--- name: "enblend" section: "1" platform: "linux" package: "enblend-enfuse" version: "4.2.r1546+hb66125418f7d-5" release: "Arch Linux 2026-11-01" platforms: ["linux"] url: "https://new.carta.tech/enblend" --- # enblend(1) manual page for enblend 4.2.1-0bcfaaf743e7 ## NAME enblend - manual page for enblend 4.2.1-0bcfaaf743e7 ## SYNOPSIS **enblend** \[*options*\] \[*--output=IMAGE*\] *INPUT*... ## DESCRIPTION Blend INPUT images into a single IMAGE. INPUT... are image filenames or response filenames. Response filenames start with an "@" character. ## OPTIONS ### Common options: - **-l**, **--levels**=*LEVELS* limit number of blending LEVELS to use (1 to 29); negative number of LEVELS decreases maximum; "auto" restores the default automatic maximization - **-o**, **--output**=*FILE* write output to FILE; default: "a.tif" - **-v**, **--verbose**\[=*LEVEL*\] verbosely report progress; repeat to increase verbosity or directly set to LEVEL - **--compression**=*COMPRESSION* set compression of output image to COMPRESSION, where COMPRESSION is: "deflate", "jpeg", "lzw", "none", "packbits", for TIFF files and 0 to 100, or "jpeg", "jpeg-arith" for JPEG files, where "jpeg" and "jpeg-arith" accept a compression level ### Advanced options: - **--blend-colorspace**=*COLORSPACE* force COLORSPACE for blending operations; Enblend uses "CIELUV" for images with ICC-profile and "IDENTITY" for those without and also for all floating-point images; other available blend color spaces are "CIELAB" and "CIECAM" - **-c**, **--ciecam** use CIECAM02 to blend colors; disable with "--no-ciecam"; note that this option will be withdrawn in favor of "--blend-colorspace" - **-d**, **--depth**=*DEPTH* set the number of bits per channel of the output image, where DEPTH is "8", "16", "32", "r32", or "r64" - **-f** WIDTHxHEIGHT\[+xXOFFSET+yYOFFSET\] manually set the size and position of the output image; useful for cropped and shifted input TIFF images, such as those produced by Nona - **-g** associated-alpha hack for Gimp (before version 2) and Cinepaint - **-w**, **--wrap**\[=*MODE*\] wrap around image boundary, where MODE is "none", "horizontal", "vertical", or "both"; default: none; without argument the option selects horizontal wrapping ### Mask generation options: - **--coarse-mask**\[=*FACTOR*\] shrink overlap regions by FACTOR to speedup mask generation; this is the default; if omitted FACTOR defaults to 8 - **--fine-mask** generate mask at full image resolution; use e.g. if overlap regions are very narrow - **--optimize** turn on mask optimization; this is the default; disable with "--no-optimize" - **--save-masks**\[=*TEMPLATE*\] save generated masks in TEMPLATE; default: "mask-%n.tif"; conversion chars: "%i": mask index, "%n": mask number, "%p": full path, "%d": dirname, "%b": basename, "%f": filename, "%e": extension; lowercase characters refer to input images uppercase to the output image - **--load-masks**\[=*TEMPLATE*\] use existing masks in TEMPLATE instead of generating them; same template characters as "--save-masks"; default: "mask-%n.tif" - **--visualize**\[=*TEMPLATE*\] save results of optimizer in TEMPLATE; same template characters as "--save-masks"; default: "vis-%n.tif" ### Expert options: - **-a**, **--pre-assemble** pre-assemble non-overlapping images; negate with "--no-pre-assemble" - **-x** checkpoint partial results - **--fallback-profile**=*PROFILE-FILE* use the ICC profile from PROFILE-FILE instead of sRGB - **--layer-selector**=*ALGORITHM* set the layer selector ALGORITHM; default: "all-layers"; available algorithms are: "all-layers": select all layers in any image; "first-layer": select only first layer in each (multi-)layer image; "last-layer": select only last layer in each (multi-)layer image; "largest-layer": select largest layer in each (multi-)layer image; "no-layer": do not select any layer from any image; - **--parameter**=*KEY1\[=VALUE1\]\[*:KEY2\[=VALUE2\]\[:...\]\] set one or more KEY-VALUE pairs ### Expert mask generation options: - **--primary-seam-generator**=*ALGORITHM* use main seam finder ALGORITHM, where ALGORITHM is "nearest-feature-transform" or "graph-cut"; default: "graph-cut" - **--image-difference**=*ALGORITHM\[*:LUMINANCE-WEIGHT\[:CHROMINANCE-WEIGHT\]\] use ALGORITHM for calculation of the difference image, where ALGORITHM is "max-hue-luminance" or "delta-e"; LUMINANCE-WEIGHT and CHROMINANCE-WEIGHT define the weights of lightness and color; default: delta-e:1:1 - **--optimizer-weights**=*DISTANCE-WEIGHT\[*:MISMATCH-WEIGHT\] set the optimizer's weigths for distance and mismatch; default: 8:1 - **--mask-vectorize**=*LENGTH* set LENGTH of single seam segment; append "%" for relative value; defaults: 4 for coarse masks and 20 for fine masks - **--anneal**=*TAU\[*:DELTAE-MAX\[:DELTAE-MIN\[:K-MAX\]\]\] set annealing parameters of optimizer strategy 1; defaults: 0.75:7000:5:32 - **--dijkstra**=*RADIUS* set search RADIUS of optimizer strategy 2; default: 25 pixels ### Information options: - **-h**, **--help** print this help message and exit - **-V**, **--version** output version information and exit - **--show-globbing-algorithms** show all globbing algorithms - **--show-image-formats** show all recognized image formats and their filename extensions - **--show-signature** show who compiled the binary when and on which machine - **--show-software-components** show the software components with which Enblend was compiled Enblend accepts arguments to any option in uppercase as well as in lowercase letters. ## ENVIRONMENT - OMP\_NUM\_THREADS The OMP\_NUM\_THREADS environment variable sets the number of threads to use in OpenMP parallel regions. If unset Enblend uses as many threads as there are CPUs. - OMP\_DYNAMIC The OMP\_DYNAMIC environment variable controls dynamic adjustment of the number of threads to use in executing OpenMP parallel regions. ## AUTHOR Written by Andrew Mihal, Christoph Spiel and others. ## REPORTING BUGS Report bugs at \. ## COPYRIGHT Copyright © 2004-2009 Andrew Mihal.\ Copyright © 2009-2017 Christoph Spiel. License GPLv2+: GNU GPL version 2 or later \\ This is free software: you are free to change and redistribute it. There is NO WARRANTY, to the extent permitted by law.