Optical Testing / OPTI513R

News Feed

• New

  • The Final exam (12:30 pm on 5/6 - 9:00 pm on 5/11 MST) will be the same take-home format. The final exam will be assigned through D2L. There will be no class on 5/6. (Note: The lecture on 5/1 will be the last class of the semester.) This schedule will provide sufficient time to the students (especially, DL students who may have work-related commitments) and also to the TA grading the exam. Thank you. 

  • I recevied a great video link and comments from a student who is willing to share it with the classmastes. LOVELY! Indeed, this is a great video showing the manufacuring and testing of the camera optical system. Please, enjoy! This class is awesome with all the contributing classmates!

"I had a video to share with you that I thought you may find interesting. It is the production process for a camera lens at Sigma in Japan. It is so interesting to see how testing is used throughout each step of the manufacturing process. While watching, I was so happy not only that I recognized the tests they were doing throughout each step, but that I also felt like I knew why they were doing them and what they were looking for. Optical testing truly is an awesome topic. I will link the video below for you to check out. Feel free to share with the rest of the class as well. 

The process of making a camera lens. The best optical equipment factory in Japan.

Some key points of interest:

- At 4:03 you can see them checking the lens with what looks like a Geneva gauge after they get rough in the basic shape.

-At 5:36 they check the lens after polishing using a very basic Fizeau test using what I believe is a sodium lamp to observe uniform fringes. The fringes are hard to spot, but do show up on the video still. 

-Then at 5:50 they move to a laser Fizeau setup to check the surface precision if the lens passed the previous test. 

- At 6:00 they are doing another interferometric test and tilt fringes can be seen on a monitor towards the bottom right of the video. I'm not sure if this is still a Fizeau setup or something else, but from the geometry and the fringes it seems like Fizeau." 

• Old

  • Welcome to the OPTI513R Optical Testing class of 2025! This will be an exciting class discussing various optical metrology topics. Let's work together, learn together, and grow! We will shape the class together in a meaningful and beautiful way. Thank you!

  • I hope you are all enjoying the opening part of the class. We took a grand overview setting up the landscape of the exciting metorlogy field and topics in the last three lectures. Also, I was impressed with the lively engagements of all of you in the classroom during the lectures and after the lectures. Of course, I have received a number of excellent messages from the Distance Learning students. What a great class of 2025! 

  • There will be no in-person class on 1/28 and 1/30 because I (and a large number of students in the class) am going to attend the SPIE Photonics West conference for my Short Course teaching and board member commitment. Instead, please, watch the lectures through D2L.arizona.edu website. (Note: The lecture videos are from the 2024 semester, so you may need to consider some outdated information. In the meantime, you can enjoy Daewook's short hair a year ago. It is always nice to see someone's past.) I am sorry for missing the in-person lectures and thank you for your kind understanding.    

  • The 2/18 class (Tue) will be canceled due to the College Industrial Affiliate (IA) Event. This is an exciting opportunity for you to connect with industry professionals, learn about potential career paths, and even explore future job or internship opportunities. The Wyant College strongly encourages you and feculty members to attend, so I will cancel the lecture. Please, engage with professionals from the various IA companies that might shape your future career in a wonderful way. 

  • HW-1 has been assigned through D2L site. Please, make sure you check the website and turn in your HW by the due date. Enjoy and learn! Thank you. 

  • Here is another nice question from a student. Let's discuss this in the class! "I also had a question regarding autocollimators. When assembling one in the lab, is a cube beam splitter preferred over a plate? My initial thought is yes. Cubes seem easier to mechanically position vs. a plate and also do not have the problem of ghosting. However, from an aberration standpoint I suppose a cube may generate more aberration, mainly astigmatism, since it is a thicker plane parallel plate than a plate beam splitter. If designing the autocollimator with a custom lens this could easily be accounted for, but if attempting it with off the shelf components would this create more of an issue with creating a good quality plane wavefront?"

  • A awesome DL student sent me a comment as follows with a great video materials explaining the autocollimator. "I have been enjoying the class very much to date. I stumbled upon some videos on autocollimators and thought I would share it with you in case it was something you would be interested in sharing with the rest of the class. The optics are presented very basic and straightforward, but I thought it was still a nice supplement to the course materials." Thank you for sharing these with your classmates. 

Autocollimators 1: Introduction

Autocollimators 2: Formula for Resolution

Autocollimators 3: Optical Design

• • Dr. Heejoo Choi will host an Invited Lecture by Dr. Michael North-Morris from 4D Technology during the class on 3/4. You will learn one of the most advanced commercially available interferometry solutions. Enjoy!

  • Daewook will be on a business trip so will not be able to teach on 3/6. The lecture will be replaced with the previous year's lecture recording. Please, watch the video via D2L. Thank you for your understanding.  

  • HW-2 will be assigned via D2L and it is due by 12:30 pm on 3/18 MST. Learn and enjoy! ;-)

  • Here is another great question from a student that I want to discuss during the class! We will talk about this. Thank you for engaging the class with such a lively communications! :) "Regarding the use of auto collimators. How do you set-up the autocollimator, a temporary fixture, to precisely align a mirror, since you will be aligning the mirror to the axis of the autocollimator itself? For example, lets say I have a telescope that consists of a TMA in the form of 3 OAPs that sits on a bench. After the third mirror, I need to fold the beam twice using 2 flat mirrors at 45 degrees. Do I add inserts to the bench and specify tight tolerance on the positions? Do autocollimators often come with some sort of calibration feature that allows you to align its axis parallel with the bench surface? Just curious how you would deal with aligning the autocollimator to some nominal global axis so that the mirror under test is aligned (with some error between the global axis/axis of the previous mirror and some error in the autocollimator itself)?"

  • I am very happy that I am receiving various emails with good questions and discussions. Indeed, this is a bi-directional class! I am sharing a discussion with a DL student. ;-)

Q: "I had a general question about alignment of optical systems. At my work, I am responsible for several CO2 laser cutting systems, all of which require periodic alignment of lens, mirrors, or gas nozzle. Since the working beam is 10600 nm, focused to a spot size of ~100 um, we cannot use the image of the beam to precisely align any of the beam delivery components (too invisible, too small to see, difficult ergonomics, and safety risk). Instead, we use mode burns on thermal paper. Are there applications for optical alignment that do not utilize the image? While I do not anticipate upending the thermal paper method of CO2 alignments, I am curious about how higher power lasers are aligned, particularly when melting the sensor is a possibility! On a separate note, a typical CO2 laser beam delivery consists of 2-3 bending mirrors (see picture below). When I heard your statement on pentaprisms, I was curious why they were not used more commonly on industrial lasers? I would guess that cost is the main factor, but are there other issues with using prisms in higher energy optics?" 

A: "1. There are so many types of optical alignment approaches. Some of them do not use "focused images" at all. For instance, we may use a Laser Tracker or Coordinate Measuring Machine. Or, my team recently published a technique using Bessel Beam as a quasi-ray to align the optical system. Sometimes, the wavelength is not visible (just like your case) so another wavelength (e.g., HeNe Laser) is being used to align the system. Every kind of alignment techniques (including the thermal paper method) have their advantages depending on applications. It is exactly why we are learning, discussing, and sharing various types of methods in this class. Regarding the pentaprism, they are heavier and more expensive. Also, they may have transmittance issues depending on the wavelength. Thus, unless you need such an exact 90 degree folding regardless of its misalignment, you do not want to use pentaprism."

• • Hello All, I hope you had a refreshing and happy spring break.

1. The video lecture replacing the 3/6 class has not been uploaded due to some technical issues. I tried to resolve it with the A/V staff, but couldn't. Because it is now one day before my class tomorrow (3/18) I decided to teach the content in the classroom tomorrow. I am very sorry for this inconvenience and appreciate your understanding. 

2. I decided to extend the HW-2 submission deadline to 9:00pm on 3/22 MST. I hope this provides you with more time to think and learn from the HW questions. 

Have a great Monday and I am looking forward to meeting you in the class. Thank you. 

• • Here is another good Q&A discussion with a student that I want to share with the class! These are all great discussions! Let's keep shaping the course together.

Q1. Parabolas are used to direct all incoming collimated light to a point at the parent focal length. In the opposite direction, we can place a light source at the parent focal length to create collimated light. If you wanted to test an imaging system's performance in a through focus fashion by using collimated light, I would imagine it's easiest to place a source at the FL of a parabola and measure through focus with a detector at the systems imaging plane. But in a case where you cannot move your detector (say you're evaluating how well you've placed your detector relative to your imaging system) you should be able to simulate the through focus measurement by moving the point source relative to the parabolas focal length. If the imaging system performs perfectly when the point source is at the parabola FL then you can assume the detector is placed perfectly (axially) behind the imaging system. THE QUESTION: if you were to try to simulate through focus conditions with collimated beam, should you move the point source along the parent focal length of the system, or the path that the gut ray takes through the imaging system (i.e. some angle set by the off axis distance and FL of the parabola)? 1a. In either case, can you verify that the light source is moving axially (i.e. no moving in x-y as it translates through optical axis z) using an alignment telescope by first viewing the source through the parabola with the alignment telescope focused at infinity. Then as the light source travels though the parabola focus, you would then re-focus the alignment telescope to view that the image of the light source has not moved in translation (x & y)? 

A1. Indeed, a parabola or OAP (Off-Axis Parabola) is often used for creating a collimated input light emulating an object at infinity (e.g., star). Yes, you may simulate such a through-focus imaging (of your optical system under test) by moving the light source source (at the focus of the parabola). However, the method has a limitation so it is not often used. You are basically creating a power aberration in your input beam. However, it doesn't directly mean you will have the power aberration on your output beam. It is because that your aberrated (i.e., power added or defocused) input beam will let it go through different sections and apertures throughout your optical system, which will add various new aberrations. That is why, if you want to carefully measure the through-focus imaging of your optical system, you may want to translate the detector plane of the optical system. Regarding the question about the alignment telescope to verify your light source is simply adding power (but no tip/tilt), yes, you can definitely measure and confirm it as you described.  

Q2. I understand that Zernike's are defined at the imaging plane while Siedel terms are defined at the pupil of a system. I didn't understand the bits about how the Zernike's tell you what to compensate by to reduce RMS WFE. Could you talk about that again and maybe distinguish what the coefficient of the Zernike is telling me and a term broken it's perspective error/correction? Not sure if that was asked properly, my apologies.

A2. As a matter of fact, Zernike can be defined anywhere as long as it has a circular domain. It can be used to describe a mirror surface, exit pupil, entrance pupil, any other mirrors or lens surfaces in the optical system because it is not specifically developed for the optical aberration theory. It is simply a very useful mathematical tool for so many applications in the world. In the meantime, the polynomial is "mathematically defined" so that it minimizes the polynomial's RMS (e.g., Z11) by compensating the order using all the lower order Zernike terms. That is why we can use the Zernike mathematical definition to get insights about minimizing the WFE (wavefront error) of the order using lower order aberrations if we use Zernike polynomials to represent the WFE of an optical system.  

Q3. When we talk about coherence length, at what point in a system is the "0" location of coherence length? In the case where light travels from left to right from a laser, it passes through a beam splitter where it moves upwards by some length L to a reference surface and then is reflected and imaged onto a sensor and the rest of the light is traveling a distance L' where it's reflected from a test surface to be recombined at the beamsplitter. Is the "zero" location of where coherence length begins defined at the beamsplitter or the test surface? Also, is coherence length defined to be the sum of the optical path of L and L'? 

A3. More exact coherence length definition and discussion requires more in-depth discussion (covered in the prerequisite OPTI505 course). However, in the context of interferometric optical testing, you may think in this way. The coherence length roughly tells you how much OPD can be still ok to make an interferometric measurement. For instance, if you have a "perfect single wavelength" which gives you an infinite coherence length, the OPD (i.e., the difference between L and L') can be infinite and you can still get a high contrast interferogram and make a measurement. If you use a laser with a shorter coherence length (e.g., 5 mm) and the OPD is 10 mm, you will not see a high contrast interferogram because your interference fringes will be washed out due to the overlaps of all wavelength-dependent fringe patterns on top of each others.

• • Here is a good question that I want to discuss in the class. Let's talk and discuss. We want to be all clear about the exciting metrology technologies! 

Q. I was curious, do the concepts of coherence length and intensity of the return of the light from the test surface apply in same fashion for ESPI as they do for standard interferometry? Or, since we're not looking at surface fringes, we're not so concerned about coherence length? We do increase the intensity of the beam significantly for ESPI, but is it still the case that when the test beam recombines with the reference beam that we want them both to be of equal intensity? 

• • The take-home midterm exam will be released via D2L at 5 pm on 3/27 MST. Please, study, learn, and master the topics as you are working on the midterm exam questions. 

  • The take-home midterm exam will be from Mar 27 - Apr 1. It is an open-book midterm exam and it will cover all the topics taught until the Mar 25 class. Thank you.