Hi Tobias, On Thu, 7 Feb 2013, Tobias Pietzsch wrote:
> Ok, I tried the things you suggested as well as the setGroupFiles( false > ) suggested by Lee. Here are the results: > > baseline (unmodified ImgOpener): > loading 50 tif images using ImageJ, 10000 other tif files in same directory > median: 141 ms > loading 50 tif images using ImgOpener, 10000 other tif files in same directory > median: 6953 ms > > with setAllowOpenFiles( false ): > loading 50 tif images using ImgOpener, 10000 other tif files in same directory > median: 5972 ms > > with setGroupFiles( false ): > loading 50 tif images using ImgOpener, 10000 other tif files in same directory > median: 2331 ms > > with both, setAllowOpenFiles( false ) and setGroupFiles( false ): > loading 50 tif images using ImgOpener, 10000 other tif files in same directory > median: 1070 ms > > with TiffReader: > loading 50 tif images using ImgOpener, 10000 other tif files in same directory > median: 5061 ms > > with TiffReader and setGroupFiles( false ): > loading 50 tif images using ImgOpener, 10000 other tif files in same directory > median: 568 ms > > > > > For comparison, the case where there are no other files in the directory > baseline (unmodified ImgOpener): > loading 50 tif images using ImageJ, 0 other tif files in same directory > median: 140 ms > loading 50 tif images using ImgOpener, 0 other tif files in same directory > median: 1804 ms > > > with TiffReader and setGroupFiles( false ): > loading 50 tif images using ImgOpener, 0 other tif files in same directory > median: 552 ms > > That's still 4x slower than ImageJ1 but already getting a lot closer. Nice! Thanks for the detailed comparison. I will add code to perform these timing tests more automatically and try to get at the root of the problem as I indicated in a private conversation previously (using OProfile). Ciao, Dscho _______________________________________________ ImageJ-devel mailing list [email protected] http://imagej.net/mailman/listinfo/imagej-devel
