Hi,

Without seeing the Merit Function, it is difficult to know what is happening. Would you be able to share your file? You can do so by pressing File..Create Archive.. and putting the resulting *.ZAR file into a *.ZIP and upload it in your reply.

When you did your manual adjustment, did both field got a smaller spot radius? By default, the fields are given equal weight, and there’s a compromise to be striked across the fields. It is possible to get a smaller spot in one field, but it would typically worsen the other one.

Take care,

David

Thank you for your reply

Sorry, the source file cannot be sent due to confidentiality。

The following is Merit Function

When I use the above Merit Function to optimize, the results obtained are as follows

There is only one variable, and that is the thickness of the 9th surface。

You might try increasing the rings and regenerate the merit function. 

I found the reason。

I added a circular obscuration in front of the lens.The software doesn't seem to consider this circular obscuration when optimizing. Therefore, the image surface position found by the software  is the best position when there is no circular obscuration limit.

When I remove this circular obscuration, the optimization result is normal。

Good job, if you want to keep the obscuration, you can also run an optimization with a Rectangular Array of rays as opposed to the Gaussian Quadrature. This gives you the opportunity to check Delete Vignetted to suppress the rays that would be terminated at the obscuration. However, it might necessitate more rays to get a good sampling.

Take care,

David

OpticStudio supports central obscurations natively, because they are so common in optical systems. In the Optimization Wizard, use the Obscuration control:

Just enter the radius of the obscuration in lens units. The docs give details. 

Also, be aware that the Zernike Annular feature exists to compute the Zernike coefficients in the presence of a central obscuration: