PageRenderTime 43ms CodeModel.GetById 34ms app.highlight 3ms RepoModel.GetById 1ms app.codeStats 0ms

Plain Text | 637 lines | 518 code | 119 blank | 0 comment | 0 complexity | e1caaadfb05d30b9e6b720399ea51e61 MD5 | raw file
  1USAGE instructions for the Independent JPEG Group's JPEG software
  4This file describes usage of the JPEG conversion programs cjpeg and djpeg,
  5as well as the utility programs jpegtran, rdjpgcom and wrjpgcom.  (See
  6the other documentation files if you wish to use the JPEG library within
  7your own programs.)
  9If you are on a Unix machine you may prefer to read the Unix-style manual
 10pages in files cjpeg.1, djpeg.1, jpegtran.1, rdjpgcom.1, wrjpgcom.1.
 15These programs implement JPEG image encoding, decoding, and transcoding.
 16JPEG (pronounced "jay-peg") is a standardized compression method for
 17full-color and gray-scale images.
 22We provide two programs, cjpeg to compress an image file into JPEG format,
 23and djpeg to decompress a JPEG file back into a conventional image format.
 25On Unix-like systems, you say:
 26	cjpeg [switches] [imagefile] >jpegfile
 28	djpeg [switches] [jpegfile]  >imagefile
 29The programs read the specified input file, or standard input if none is
 30named.  They always write to standard output (with trace/error messages to
 31standard error).  These conventions are handy for piping images between
 34On most non-Unix systems, you say:
 35	cjpeg [switches] imagefile jpegfile
 37	djpeg [switches] jpegfile  imagefile
 38i.e., both the input and output files are named on the command line.  This
 39style is a little more foolproof, and it loses no functionality if you don't
 40have pipes.  (You can get this style on Unix too, if you prefer, by defining
 41TWO_FILE_COMMANDLINE when you compile the programs; see install.txt.)
 43You can also say:
 44	cjpeg [switches] -outfile jpegfile  imagefile
 46	djpeg [switches] -outfile imagefile  jpegfile
 47This syntax works on all systems, so it is useful for scripts.
 49The currently supported image file formats are: PPM (PBMPLUS color format),
 50PGM (PBMPLUS gray-scale format), BMP, Targa, and RLE (Utah Raster Toolkit
 51format).  (RLE is supported only if the URT library is available.)
 52cjpeg recognizes the input image format automatically, with the exception
 53of some Targa-format files.  You have to tell djpeg which format to generate.
 55JPEG files are in the defacto standard JFIF file format.  There are other,
 56less widely used JPEG-based file formats, but we don't support them.
 58All switch names may be abbreviated; for example, -grayscale may be written
 59-gray or -gr.  Most of the "basic" switches can be abbreviated to as little as
 60one letter.  Upper and lower case are equivalent (-BMP is the same as -bmp).
 61British spellings are also accepted (e.g., -greyscale), though for brevity
 62these are not mentioned below.
 67The basic command line switches for cjpeg are:
 69	-quality N[,...]  Scale quantization tables to adjust image quality.
 70			Quality is 0 (worst) to 100 (best); default is 75.
 71			(See below for more info.)
 73	-grayscale	Create monochrome JPEG file from color input.
 74			Be sure to use this switch when compressing a grayscale
 75			BMP file, because cjpeg isn't bright enough to notice
 76			whether a BMP file uses only shades of gray.  By
 77			saying -grayscale, you'll get a smaller JPEG file that
 78			takes less time to process.
 80	-rgb		Create RGB JPEG file.
 81			Using this switch suppresses the conversion from RGB
 82			colorspace input to the default YCbCr JPEG colorspace.
 83			Use this switch in combination with the -block N
 84			switch (see below) for lossless JPEG coding.
 86	-optimize	Perform optimization of entropy encoding parameters.
 87			Without this, default encoding parameters are used.
 88			-optimize usually makes the JPEG file a little smaller,
 89			but cjpeg runs somewhat slower and needs much more
 90			memory.  Image quality and speed of decompression are
 91			unaffected by -optimize.
 93	-progressive	Create progressive JPEG file (see below).
 95	-scale M/N	Scale the output image by a factor M/N.  Currently
 96			supported scale factors are M/N with all N from 1 to
 97			16, where M is the destination DCT size, which is 8 by
 98			default (see -block N switch below).
100	-targa		Input file is Targa format.  Targa files that contain
101			an "identification" field will not be automatically
102			recognized by cjpeg; for such files you must specify
103			-targa to make cjpeg treat the input as Targa format.
104			For most Targa files, you won't need this switch.
106The -quality switch lets you trade off compressed file size against quality of
107the reconstructed image: the higher the quality setting, the larger the JPEG
108file, and the closer the output image will be to the original input.  Normally
109you want to use the lowest quality setting (smallest file) that decompresses
110into something visually indistinguishable from the original image.  For this
111purpose the quality setting should be between 50 and 95; the default of 75 is
112often about right.  If you see defects at -quality 75, then go up 5 or 10
113counts at a time until you are happy with the output image.  (The optimal
114setting will vary from one image to another.)
116-quality 100 will generate a quantization table of all 1's, minimizing loss
117in the quantization step (but there is still information loss in subsampling,
118as well as roundoff error).  This setting is mainly of interest for
119experimental purposes.  Quality values above about 95 are NOT recommended for
120normal use; the compressed file size goes up dramatically for hardly any gain
121in output image quality.
123In the other direction, quality values below 50 will produce very small files
124of low image quality.  Settings around 5 to 10 might be useful in preparing an
125index of a large image library, for example.  Try -quality 2 (or so) for some
126amusing Cubist effects.  (Note: quality values below about 25 generate 2-byte
127quantization tables, which are considered optional in the JPEG standard.
128cjpeg emits a warning message when you give such a quality value, because some
129other JPEG programs may be unable to decode the resulting file.  Use -baseline
130if you need to ensure compatibility at low quality values.)
132The -quality option has been extended in IJG version 7 for support of separate
133quality settings for luminance and chrominance (or in general, for every
134provided quantization table slot).  This feature is useful for high-quality
135applications which cannot accept the damage of color data by coarse
136subsampling settings.  You can now easily reduce the color data amount more
137smoothly with finer control without separate subsampling.  The resulting file
138is fully compliant with standard JPEG decoders.
139Note that the -quality ratings refer to the quantization table slots, and that
140the last value is replicated if there are more q-table slots than parameters.
141The default q-table slots are 0 for luminance and 1 for chrominance with
142default tables as given in the JPEG standard.  This is compatible with the old
143behaviour in case that only one parameter is given, which is then used for
144both luminance and chrominance (slots 0 and 1).  More or custom quantization
145tables can be set with -qtables and assigned to components with -qslots
146parameter (see the "wizard" switches below).
147CAUTION: You must explicitly add -sample 1x1 for efficient separate color
148quality selection, since the default value used by library is 2x2!
150The -progressive switch creates a "progressive JPEG" file.  In this type of
151JPEG file, the data is stored in multiple scans of increasing quality.  If the
152file is being transmitted over a slow communications link, the decoder can use
153the first scan to display a low-quality image very quickly, and can then
154improve the display with each subsequent scan.  The final image is exactly
155equivalent to a standard JPEG file of the same quality setting, and the total
156file size is about the same --- often a little smaller.
158Switches for advanced users:
160	-arithmetic	Use arithmetic coding.  CAUTION: arithmetic coded JPEG
161			is not yet widely implemented, so many decoders will
162			be unable to view an arithmetic coded JPEG file at
163			all.
165	-block N	Set DCT block size.  All N from 1 to 16 are possible.
166			Default is 8 (baseline format).
167			Larger values produce higher compression,
168			smaller values produce higher quality
169			(exact DCT stage possible with 1 or 2; with the
170			default quality of 75 and default Luminance qtable
171			the DCT+Quantization stage is lossless for N=1).
172			CAUTION: An implementation of the JPEG SmartScale
173			extension is required for this feature.  SmartScale
174			enabled JPEG is not yet widely implemented, so many
175			decoders will be unable to view a SmartScale extended
176			JPEG file at all.
178	-dct int	Use integer DCT method (default).
179	-dct fast	Use fast integer DCT (less accurate).
180	-dct float	Use floating-point DCT method.
181			The float method is very slightly more accurate than
182			the int method, but is much slower unless your machine
183			has very fast floating-point hardware.  Also note that
184			results of the floating-point method may vary slightly
185			across machines, while the integer methods should give
186			the same results everywhere.  The fast integer method
187			is much less accurate than the other two.
189	-nosmooth	Don't use high-quality downsampling.
191	-restart N	Emit a JPEG restart marker every N MCU rows, or every
192			N MCU blocks if "B" is attached to the number.
193			-restart 0 (the default) means no restart markers.
195	-smooth N	Smooth the input image to eliminate dithering noise.
196			N, ranging from 1 to 100, indicates the strength of
197			smoothing.  0 (the default) means no smoothing.
199	-maxmemory N	Set limit for amount of memory to use in processing
200			large images.  Value is in thousands of bytes, or
201			millions of bytes if "M" is attached to the number.
202			For example, -max 4m selects 4000000 bytes.  If more
203			space is needed, temporary files will be used.
205	-verbose	Enable debug printout.  More -v's give more printout.
206	or  -debug	Also, version information is printed at startup.
208The -restart option inserts extra markers that allow a JPEG decoder to
209resynchronize after a transmission error.  Without restart markers, any damage
210to a compressed file will usually ruin the image from the point of the error
211to the end of the image; with restart markers, the damage is usually confined
212to the portion of the image up to the next restart marker.  Of course, the
213restart markers occupy extra space.  We recommend -restart 1 for images that
214will be transmitted across unreliable networks such as Usenet.
216The -smooth option filters the input to eliminate fine-scale noise.  This is
217often useful when converting dithered images to JPEG: a moderate smoothing
218factor of 10 to 50 gets rid of dithering patterns in the input file, resulting
219in a smaller JPEG file and a better-looking image.  Too large a smoothing
220factor will visibly blur the image, however.
222Switches for wizards:
224	-baseline	Force baseline-compatible quantization tables to be
225			generated.  This clamps quantization values to 8 bits
226			even at low quality settings.  (This switch is poorly
227			named, since it does not ensure that the output is
228			actually baseline JPEG.  For example, you can use
229			-baseline and -progressive together.)
231	-qtables file	Use the quantization tables given in the specified
232			text file.
234	-qslots N[,...] Select which quantization table to use for each color
235			component.
237	-sample HxV[,...]  Set JPEG sampling factors for each color component.
239	-scans file	Use the scan script given in the specified text file.
241The "wizard" switches are intended for experimentation with JPEG.  If you
242don't know what you are doing, DON'T USE THEM.  These switches are documented
243further in the file wizard.txt.
248The basic command line switches for djpeg are:
250	-colors N	Reduce image to at most N colors.  This reduces the
251	or -quantize N	number of colors used in the output image, so that it
252			can be displayed on a colormapped display or stored in
253			a colormapped file format.  For example, if you have
254			an 8-bit display, you'd need to reduce to 256 or fewer
255			colors.  (-colors is the recommended name, -quantize
256			is provided only for backwards compatibility.)
258	-fast		Select recommended processing options for fast, low
259			quality output.  (The default options are chosen for
260			highest quality output.)  Currently, this is equivalent
261			to "-dct fast -nosmooth -onepass -dither ordered".
263	-grayscale	Force gray-scale output even if JPEG file is color.
264			Useful for viewing on monochrome displays; also,
265			djpeg runs noticeably faster in this mode.
267	-scale M/N	Scale the output image by a factor M/N.  Currently
268			supported scale factors are M/N with all M from 1 to
269			16, where N is the source DCT size, which is 8 for
270			baseline JPEG.  If the /N part is omitted, then M
271			specifies the DCT scaled size to be applied on the
272			given input.  For baseline JPEG this is equivalent to
273			M/8 scaling, since the source DCT size for baseline
274			JPEG is 8.  Scaling is handy if the image is larger
275			than your screen; also, djpeg runs much faster when
276			scaling down the output.
278	-bmp		Select BMP output format (Windows flavor).  8-bit
279			colormapped format is emitted if -colors or -grayscale
280			is specified, or if the JPEG file is gray-scale;
281			otherwise, 24-bit full-color format is emitted.
283	-gif		Select GIF output format.  Since GIF does not support
284			more than 256 colors, -colors 256 is assumed (unless
285			you specify a smaller number of colors).  If you
286			specify -fast, the default number of colors is 216.
288	-os2		Select BMP output format (OS/2 1.x flavor).  8-bit
289			colormapped format is emitted if -colors or -grayscale
290			is specified, or if the JPEG file is gray-scale;
291			otherwise, 24-bit full-color format is emitted.
293	-pnm		Select PBMPLUS (PPM/PGM) output format (this is the
294			default format).  PGM is emitted if the JPEG file is
295			gray-scale or if -grayscale is specified; otherwise
296			PPM is emitted.
298	-rle		Select RLE output format.  (Requires URT library.)
300	-targa		Select Targa output format.  Gray-scale format is
301			emitted if the JPEG file is gray-scale or if
302			-grayscale is specified; otherwise, colormapped format
303			is emitted if -colors is specified; otherwise, 24-bit
304			full-color format is emitted.
306Switches for advanced users:
308	-dct int	Use integer DCT method (default).
309	-dct fast	Use fast integer DCT (less accurate).
310	-dct float	Use floating-point DCT method.
311			The float method is very slightly more accurate than
312			the int method, but is much slower unless your machine
313			has very fast floating-point hardware.  Also note that
314			results of the floating-point method may vary slightly
315			across machines, while the integer methods should give
316			the same results everywhere.  The fast integer method
317			is much less accurate than the other two.
319	-dither fs	Use Floyd-Steinberg dithering in color quantization.
320	-dither ordered	Use ordered dithering in color quantization.
321	-dither none	Do not use dithering in color quantization.
322			By default, Floyd-Steinberg dithering is applied when
323			quantizing colors; this is slow but usually produces
324			the best results.  Ordered dither is a compromise
325			between speed and quality; no dithering is fast but
326			usually looks awful.  Note that these switches have
327			no effect unless color quantization is being done.
328			Ordered dither is only available in -onepass mode.
330	-map FILE	Quantize to the colors used in the specified image
331			file.  This is useful for producing multiple files
332			with identical color maps, or for forcing a predefined
333			set of colors to be used.  The FILE must be a GIF
334			or PPM file.  This option overrides -colors and
335			-onepass.
337	-nosmooth	Don't use high-quality upsampling.
339	-onepass	Use one-pass instead of two-pass color quantization.
340			The one-pass method is faster and needs less memory,
341			but it produces a lower-quality image.  -onepass is
342			ignored unless you also say -colors N.  Also,
343			the one-pass method is always used for gray-scale
344			output (the two-pass method is no improvement then).
346	-maxmemory N	Set limit for amount of memory to use in processing
347			large images.  Value is in thousands of bytes, or
348			millions of bytes if "M" is attached to the number.
349			For example, -max 4m selects 4000000 bytes.  If more
350			space is needed, temporary files will be used.
352	-verbose	Enable debug printout.  More -v's give more printout.
353	or  -debug	Also, version information is printed at startup.
358Color GIF files are not the ideal input for JPEG; JPEG is really intended for
359compressing full-color (24-bit) images.  In particular, don't try to convert
360cartoons, line drawings, and other images that have only a few distinct
361colors.  GIF works great on these, JPEG does not.  If you want to convert a
362GIF to JPEG, you should experiment with cjpeg's -quality and -smooth options
363to get a satisfactory conversion.  -smooth 10 or so is often helpful.
365Avoid running an image through a series of JPEG compression/decompression
366cycles.  Image quality loss will accumulate; after ten or so cycles the image
367may be noticeably worse than it was after one cycle.  It's best to use a
368lossless format while manipulating an image, then convert to JPEG format when
369you are ready to file the image away.
371The -optimize option to cjpeg is worth using when you are making a "final"
372version for posting or archiving.  It's also a win when you are using low
373quality settings to make very small JPEG files; the percentage improvement
374is often a lot more than it is on larger files.  (At present, -optimize
375mode is always selected when generating progressive JPEG files.)
377GIF input files are no longer supported, to avoid the Unisys LZW patent.
378(Conversion of GIF files to JPEG is usually a bad idea anyway.)
383To get a quick preview of an image, use the -grayscale and/or -scale switches.
384"-grayscale -scale 1/8" is the fastest case.
386Several options are available that trade off image quality to gain speed.
387"-fast" turns on the recommended settings.
389"-dct fast" and/or "-nosmooth" gain speed at a small sacrifice in quality.
390When producing a color-quantized image, "-onepass -dither ordered" is fast but
391much lower quality than the default behavior.  "-dither none" may give
392acceptable results in two-pass mode, but is seldom tolerable in one-pass mode.
394If you are fortunate enough to have very fast floating point hardware,
395"-dct float" may be even faster than "-dct fast".  But on most machines
396"-dct float" is slower than "-dct int"; in this case it is not worth using,
397because its theoretical accuracy advantage is too small to be significant
398in practice.
400Two-pass color quantization requires a good deal of memory; on MS-DOS machines
401it may run out of memory even with -maxmemory 0.  In that case you can still
402decompress, with some loss of image quality, by specifying -onepass for
403one-pass quantization.
405To avoid the Unisys LZW patent, djpeg produces uncompressed GIF files.  These
406are larger than they should be, but are readable by standard GIF decoders.
411If more space is needed than will fit in the available main memory (as
412determined by -maxmemory), temporary files will be used.  (MS-DOS versions
413will try to get extended or expanded memory first.)  The temporary files are
414often rather large: in typical cases they occupy three bytes per pixel, for
415example 3*800*600 = 1.44Mb for an 800x600 image.  If you don't have enough
416free disk space, leave out -progressive and -optimize (for cjpeg) or specify
417-onepass (for djpeg).
419On MS-DOS, the temporary files are created in the directory named by the TMP
420or TEMP environment variable, or in the current directory if neither of those
421exist.  Amiga implementations put the temp files in the directory named by
422JPEGTMP:, so be sure to assign JPEGTMP: to a disk partition with adequate free
425The default memory usage limit (-maxmemory) is set when the software is
426compiled.  If you get an "insufficient memory" error, try specifying a smaller
427-maxmemory value, even -maxmemory 0 to use the absolute minimum space.  You
428may want to recompile with a smaller default value if this happens often.
430On machines that have "environment" variables, you can define the environment
431variable JPEGMEM to set the default memory limit.  The value is specified as
432described for the -maxmemory switch.  JPEGMEM overrides the default value
433specified when the program was compiled, and itself is overridden by an
434explicit -maxmemory switch.
436On MS-DOS machines, -maxmemory is the amount of main (conventional) memory to
437use.  (Extended or expanded memory is also used if available.)  Most
438DOS-specific versions of this software do their own memory space estimation
439and do not need you to specify -maxmemory.
444jpegtran performs various useful transformations of JPEG files.
445It can translate the coded representation from one variant of JPEG to another,
446for example from baseline JPEG to progressive JPEG or vice versa.  It can also
447perform some rearrangements of the image data, for example turning an image
448from landscape to portrait format by rotation.
450jpegtran works by rearranging the compressed data (DCT coefficients), without
451ever fully decoding the image.  Therefore, its transformations are lossless:
452there is no image degradation at all, which would not be true if you used
453djpeg followed by cjpeg to accomplish the same conversion.  But by the same
454token, jpegtran cannot perform lossy operations such as changing the image
457jpegtran uses a command line syntax similar to cjpeg or djpeg.
458On Unix-like systems, you say:
459	jpegtran [switches] [inputfile] >outputfile
460On most non-Unix systems, you say:
461	jpegtran [switches] inputfile outputfile
462where both the input and output files are JPEG files.
464To specify the coded JPEG representation used in the output file,
465jpegtran accepts a subset of the switches recognized by cjpeg:
466	-optimize	Perform optimization of entropy encoding parameters.
467	-progressive	Create progressive JPEG file.
468	-arithmetic	Use arithmetic coding.
469	-restart N	Emit a JPEG restart marker every N MCU rows, or every
470			N MCU blocks if "B" is attached to the number.
471	-scans file	Use the scan script given in the specified text file.
472See the previous discussion of cjpeg for more details about these switches.
473If you specify none of these switches, you get a plain baseline-JPEG output
474file.  The quality setting and so forth are determined by the input file.
476The image can be losslessly transformed by giving one of these switches:
477	-flip horizontal	Mirror image horizontally (left-right).
478	-flip vertical		Mirror image vertically (top-bottom).
479	-rotate 90		Rotate image 90 degrees clockwise.
480	-rotate 180		Rotate image 180 degrees.
481	-rotate 270		Rotate image 270 degrees clockwise (or 90 ccw).
482	-transpose		Transpose image (across UL-to-LR axis).
483	-transverse		Transverse transpose (across UR-to-LL axis).
485The transpose transformation has no restrictions regarding image dimensions.
486The other transformations operate rather oddly if the image dimensions are not
487a multiple of the iMCU size (usually 8 or 16 pixels), because they can only
488transform complete blocks of DCT coefficient data in the desired way.
490jpegtran's default behavior when transforming an odd-size image is designed
491to preserve exact reversibility and mathematical consistency of the
492transformation set.  As stated, transpose is able to flip the entire image
493area.  Horizontal mirroring leaves any partial iMCU column at the right edge
494untouched, but is able to flip all rows of the image.  Similarly, vertical
495mirroring leaves any partial iMCU row at the bottom edge untouched, but is
496able to flip all columns.  The other transforms can be built up as sequences
497of transpose and flip operations; for consistency, their actions on edge
498pixels are defined to be the same as the end result of the corresponding
499transpose-and-flip sequence.
501For practical use, you may prefer to discard any untransformable edge pixels
502rather than having a strange-looking strip along the right and/or bottom edges
503of a transformed image.  To do this, add the -trim switch:
504	-trim		Drop non-transformable edge blocks.
505Obviously, a transformation with -trim is not reversible, so strictly speaking
506jpegtran with this switch is not lossless.  Also, the expected mathematical
507equivalences between the transformations no longer hold.  For example,
508"-rot 270 -trim" trims only the bottom edge, but "-rot 90 -trim" followed by
509"-rot 180 -trim" trims both edges.
511If you are only interested in perfect transformation, add the -perfect switch:
512	-perfect	Fails with an error if the transformation is not
513			perfect.
514For example you may want to do
515  jpegtran -rot 90 -perfect foo.jpg || djpeg foo.jpg | pnmflip -r90 | cjpeg
516to do a perfect rotation if available or an approximated one if not.
518We also offer a lossless-crop option, which discards data outside a given
519image region but losslessly preserves what is inside.  Like the rotate and
520flip transforms, lossless crop is restricted by the current JPEG format: the
521upper left corner of the selected region must fall on an iMCU boundary.  If
522this does not hold for the given crop parameters, we silently move the upper
523left corner up and/or left to make it so, simultaneously increasing the region
524dimensions to keep the lower right crop corner unchanged.  (Thus, the output
525image covers at least the requested region, but may cover more.)
527The image can be losslessly cropped by giving the switch:
528	-crop WxH+X+Y	Crop to a rectangular subarea of width W, height H
529			starting at point X,Y.
531Other not-strictly-lossless transformation switches are:
533	-grayscale	Force grayscale output.
534This option discards the chrominance channels if the input image is YCbCr
535(ie, a standard color JPEG), resulting in a grayscale JPEG file.  The
536luminance channel is preserved exactly, so this is a better method of reducing
537to grayscale than decompression, conversion, and recompression.  This switch
538is particularly handy for fixing a monochrome picture that was mistakenly
539encoded as a color JPEG.  (In such a case, the space savings from getting rid
540of the near-empty chroma channels won't be large; but the decoding time for
541a grayscale JPEG is substantially less than that for a color JPEG.)
543	-scale M/N	Scale the output image by a factor M/N.
544Currently supported scale factors are M/N with all M from 1 to 16, where N is
545the source DCT size, which is 8 for baseline JPEG.  If the /N part is omitted,
546then M specifies the DCT scaled size to be applied on the given input.  For
547baseline JPEG this is equivalent to M/8 scaling, since the source DCT size
548for baseline JPEG is 8.  CAUTION: An implementation of the JPEG SmartScale
549extension is required for this feature.  SmartScale enabled JPEG is not yet
550widely implemented, so many decoders will be unable to view a SmartScale
551extended JPEG file at all.
553jpegtran also recognizes these switches that control what to do with "extra"
554markers, such as comment blocks:
555	-copy none	Copy no extra markers from source file.  This setting
556			suppresses all comments and other excess baggage
557			present in the source file.
558	-copy comments	Copy only comment markers.  This setting copies
559			comments from the source file, but discards
560			any other inessential (for image display) data.
561	-copy all	Copy all extra markers.  This setting preserves
562			miscellaneous markers found in the source file, such
563			as JFIF thumbnails, Exif data, and Photoshop settings.
564			In some files these extra markers can be sizable.
565The default behavior is -copy comments.  (Note: in IJG releases v6 and v6a,
566jpegtran always did the equivalent of -copy none.)
568Additional switches recognized by jpegtran are:
569	-outfile filename
570	-maxmemory N
571	-verbose
572	-debug
573These work the same as in cjpeg or djpeg.
578The JPEG standard allows "comment" (COM) blocks to occur within a JPEG file.
579Although the standard doesn't actually define what COM blocks are for, they
580are widely used to hold user-supplied text strings.  This lets you add
581annotations, titles, index terms, etc to your JPEG files, and later retrieve
582them as text.  COM blocks do not interfere with the image stored in the JPEG
583file.  The maximum size of a COM block is 64K, but you can have as many of
584them as you like in one JPEG file.
586We provide two utility programs to display COM block contents and add COM
587blocks to a JPEG file.
589rdjpgcom searches a JPEG file and prints the contents of any COM blocks on
590standard output.  The command line syntax is
591	rdjpgcom [-raw] [-verbose] [inputfilename]
592The switch "-raw" (or just "-r") causes rdjpgcom to also output non-printable
593characters in comments, which are normally escaped for security reasons.
594The switch "-verbose" (or just "-v") causes rdjpgcom to also display the JPEG
595image dimensions.  If you omit the input file name from the command line,
596the JPEG file is read from standard input.  (This may not work on some
597operating systems, if binary data can't be read from stdin.)
599wrjpgcom adds a COM block, containing text you provide, to a JPEG file.
600Ordinarily, the COM block is added after any existing COM blocks, but you
601can delete the old COM blocks if you wish.  wrjpgcom produces a new JPEG
602file; it does not modify the input file.  DO NOT try to overwrite the input
603file by directing wrjpgcom's output back into it; on most systems this will
604just destroy your file.
606The command line syntax for wrjpgcom is similar to cjpeg's.  On Unix-like
607systems, it is
608	wrjpgcom [switches] [inputfilename]
609The output file is written to standard output.  The input file comes from
610the named file, or from standard input if no input file is named.
612On most non-Unix systems, the syntax is
613	wrjpgcom [switches] inputfilename outputfilename
614where both input and output file names must be given explicitly.
616wrjpgcom understands three switches:
617	-replace		 Delete any existing COM blocks from the file.
618	-comment "Comment text"	 Supply new COM text on command line.
619	-cfile name		 Read text for new COM block from named file.
620(Switch names can be abbreviated.)  If you have only one line of comment text
621to add, you can provide it on the command line with -comment.  The comment
622text must be surrounded with quotes so that it is treated as a single
623argument.  Longer comments can be read from a text file.
625If you give neither -comment nor -cfile, then wrjpgcom will read the comment
626text from standard input.  (In this case an input image file name MUST be
627supplied, so that the source JPEG file comes from somewhere else.)  You can
628enter multiple lines, up to 64KB worth.  Type an end-of-file indicator
629(usually control-D or control-Z) to terminate the comment text entry.
631wrjpgcom will not add a COM block if the provided comment string is empty.
632Therefore -replace -comment "" can be used to delete all COM blocks from a
635These utility programs do not depend on the IJG JPEG library.  In
636particular, the source code for rdjpgcom is intended as an illustration of
637the minimum amount of code required to parse a JPEG file header correctly.