Based on [SPICE documentation](https://pds.nasa.gov/data/go-j-pws-2-redr-lpw-sa-full-v1.0/gomw_5001/document/spice/ck.htm), C-matrix (camera matrix) transforms the coordinates of a point in a reference frame (like J2000) into the instrument fixed coordinates: ```text [ x_inst] [ ] [ x_J2000 ] | y_inst| = | C-matrix | | y_J2000 | [ z_inst] [ ] [ z_J2000 ] ``` The transpose of a C-matrix rotates vectors from the instrument-fixed frame to the base frame: ```text [ x_J2000] [ ]T [ x_inst ] | y_J2000| = | C-matrix | | y_inst | [ z_J2000] [ ] [ z_inst ] ``` Currently, the `planetary_coverage.spice.toolbox.attitude()` is computing the transpose of the C-matrix and not the C-matrix: ```python sp.pxform(str(observer.frame), ref, time) ``` whereas it should be: ```python sp.pxform(ref, str(observer.frame), time) ``` A better docstring should describe this matrix with the example above. Currently, the trajectory objects provide a `sc_attitude` property base on `planetary_coverage.spice.toolbox.attitude()`. It should be updated to take the transpose of the results. Additionally, this `sc_attitude` is computed for the observer (spacecraft or instrument) frame. This is not consistent with the `sc_` prefix. We recommend to depreciate this property (with a warning) and provide a generic `attitude` property that takes into account the nature of the observer. This will also affect the `radec` values that are also observer specific. /cc @lpenasa