Narrow lepton Field of View

[Bill W]

From another thread..

My problem is that I am limited to high fov and low resolution with my phone. I can increase the resolution partially by super resolution. But lowering the fov will need a lens. And the offer I got from TPLogic is kinda expensive and might not work. So diy and a ZnSe solution might help to lower fov and get detail at range - for better stiching. A tripod with a panorama mount that fits my phone could help.... That might be printable.

I suggest you consider starting a new topic (something like 'Narrowing Lepron FOV'), but I am also happy to continue the discussion here if you prefer.

I know very little about the Lepton. I assume it's not possible to remove the lens assembly. That really only leaves you the option of making a telescope arrangement - which can work well. Try using a reflector telescope and putting a germanium or ZnSe lens in place of the eyepiece. You may be able to use it to form the light so it can be seen by the presumably fixed-focus Lepton. I use a simple astronomical telescope like this, except I do not need the extra lens for my setup. I suggest using a hot target such as a soldering iron a few tens of metres from the telescope. Set it up with its original eyepiece so you can be certain it's pointing accurately. Then start playing with lenses and the position of the Lepton to see if you can get anything resembling an image. I expect you so get poor sensitivity and high magnification - but once you have demonstrated that it works you can start optimising different bits.

I have taken a lepton apart and am indeed going to be trying out a number of other lenses on it.  Summer has got in the way, but yes the lens can come off although I might also have to sacrifice the shutter at least for now.
Next step is making up a good mounting to line up the available lenses and see what works, and also making the battery connection and the output interface to the phone more robust with the case out of the way.

The biggest issue is that the lepton is 12µm pixels which will be way beyond the design intent of the available lenses.  I expect some bad surprises but perhaps some good ones too.

The first good surprise was that just refocussing the lens supplied helped quite a bit.  I suspect the Chinese factory have been screwing them in fully regardless.


Bill

[Vipitis]

Hey Bill,

Thanks for starting this topic up again.

The Lepton I have is integrated in my phone and I won't be taking it apart.

I reached out to CAT, their service partner and Bullit but I got no information whatsoever on the lens used to cover the Lepton. I suspect it's similar to what is used in the flir one. And that would be silicon.

I was about to do some test to confirm the fov of the sensor and do measurements with callipers on my own.

From what I got from TPL, I can say that their adapter has an opening for the Lepton lens, but not for my phone and therefore would result in large vignette and loss of pixels.

I plan to buy lenses and print the mounting solution on my own, similar to what Boris achieved.

I was going to make a thread to compile all attempts, like some links to Boris and his YouTube, the Fresnel attempts by Ultrapurple and others as a knowledge base.

I got something to take care off ATM and will do some later this day.

Attachment: little friend in my basement trying to hide - I will "take care" of him/her now!

[Fraser]

Please do not kill the mouse !

I always live capture them and release a good distance away from my house.
I kill only when there is no other option. It is good Karma Please do have mercy on the little guy and live capture it.

Fraser

[Vipitis]

So let's start this.

I will collect a few links to different experiments, people and examples of building custom lenses for thermal imagers. The topic might be lepton, but the examples I will present include Seek, Flir One, Therm App and other cameras. In the end it should work for any camera out there anyways.

Uho Boris
Boris has built quite a few modifications for Seek and Flir One. His YouTube channel has a few videos on his creation. He is active on this forum under the name Uho
Relevant projects include:

x6 Telephoto lens seen in
This is an auxilary telescope made for the Flir One v2 smartphone dongle. It uses 2 lenses(as far as I understand it) and is mounted on a selfmade case with a thread.

T10 upgrade as seen in this post
This is an additional lens to the T10 imager by Torres Pines Logic. It is made of at least 2 germanium lenses and 3D printed parts.


***

I will do the next references another time.

[Fraser]

Some information on decreasing a thermal cameras Field of View.....

Refracting telescopes are used as an auxiliary lens to decrease the field of view on some thermal cameras.

Refracting telescopes commonly come in two types, Keplerian and Galilean.

The Keplerian is commonly used in astronomical applications but may be used in front of a thermal cameras objective provided the correct materials are used for the lens elements. Only two lens elements are needs, both Bi-Convex. The disadvantage of the Keplerian telescope is that it is inverting so the image appears on the camera upside down and reversed left to right.

The Galilean is also used in astonomical applications and it has the advantage of being non-inverting. The telescope comprises two lens elements, but one is bi-convex whereas the other is bi-concave. The Galilean suffers from relatively poor field of view.

Keplerian telescopes are relatively easy to construct providing suitable lenses are available. Such is not normally a problem with visible light lenses, but lenses suitable for thermal imaging wavelengths are harder to source at reasonable cost.

How to select a lens set for a thermal telescope......


In the Keplerian telescope there is a front lens of long focal length and a rear lens of short focal length. The front lens should be as large diameter as practical in order to collect plenty of energy. The image energy is inverted and passes to the rear lens where it is converted to parallel energy beams to illuminate the thermal cameras objective lens. The magnification factor (and subsequent reduction in camera FOV) is calculated by dividing the front lens Focal Length by the rear lens Focal Length.

For example:

Front lens FL = 150mm
Rear lens FL = 50mm

150/50 = x3 magnification and a reduction in FOV by a factor of 3. A 36 degree FOV Camera objective would be reduced to 12 degrees.

This may sound great but it is rarely so simple. The selection of the two lenses is essential to decent performance. A large front lens is desirable, as already stated, otherwise the telescope will have high losses. The lens length must also be considered. The refractive index of the lens element material will directly effect the length of the lens for a given magnification. Germanium has a very high refractive index of ~4, whereas Zinc Selenide is closer to visible light lens refractive index of ~2.4.  A x3 telescope that uses Germanium lenses can be shorter than one that uses Zinc Selenide lenses of the same dimensions.

The Galilean telescope uses a Bi-Convex front lens and a Bi-Concave rear lens. Finding a suitable Bi-Convex lens for thermal imaging wavelengths can be challenging, but finding a suitable Bi-Concave lens can be even more so !

The magnification calculation is detailed here:

http://www.citycollegiate.com/geometricaloptics2.htm

I have detailed telescopes that are placed in front of the thermal cameras original objective lens. This basically acts like an optical translator for the objective to produce a magnified image of smaller Field of View.

There is also the option to remove the thermal cameras original lens block and replace it with a lens block that provides a greater distance imaging capability. Such lenses can often be found on the secondary market and are often a precision assembly incorporating quality Germanium, ZnSe or Chalcogenide IR Glass lenses. The Focal Length of the lens needs to taken into consideration and some form of mount devised to hold the lens block at the correct back focus distance from the microbolometer.

Now a word about lens blocks. I use the term lens block to describe a complete lens assembly incorporating all required lens elements. This avoids confusion when talking about "lenses" and "lens elements" which can be quite different !

A lens block is designed to match the sensor that it illuminates. This needs to be considered when sourcing a lens block on the secondary market. If a lens block is designed to illuminate a thermal sensor of area "X", if it is placed in front of a sensor of area 0.25 X, the field of view specification for the lens block will be reduced by a factor of 4. Many thermal camera lens blocks sourced from older thermal cameras are a bit of a mixed bag. Whilst they will likely contain decent quality Germanium lens but will have been designed to illuminate a sensor with large pixels compared to more modern microbolometers. This directly effects the size of the sensor array and so, how the lens block illuminates it. Over illumination of the sensor array is common in such mixes of older lens block with modern microbolometer. This fact needs to be considered when selecting a certain FOV lens block for a modern sensor array. The up side of this over illumination situation is that the lens block provides a telephoto characteristic The Lepton microbolometer area is tiny when compared to the relatively large BST imaging sensors from around the year 2000. The BST sensor pixel array has a corner to corner dimension of around 15mm (best guess). Bill W will know a more accurate figure.

Another important factor to consider when planning to use a lens block designed for older thermal cameras, such as the BST based units, is pixel size and lens resolution. 50um pixels were common and are huge when compared to modern 17um and 12um pixels. The older lens block resolution was chosen to match the large pixels sizes of the time. Placing an older lens block, designed for 50um pixels, in front of a 12um pixel Lepton microbolometer is a bit of an unknown. Depending upon the lens element quality, the resolution may, or may not be adequate for decent imaging with such a small pixel size.

Older lens blocks will likely be physically larger than more modern lens blocks that are designed for the smaller sensor arrays. It is easy to end up with a huge lens with a small thermal camera like the Lepton hanging off of its rear ! Such could be an issue in Drone applications where weight can be an issue. Germanium lens elements are relatively heavy.

Finally, if considering the purchase of a lens block on the secondary market, ensure that it is designed for use in the Long Wave thermal spectrum as a "Short Wave only" lens will not work with a microbolometer sensor. Some dual band lenses are available that work in both LW and SW. They will be marked as such.

I hope this helps those unfamiliar with lenses, lens blocks and auxiliary telescopes.

Fraser

[Vipitis]

Hey Fraser,

this is a very helpful addition. I am stuck, trying to figure out, what "lensblock" CAT used in my phone. I emailed them and the service center, but  all responses are negative so far, and I don't see any more options. While looking through local e-bay listings and federal auctions, I found someone selling their CAT S60 with a broken lens block. I am tempted to contact them and ask if they can provide a few better pictures and measure the thickness. I also believe the lensblock used in the CAT S60 is similar to the one used in the Flir One. By my own measurements the outer diameter on the top is slightly lower then 9mm and by doing some really sketchy experiments with calipers, I estimated the effective sensor size was like 6mm x 4mm ontop if it. I want to do proper test with 3D printed parts later. Given the vague information from CAT regrading the FoV, I could then estimate the thickness of the lensblock. The TPL unit I got offered had an opening of about 3.6mm so I will likely end up with vignette if I go down the easy route.
Anyways, If I manage to get a rear lens larger then 9mm and mount it close enough, I should have no issues with vignette. I will look into lens offers and some proper calculations in the following days, but I will also continue to list some references on different projects I have seen around here and elsewhere.

[Fraser]

To check for the minimum exit pupil size needed for an auxiliary lens, just cut or drill a suitably sized hole in a piece of card from a cereal box and place it over the lens aperture. Start with a 8mm hole and work your way down until vignetting occurs.

This is a very rough test but should serve your needs. The Lepton and its deployment in the CAT phone should be almost identical to that found in the original iphone FLIR One unit.

Fraser

[Fraser]

Inside the CAT60, showing the whole optical sub assembly 

Fraser

[IwuzBornanerd]

Please do not kill the mouse !

I always live capture them and release a good distance away from my house.
I kill only when there is no other option. It is good Karma Please do have mercy on the little guy and live capture it.

Fraser

Pardon me for continuing the off-topic, but I suspect others will appreciate the comic relief.    My favorite story about this very suggestion was told by one of my former co-workers who said while he was in college he caught a mouse in a trap in his apartment, but it was not dead, so he decided to be nice & do exactly what Fraser suggested.  But first he took a permanent marker & put a big red stripe down the mouse's back.  Then he took it out & carried it past several buildings & released it in the snow.  A few hours later there was a mouse in the trap again & it had that red stripe down it's back!  It wasn't so lucky that time.  He says "Mice is stupid." 

A similar story was told by another former co-worker who live-trapped chipmunks & took them to one of the local city parks.  On 1 or 2 occasions while he watched the rodent scurry away a hawk swooped down & grabbed it.

<edit>
I apologize to the owner of this thread for "posing a question" here.  I have deleted the last 2 lines of the original post in the hope that doing so prevents this thread from deteriorating into an off-topic discussion that belongs elsewhere.  The intent of my post was merely to provide a couple stories about live releases "gone bad".

[LesioQ]

I'd say that it's not our (human) duty to decide of life & death to lower species. Let their own karma take over.

[Vipitis]

You know, I missed that post. But here is the result.

Little mouse slipped below he door while I was drumming in the basement. I observed as she got closer, seemingly eating bits of popcorn I splitted on the carpet some time ago. She was really shy, and run away with every movement I did. I pulled out my phone and use the thermal camera to locate it below a cupboard. I went upstairs and grabbed the empty bucket of popcorn.

After a few observations of what routes she took as I chased her, I found out that she seem to not see me while in a dark area behind a deskplate and she came as close as seen in the image posed earlier. I put the bucket on one end and managed to get her behind a desk, a little. Bit of waiting and she was in the bucket, it was tall enough with slippery walls - she couldn't escape and was trapped. I put the lid on the bucket, and took it with me by bike. I released her on the opposite side of a river. I haven't marked it.

Attached are pictures of it in the bucket.

[Fraser]

Wonderful

Thank you !

Fraser

[Uho]

I use three lenses in the latest models of my telephoto lenses. Two germanium and one zinc selenide.

[Fraser]

Uho,

Nice work. I presume you are adding an erection lens to the Keplerian telescope, hence three lenses.

How easy was it to find the Germanium lenses ? Such are normally very expensive. I found some cheaper lenses but they were not coated. The same supplier could provide AR coated lenses but the price sky rocketed.

Fraser

[Uho]

Hi Fraser,
 I bought wholesale on the lens sale. It is very difficult to buy a lens with a small focus. The right lens with a negative focus is almost impossible to buy. The third lens is used to reduce distortion. Distortion can be reduced by placing a negative lens or lens with a different refractive index.

[Fraser]

Uho,

Thank you for the explanation. I am no expert on optics )

I have some nice 12mm diameter / 15mm FL Germanium lenses that a fellow forum member gifted to me for experiments. He also sells them on eBay. Very nice quality and with the correct AR coating. The small diameter limits their use with my cameras as most have large objectives. I shall have to do some Keplerian telescope experiments with them soon.

I managed to purchase some 25mm diameter bi-concave ZnSe lenses from a chap in Israel. Made by Coherant and nice quality. I was intending using them in a Galilean telescope but have not gorvaround to trying them yet.

One of my ongoing telescope projects is the conversion of the Inframetrics X3 LW telescopes to provide a non inverted image. Those telescopes contain 4 large Germanium lens elements as the rear two are collimators. If nothing else, the front two lens elementsare perfect for a DIY Keplerian telescope

I disassembled a lens block from an EEV ARGUS 2 with the intention of using the lens elements in experiments. After testing I found that the front lens is great asca Close-up lens as it is a lovely positive meniscus type, much like reading glasses ! Sadly this is the lens that gets water damaged on these cameras. The other lenses in the block are very thick and VERY short FL. too short to use as a Close-up lens. Maybe as a thermal microscope lens though??.

My best source of lenses for experimentation remains the old LW inframetrics telescopes. Sadly they are relatively scarce and their prices on eBay have risen considerably in recent years. Others have discovered the usefulness and quality of these telescope lenses.

Fraser

[Uho]

I did a lens design for the E4 telephoto lens. To understand the problem, the thermal imager brand is not relevant. Distortion is the biggest problem. The second problem is the high price of lenses. The person who bought the lepton will not spend a lot of money on the telephoto lens. Sometimes I do not need lenses. I make a telephoto lens with a mount. I sell at the minimum price of $ 180. This is practically the cost of materials. So there is no simple and cheap telephoto lens.

[Uho]

Vipitis,
 Did you break the lens? If so, I can look  my lens. At me where that lays with Flir One.

[Fraser]

Nice work Uho. You are obviously very comfortable working in the optics domain

Can I ask what ray tracing program you are using to plot the lens beam paths. I have a public domain program but I think yours looks better.

Fraser

[Uho]

I had to learn a bit of optics. I use the CODE V program to test my ideas.

[Vipitis]

Hey Uho,

those are some really nice examples. I have not broken any lens, my phone is still intact. Fraser showed some images of a tear down from my phone, I have contacted azrepair.eu and asked them if they know more about the lensblock used. Given the price an availability of Germanium or ZnSe lens elements, I am still considering the offer I got from TPL. as far as I can tell from the drawing that they send me, it might work, if I manage to mount it close enough. I can turn the image by 180° in my capturing software. They want 200$, I think with added shipping, tax and import tax this could be very expensive as well so I am not giving up on the DIY idea.

I will attach the drawing and a screenshot of our email conversation, as they told me to not share it but then allowed it again.

[Uho]

The problem of convergence is solved by a window of germanium of the required thickness.

[Bill W]

The simplest to understand would be to put a second FLIRone lens front-to-front with the existing lens.  This creates a second image plane at the focus of the additional lens, out in free space in front of the system.

You can then put any lens you like on the front focussed to that image plane.
Equally at that plane you could put an object which you will then image at a 1:1 ratio.

This way you know that the image plane is flat and of acceptable distortion.

Bill

[Uho]

More details about the x6 telephoto lens for Flir one in this video. It is possible to include subtitles.
https://youtu.be/pc7ZwEyY5rA