-
Notifications
You must be signed in to change notification settings - Fork 1
Transform_downsample_v31
Downsample a micrograph. For volumes use xmipp_transform_geometry. There are several downsampling methods. The most general and recommended is Fourier. Fourier downsampling puts a window in Fourier space. This is the best downsampling that can be performed. Altermatively, smoothing makes color dithering which is pretty good for visualization, but it modifies the particle spectrum. Binning with a rectangle kernel modifies the spectrum of the micrographs and is not recommended. You may see the effects of the different downsampling schemes at this article.
The downsampling factor (--step) is the factor by which the micrograph will be reduced. For instance, a downsampling by 2 will reduce the image size to one half. Using Fourier and smooth you may use non-integer downsampling factors, and the image size will be reduced by 1/factor
See also Transform_geometry_v3
Parameters
--mode <modeoverwrite> where <mode> can be:
-
- $
--label <image_labelimage> $-o, --output <output_file> $--oroot <root> $--save_metadata_stack <output_md> $: Store the original image filename in the output metadata in column imageOriginal. $: Preserve the columns from the input metadata. Some of the column values can be changed by the program.$``: Downsampling factor. factor=2 reduces the image size to one half. Fourier and smooth support non-integer downsampling factors. Rectangular binning must use integer factors. $ --method <mthfourier> where <mth> can be:
xmipp_transform_downsample -i micrograph.tif -o downsampledMicrograph.tif --step 2