.\" DO NOT MODIFY THIS FILE! It was generated by help2man 1.49.3. .TH TJBENCH "1" "April 2024" "tjbench 2.1.5" "User Commands" .SH NAME tjbench \- benchmark the performance of libjpeg-turbo .SH DESCRIPTION tjbench is a JPEG decompression/compression benchmark part of libjpeg-turbo. .PP .SH USAGE .IP tjbench [options] .IP tjbench [options] .SH OPTIONS \fB\-alloc\fR = Dynamically allocate JPEG buffers \fB\-bmp\fR = Use Windows Bitmap format for output images [default = PPM] \fB\-bottomup\fR = Use bottom\-up row order for packed\-pixel source/destination buffers \fB\-tile\fR = Compress/transform the input image into separate JPEG tiles of varying .IP sizes (useful for measuring JPEG overhead) .PP \fB\-rgb\fR, \fB\-bgr\fR, \fB\-rgbx\fR, \fB\-bgrx\fR, \fB\-xbgr\fR, \fB\-xrgb\fR = .IP Use the specified pixel format for packed\-pixel source/destination buffers [default = BGR] .PP \fB\-cmyk\fR = Indirectly test YCCK JPEG compression/decompression .IP (use the CMYK pixel format for packed\-pixel source/destination buffers) .PP \fB\-fastupsample\fR = Use the fastest chrominance upsampling algorithm available \fB\-fastdct\fR = Use the fastest DCT/IDCT algorithm available \fB\-accuratedct\fR = Use the most accurate DCT/IDCT algorithm available \fB\-progressive\fR = Use progressive entropy coding in JPEG images generated by .IP compression and transform operations .PP \fB\-subsamp\fR = When compressing, use the specified level of chrominance .IP subsampling ( = 444, 422, 440, 420, 411, or GRAY) [default = test Grayscale, 4:2:0, 4:2:2, and 4:4:4 in sequence] .PP \fB\-quiet\fR = Output results in tabular rather than verbose format \fB\-yuv\fR = Compress from/decompress to intermediate planar YUV images \fB\-yuvpad\fR

= The number of bytes by which each row in each plane of an .IP intermediate YUV image is evenly divisible (must be a power of 2) [default = 1] .PP \fB\-scale\fR M/N = When decompressing, scale the width/height of the JPEG image by a .IP factor of M/N (M/N = 2/1, 15/8, 7/4, 13/8, 3/2, 11/8, 5/4, 9/8, 1/1, 7/8, 3/4, 5/8, 1/2, 3/8, 1/4, or 1/8) .PP \fB\-hflip\fR, \fB\-vflip\fR, \fB\-transpose\fR, \fB\-transverse\fR, \fB\-rot90\fR, \fB\-rot180\fR, \fB\-rot270\fR = .IP Perform the specified lossless transform operation on the input image prior to decompression (these operations are mutually exclusive) .PP \fB\-grayscale\fR = Transform the input image into a grayscale JPEG image prior to .IP decompression (can be combined with the other transform operations above) .PP \fB\-copynone\fR = Do not copy any extra markers (including EXIF and ICC profile data) .IP when transforming the input image .PP \fB\-benchtime\fR = Run each benchmark for at least seconds [default = 5.0] \fB\-warmup\fR = Run each benchmark for seconds [default = 1.0] prior to .IP starting the timer, in order to prime the caches and thus improve the consistency of the benchmark results .PP \fB\-componly\fR = Stop after running compression tests. Do not test decompression. \fB\-nowrite\fR = Do not write reference or output images (improves consistency of .IP benchmark results) .PP \fB\-limitscans\fR = Refuse to decompress or transform progressive JPEG images that .IP have an unreasonably large number of scans .PP \fB\-stoponwarning\fR = Immediately discontinue the current .IP compression/decompression/transform operation if a warning (non\-fatal error) occurs .PP NOTE: If the quality is specified as a range (e.g. 90\-100), a separate test will be performed for all quality values in the range. .SH COMMENT This manual page was written by Mathieu Malaterre for the Debian GNU/Linux system, but may be used by others. .SH COPYRIGHT Copyright (C)2009-2019, 2021 D. R. Commander. All Rights Reserved.