GDAL is probably installed on your system, you need it for QGIS :) something like this should do the trick in a linux bash or from the osgeo4w embedded shells (either windows or msys ) gdalbuildvrt all_ecw.vrt *.tiff -a_srs "EPSG:yourSridHere" See http://geoserver.geo-solutions.it/edu/en/raster_data/advanced_gdal/example1.html for instance
2017-10-05 14:19 GMT+02:00 Patrick Dunford <blackwhite...@gmail.com>: > I probably don't have the technical means to use that. > > On 06/10/17 01:12, Régis Haubourg wrote: > > here it is: http://www.gdal.org/gdal_vrttut.html > > you can assemble raster of vector files, apply filters, reprojections, > build pyramids to avoid fetching the lowest resolution data. This is pretty > efficient. > > Cheers > Régis > > 2017-10-05 14:10 GMT+02:00 Patrick Dunford <blackwhite...@gmail.com>: > >> I don't know what this is. >> >> On 06/10/17 01:08, Régis Haubourg wrote: >> >> Hi Patrick , >> did you consider using GDAL VRT to avoid opening so much files? >> In my experience, this works well. >> Régis >> >> >> 2017-10-05 13:48 GMT+02:00 Patrick Dunford <blackwhite...@gmail.com>: >> >>> Some time ago in a discussion of a particular bug a contributor >>> expressed concern that the refresh of background rasters (aerial >>> photography) in Windows was too slow. >>> >>> Maybe this is the reason that recent versions of master appear to be >>> loading all of the background imagery into memory (I use a master from >>> January this year to work around issues with later ones, and that master >>> does not have this feature). >>> >>> Unfortunately if there are a lot of rasters then the memory demand is >>> excessive and unsurprisingly slows down the computer negating any purported >>> benefit of caching. >>> >>> As an example a project I am currently working on has about 900 aerial >>> photo images (GeoJpeg). When the layer is turned on for display, Qgis >>> requires about 46 GB of virtual memory. Since my computer only has 24 GB of >>> physical memory, it is required to dip into the swap space considerably. >>> Even with 60 GB of swap space on an SSD, the swapping needed to refresh the >>> canvas is substantial and dramatically reduces performance resulting in >>> substantial delays. Compare with the January master referred to above which >>> only requires about 7 GB of virtual memory total with the aerial photo >>> layer displayed. The time needed to refresh the canvas is less than 1 >>> second, most of the time. >>> >>> I know that the canvas refresh in Windows with aerial photos can be >>> substantially slower than in Linux. This does not affect me, because I >>> don't use Windows now that I have a stable platform for running an older >>> Linux master alongside the most recent one. What I do know is that the >>> memory demands are making it difficult to evaluate the recent masters. I >>> need some kind of setting to turn this caching off. With the aerial photo >>> layer turned off, the memory usage of the current master is about the same >>> as the old one, and it's much quicker to update. >>> >>> >>> _______________________________________________ >>> QGIS-Developer mailing list >>> QGIS-Developer@lists.osgeo.org >>> List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer >>> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer >> >> >> >> > >
_______________________________________________ QGIS-Developer mailing list QGIS-Developer@lists.osgeo.org List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer
