dwebp - decompress a WebP file to an image file
This manual page documents the dwebp
decompresses WebP files into PNG, PAM, PPM or PGM images.
The basic options are:
- Print usage summary.
- Print the version number (as major.minor.revision) and
- -o string
- Specify the name of the output file (as PNG format by
default). Using "-" as output name will direct output to
- -- string
- Explicitly specify the input file. This option is useful if
the input file starts with an '-' for instance. This option must appear
last. Any other options afterward will be ignored. If the input
file is "-", the data will be read from stdin instead of
- Change the output format to uncompressed BMP.
- Change the output format to uncompressed TIFF.
- Change the output format to PAM (retains alpha).
- Change the output format to PPM (discards alpha).
- Change the output format to PGM. The output consists of
luma/chroma samples instead of RGB, using the IMC4 layout. This option is
mainly for verification and debugging purposes.
- Change the output format to raw YUV. The output consists of
luma/chroma-U/chroma-V samples instead of RGB, saved sequentially as
individual planes. This option is mainly for verification and debugging
- Don't use the fancy upscaler for YUV420. This may lead to
jaggy edges (especially the red ones), but should be faster.
- Don't use the in-loop filtering process even if it is
required by the bitstream. This may produce visible blocks on the
non-compliant output, but it will make the decoding faster.
- -dither strength
- Specify a dithering strength between 0 and 100.
Dithering is a post-processing effect applied to chroma components in
lossy compression. It helps by smoothing gradients and avoiding banding
- If the compressed file contains a transparency plane that
was quantized during compression, this flag will allow dithering the
reconstructed plane in order to generate smoother transparency
- Disable all dithering (default).
- Use multi-threading for decoding, if possible.
- -crop x_position y_position width height
- Crop the decoded picture to a rectangle with top-left
corner at coordinates ( x_position, y_position) and size
width x height. This cropping area must be fully contained
within the source rectangle. The top-left corner will be snapped to even
coordinates if needed. This option is meant to reduce the memory needed
for cropping large images. Note: the cropping is applied before any
- Flip decoded image vertically (can be useful for OpenGL
textures for instance).
- -resize, -scale width height
- Rescale the decoded picture to dimension width x
height. This option is mostly intended to reducing the memory
needed to decode large images, when only a small version is needed
(thumbnail, preview, etc.). Note: scaling is applied after
cropping. If either (but not both) of the width or height
parameters is 0, the value will be calculated preserving the
- Do not print anything.
- Print extra information (decoding time in particular).
- Disable all assembly optimizations.
Please report all bugs to the issue tracker: https://bugs.chromium.org/p/webp
Patches welcome! See this page to get started:
dwebp picture.webp -o output.png
dwebp picture.webp -ppm -o output.ppm
dwebp -o output.ppm -- ---picture.webp
cat picture.webp | dwebp -o - -- - > output.ppm
is a part of libwebp and was written by the WebP team.
The latest source tree is available at
This manual page was written by Pascal Massimino
<firstname.lastname@example.org>, for the Debian project (and may be used by
Please refer to http://developers.google.com/speed/webp/ for additional