[gdal-dev] WKT for epsg:3857 generated by epsg_tr.py

2012-12-14 Thread Pepijn Van Eeckhoudt
When running 'epsg_tr.py -wkt 3857', gdal produces the WKT listed below. I believe the projection that is stated in the WKT is incorrect though. The combination of WGS 84 as base geogcs with Mercator_1SP as projection means that the regular ellipsoidal Mercator projection will be used which is

Re: [gdal-dev] WKT for epsg:3857 generated by epsg_tr.py

2012-12-14 Thread Even Rouault
Selon Pepijn Van Eeckhoudt pep...@vaneeckhoudt.net: When running 'epsg_tr.py -wkt 3857', gdal produces the WKT listed below. I believe the projection that is stated in the WKT is incorrect though. The combination of WGS 84 as base geogcs with Mercator_1SP as projection means that the regular

Re: [gdal-dev] WKT for epsg:3857 generated by epsg_tr.py

2012-12-14 Thread Pepijn Van Eeckhoudt
On 14-12-12 10:10, Even Rouault wrote: Selon Pepijn Van Eeckhoudt pep...@vaneeckhoudt.net: I also believe that the current WKT definition is known to be a bit hackish (is there some official/generally agreed name for the projection that should be used ?). But you are probably using not the most

Re: [gdal-dev] WKT for epsg:3857 generated by epsg_tr.py

2012-12-14 Thread Pepijn Van Eeckhoudt
On 14-12-12 10:10, Even Rouault wrote: I also believe that the current WKT definition is known to be a bit hackish (is there some official/generally agreed name for the projection that should be used ?). But you are probably using not the most recent version of GDAL, since recent versions

[gdal-dev] Can GDAL be configured to do direct I/O? (bypass cache, control blocksize, etc)

2012-12-14 Thread Ethan Alpert
I'm running in to a situation where I believe the filesystem cache settings are screwing things up. Based on iozone tests I'm not getting near the throughput I should for large 4gb+ files with multiple writing processes. It is a complex interaction of the number of dirty pages in filesystem cache

Re: [gdal-dev] Can GDAL be configured to do direct I/O? (bypass cache, control blocksize, etc)

2012-12-14 Thread Ivan Lucena
Interesting: http://www.centos.org/docs/4/html/rh-gfs-en-6.1/s1-manage-direct-io.html ---Original Message--- From: Ethan Alpert ealp...@digitalglobe.com To: gdal-dev@lists.osgeo.org Subject: [gdal-dev] Can GDAL be configured to do direct I/O? (bypass cache, control blocksize,

Re: [gdal-dev] Can GDAL be configured to do direct I/O? (bypass cache, control blocksize, etc)

2012-12-14 Thread Even Rouault
Selon Ethan Alpert ealp...@digitalglobe.com: I'm running in to a situation where I believe the filesystem cache settings are screwing things up. Based on iozone tests I'm not getting near the throughput I should for large 4gb+ files with multiple writing processes. It is a complex interaction

[gdal-dev] Working with gdalwarp -dstalpha to add transparency to a GeoTIFF

2012-12-14 Thread Alexandre Leroux
Hi list, This should simple, but I failed to make it work. I simply want a black pixels in a RGB GeoTIFF to become transparent or white. Using GDAL 1.6.3. I'm trying gdalwarp -srcnodata 0 -dstalpha in.tif out.tif I also tried (since it's RGB) gdalwarp -srcnodata 0 0 0

Re: [gdal-dev] Overviews using GDAL

2012-12-14 Thread Dmitry Baryshnikov
Hi, GDALDataset::RasterIO pass the request off to each band objects rasterio methods with appropriate arguments In GDALRasterBand::RasterIO (http://www.gdal.org/classGDALRasterBand.html#a5497e8d29e743ee9177202cb3f61c3c7) Some formats may efficiently implement decimation into a buffer by

[gdal-dev] how to gdalwarp 4-band orthoimagery

2012-12-14 Thread Smith, Michael
Seems like this is a straightforward issue but finding it difficult to solve online. I have a new crop of 4-band orthoimages, as GeoTIFFs. I want to create mosaics of some tiles. There is a slight pixel overlap at the edges. When I run gdalwarp it always treats the 4th band as alpha,

Re: [gdal-dev] how to gdalwarp 4-band orthoimagery

2012-12-14 Thread Even Rouault
Le vendredi 14 décembre 2012 19:17:07, Smith, Michael a écrit : Seems like this is a straightforward issue but finding it difficult to solve online. I have a new crop of 4-band orthoimages, as GeoTIFFs. I want to create mosaics of some tiles. There is a slight pixel overlap at the

Re: [gdal-dev] how to gdalwarp 4-band orthoimagery

2012-12-14 Thread Smith, Michael
Thanks Evan BUT Firstly in my binaries-only install I don't see any .cpp files And secondly the other method (using VRTs) how would one then automate that (to do, say 1000 tiles in a script). The editing of the VRT would be the tricky part. == Michael

Re: [gdal-dev] how to gdalwarp 4-band orthoimagery

2012-12-14 Thread Even Rouault
And secondly the other method (using VRTs) how would one then automate that (to do, say 1000 tiles in a script). The editing of the VRT would be the tricky part. sed s/ColorInterpAlpha\/ColorInterp// should do it But actually I think that if you just want to mosaic files in the same

[gdal-dev] Geosoft format?

2012-12-14 Thread Tyler Mitchell
Anyone looked at the trying to read in the geosoft.com file formats (.gdb) used in their mineral and oil/gas exploration software. They already have ERMapper and ArcGIS extensions as well as an SDK, though I don't know how deep it goes..

Re: [gdal-dev] how to gdalwarp 4-band orthoimagery

2012-12-14 Thread Ivan Lucena
gdal_translate -of VRT input.tif output.vrt -b 1 -b 2 -b 3 Mike, you can try to adapt that call to a for loop if it does what you expect. ---Original Message--- From: Smith, Michael michael.sm...@maine.gov To: gdal-dev@lists.osgeo.org gdal-dev@lists.osgeo.org Subject: Re: