I’m trying to estimate the angle of the tilt axis relative to the image axes (not the tilt angle!) based on the CTF estimation outputs. I’m digging through the data, and the only column that seems to be related is ‘ctf_stats/df_tilt_normal’. Is that what I need? If so, how can I interpret it? It’s 2 numbers, not normalized that I can see, and seemingly very variable…

Hi @brisvag,

Apologies for the delayed response. If you’re looking for the physical tilt of the micrograph (I assume not “beam tilt” or “coma” aberration, which is a distinct phenomenon), this can be estimated from the defocus landscape fitted during Patch CTF estimation. You’re correct that `ctf_stats/df_tilt_normal`

contains information about the tilt; for a given micrograph, `[df_tilt_normal[0], df_tilt_normal[1], -1]`

is a normal vector to the plane in x, y, z coordinates.

The cosine of the tilt angle can then be found by normalizing this vector to unit norm, then taking the dot product with a unit normal in the -z direction (i.e. [0, 0, -1]). Let me know if this makes sense – I can provide additional clarification if it’s unclear

best,

Michael