Changes between Version 8 and Version 10 of Ticket #5183


Ignore:
Timestamp:
Jul 1, 2014, 6:40:29 AM (10 years ago)
Author:
swalter75
Comment:

Legend:

Unmodified
Added
Removed
Modified

  • Ticket #5183

    • Property Priority normal → high

  • Ticket #5183 – Description

    v8 v10  
    1 When converting Int16 data to JP2KAK with the -co "QUALITY=100" option, strange artifacts are created in the nodata values and in some cases also in the image data (not in the example below). Using the conversion without the lossless option everything works as expected.
     1When converting Int16 data to JP2KAK with the -co "QUALITY=100" option, scrambled artifacts are created in the no_data values and in some cases also in the image data. Using the conversion without the lossless option everything works as expected.
    22
    3 For example, the problem can be inspected with this xample data (Mars elevation):
     3For example, the problem can be inspected with this example data (Mars elevation):
    44
    5 https://triton.planet.fu-berlin.de/h0533_0000.dt4.53.tif
     5http://maps.planet.fu-berlin.de/2956/h2956_0000.da4.50.tif
    66
    77The following command gives strange results:
    88
    9 gdal_translate -ot Int16 -of JP2KAK -co "QUALITY=100" h0533_0000.dt4.53.tif h0533_0000.dt4.53.jp2
     9gdal_translate -ot Int16 -of JP2KAK -co "QUALITY=100" h2956_0000.da4.50.tif h2956_0000.da4.50.jp2
    1010
    11 Here is a link to the resulting problematic image:
    12 
    13 https://triton.planet.fu-berlin.de/h0533_0000.dt4.53.jp2
    14 
    15 This command gives expected results:
    16 
    17 gdal_translate -ot Int16 -of JP2KAK h0533_0000.dt4.53.tif h0533_0000.dt4.53.jp2
    18 
    19 The problem has already been described in this thread, but never been addresses (as far as I know):
    20 
    21 http://lists.osgeo.org/pipermail/gdal-dev/2011-February/027900.html
    22 
     11These problems occur not only in the no_data areas, but also in the valid data borders close to the no_data.