I have found that the pixel type resulted from running a raster related tool on ArcToolBox on ArcGIS 10.5 was different from that from running exactly the same tool on ArcGIS 10.1 and ArcGIS 10.2.

For example, the "Feature to Raster" conversion tool on ArcGIS 10.5 generates a raster with pixel type of unsigned integer with a pixel depth of 1 Bit while the same tool on ArcGIS 10.1 achieves a raster output with pixel type of unsigned integer with a pixel depth of 8 Bit.

How does that happen given with the same input feature class? Is there a way to configure the pixel type and depth as a default setting on ArcGIS 10.5?

  • Is this causing any problems? Feature to raster chooses the raster pixel type based on the chosen field and the min/max values contained. For a 1 bit image only values of 0 or 1 should be present but an 8 bit can contain values between 0 and 255 though it's not unusual to find a raster generated in 8 bit containing values of only 0 and 1; you must of course take into account the limitations of the raster format, some formats do not support true 1 bit and store 0/1 in a byte desktop.arcgis.com/en/arcmap/10.3/manage-data/raster-and-images/… – Michael Stimson Jun 7 '17 at 3:33
  • Where can we configure the default pixel type and depth values for raster calculation on ArcGIS? – alextc Jun 7 '17 at 4:08
  • Related: gis.stackexchange.com/q/85657/8104 – Aaron Jun 7 '17 at 4:22
  • 1
    I have seen much confusion and experienced a great deal of frustration on this subject. There is nowhere that I've found to configure the default raster pixel type. If I end up with a pixel type that isn't suitable I convert the raster to a suitable raster with Copy Raster help.arcgis.com/en/arcgisdesktop/10.0/help/index.html#//… where you can configure the pixel type provided the pixel type is supported by ArcGIS and the raster format. – Michael Stimson Jun 7 '17 at 4:22
  • The field has the <Null> value in the input feature class. That is expected. This generated an empty raster with the pixel type of 1 bit unsigned integer on 10.5 while an empty raster with the pixel type of 8 bit unsigned integer on 10.1 even if the input feature classes are the same. Very weird. – alextc Jun 9 '17 at 0:39

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.