Hi Steven,

Interesting quesiton. I don’t think that there’s a feature readily available to calculate the depth of field. You could calculate it based on the NA of your lens (using a formula such as this one).

You can also make use of the Analyze..Rays & Spots..Through Focus Spot Diagram, which displays a sequence of spot diagram taken at different axial positions around the focus.

Ultimately, you could also leverage the ZOS-API to produce a 3D stack of PSFs across focus, which would probably give you the most accurate depiction of your depth of field in my opinion (but could be an overkill).

I guess the problem is a little bit the same as for the resolution. You won’t find an analysis feature called resolution because it doesn’t have a strict definition. People use different criterion, or sometimes even make up their own. What is your definition of depth of field?

Take care,

David

Hi all - this was originally posted to the Chinese forums. However, since this post is in English, I am moving it to the English forums. @steven.Ou - I noticed you are located in China. Feel free to post in either English or Chinese. We have folks in all regions monitoring the Community for questions in any supported language . 

Thanks for your enlightenment, David!

Hello Steven,

I think you could possibly use the Universal Plot facility located in the Analyze part. In the X data, you could fill the range of object shift ( thickness of the object ) and in the Y you could use a merit data such as TRCX defined in a merit function. The data are available in the same facility and you could iterate with several examples and plot them into another framework for comparison. The advantage of the approach in the object side is that you have a direct comparison on the image in including the magnification and the NA impact. 

Best regard

Pascale

There’s also the RMS versus Focus Plots:

Hello,

Another  way  to consider the depth of field in the object space is to use the tolerancing facility. In the Tolerance Data Editor,  the test wavelength has to be defined by using TWAV and the range of the object displacement using  TTHI  aroung the nominal value of the object. The compensator and other variable have to be removed. The tolerancing can be launched in chosen the criterion of interest such as MTF, RMS spot radius…. The result can be exported in other application.

Regards, Pascale

Hi Steven,

DOF depends on your entire system (also the pixel size of your detector is relevant). For a large DOF your system should not be limited by optical resolution but rather by the “digital resolution” of the sensor chip (circle of confusion is small compared to pixel size). E.g., if your optical system is diffraction limited and the Airy disk radius is 3µm (=Abbe diffraction limit) the image space MTF frequency corresponding to this optical resolution limit is ca. 330lp/mm (1/0.003 → expected MTF value is ~10%). 

With a detector pixel size of e.g. 5µm the system resolution is determined by the detector (“digital resolution”)→ MTF corresponding to the digital resolution with 5µm pixel size is 1/(2x0.005) = 100lp/mm

You can now read out the MTF at 100lp/mm with the optimization operand MTFA while scanning the object position with an universal plot 1D. This gives you “MTF@100lp/mm vs. object position“. I would determine the DOF by the range defined by MTF@100lp/mm is > 10%

You can also use the “MTF vs. focus“ analysis in OpticStudio which gives the same results but be careful because the “Delta Focus” value is the image/detector position and not the object position.

Regards, Norbert

Hi, @Norbert.Reitinger I have the same issue with measuring the depth of field of a given system in zemax. I have been reading one paper  and it is possible to know the depth of field using FFT Through focus MTF in the analyze button. However, the delta focus still confused me. I would be happy if you could explain what parameters are used to defocus the object in this regard.

Best regards,

Kibru

Hi @tk1234 ,

The MTF evaluation takes place in the image space. The “FFT through focus MTF” varies the image/detector position and not the object position. In the paper you mentioned, they are using a thin lens design with magnification 1 → “For the Zemax simulation, a point object is placed at a distance e 2 f (120 mm) to the left of a thin lens, the image is formed at e 2 f (120mm) to the right of the lens, and the thin lens magnification is unity”

For this case a variation in image/detector position is nearly the same as a variation in object position (if you are not too far away from the focus position).

If you don’t have a magnification of 1 you can still use the “FFT through focus MTF” but you must convert the shift in image/detector space that you get from OpticStudio to the corresponding shift in object space.

Or in other words: OpticStudio scans the image/detector position for a fixed object position and what you want is the MTF at a fixed image/detector position while you scan the object position.

This is the reason why I prefer the approach with the universal plot I described above.

Regards, Norbert

@Norbert.Reitinger Thank you for your clear and helpful response. Your explanation was exactly what I needed. 

Take care,

Kibru