I think that’s what the default merit functions do (Optimization Wizard). They are a series of merit functions that can be used from any point in the design process, from initial flat-surfaces-only through to final design and tolerancing. Does that help, or are you thinking of something else?

Thank you for a very straight forward answer Mark, and yes, I totally agree that the Optimization Wizard give the default and basic merit functions. Which is of course very useful btw for beginners, such as myself, designing simple lens system.

I’ve seen merit functions that are very sophisticated, with multiple operands, depending on the optical system and problem being solved. Lens designers would keep this and perhaps share this but limited within the company or team they’re working with. I know this will be unique for every lens design but it would be very helpful for new lens designers to understand what combination of operands works in order to solve a particular problem. For example, I would like to design an automotive camera lens with a custom defined distortion / field of view with high MTF, and at the same time controlling CRA, chromatics aberrations, RI and its corresponding lens dimension. This problem cannot be simply solved using Optimization Wizard functions so I’m wondering if there are sample set of merit functions that we can use as reference to solve a particular problem. It may be spoon feeding but it will greatly help somebody learning lens design on his own. 

Hi Jay,

Interesting questions. The default MF are not for just beginners, or simple systems though. They target the basic image quality of the system (RMS spot, RMS wavefront, Contrast at a specified frequency) and can be used for ANY system, whether close to a good solution or far from it. They are the bedrock of getting good Image quality.

Then, you can also add other operands to tailor the design further. A smartphone lens would still have a bedrock spot or wavefront image quality, but you’d then add in other operands for chief ray angle, image size, distortion, RI etc.

Its often the case that you do a two-step optimization, the first getting a system to be ‘close enough’ and then the second, once the system is close enough, can then tailor the design to a specific set of conditions.

• Mark

Hello Mark,

Thank you very much for the answer. This is the kind of guidance that I need. But I guess the challenging part would be to understand how to use the operands and combining the appropriate ones. The operand descriptions is helpful though but it seems it requires experience to fully grasp the techniques in using them. I’ve undergone Zemax training and have been trying out simple optimization process but there are problems that are really difficult to solve on my own. 

Jay