As mentionned here: the tiff reader reads the data as point data (and interpolates it) the same happens with the image reader (when reading raw data the same happens). when originally this data was voxel information, not point data. the feature request is to have a toggle on the image reader to be able to set it as cell data instead of point data.
2 Likes
Sounds perfectly reasonnable to me.
1 Like
should i open an issue?
1 Like
Rather than add a feature to multiple image readers, a better solution may be to instead provide a filter that generally converts an image to its dual by growing or shrinking its point dimensions by 1 and swapping the point data to cell data or vice versa. It would be less overall work and could be used as a better alternative to PointDataToCellData or CellDataToPointData because it wouldn’t blur the data.
1 Like
True it is another alternative, but the thing, is that the image readers are not returning the correct behavior, dont ask me how, but my data the boundaries are read correctly (i mean it has the desired voxel dimensions) but not the correct values, what you suggest would increase the memory consomption x2 as you need to read it and the conversion (and not sure if that would give the huge ram spikes that i have been having with image structured data ) furthermore, it will give a 1 less voxel in each dimension which is the kind of thing that an issue will be open a few years down the road saying why my data is smaller that it should be?