.\" Hey, EMACS: -*- nroff -*- .TH SJPEG 1 "Feb 26, 2018" .SH NAME sjpeg \- compress a PNG, JPEG or PPM image file to a baseline JPEG file .SH SYNOPSIS .B sjpeg .RI [ options ] " input_file \-o output_file.jpg .br .SH DESCRIPTION This manual page documents the .B sjpeg command. .PP \fBsjpeg\fP compresses a PNG, JPEG or PPM image or recompresses a JPEG image, using the baseline JPEG format. .SH OPTIONS The basic options are: .TP .BI \-o " string Specify the name of the output JPEG file. .TP .B \-h, \-help A summary of all the possible options. .TP .B \-version Print the version number (as major.minor.revision) and exit. .TP .BI \-q " float Specify the compression factor between 0 and 100. The default is 75. A small factor produces a smaller file with lower quality. Best quality is achieved by using a value of 100. If the input source is a JPEG file, the compression factor will be limited to never exceed the source's quality unless the -no_limit option is used. This option is disabled by subsequent use (in parsing order) of the \-r option. .TP .BI \-r " float Specify the reduction factor for JPEG source. The default is 100 (no reduction). A small reduction factor produces a smaller file with lower quality. When using a value of 100, the source JPEG file is only recompressed with closed to no quality loss (but file size will likely be smaller). This option is disabled by subsequent use (in parsing order) of the \-q option. .TP .BI \-size " int Specify a target size to try and reach using several passes. By default, the internal search loop will try to get as close as possible to the specified size. This option has precedence over the \-pnsr one. See also the \-pass option. .TP .BI \-psnr " float Specify a target PSNR to try and reach using several passes. By default, the internal search loop will try to get as close as possible to the specified distortion. Note that the Peak-Signal-to-Noise is computed using the YUV samples, not the RGB ones. See also the \-pass option. .TP .BI \-tolerance " float Tolerance (in percent) used during convergence to target value when \-psnr or \-size option is used. .TP .BI \-qmin " int Minimum acceptable quality factor used during the search (when \-psnr or \-size option is used). The search will halt if the quality adjustment would drop lower than the \fBqmin\fP value. .TP .BI \-qmax " int Maximum acceptable quality factor used during the search (when \-psnr or \-size option is used). The search will avoid using a quality factor larger than the \fBqmax\fP value. .TP .BI \-pass " int Specify the number of internal loop to use for convergence toward a target size or distortion. If none is specified but \-size or \-psnr option is used, then a default value of \fB10\fP is used. .TP .B \-yuv_mode " int Specify the YUV color space method to use. Possible values are .IP 0: automatic decision (default) .IP 1: use YUV 4:2:0 downsampling .IP 2: use 'Sharp' YUV 4:2:0 downsampling - slower but more accurate .IP 3: use YUV 4:4:4 (full resolution for U/V chroma planes) .TP .B \-444 Shortcut for \-yuv_mode 3 .TP .B \-sharp Shortcut for \-yuv_mode 2 .TP .B \-420 Shortcut for \-yuv_mode 1 .TP .B \-estimate If the source is a JPEG file, print the estimated quality and exit. .TP .B \-i, \-info Print some information about the input file and exit. .TP .B \-quiet Disable information message (error messages are still printed). .TP .B \-short Only display short information about the output result. .TP .B \-crc Only display the CRC32 checksum of the output and exit. .TP .B \-no_limit Disable the quantization limit (in case of JPEG input file) allowing to recompress the source at a higher quality factor (result is uncertain, though!). .TP .B \-no_adapt Disable adaptive quantization optimization (faster processing, larger file) .TP .B \-trellis Enable trellis-based quantization (slower processing, but produces file optimized for rate-distortion) .TP .B \-no_optim Disable Huffman code optimization (faster processing, larger file) .TP .B \-adapt_bias Disable the adaptive bias optimization. .TP .B \-no_metadata Disable transfer of metadata from source to JPEG (saves file size, but discards information). .SH BUGS Please report any bugs to the SJPEG discussion list: https://groups.google.com/forum/#!forum/sjpeg The issue tracker is located at: https://github.com/webmproject/sjpeg/issues .br Patches welcome! See this page to get started: http://www.webmproject.org/code/contribute/submitting-patches/ .SH EXAMPLES sjpeg \-q 50 picture.png \-o picture.jpg .br sjpeg picture.jpg \-r 90 \-o optimized_jpeg.jpg .SH AUTHORS \fBsjpeg\fP was written by Pascal Massimino. .br The latest source tree is available at http://www.webmproject.org/code .PP This manual page was written by Pascal Massimino , for the Debian project (and may be used by others). .SH SEE ALSO .BR cjpeg (1), djpeg(1) .br .SS Input file format details PPM: http://netpbm.sourceforge.net/doc/ppm.html