pamtojpeg2k

Updated: 1 November 2022
Table Of Contents

NAME

pamtojpeg2k - convert PAM/PNM image to a JPEG-2000 code stream

SYNOPSIS

pamtojpeg2k [-imgareatlx=column] [-imgareatly=row] [-tilegrdtlx=column] [-tilegrdtly=row] [-tilewidth=columns] [-tileheight=rows] [-prcwidth=columns] [-prcheight=rows] [-cblkwidth=columns] [-cblkheight=rows] [-mode={integer|int|real}] [-compression=ratio] [-size=bytes] [-ilyrrates=[realnumber[,realnumber, ...]] [-numrlvls=number] [-progression={lrcp|rlcp|rpcl|pcrl|cprl}] [-numgbits=number] [-nomct] [-sop] [-eph] [-lazy] [-termall] [-segsym] [-vcausal] [-pterm] [-resetprob] [-verbose] [-debuglevel=number] filename

Minimum unique abbreviation of option is acceptable. You may use double hyphens instead of single hyphen to denote options. You may use white space in place of the equals sign to separate an option name from its value.

DESCRIPTION

This program is part of Netpbm.

pamtojpeg2k converts the named PBM, PGM, PPM, or PAM file, or Standard Input if no file is named, to a JPEG-2000 code stream (JPC) file on Standard Output.

The JPEG-2000 specification specifies two separate formats: JP2 and JPEG-2000 code stream (JPC). JP2 represents a visual image quite specifically, whereas JPC is a more or less arbitrary array of codes. pamtojpeg2k can't produce a JP2, but the JPC image that pamtojpeg2k produces is very similar to a JP2 if the input is a PBM, PGM, or PPM image or equivalent PAM image. One difference is that the RGB intensity values in a JP2 are SRGB values, while pamtojpeg2k produces ITU-R Recommendation BT.709 values. Those are very similar, but not identical. Another difference is that a JP2 can contain extra information about an image that JPC cannot.

When the input is a PAM image other than a PBM, PGM, or PPM equivalent, the JPC raster produced contains whatever the PAM raster does. It can have any number of planes with any meanings; the planes are in the same order in the JPC output as in the PAM input.

A JPC image has a "precision," which is the number of bits used for each code (in Netpbm lingo, "sample"). Actually, it has a separate precision for each component. pamtojpeg2k uses for the precision of every component the least number of bits that can represent the maxval of the input image. A JPC image does not have an independent concept of maxval; the maxval of a JPC sample is the maximum value that the number of bits specified by the precision can represent in pure binary code. E.g. if the precision is 4, the maxval is 15. pamtojpeg2k does of course scale the sample values from the input maxval to the output maxval. Example: The input maxval is 99. This means JPC precision is 7 bits and the JPC maxval is 127. A sample value of 33 in the input becomes a sample value of 43 in the output.

pamtojpeg2k generates the JPC output with the Jasper JPEG-2000 library. See documentation of the library for details on what pamtojpeg2k produces. Note that the Jasper library contains facilities for reading PNM images, but pamtojpeg2k does not use those. It uses the Netpbm library instead. Note that the makers of the Jasper library write it "JasPer," but Netpbm documentation follows standard American English typography rules, which don't allow that kind of capitalization.

Use jpeg2ktopam to convert in the other direction.

The program jasper, which is packaged with the Jasper JPEG-2000 library, also converts between JPEG-2000 and PNM formats. Because it's packaged with the library, it may exploit it better, especially recently added features. However, since it does not use the Netpbm library to read and write the Netpbm formats, it doesn't do as good a job on that side.

Another format with goals similar to those of JPEG-2000 but that allows for faster encoding and decoding, is JPEG-LS. CharLS is a package of software for using JPEG-LS.

OPTIONS

In addition to the options common to all programs based on libnetpbm (most notably -quiet, see Common Options), pamtojpeg2k recognizes the following command line options:

Jasper Library Options

These options are identical in name and function to options that the Jasper library JPC encoder subroutine takes. See Jasper documentation for details.

-imgareatlx=column
-imgareatly=row
-tilegrdtlx=column
-tilegrdtly=row
-tilewidth=columns
-tileheight=rows
-prcwidth=columns
-prcheight=rows
-cblkwidth=columns
-cblkheight=rows
-mode={integer|int|real}
-ilyrrates=[realnumber[,realnumber, ...]]
-numrlvls=number
-progression={lrcp|rlcp|rpcl|pcrl|cprl}
-numgbits=number
-nomct
-sop
-eph
-lazy
-termall
-segsym
-vcausal
-pterm
-resetprob

-ilyrrates ("intermediate layer rates") lets you control the compression at each layer. The compressed image is arranged in layers, so if you transmit it somewhere serially, enough information to make a low quality image arrives soon, then later more information arrives to improve the quality of the image, and so on until the entire image arrives. The value of this option is a list of ascending fractions, such as ".1,.3,.5". Each fraction says the size of the layers up to and including that layer should add up to that fraction of the size of the input image. If the size of the entire image is limited by -compression or -size, none of these numbers may exceed that limit.

Other Options

-compression=ratio
ratio is a floating point number that specifies the compression ratio. pamtojpeg2k will adjust quality as necessary to ensure that you get this compression ratio. E.g. 4 means the output will be about one fourth the size in bytes of the input file.

The ratio concerns just the raster part of the image, with the denominator being what the raster would take if it were encoded the most naive way possible (e.g. 3 bytes per pixel in 8-bit-per-sample RGB). It does, however, include metadata that is part of the compressed raster. Because of that, it may not be possible to give you your requested compression ratio at any quality. If it isn't, pamtojpeg2k fails with a message saying so.

If you don't specify this option, pamtojpeg2k gives you the best compression it can without losing any quality. Because of the metadata issue described above, this may mean, for a small image, the image actually expands.

This option controls the "rate" option of the Jasper library. Note that though the Jasper library takes a compression factor, this option specifies a compression ratio. The compression factor is the multiplicative inverse of (1 divided by) the compression ratio.

You may not specify this with -size.

Before Netpbm 10.61 (December 2012), the default was a compression ratio of 1, and if pamtojpeg2k could not make the output that small, it just made it as small as it could, with zero quality. You know this is happening when you see the warning message, "empty layer generated."

-size=bytes
This option specifies the maximum size in bytes you want the output image to have. This size is all-in, including headers and trailers and other metadata. pamtojpeg2k will omit as much information as necessary to get under this limit.

It is possible to specify a value impossibly small, for example a value that doesn't even leave room for the image header. If you do this, the program fails with an explanatory message.

This option controls the "rate" option of the Jasper library.

You may not specify this with -compression.

This option was new in Netpbm 11.1 (December 2022).

-verbose
This option causes pamtojpeg2k to issue informational messages about the conversion process.
-debuglevel=number
This option controls debug messages from the Jasper library. pamtojpeg2k passes number as the debug level to the Jasper JPC encoder.

EXAMPLES

This example compresses losslessly.

  pamtojpeg2k myimg.ppm >myimg.jpc
jpeg2ktopam will recreate myimg.ppm exactly.

This example compresses the file to one tenth its original size, throwing away information as necessary.

  pamtojpeg2k -compression=10 myimg.pgm >myimg.jpc

ABOUT JPEG-2000

JPEG-2000 is a format that compresses a visual image (or a similar set of data) into a minimal number of bytes for storage or transmission. In that, its goal is similar to JPEG. It has two main differences from JPEG.

One difference is that it does a much better job on most images of throwing out information in order to achieve a smaller output. That means when you reconstruct the image from the resulting compressed file, it looks a lot closer to the image you started with JPEG-2000 than with JPEG, for the same compressed file size. Or, looked at another way, with JPEG-2000 you get a much smaller file than with JPEG for the same image quality.

The second difference is that with JPEG-2000, you decide how much compression you want and the compressor adjusts the quality to meet your requirement, whereas with JPEG, you decide how much quality you want and the compressor adjusts the size of the output to meet your requirement. I.e. with JPEG-2000, the quality of the result depends on the compressibility of the input, but with JPEG, the size of the result depends on the compressibility of the input.

With JPEG-2000, you can specify lossless compression, thus making it compete with GIF and PNG. With standard JPEG, you always lose something. (There are rumored to be variations of JPEG around that are lossless, though).

JPEG is much older than JPEG-2000 and far more popular. JPEG is one of the half dozen most popular graphics formats and virtually all graphics facilities understand it. JPEG-2000 is virtually unknown.

There is no compatibility between JPEG and JPEG-2000. Programs that read JPEG do not automatically read JPEG-2000 and vice versa.

SEE ALSO

jpeg2ktopam, pnmtojpeg, ppm, pgm, pbm, pam,

HISTORY

pamtojpeg2k was added to Netpbm in Release 10.12 (November 2002).


Table Of Contents