Table of Contents


marcvt- format conversion for images collected on marresearch Image Plate Detectors and marCCD detectors.


marcvt [ -h ] [ --help ] [ -f ] [ --force ] [ -l ] [ --list ] [ -r ] [ --recursive ] [ -s ] [ --swap ] [ -d ] [ --delete ] [ -v ] [ --verbose ] [ --flip[v] ] [ -j N ] [ --jump N ] [ --add N ] [ -crop X*Y[+x+y] ] [ -cut CUT ] [ -reduce N ] [ --rotate N ] [ --scale N ] [ -type TYPE ] [ -x NX ] [ --xaxis NX ] [ -y NY ] [ --yaxis NY ] [ -o directory ] -out FORMAT or [ -mar345 ] [ -pck ] [ -image ] [ -cbf ] [ -cif ] [ -raw32 ] [ -raw16 ] [ -raw8 ] [ -tiff ] [ -png ] [ -jpg ] [ -ascii ] file(s)
Additional TIFF/PNG/JPG file output options: -rainbow -blue -min MIN -max

MAX -colors N


marcvt converts image formats from marresearch/marXperts Imaging Plate Detector Systems. Supported input formats are: "image" (300 mm uncompressed images), "pck" (300 mm compressed images), "mar345" (345 mm compressed images) and images created by the marCCD CCD-detector. marcvt >= 4.0 also supports CBF/imgCIF formatted images that have been produced by mar programs (marcvt is NOT intended to be a general utility for handling CBF/imgCIF files). It is also possible to save binary images into plain ASCII-files. Other supported formats for output are images containing plain arrays with 8-bit, 16-bit or 32-bit integers, without headers or trailers. These raw binary data can easily be read by any other image processing program.

Note, that it is not possible to obtain a complete image header when converting mar300 formats into the mar345 format, since the new mar345 header store information that is not used by the mar300 headers.


--add N
Add a value of N to all pixels in file.
-cbf, -cif
Formats "cbf" an "cif" are the new "crystallographic binary file" conventions created by the CBF/imgCIF committee. "CBF" and "imgCIF" differ by the way the binary data are encoded. "CBF" uses a binary encoding while "imgCIF" is ASCII-encoded. When converting into "cbf" and "cif" files, data compression is always used. Data are stored as 32-bit integers.
--crop X*Y[+x+y]
Crop input image to X*Y pixels with offsets x and y in horizontal and vertical direction, respectively.
-cut N
Cutoff marCCD images to N*N pixels (e.g. 2000)
-d, --delete
Deletes input image after successful conversion. The default is to keep the input file.
-f, --force
Forces image output, even if a file with the same name already exists. The default is to skip output.
Flip image horizontally (vertically) on output
-h, --help
Provides a summary of the options.
-j, --jump N
Skip first N bytes of i/p-file
-l, --list
Lists the selected files (directories) only, no output.
Output images will go into directory DIRECTORY. By default, they will be in the same directory as the input images.
Input images are converted into format FORMAT. This option is compulsory! Allowed formats are: "mar345", "pck", "image", "cbf", "cif", "raw32", "raw16", "raw8", "tiff", "png", and "ascii".
-r, --recursive
If the file argument (see below) is a directory, all images in this directory and its subdirectories will be converted. This is a very fast way of converting large numbers of images scattered in many directories.
-raw32, -raw16 and -raw8
Format "raw32" is an array of 32-bit integers, "raw16" only 16-bit and "raw8" only 8-bit. Note, that values > 16-bit will be truncated to 65535 in "raw16" format. For "raw8", all intensity values will be divided by 256. Values > 16-bit will also be truncated. The "raw32"-format, however, does not truncate the data. Saturations will be marked as 999999.
-reduce N
Reduce image size by a factor of N (integer number), e.g. 5. This is helpful when making thumbnail images in png/jpg/tiff formats out of diffraction images.
--rotate N
Rotate image counter-/clockwise by N deg.
-s, --swap
Force swapping of bytes in input file
--scale N
Apply multiplication factor of N to all pixels in file.
-tiff, -png and -jpg
When producing TIFF/PNG/JPG files, data will be modified to give an image with reasonable colors. By default, the program calculates a minimum and maximum threshold from a pixel value histogram. Values < minimum will get color white, values > maximum will get color black. 256 colors shades will be distributed in between the min. and max. The colorscheme can be altered to give a rainbow spectrum (-rainbow) or to use shades of blue (-blue) instead of grey. The number of colors may be altered with option -colors N where N defaults to 256. The min. and max. threshold may be set explicitely with option -min MIN and -max MAX.
-type TYPE
Force input of type TYPE: see -out
-v, --verbose
Writes more information to stdout.
-x, --xaxis NX
No. of elements of fast varying axis in i/p-file
-y, --yaxis NY
No. of elements of slow varying axis in i/p-file
Wild cards are allowed. If no input format is given by option "-in", the corresponding file name extension is used to determine the input format.


marcvt -image x_001.mar1200
Converts x_001.mar1200 into x_001.image
marcvt -o /data -pck -r /usr/people/mar/data
Converts all images found in /usr/people/mar/data and subdirecties into pck format and writes output into directory /data.
marcvt -cbf *.mar1200
Converts all images with extension .mar1200 in the current directory into the corresponding "CBF"-style file. The extension will be .cbf1200.

See Also

mar345_formats, mar300_formats, marpack, marunpack


Claudio Klein, marXperts GmbH, Norderstedt, Germany


© Copyright 2000-2015 marXperts GmbH, Norderstedt,



marXperts GmbHPhone: +49 - (40) - 529 884-0
Werkstr. 3 FAX: +49 - (40) - 529 884-20
D-22844 Norderstedt -

Table of Contents