xresolution=xres]
31
32 [-yresolution=yres] [-resolutionunit={inch | centimeter | none | in |
33 cm | no}]
34
35 [-indexbits=[1[2[4[8]]]]]
36
37 [-append]
38
39 [-tag=taglist]
40
41 [pamfile]
42
43 You can use the minimum unique abbreviation of the options. You can
44 use two hyphens instead of one. You can separate an option name from
45 its value with white space instead of an equals sign.
46
47
49 This program is part of Netpbm(1).
50
51 pamtotiff reads a PNM or PAM image as input and produces a TIFF file as
52 output.
53
54 Actually, it handles multi-image Netpbm streams, producing multi-image
55 TIFF streams (i.e. a TIFF stream with multiple 'directories'). But
56 before Netpbm 10.27 (March 2005), it ignored all but the first Netpbm
57 image in the input stream.
58
59
60 The Output File
61 The output goes to Standard Output. pamtotiff approaches this output
62 file differently from Unix and Netpbm convention. This is entirely due
63 to pamtotiff's use of the TIFF library to do all TIFF output.
64
65
66
67 · The output file must be seekable. pamtotiff does not write it
68 sequentially. Therefore, you can't use a pipe; you can't pipe
69 the output of pamtotiff to some other program. But any regular
70 file should work.
71
72
73 · If the output file descriptor is readable, you must either spec‐
74 ify -append so as to add to the existing file, or make sure the
75 file is empty. Otherwise, pamtotiff will fail with an unhelpful
76 message telling you that a TIFF library function failed to open
77 the TIFF output stream.
78
79
80 · If you are converting multiple images (your input stream con‐
81 tains multiple images), the output file must be both readable
82 and writable.
83
84
85
86 If you're using a Unix command shell to run pamtotiff, you use facili‐
87 ties of your shell to set up Standard Output. In Bash, for example,
88 you would set up a write-only Standard Output to the file /tmp/myim‐
89 age.tiff like this:
90
91 $ pamtotiff myimage.pnm >/tmp/myimage.tiff
92
93 In Bash, you would set up a read/write Standard Output to the file
94 /tmp/myimage.tiff like this:
95
96 $ pamtotiff myimage.pnm 1<>/tmp/myimage.tiff
97
98
100 Compression
101 By default, pamtotiff creates a TIFF file with no compression. This is
102 your best bet most of the time. If you want to try another compression
103 scheme or tweak some of the other even more obscure output options,
104 there are a number of options which which to play.
105
106 Before Netpbm 8.4 (April 2000), the default was to use LZW compression.
107 But then new releases of the TIFF library started omitting the LZW com‐
108 pression capability due to concern about patents on LZW. So since
109 then, the default has been no compression. The LZW patents have now
110 expired and new TIFF libraries do LZW, but the pamtotiff behavior
111 remains the same for compatibility with older TIFF libraries and appli‐
112 cations of pamtotiff.
113
114 The -none, -packbits, -lzw, -g3, -g4, -flate, and -adobeflate options
115 are used to override the default and set the compression scheme used in
116 creating the output file. The CCITT Group 3 and Group 4 compression
117 algorithms can be used only with bilevel data. The -2d and -fill
118 options are meaningful only with Group 3 compression: -2d requests
119 2-dimensional encoding, while -fill requests that each encoded scanline
120 be zero-filled to a byte boundary. The -predictor option is meaningful
121 only with LZW compression: a predictor value of 2 causes each scanline
122 of the output image to undergo horizontal differencing before it is
123 encoded; a value of 1 forces each scanline to be encoded without dif‐
124 ferencing. By default, pamtotiff creates a TIFF file with msb-to-lsb
125 fill order. The -msb2lsb and -lsb2msb options are used to override the
126 default and set the fill order used in creating the file.
127
128 With some older TIFF libraries, -lzw doesn't work because the TIFF
129 library doesn't do LZW compression. This is because of concerns about
130 Unisys's patent on LZW which was then in force. Actually, with very
131 old TIFF libraries, -lzw works because no distributors of the TIFF
132 library were sensitive yet to the patent issue.
133
134 -flate chooses 'flate' compression, which is the patent-free compres‐
135 sion common in the Unix world implemented by the 'Z' library. It is
136 what the PNG format uses.
137
138
139
140 Fill Order
141 The -msb2lsb and lsb2msb options control the fill order.
142
143 The fill order is the order in which pixels are packed into a byte in
144 the Tiff raster, in the case that there are multiple pixels per byte.
145 msb-to-lsb means that the leftmost columns go into the most significant
146 bits of the byte in the Tiff image. However, there is considerable
147 confusion about the meaning of fill order. Some believe it means
148 whether 16 bit sample values in the Tiff image are little-endian or
149 big-endian. This is totally erroneous (The endianness of integers in a
150 Tiff image is designated by the image's magic number). However,
151 ImageMagick and older Netpbm both have been known to implement that
152 interpretation. 2001.09.06.
153
154 If the image does not have sub-byte pixels, these options have no
155 effect other than to set the value of the FILLORDER tag in the Tiff
156 image (which may be useful for those programs that misinterpret the tag
157 with reference to 16 bit samples).
158
159
160 Color Space
161 -color tells pamtotiff to produce a color, as opposed to grayscale,
162 TIFF image if the input is PPM, even if it contains only shades of
163 gray. Without this option, pamtotiff produces a grayscale TIFF image
164 if the input is PPM and contains only shades of gray, and at most 256
165 shades. Otherwise, it produces a color TIFF output. For PBM and PGM
166 input, pamtotiff always produces grayscale TIFF output and this option
167 has no effect.
168
169 The -color option can prevent pamtotiff from making two passes through
170 the input file, thus improving speed and memory usage. See Multiple
171 Passes ⟨#multipass⟩ .
172
173 -truecolor tells pamtotiff to produce the 24-bit RGB form of TIFF out‐
174 put if it is producing a color TIFF image. Without this option, pamto‐
175 tiff produces a colormapped (paletted) TIFF image unless there are more
176 than 256 colors (and in the latter case, issues a warning).
177
178 The -truecolor option can prevent pamtotiff from making two passes
179 through the input file, thus improving speed and memory usage. See
180 Multiple Passes ⟨#multipass⟩ .
181
182 The -color and -truecolor options did not exist before Netpbm 9.21
183 (December 2001).
184
185 If pamtotiff produces a grayscale TIFF image, this option has no
186 effect.
187
188 The -minisblack and -miniswhite options force the output image to have
189 a 'minimum is black' or 'minimum is white' photometric, respectively.
190 If you don't specify either, pamtotiff uses minimum is black except
191 when using Group 3 or Group 4 compression, in which case pamtotiff fol‐
192 lows CCITT fax standards and uses 'minimum is white.' This usually
193 results in better compression and is generally preferred for bilevel
194 coding.
195
196 Before February 2001, pamtotiff always produced 'minimum is black,' due
197 to a bug. In either case, pamtotiff sets the photometric interpreta‐
198 tion tag in the TIFF output according to which photometric is actually
199 used.
200
201 The -indexbits option is meaningful only for a colormapped (paletted)
202 image. In this kind of image, the raster contains values which are
203 indexes into a table of colors, with the indexes normally taking less
204 space that the color description in the table. pamtotiff can generate
205 indexes of 1, 2, 4, or 8 bits. By default, it will use 8, because many
206 programs that interpret TIFF images can't handle any other width.
207
208 But if you have a small number of colors, you can make your image con‐
209 siderably smaller by allowing fewer than 8 bits per index, using the
210 -indexbits option. The value is a comma-separated list of the bit
211 widths you allow. pamtotiff chooses the smallest width you allow that
212 allows it to index the entire color table. If you don't allow any such
213 width, pamtotiff fails. Normally, the only useful value for this
214 option is 1,2,4,8, because a program either understands the 8 bit width
215 (default) or understands them all.
216
217 In a Baseline TIFF image, according to the 1992 TIFF 6.0 specification,
218 4 and 8 are the only valid widths. There are no formal standards that
219 allow any other values.
220
221 This option was added in June 2002. Before that, only 8 bit indices
222 were possible.
223
224
225 Extra Tags
226 There are lots of tag types in the TIFF format that don't correspond to
227 any information in the PNM format or to anything in the conversion
228 process. For example, a TIFF_ARTIST tag names the artist who created
229 the image.
230
231 You can tell pamtotiff explicitly to include tags such as this in its
232 output with the -tag option. You identify a list of tag types and val‐
233 ues and pamtotiff includes a tag in the output for each item in your
234 list.
235
236 The value of -tag is the list of tags, like this example:
237
238 <code>
239 -tag=subfiletype=reducedimage,documentname=Fred,xposition=25
240 </code>
241
242 As you see, it is a list of tag specifications separated by commas.
243 Each tag specification is a name and a value separated by an equal
244 sign. The name is the name of the tag type, except in arbitrary
245 upper/lower case. One place to see the names of TIFF tag types is in
246 the TIFF library's tiff.h file, where there is a macro defined for each
247 consisting of 'TIFF_' plus the name. E.g. for the SUBFILETYPE tag
248 type, there is a macro TIFF_SUBFILETYPE.
249
250 The format of the value specification for a tag (stuff after the equal
251 sign) depends upon what kind of value the tag type has:
252
253
254
255 · Integer: a decimal number
256
257
258 · Floating point number: a decimal number
259
260
261 · String: a string
262
263
264 · Enumerated (For example, a 'subfiletype' tag takes an enumerated
265 value. Its possible values are REDUCEDIMAGE, PAGE, and MASK.):
266 The name of the value. You can see the possible value names in
267 the TIFF library's tiff.h foile, where there is a macro defined
268 for each consisting of a qualifier plus the value name. E.g.
269 for the REDUCEDIMAGE value of a SUBFILETYPE tag, you see the
270 macro FILETYPE_REDUCEDIMAGE.
271
272 The TIFF format assigns a unique number to each enumerated value
273 and you can specify that number, in decimal, as an alternative.
274 This is useful if you are using an extension of TIFF that pamto‐
275 tiff doesn't know about.
276
277
278
279 If you specify a tag type with -tag that is not independent of the con‐
280 tent of your PNM source image and pamtotiff's conversion process (i.e.
281 a tag type in which pamtotiff is interested), pamtotiff fails. For
282 example, you cannot specify an IMAGEWIDTH tag with -tag, because pamto‐
283 tiff generates an IMAGEWIDTH tag that gives the actual width of the
284 image.
285
286 -tag was new in Netpbm 10.31 (December 2005).
287
288
289 Other
290 You can use the -rowsperstrip option to set the number of rows (scan‐
291 lines) in each strip of data in the output file. By default, the out‐
292 put file has the number of rows per strip set to a value that will
293 ensure each strip is no more than 8 kilobytes long.
294
295 The -append option tells pamtotiff to add images to the existing output
296 file (a TIFF file may contain multiple images) instead of the default,
297 which is to replace the output file.
298
299 -append was new in Netpbm 10.27 (March 2005).
300
301
302
304 There are myriad variations of the TIFF format, and this program gener‐
305 ates only a few of them. pamtotiff creates a grayscale TIFF file if
306 its input is a PBM (monochrome) or PGM (grayscale) or equivalent PAM
307 file. pamtotiff also creates a grayscale file if it input is PPM
308 (color) or equivalent PAM, but there is only one color in the image.
309
310 If the input is a PPM (color) file and there are 256 colors or fewer,
311 but more than 1, pamtotiff generates a color palette TIFF file. If
312 there are more colors than that, pamtotiff generates an RGB (not RGBA)
313 single plane TIFF file. Use pnmtotiffcmyk to generate the cyan-
314 magenta-yellow-black ink color separation TIFF format.
315
316 The number of bits per sample in the TIFF output is determined by the
317 maxval of the Netpbm input. If the maxval is less than 256, the bits
318 per sample in the output is the smallest number that can encode the
319 maxval. If the maxval is greater than or equal to 256, there are 16
320 bits per sample in the output.
321
322
323 Extra Channels
324 Like most Netpbm programs, pamtotiff's function is mostly undefined if
325 the input is PAM image with tuple type other than BLACKANDWHITE,
326 GRAYSCALE, or RGB. Most of the statements in this manual assume the
327 input is not such an exotic PAM. But there is a little defined pro‐
328 cessing of other PAM subformats.
329
330 pamtotiff assumes any 1 plane PAM image is BLACKANDWHITE or GRAYSCALE
331 (and doesn't distinguish between those two).
332
333 pamtotiff assumes a PAM with more than 1 plane is of tuple type RGB
334 except with that number of planes instead of 3. pamtotiff doesn't
335 really understand red, green, and blue, so it has no trouble with a
336 2-component or 5-component color space. The TIFF format allows an
337 arbitrary number of color compoonents, so pamtotiff simply maps the PAM
338 planes directly to TIFF color components. I don't know if the meanings
339 of 5 components in a TIFF image are standard at all, but the function
340 is there if you want to use it.
341
342 Note that pamtotiff may generate either a truecolor or colormapped
343 image with an arbitrary number of color components. In the truecolor
344 case, the raster has that number of planes. In the colormapped case,
345 the raster has of course 1 plane, but the color map has all the color
346 components in it.
347
348 The most common reason for a PAM to have extra planes is when the tuple
349 type is xxx_ALPHA, which means the highest numbered plane is a trans‐
350 parency plane (alpha channel). At least one user found that a TIFF
351 with an extra plane for transparency was useful.
352
353 Note that the grayscale detection works on N-component colors, so if
354 your planes aren't really color components, you'll want to disable this
355 via the -color option.
356
357
358
359 Multiple Passes
360 pamtotiff reads the input image once if it can, and otherwise twice.
361 It needs that second pass (which happens before the main pass, of
362 course) to analyze the colors in the image and generate a color map
363 (palette) and determine if the image is grayscale. So the second pass
364 happens only when the input is PPM. And you can avoid it then by spec‐
365 ifying both the -truecolor and -color options.
366
367 If the input image is small enough to fit in your system's file cache,
368 the second pass is very fast. If not, it requires reading from disk
369 twice, which can be slow.
370
371 When the input is from a file that cannot be rewound and reread, pamto‐
372 tiff reads the entire input image into a temporary file which can, and
373 works from that. Even if it needs only one pass.
374
375 Before Netpbm 9.21 (December 2001), pamtotiff always read the entire
376 image into virtual memory and then did one, two, or three passes
377 through the memory copy. The -truecolor and -color options did not
378 exist. The passes through memory would involve page faults if the
379 entire image did not fit into real memory. The image in memory
380 required considerably more memory (12 bytes per pixel) than the cached
381 file version of the image would.
382
383
384
385 Resolution
386 A Tiff image may contain information about the resolution of the image,
387 which means how big in real dimensions (centimeters, etc.) each pixel
388 in the raster is. That information is in the TIFF XRESOLUTION, YRESO‐
389 LUTION, and RESOLUTIONUNIT tags. By default, pamtotiff does not
390 include any tags of these types, but you can specify them with the
391 -tags option.
392
393 There are also options -xresolution, -yresolution, and -resolutionunit,
394 but those are obsolete. Before -tags existed (before Netpbm 10.31
395 (December 2005), they were the only way.
396
397 Note that the number of pixels in the image and how much information
398 each contains is determined independently from the setting of the reso‐
399 lution tags. The number of pixels in the output is the same as in the
400 input, and each pixel contains the same information. For your resolu‐
401 tion tags to be meaningful, they have to consistent with whatever cre‐
402 ated the PNM input. E.g. if a scanner turned a 10 centimeter wide
403 image into a 1000 pixel wide PNM image, then your horizontal resolution
404 is 100 pixels per centimeter, and if your XRESOLUTION tag says anything
405 else, something that prints your TIFF image won't print the proper 10
406 centimeter image.
407
408 The value of the XRESOLUTION tag is a floating point decimal number
409 that tells how many pixels there are per unit of distance in the hori‐
410 zontal direction. -yresolution is analogous for the vertical direc‐
411 tion.
412
413 The unit of distance is given by the value of the RESOLUTIONUNIT
414 option. That value is either INCH, CENTIMETER, or NONE. NONE means
415 the unit is arbitrary or unspecified. This could mean that the creator
416 and user of the image have a separate agreement as to what the unit is.
417 But usually, it just means that the horizontal and vertical resolution
418 values cannot be used for anything except to determine aspect ratio
419 (because even though the unit is arbitrary or unspecified, it has to be
420 the same for both resolution numbers).
421
422 If you don't use a -tag option to specify the resolution tag and use
423 the obsolete options instead, note the following:
424
425
426
427 · If you don't include an specify -xresolution, the Tiff image
428 does not contain horizontal resolution information. Likewise
429 for -yresolution. If you don't specify -resolutionunit, the
430 default is inches.
431
432
433 · Before Netpbm 10.16 (June 2003), -resolutionunit did not exist
434 and the resolution unit was always inches.
435
436
437
438
440 pamtotiff was originally pnmtotiff and did not handle PAM input. It
441 was extended and renamed in Netpbm 10.30 (October 2005).
442
443
444
446 tifftopnm(1), pnmtotiffcmyk(1), pamdepth(1), pamtopnm(1), pam(1)
447
448
450 Derived by Jef Poskanzer from ras2tiff.c, which is Copyright (c) 1990
451 by Sun Microsystems, Inc. Author: Patrick J. Naughton
452 (naughton@wind.sun.com).
453
454
455
456netpbm documentation 27 March 2005 Pamtotiff User Manual(0)