gdaladdo - Man Page

• Build overview images of a raster dataset using the "average" [r]esampling method: gdaladdo -r average path/to/input.tif

tldr.sh

gdaladdo [--help] [--help-general]
         [-r {nearest|average|rms|gauss|bilinear|cubic|cubicspline|lanczos|average_mp|average_magphase|mode}]
         [-ro] [-clean] [-q] [-oo <NAME>=<VALUE>]... [-minsize <val>]
         [--partial-refresh-from-source-timestamp]
         [--partial-refresh-from-projwin <ulx> <uly> <lrx> <lry>]
         [--partial-refresh-from-source-extent <filename1>[,<filenameN>]...]
         <filename> [<levels>]...

The gdaladdo utility can be used to build or rebuild overview images for most supported file formats with one of several downsampling algorithms.

--help

Show this help message and exit

--help-general

Gives a brief usage message for the generic GDAL commandline options and exit.

-r {nearest|average|rms|gauss|bilinear|cubic|cubicspline|lanczos|average_magphase|mode}

Select a resampling algorithm. The default is nearest, which is generally not appropriate if sub-pixel accuracy is desired.

Starting with GDAL 3.9, when refreshing existing TIFF overviews, the previously used method, as noted in the RESAMPLING metadata item of the overview, will be used if -r is not specified.

The available methods are:

nearest applies a nearest neighbour (simple sampling) resampler.

average computes the average of all non-NODATA contributing pixels. Starting with GDAL 3.1, this is a weighted average taking into account properly the weight of source pixels not contributing fully to the target pixel.

rms computes the root mean squared / quadratic mean of all non-NODATA contributing pixels (GDAL >= 3.3)

gauss applies a Gaussian kernel before computing the overview, which can lead to better results than simple averaging in e.g case of sharp edges with high contrast or noisy patterns. The advised level values should be 2, 4, 8, ... so that a 3x3 resampling Gaussian kernel is selected.

bilinear applies a bilinear convolution kernel.

cubic applies a cubic convolution kernel.

cubicspline applies a B-Spline convolution kernel.

lanczos applies a Lanczos windowed sinc convolution kernel.

average_magphase averages complex data in mag/phase space.

mode selects the value which appears most often of all the sampled points.

-b <band>

Select an input band band for overview generation. Band numbering starts from 1. Multiple -b switches may be used to select a set of input bands to generate overviews.

-ro

open the dataset in read-only mode, in order to generate external overview (for GeoTIFF especially).

-clean

remove all overviews.

-oo <NAME>=<VALUE>

Dataset open option (format specific)

-minsize <val>

Maximum width or height of the smallest overview level. Only taken into account if explicit levels are not specified. Defaults to 256.

New in version 2.3.

--partial-refresh-from-source-timestamp

New in version 3.8.

This option performs a partial refresh of existing overviews, when <filename> is a VRT file with an external overview. It checks the modification timestamp of all the sources of the VRT and regenerate the overview for areas corresponding to sources whose timestamp is more recent than the external overview of the VRT. By default all existing overview levels will be refreshed, unless explicit levels are specified.

--partial-refresh-from-projwin <ulx> <uly> <lrx> <lry>

New in version 3.8.

This option performs a partial refresh of existing overviews, in the region of interest specified by georeference coordinates where <ulx> is the X value of the upper left corner, <uly> is the Y value of the upper left corner, <lrx> is the X value of the lower right corner and <lry> is the Y value of the lower right corner. By default all existing overview levels will be refreshed, unless explicit levels are specified.

--partial-refresh-from-source-extent <filename1>[,<filenameN>]...

New in version 3.8.

This option performs a partial refresh of existing overviews, in the region of interest specified by one or several filenames (names separated by comma). Note that the filenames are only used to determine the regions of interest to refresh. The reference source pixels are the one of the main dataset. By default all existing overview levels will be refreshed, unless explicit levels are specified.

<filename>

The file to build overviews for (or whose overviews must be removed).

<levels>

A list of integral overview levels to build. Ignored with -clean option.

New in version 2.3: Levels are no longer required to build overviews. In which case, appropriate overview power-of-two factors will be selected until the smallest overview is smaller than the value of the -minsize switch.

Starting with GDAL 3.9, if there are already existing overviews, the corresponding levels will be used to refresh them if no explicit levels are specified.

gdaladdo will honour properly NODATA_VALUES tuples (special dataset metadata) so that only a given RGB triplet (in case of a RGB image) will be considered as the nodata value and not each value of the triplet independently per band.

Selecting a level value like 2 causes an overview level that is 1/2 the resolution (in each dimension) of the base layer to be computed.  If the file has existing overview levels at a level selected, those levels will be recomputed and rewritten in place.

For internal GeoTIFF overviews (or external overviews in GeoTIFF format), note that -clean does not shrink the file. A later run of gdaladdo with overview levels will cause the file to be expanded, rather than reusing the space of the previously deleted overviews. If you just want to change the resampling method on a file that already has overviews computed, you don't need to clean the existing overviews.

Some format drivers do not support overviews at all.  Many format drivers store overviews in a secondary file with the extension .ovr that is actually in TIFF format.  By default, the GeoTIFF driver stores overviews internally to the file operated on (if it is writable), unless the -ro flag is specified.

Most drivers also support an alternate overview format using Erdas Imagine format.  To trigger this use the USE_RRD=YES configuration option.  This will place the overviews in an associated .aux file suitable for direct use with Imagine or ArcGIS as well as GDAL applications.  (e.g. --config USE_RRD YES)

External overviews created in TIFF format may be compressed using the COMPRESS_OVERVIEW configuration option.  All compression methods, supported by the GeoTIFF driver, are available here. (e.g. --config COMPRESS_OVERVIEW DEFLATE). The photometric interpretation can be set with the PHOTOMETRIC_OVERVIEW =RGB/YCBCR/... configuration option, and the interleaving with the INTERLEAVE_OVERVIEW =PIXEL/BAND configuration option.

Since GDAL 3.6, COMPRESS_OVERVIEW and INTERLEAVE_OVERVIEW are honoured when creating internal overviews of TIFF files.

For JPEG compressed external and internal overviews, the JPEG quality can be set with --config JPEG_QUALITY_OVERVIEW value.

For WEBP compressed external and internal overviews, the WEBP quality level can be set with --config WEBP_LEVEL_OVERVIEW value. If not set, will default to 75.

For WEBP compressed external and internal overviews, the WEBP lossless/lossy switch can be set with --config WEBP_LOSSLESS_OVERVIEW value. If not set, will default to NO (lossy). Added in GDAL 3.6.0

For LERC compressed external and internal overviews, the max error threshold can be set with --config MAX_Z_ERROR_OVERVIEW value. If not set, will default to 0 (lossless). Added in GDAL 3.4.1

For DEFLATE or LERC_DEFLATE compressed external and internal overviews, the compression level can be set with --config ZLEVEL_OVERVIEW value. If not set, will default to 6. Added in GDAL 3.4.1

For ZSTD or LERC_ZSTD compressed external and internal overviews, the compression level can be set with --config ZSTD_LEVEL_OVERVIEW value. If not set, will default to 9. Added in GDAL 3.4.1

For JPEG-XL compressed external and internal overviews, the following settings can be set since GDAL 3.9.0:

• Whether compression should be lossless with --config JXL_LOSSLESS_OVERVIEW YES|NO. Default is YES
• Level of effort with --config JXL_EFFORT_OVERVIEW value, with value between 1(fast) and 9(flow). Default is 5
• Distance level for lossy compression with --config JXL_DISTANCE_OVERVIEW value, with value: 0=mathematically lossless, 1.0=visually lossless, usual range [0.5,3]. Default is 1.0. Ignored if JXL_LOSSLESS_OVERVIEW is YES
• Distance level for lossy compression of alpha channel with --config JXL_ALPHA_DISTANCE_OVERVIEW value, with value: 0=mathematically lossless, 1.0=visually lossless, usual range [0.5,3]. Default is the same value as JXL_DISTANCE_OVERVIEW. Ignored if JXL_LOSSLESS_OVERVIEW is YES

For LZW, ZSTD or DEFLATE compressed external overviews, the predictor value can be set with --config PREDICTOR_OVERVIEW 1|2|3.

To produce the smallest possible JPEG-In-TIFF overviews, you should use:

--config COMPRESS_OVERVIEW JPEG --config PHOTOMETRIC_OVERVIEW YCBCR --config INTERLEAVE_OVERVIEW PIXEL

External overviews can be created in the BigTIFF format by using the BIGTIFF_OVERVIEW configuration option: --config BIGTIFF_OVERVIEW {IF_NEEDED|IF_SAFER|YES|NO}.

The default value is IF_SAFER starting with GDAL 2.3.0 (previously was IF_NEEDED). The behavior of this option is exactly the same as the BIGTIFF creation option documented in the GeoTIFF driver documentation.

• YES forces BigTIFF.
• NO forces classic TIFF.
• IF_NEEDED will only create a BigTIFF if it is clearly needed (uncompressed, and overviews larger than 4GB).
• IF_SAFER will create BigTIFF if the resulting file might exceed 4GB.

Sparse GeoTIFF overview files (that is tiles which are omitted if all their pixels are at the nodata value, when there's one, or at 0 otherwise) can be obtained with --config SPARSE_OK_OVERVIEW ON. Added in GDAL 3.4.1

See the documentation of the GTiff -- GeoTIFF File Format driver for further explanations on all those options.

--config GDAL_TIFF_OVR_BLOCKSIZE <size>

Example: --config GDAL_TIFF_OVR_BLOCKSIZE 256

Default value is 128, or starting with GDAL 3.1, if creating overviews on a tiled GeoTIFF file, the tile size of the full resolution image. Note: without this setting, the file can have the full resolution image with a blocksize different from overviews blocksize.(e.g. full resolution image at blocksize 256, overviews at blocksize 128)

Invalid values in source pixels, either identified through a nodata value metadata set on the source band, a mask band, an alpha band will not be used during resampling.

The details of how it is taken into account depends on the resampling kernel:

• for nearest resampling, for each target pixel, one of the potential contributing source pixels is selected (in an implementation specific way). Its value is used as it, be it valid or invalid.
• for bilinear, cubic, cubicspline and lanczos, for each target pixel, the weights of contributing source pixels is set to zero to ignore them when they are masked. There is an extra specificity for cubic: given that some of the weights in the kernel are negative, such strategy could lead to strong overshoot/undershoot when there is an alternance of valid and invalid pixels. Consequently, if any of the horizontal or vertical direction, if the maximum number of valid source pixels in each dimension is less than the radius of the resampling kernel, the target pixel is considered as nodata.
• for the other resampling methods, source pixels contributing to the target pixel are ignored if masked. Only the valid ones are taken into account. If there are none, the target pixel is considered as nodata.

New in version 3.2.

The GDAL_NUM_THREADS configuration option can be set to ALL_CPUS or a integer value to specify the number of threads to use for overview computation.

Functionality of this utility can be done from C with GDALBuildOverviews().

Create overviews, embedded in the supplied TIFF file, with automatic computation of levels (GDAL 2.3 or later)

gdaladdo -r average abc.tif

Create overviews, embedded in the supplied TIFF file:

gdaladdo -r average abc.tif 2 4 8 16

Create an external compressed GeoTIFF overview file from the ERDAS .IMG file:

gdaladdo -ro --config COMPRESS_OVERVIEW DEFLATE erdas.img 2 4 8 16

Create an external JPEG-compressed GeoTIFF overview file from a 3-band RGB dataset (if the dataset is a writable GeoTIFF, you also need to add the -ro option to force the generation of external overview):

gdaladdo --config COMPRESS_OVERVIEW JPEG --config PHOTOMETRIC_OVERVIEW YCBCR
         --config INTERLEAVE_OVERVIEW PIXEL rgb_dataset.ext 2 4 8 16

Create an Erdas Imagine format overviews for the indicated JPEG file:

gdaladdo --config USE_RRD YES airphoto.jpg 3 9 27 81

Create overviews for a specific subdataset, like for example one of potentially many raster layers in a GeoPackage (the "filename" parameter must be driver prefix, filename and subdataset name, like e.g. shown by gdalinfo):

gdaladdo GPKG:file.gpkg:layer

Refresh overviews of a VRT file, for sources that have been modified after the .vrt.ovr generation:

gdalbuildvrt my.vrt tile1.tif tile2.tif                          # create VRT
gdaladdo -r cubic my.vrt                                         # initial overview generation
touch tile1.tif                                                  # simulate update of one of the source tiles
gdaladdo --partial-refresh-from-source-timestamp -r cubic my.vrt # refresh overviews

Refresh overviews of a TIFF file:

gdalwarp -overwrite tile1.tif tile2.tif mosaic.tif                      # create mosaic
gdaladdo -r cubic mosaic.tif                                            # initial overview generation
touch tile1.tif                                                         # simulate update of one of the source tiles
gdalwarp tile1.tif mosaic.tif                                           # update mosaic
gdaladdo --partial-refresh-from-source-extent tile1.tif -r cubic my.vrt # refresh overviews

Frank Warmerdam <warmerdam@pobox.com>, Silke Reimer <silke@intevation.de>

1998-2024