blob: e78b26c22053cb10827cd058424e08d36cd24882 [file] [log] [blame]
USAGE instructions for the Independent JPEG Group's JPEG software
This file describes usage of the JPEG conversion programs cjpeg and djpeg,
as well as the utility programs jpegtran, rdjpgcom and wrjpgcom. (See
the other documentation files if you wish to use the JPEG library within
your own programs.)
If you are on a Unix machine you may prefer to read the Unix-style manual
pages in files cjpeg.1, djpeg.1, jpegtran.1, rdjpgcom.1, wrjpgcom.1.
These programs implement JPEG image encoding, decoding, and transcoding.
JPEG (pronounced "jay-peg") is a standardized compression method for
full-color and gray-scale images.
We provide two programs, cjpeg to compress an image file into JPEG format,
and djpeg to decompress a JPEG file back into a conventional image format.
On Unix-like systems, you say:
cjpeg [switches] [imagefile] >jpegfile
djpeg [switches] [jpegfile] >imagefile
The programs read the specified input file, or standard input if none is
named. They always write to standard output (with trace/error messages to
standard error). These conventions are handy for piping images between
On most non-Unix systems, you say:
cjpeg [switches] imagefile jpegfile
djpeg [switches] jpegfile imagefile
i.e., both the input and output files are named on the command line. This
style is a little more foolproof, and it loses no functionality if you don't
have pipes. (You can get this style on Unix too, if you prefer, by defining
TWO_FILE_COMMANDLINE when you compile the programs; see install.txt.)
You can also say:
cjpeg [switches] -outfile jpegfile imagefile
djpeg [switches] -outfile imagefile jpegfile
This syntax works on all systems, so it is useful for scripts.
The currently supported image file formats are: PPM (PBMPLUS color format),
PGM (PBMPLUS gray-scale format), BMP, Targa, and RLE (Utah Raster Toolkit
format). (RLE is supported only if the URT library is available.)
cjpeg recognizes the input image format automatically, with the exception
of some Targa-format files. You have to tell djpeg which format to generate.
JPEG files are in the defacto standard JFIF file format. There are other,
less widely used JPEG-based file formats, but we don't support them.
All switch names may be abbreviated; for example, -grayscale may be written
-gray or -gr. Most of the "basic" switches can be abbreviated to as little as
one letter. Upper and lower case are equivalent (-BMP is the same as -bmp).
British spellings are also accepted (e.g., -greyscale), though for brevity
these are not mentioned below.
The basic command line switches for cjpeg are:
-quality N[,...] Scale quantization tables to adjust image quality.
Quality is 0 (worst) to 100 (best); default is 75.
(See below for more info.)
-grayscale Create monochrome JPEG file from color input.
Be sure to use this switch when compressing a grayscale
BMP file, because cjpeg isn't bright enough to notice
whether a BMP file uses only shades of gray. By
saying -grayscale, you'll get a smaller JPEG file that
takes less time to process.
-optimize Perform optimization of entropy encoding parameters.
Without this, default encoding parameters are used.
-optimize usually makes the JPEG file a little smaller,
but cjpeg runs somewhat slower and needs much more
memory. Image quality and speed of decompression are
unaffected by -optimize.
-progressive Create progressive JPEG file (see below).
-scale M/N Scale the output image by a factor M/N. Currently
supported scale factors are 8/N with all N from 1 to
-targa Input file is Targa format. Targa files that contain
an "identification" field will not be automatically
recognized by cjpeg; for such files you must specify
-targa to make cjpeg treat the input as Targa format.
For most Targa files, you won't need this switch.
The -quality switch lets you trade off compressed file size against quality of
the reconstructed image: the higher the quality setting, the larger the JPEG
file, and the closer the output image will be to the original input. Normally
you want to use the lowest quality setting (smallest file) that decompresses
into something visually indistinguishable from the original image. For this
purpose the quality setting should be between 50 and 95; the default of 75 is
often about right. If you see defects at -quality 75, then go up 5 or 10
counts at a time until you are happy with the output image. (The optimal
setting will vary from one image to another.)
-quality 100 will generate a quantization table of all 1's, minimizing loss
in the quantization step (but there is still information loss in subsampling,
as well as roundoff error). This setting is mainly of interest for
experimental purposes. Quality values above about 95 are NOT recommended for
normal use; the compressed file size goes up dramatically for hardly any gain
in output image quality.
In the other direction, quality values below 50 will produce very small files
of low image quality. Settings around 5 to 10 might be useful in preparing an
index of a large image library, for example. Try -quality 2 (or so) for some
amusing Cubist effects. (Note: quality values below about 25 generate 2-byte
quantization tables, which are considered optional in the JPEG standard.
cjpeg emits a warning message when you give such a quality value, because some
other JPEG programs may be unable to decode the resulting file. Use -baseline
if you need to ensure compatibility at low quality values.)
The -quality option has been extended in IJG version 7 for support of separate
quality settings for luminance and chrominance (or in general, for every
provided quantization table slot). This feature is useful for high-quality
applications which cannot accept the damage of color data by coarse
subsampling settings. You can now easily reduce the color data amount more
smoothly with finer control without separate subsampling. The resulting file
is fully compliant with standard JPEG decoders.
Note that the -quality ratings refer to the quantization table slots, and that
the last value is replicated if there are more q-table slots than parameters.
The default q-table slots are 0 for luminance and 1 for chrominance with
default tables as given in the JPEG standard. This is compatible with the old
behaviour in case that only one parameter is given, which is then used for
both luminance and chrominance (slots 0 and 1). More or custom quantization
tables can be set with -qtables and assigned to components with -qslots
parameter (see the "wizard" switches below).
CAUTION: You must explicitely add -sample 1x1 for efficient separate color
quality selection, since the default value used by library is 2x2!
The -progressive switch creates a "progressive JPEG" file. In this type of
JPEG file, the data is stored in multiple scans of increasing quality. If the
file is being transmitted over a slow communications link, the decoder can use
the first scan to display a low-quality image very quickly, and can then
improve the display with each subsequent scan. The final image is exactly
equivalent to a standard JPEG file of the same quality setting, and the total
file size is about the same --- often a little smaller.
Switches for advanced users:
-dct int Use integer DCT method (default).
-dct fast Use fast integer DCT (less accurate).
-dct float Use floating-point DCT method.
The float method is very slightly more accurate than
the int method, but is much slower unless your machine
has very fast floating-point hardware. Also note that
results of the floating-point method may vary slightly
across machines, while the integer methods should give
the same results everywhere. The fast integer method
is much less accurate than the other two.
-nosmooth Don't use high-quality downsampling.
-restart N Emit a JPEG restart marker every N MCU rows, or every
N MCU blocks if "B" is attached to the number.
-restart 0 (the default) means no restart markers.
-smooth N Smooth the input image to eliminate dithering noise.
N, ranging from 1 to 100, indicates the strength of
smoothing. 0 (the default) means no smoothing.
-maxmemory N Set limit for amount of memory to use in processing
large images. Value is in thousands of bytes, or
millions of bytes if "M" is attached to the number.
For example, -max 4m selects 4000000 bytes. If more
space is needed, temporary files will be used.
-verbose Enable debug printout. More -v's give more printout.
or -debug Also, version information is printed at startup.
The -restart option inserts extra markers that allow a JPEG decoder to
resynchronize after a transmission error. Without restart markers, any damage
to a compressed file will usually ruin the image from the point of the error
to the end of the image; with restart markers, the damage is usually confined
to the portion of the image up to the next restart marker. Of course, the
restart markers occupy extra space. We recommend -restart 1 for images that
will be transmitted across unreliable networks such as Usenet.
The -smooth option filters the input to eliminate fine-scale noise. This is
often useful when converting dithered images to JPEG: a moderate smoothing
factor of 10 to 50 gets rid of dithering patterns in the input file, resulting
in a smaller JPEG file and a better-looking image. Too large a smoothing
factor will visibly blur the image, however.
Switches for wizards:
-arithmetic Use arithmetic coding. CAUTION: arithmetic coded JPEG
is not yet widely implemented, so many decoders will
be unable to view an arithmetic coded JPEG file at
-baseline Force baseline-compatible quantization tables to be
generated. This clamps quantization values to 8 bits
even at low quality settings. (This switch is poorly
named, since it does not ensure that the output is
actually baseline JPEG. For example, you can use
-baseline and -progressive together.)
-qtables file Use the quantization tables given in the specified
text file.
-qslots N[,...] Select which quantization table to use for each color
-sample HxV[,...] Set JPEG sampling factors for each color component.
-scans file Use the scan script given in the specified text file.
The "wizard" switches are intended for experimentation with JPEG. If you
don't know what you are doing, DON'T USE THEM. These switches are documented
further in the file wizard.txt.
The basic command line switches for djpeg are:
-colors N Reduce image to at most N colors. This reduces the
or -quantize N number of colors used in the output image, so that it
can be displayed on a colormapped display or stored in
a colormapped file format. For example, if you have
an 8-bit display, you'd need to reduce to 256 or fewer
colors. (-colors is the recommended name, -quantize
is provided only for backwards compatibility.)
-fast Select recommended processing options for fast, low
quality output. (The default options are chosen for
highest quality output.) Currently, this is equivalent
to "-dct fast -nosmooth -onepass -dither ordered".
-grayscale Force gray-scale output even if JPEG file is color.
Useful for viewing on monochrome displays; also,
djpeg runs noticeably faster in this mode.
-scale M/N Scale the output image by a factor M/N. Currently
supported scale factors are M/8 with all M from 1 to
16. If the /N part is omitted, then M specifies the
DCT scaled size to be applied on the given input,
which is currently equivalent to M/8 scaling, since
the source DCT size is currently always 8.
Scaling is handy if the image is larger than your
screen; also, djpeg runs much faster when scaling
down the output.
-bmp Select BMP output format (Windows flavor). 8-bit
colormapped format is emitted if -colors or -grayscale
is specified, or if the JPEG file is gray-scale;
otherwise, 24-bit full-color format is emitted.
-gif Select GIF output format. Since GIF does not support
more than 256 colors, -colors 256 is assumed (unless
you specify a smaller number of colors). If you
specify -fast, the default number of colors is 216.
-os2 Select BMP output format (OS/2 1.x flavor). 8-bit
colormapped format is emitted if -colors or -grayscale
is specified, or if the JPEG file is gray-scale;
otherwise, 24-bit full-color format is emitted.
-pnm Select PBMPLUS (PPM/PGM) output format (this is the
default format). PGM is emitted if the JPEG file is
gray-scale or if -grayscale is specified; otherwise
PPM is emitted.
-rle Select RLE output format. (Requires URT library.)
-targa Select Targa output format. Gray-scale format is
emitted if the JPEG file is gray-scale or if
-grayscale is specified; otherwise, colormapped format
is emitted if -colors is specified; otherwise, 24-bit
full-color format is emitted.
Switches for advanced users:
-dct int Use integer DCT method (default).
-dct fast Use fast integer DCT (less accurate).
-dct float Use floating-point DCT method.
The float method is very slightly more accurate than
the int method, but is much slower unless your machine
has very fast floating-point hardware. Also note that
results of the floating-point method may vary slightly
across machines, while the integer methods should give
the same results everywhere. The fast integer method
is much less accurate than the other two.
-dither fs Use Floyd-Steinberg dithering in color quantization.
-dither ordered Use ordered dithering in color quantization.
-dither none Do not use dithering in color quantization.
By default, Floyd-Steinberg dithering is applied when
quantizing colors; this is slow but usually produces
the best results. Ordered dither is a compromise
between speed and quality; no dithering is fast but
usually looks awful. Note that these switches have
no effect unless color quantization is being done.
Ordered dither is only available in -onepass mode.
-map FILE Quantize to the colors used in the specified image
file. This is useful for producing multiple files
with identical color maps, or for forcing a predefined
set of colors to be used. The FILE must be a GIF
or PPM file. This option overrides -colors and
-nosmooth Don't use high-quality upsampling.
-onepass Use one-pass instead of two-pass color quantization.
The one-pass method is faster and needs less memory,
but it produces a lower-quality image. -onepass is
ignored unless you also say -colors N. Also,
the one-pass method is always used for gray-scale
output (the two-pass method is no improvement then).
-maxmemory N Set limit for amount of memory to use in processing
large images. Value is in thousands of bytes, or
millions of bytes if "M" is attached to the number.
For example, -max 4m selects 4000000 bytes. If more
space is needed, temporary files will be used.
-verbose Enable debug printout. More -v's give more printout.
or -debug Also, version information is printed at startup.
Color GIF files are not the ideal input for JPEG; JPEG is really intended for
compressing full-color (24-bit) images. In particular, don't try to convert
cartoons, line drawings, and other images that have only a few distinct
colors. GIF works great on these, JPEG does not. If you want to convert a
GIF to JPEG, you should experiment with cjpeg's -quality and -smooth options
to get a satisfactory conversion. -smooth 10 or so is often helpful.
Avoid running an image through a series of JPEG compression/decompression
cycles. Image quality loss will accumulate; after ten or so cycles the image
may be noticeably worse than it was after one cycle. It's best to use a
lossless format while manipulating an image, then convert to JPEG format when
you are ready to file the image away.
The -optimize option to cjpeg is worth using when you are making a "final"
version for posting or archiving. It's also a win when you are using low
quality settings to make very small JPEG files; the percentage improvement
is often a lot more than it is on larger files. (At present, -optimize
mode is always selected when generating progressive JPEG files.)
GIF input files are no longer supported, to avoid the Unisys LZW patent.
(Conversion of GIF files to JPEG is usually a bad idea anyway.)
To get a quick preview of an image, use the -grayscale and/or -scale switches.
"-grayscale -scale 1/8" is the fastest case.
Several options are available that trade off image quality to gain speed.
"-fast" turns on the recommended settings.
"-dct fast" and/or "-nosmooth" gain speed at a small sacrifice in quality.
When producing a color-quantized image, "-onepass -dither ordered" is fast but
much lower quality than the default behavior. "-dither none" may give
acceptable results in two-pass mode, but is seldom tolerable in one-pass mode.
If you are fortunate enough to have very fast floating point hardware,
"-dct float" may be even faster than "-dct fast". But on most machines
"-dct float" is slower than "-dct int"; in this case it is not worth using,
because its theoretical accuracy advantage is too small to be significant
in practice.
Two-pass color quantization requires a good deal of memory; on MS-DOS machines
it may run out of memory even with -maxmemory 0. In that case you can still
decompress, with some loss of image quality, by specifying -onepass for
one-pass quantization.
To avoid the Unisys LZW patent, djpeg produces uncompressed GIF files. These
are larger than they should be, but are readable by standard GIF decoders.
If more space is needed than will fit in the available main memory (as
determined by -maxmemory), temporary files will be used. (MS-DOS versions
will try to get extended or expanded memory first.) The temporary files are
often rather large: in typical cases they occupy three bytes per pixel, for
example 3*800*600 = 1.44Mb for an 800x600 image. If you don't have enough
free disk space, leave out -progressive and -optimize (for cjpeg) or specify
-onepass (for djpeg).
On MS-DOS, the temporary files are created in the directory named by the TMP
or TEMP environment variable, or in the current directory if neither of those
exist. Amiga implementations put the temp files in the directory named by
JPEGTMP:, so be sure to assign JPEGTMP: to a disk partition with adequate free
The default memory usage limit (-maxmemory) is set when the software is
compiled. If you get an "insufficient memory" error, try specifying a smaller
-maxmemory value, even -maxmemory 0 to use the absolute minimum space. You
may want to recompile with a smaller default value if this happens often.
On machines that have "environment" variables, you can define the environment
variable JPEGMEM to set the default memory limit. The value is specified as
described for the -maxmemory switch. JPEGMEM overrides the default value
specified when the program was compiled, and itself is overridden by an
explicit -maxmemory switch.
On MS-DOS machines, -maxmemory is the amount of main (conventional) memory to
use. (Extended or expanded memory is also used if available.) Most
DOS-specific versions of this software do their own memory space estimation
and do not need you to specify -maxmemory.
jpegtran performs various useful transformations of JPEG files.
It can translate the coded representation from one variant of JPEG to another,
for example from baseline JPEG to progressive JPEG or vice versa. It can also
perform some rearrangements of the image data, for example turning an image
from landscape to portrait format by rotation.
jpegtran works by rearranging the compressed data (DCT coefficients), without
ever fully decoding the image. Therefore, its transformations are lossless:
there is no image degradation at all, which would not be true if you used
djpeg followed by cjpeg to accomplish the same conversion. But by the same
token, jpegtran cannot perform lossy operations such as changing the image
jpegtran uses a command line syntax similar to cjpeg or djpeg.
On Unix-like systems, you say:
jpegtran [switches] [inputfile] >outputfile
On most non-Unix systems, you say:
jpegtran [switches] inputfile outputfile
where both the input and output files are JPEG files.
To specify the coded JPEG representation used in the output file,
jpegtran accepts a subset of the switches recognized by cjpeg:
-optimize Perform optimization of entropy encoding parameters.
-progressive Create progressive JPEG file.
-restart N Emit a JPEG restart marker every N MCU rows, or every
N MCU blocks if "B" is attached to the number.
-arithmetic Use arithmetic coding.
-scans file Use the scan script given in the specified text file.
See the previous discussion of cjpeg for more details about these switches.
If you specify none of these switches, you get a plain baseline-JPEG output
file. The quality setting and so forth are determined by the input file.
The image can be losslessly transformed by giving one of these switches:
-flip horizontal Mirror image horizontally (left-right).
-flip vertical Mirror image vertically (top-bottom).
-rotate 90 Rotate image 90 degrees clockwise.
-rotate 180 Rotate image 180 degrees.
-rotate 270 Rotate image 270 degrees clockwise (or 90 ccw).
-transpose Transpose image (across UL-to-LR axis).
-transverse Transverse transpose (across UR-to-LL axis).
The transpose transformation has no restrictions regarding image dimensions.
The other transformations operate rather oddly if the image dimensions are not
a multiple of the iMCU size (usually 8 or 16 pixels), because they can only
transform complete blocks of DCT coefficient data in the desired way.
jpegtran's default behavior when transforming an odd-size image is designed
to preserve exact reversibility and mathematical consistency of the
transformation set. As stated, transpose is able to flip the entire image
area. Horizontal mirroring leaves any partial iMCU column at the right edge
untouched, but is able to flip all rows of the image. Similarly, vertical
mirroring leaves any partial iMCU row at the bottom edge untouched, but is
able to flip all columns. The other transforms can be built up as sequences
of transpose and flip operations; for consistency, their actions on edge
pixels are defined to be the same as the end result of the corresponding
transpose-and-flip sequence.
For practical use, you may prefer to discard any untransformable edge pixels
rather than having a strange-looking strip along the right and/or bottom edges
of a transformed image. To do this, add the -trim switch:
-trim Drop non-transformable edge blocks.
Obviously, a transformation with -trim is not reversible, so strictly speaking
jpegtran with this switch is not lossless. Also, the expected mathematical
equivalences between the transformations no longer hold. For example,
"-rot 270 -trim" trims only the bottom edge, but "-rot 90 -trim" followed by
"-rot 180 -trim" trims both edges.
If you are only interested in perfect transformation, add the -perfect switch:
-perfect Fails with an error if the transformation is not
For example you may want to do
jpegtran -rot 90 -perfect foo.jpg || djpeg foo.jpg | pnmflip -r90 | cjpeg
to do a perfect rotation if available or an approximated one if not.
We also offer a lossless-crop option, which discards data outside a given
image region but losslessly preserves what is inside. Like the rotate and
flip transforms, lossless crop is restricted by the current JPEG format: the
upper left corner of the selected region must fall on an iMCU boundary. If
this does not hold for the given crop parameters, we silently move the upper
left corner up and/or left to make it so, simultaneously increasing the region
dimensions to keep the lower right crop corner unchanged. (Thus, the output
image covers at least the requested region, but may cover more.)
The image can be losslessly cropped by giving the switch:
-crop WxH+X+Y Crop to a rectangular subarea of width W, height H
starting at point X,Y.
Another not-strictly-lossless transformation switch is:
-grayscale Force grayscale output.
This option discards the chrominance channels if the input image is YCbCr
(ie, a standard color JPEG), resulting in a grayscale JPEG file. The
luminance channel is preserved exactly, so this is a better method of reducing
to grayscale than decompression, conversion, and recompression. This switch
is particularly handy for fixing a monochrome picture that was mistakenly
encoded as a color JPEG. (In such a case, the space savings from getting rid
of the near-empty chroma channels won't be large; but the decoding time for
a grayscale JPEG is substantially less than that for a color JPEG.)
jpegtran also recognizes these switches that control what to do with "extra"
markers, such as comment blocks:
-copy none Copy no extra markers from source file. This setting
suppresses all comments and other excess baggage
present in the source file.
-copy comments Copy only comment markers. This setting copies
comments from the source file, but discards
any other inessential (for image display) data.
-copy all Copy all extra markers. This setting preserves
miscellaneous markers found in the source file, such
as JFIF thumbnails, Exif data, and Photoshop settings.
In some files these extra markers can be sizable.
The default behavior is -copy comments. (Note: in IJG releases v6 and v6a,
jpegtran always did the equivalent of -copy none.)
Additional switches recognized by jpegtran are:
-outfile filename
-maxmemory N
These work the same as in cjpeg or djpeg.
The JPEG standard allows "comment" (COM) blocks to occur within a JPEG file.
Although the standard doesn't actually define what COM blocks are for, they
are widely used to hold user-supplied text strings. This lets you add
annotations, titles, index terms, etc to your JPEG files, and later retrieve
them as text. COM blocks do not interfere with the image stored in the JPEG
file. The maximum size of a COM block is 64K, but you can have as many of
them as you like in one JPEG file.
We provide two utility programs to display COM block contents and add COM
blocks to a JPEG file.
rdjpgcom searches a JPEG file and prints the contents of any COM blocks on
standard output. The command line syntax is
rdjpgcom [-raw] [-verbose] [inputfilename]
The switch "-raw" (or just "-r") causes rdjpgcom to also output non-printable
characters in comments, which are normally escaped for security reasons.
The switch "-verbose" (or just "-v") causes rdjpgcom to also display the JPEG
image dimensions. If you omit the input file name from the command line,
the JPEG file is read from standard input. (This may not work on some
operating systems, if binary data can't be read from stdin.)
wrjpgcom adds a COM block, containing text you provide, to a JPEG file.
Ordinarily, the COM block is added after any existing COM blocks, but you
can delete the old COM blocks if you wish. wrjpgcom produces a new JPEG
file; it does not modify the input file. DO NOT try to overwrite the input
file by directing wrjpgcom's output back into it; on most systems this will
just destroy your file.
The command line syntax for wrjpgcom is similar to cjpeg's. On Unix-like
systems, it is
wrjpgcom [switches] [inputfilename]
The output file is written to standard output. The input file comes from
the named file, or from standard input if no input file is named.
On most non-Unix systems, the syntax is
wrjpgcom [switches] inputfilename outputfilename
where both input and output file names must be given explicitly.
wrjpgcom understands three switches:
-replace Delete any existing COM blocks from the file.
-comment "Comment text" Supply new COM text on command line.
-cfile name Read text for new COM block from named file.
(Switch names can be abbreviated.) If you have only one line of comment text
to add, you can provide it on the command line with -comment. The comment
text must be surrounded with quotes so that it is treated as a single
argument. Longer comments can be read from a text file.
If you give neither -comment nor -cfile, then wrjpgcom will read the comment
text from standard input. (In this case an input image file name MUST be
supplied, so that the source JPEG file comes from somewhere else.) You can
enter multiple lines, up to 64KB worth. Type an end-of-file indicator
(usually control-D or control-Z) to terminate the comment text entry.
wrjpgcom will not add a COM block if the provided comment string is empty.
Therefore -replace -comment "" can be used to delete all COM blocks from a
These utility programs do not depend on the IJG JPEG library. In
particular, the source code for rdjpgcom is intended as an illustration of
the minimum amount of code required to parse a JPEG file header correctly.