[Gdal-dev] Strange GeoTIFF error/gdal_merge.py

Frank Warmerdam warmerdam at pobox.com
Thu Mar 18 12:39:36 EST 2004


Markus Neteler wrote:
> Hello Manuel,
> 
> On Thu, Mar 18, 2004 at 05:23:14PM +0100, Manuel Massing wrote:
> 
>>Hello Markus,
>>
>>I had the same problem when creating a TIFF file and
>>merging datasets which did not cover the entire domain. The problem 
>>seems to stem from uninitialised sections of the output file,
>>so a workaround is to first initialise the whole dataset with e.g. 
>>nodata or zero values, and then proceed with merging.
>>
>>I am not sure if this occurs only with tiled datasets, but it might
>>be an index initialisation bug of some sort? 
> 
> 
> In fact there are no data areas in the final mosaic as the
> tiles do not cover the entire region. So maybe the problem
> is as you describe (unfortunately I have no idea how to change
> gdal_merge.py to first initialise the whole dataset).

Markus,

It had been my intention that GDAL would gracefully handle strips or
tiles that have an offset and size of zero as "auto-nodata".  Apparently
I did not complete that properly.   I will look into this.  Currently
preinitializing to zero should resolve the problem but I realize that is
not so easy to do.

Best regards,
-- 
---------------------------------------+--------------------------------------
I set the clouds in motion - turn up   | Frank Warmerdam, warmerdam at pobox.com
light and sound - activate the windows | http://pobox.com/~warmerdam
and watch the world go round - Rush    | Geospatial Programmer for Rent




More information about the Gdal-dev mailing list