Hi Markus,

Thanks for your post! With respect to your question, it’s a bit hard to diagnose without access to your file. Would that be something you could share either through this post or by opening a ticket by sending an email to support@zemax.com?

If not, no worries. A hint might be the fact that your System Efficiency is relatively equal between your nominal and tilted setups for your file. The System Efficiency is defined by the ratio of energy from the source input (modeled either by a fiber definition in the SMF Coupling tool or as your rays directly from your Object plane) to that which is able to exit through the system. It should be taking into account things like vignetting, your entrance pupil size and position, and more details. So, one thing that could be changing is something like your input ray bundle or entrance pupil size as you’re tilting your lens system.

Just to clarify, what are your settings for the system? Do you have specified an Aperture Type of Entrance Pupil Diameter, or is it set to something like Float By Stop Size? Is your STOP defined somewhere in the middle of the system, and do you have Ray Aiming turned on?

If you want to fix the input aperture of the system to be determined by your input ray bundle, you could consider defining a “dummy STOP” surface in front of the system to act as an Entrance Pupil. You should define this before any Coordinate Break surface. This would set the amount of energy that is coupled into the system entrance as determined by the size of the Entrance Pupil. Any subsequent losses would be from surface apertures vignetting the input beam bundle, reflectance losses if you’re accounting for those, etc. I’ve made a dummy test file below to illustrate this with a Paraxial lens:

The comparison to your file is a bit different since I clearly have different receiver efficiency values between the configurations, but it does demonstrate a difference in system efficiency. Let us know how these thoughts work out or if you have more questions. Thanks!