Hi Asuku!

In that article, they modelled each part separately and then did a non-sequential model.

You could create the two systems in a multi-configuration file but it will take a bit of time to set up. 

Hi Asuka,

You can definitely do it either way, as Sandrine says. My personal preference is to do it as a single sequential file. Build the file out as far as the reflecting surface, and then use the Make Double Pass tool to create the reflection. Edit out any second-pass surfaces you don’t need and then carry on.

Once you have built out the whole system, if you wish you could use the Make Conjugate tool in the Multi-Configuration editor

to create configurations which represent the various sub-systems of your whole system. Remember also the merit function operand IMSF that lets you redefine the image surface on the fly, so if you wanted to optimize for image quality at the intermediate image and final image simultaneously, you could have a merit function like

IMSF {intermediate image surface}

RMS Wavefront (or whatever)

IMSF {final surface}

RMS Wavefront or whatever

If you do decide to go down the separate systems path, remember that not only is the IMA of the first system the OBJ of the second, but the exit pupil of the first system is the entrance pupil of the second system. It is also harder to optimize the common optics as they appear in two separate files with no relation to each other. That’s why I prefer to do it as a single file.

• Mark