UV Macro Photography for Skin

[PDL1]

 

 

 

Hello! I'm currently looking into get UV reflectance macro photography for pigmented (or hypopigmented) skin lesions. I was hoping to get some general tips and common pitfalls that you've seen. I've looked through some of the UV portraits, and I haven't immediately come across some examples. I've also read some of the previous dermatology posts on here (like this one, which was a helpful starting point!).

 

I've got access to the following equipment:

• Canon EOS RP by Kolari 
• Canon EOS T100 by Kolari
• KV-FL1 Multispectral Flash
• Yongnuo 50mm f/1.8 EF Lens
• Rayfact 105 Lens
• Kolari Vision 52mm UV Bandpass Filter
• Macro Adapter Canon EF Lens Extension Tube x3

 

Currently in process of getting this UV ring light from MaxMax: https://maxmax.com/forensics/lights/uv-ring-light

 

I also have the following IR filters:

 

• 590nm Infrared – Gen 3
• 665nm Infrared – Gen 3
• 720nm Infrared – Gen 3
• 850nm Infrared – Gen 3
• UV/IR Cut Hot Mirror Pro 

 

If you have any suggestions, I'd be open to hearing them! My goal is to get really high quality images sufficient for publication. The main pitfalls that I've come across thus far are adequate lighting for indoor photography. The Kolari flash has been really useful, but it's hard to focus without an ambient light source and the torch on the Kolari flash isn't movable. I've also noted some difficulty with getting the entire field in focus. Any thoughts?

 

[Andy Perrin]

Talk to JMC (Jonathan), PDL1. Skin imaging is his main focus, so he is probably the most knowledgeable person around here.

 

That ring light is very narrowband. I don't know if that matters for your purposes, but you won't get much color in the images.

[Andrea B.]

It looks like you have put together a good set of gear.

 

PDL1 writes:  ......it's hard to focus without an ambient light source.

 

To focus the lens, we use a handheld UV-flashlight (UV-torch) which will illuminate an area so that focus can be made on that area in the live view LCD screen. We usually just use a common 365 nm UV-flashlight bought on Amazon or Ebay. You might also need some kind of clamp to hold the flashlight in place (once you have positioned it usefully) while making focus if hand-holding is not feasible. But the point is you need to be able to move the UV-flashlight around to position it usefully.

This link mentions current favorite UV-flashlights:  Best Basic Gear: Goggles, BG Filters, Torches

 

I don't know what you know! So I'll make some suggestions -- which you might already know. 

 

1) PDL1 writes:  ......some difficulty with getting the entire field in focus.

If you are making a close-up/macro photo, remember that the closer you get to the subject, the less depth of field you will have. You could learn how to make a depth-of-field stack to produce a photo with increased depth of field. This is created from a series of photos made at different, very small distances from the subject being photographed. The series is then blended into one photo with increased depth-of-field using an app like Helicon Focus. Another name for this technique is focus bracketing. There are mechanized platforms which can be used for focus bracketing.

 

2) Remember that diffraction also affects the sharpness of a photo by blurring small details. In UV light you can usually go to f/11 with almost no diffraction blur. Any accompanying Visible photo of the same subject will exhibit diffraction at f/11, maybe already at f/8. Diffraction is worst with the long rays of IR light. It is difficult to make IR macros which show all details. 

 

3) The Rayfact 105/4.5 will need to be on an extension tube in order to get close to a subject for macro work. I see that you have one. "-)

 

4) Practice. It isn't going to go well the first 50 times or so!! Especially if you want to focus stack 20 images or so. Just set up and photograph any little thing until you get the hang of this. I've gotten excellent detail in my botanical work using a UV-flashlight to focus and a UV-flash to illuminate. Birna has some superb macro work in UV with more specialized gear. And our member Jonathan has practical experience in your field. 

 

5) UV-BLOCKING GOGGLES. You do have some, yes? I surely do hope so. Get a few pairs so that both you and your subject can wear them. UV-flash is a mean thing on the eyes.  [UV SAFETY] UV and Your Eyes :: UV Safety Reference 

[Andrea B.]

6) The more megapixels you can lay over a small skin lesion, the more detail you will be able to photograph. So if details are important to your research photos, look at the pixel density of your cameras.

• The Canon RP has 26.2 megapixels with a pixel density of 3.05 MP/cm².
• The Canon T100 (4000D) has 18 megapixels with a pixel density of 5.43 MP/cm².

In theory, you should see slightly better detail with the T100. In practice, there is a lot which goes into the making of a macro photo that can affect detail representation.

 

 

7) Sometimes it is useful to fire the UV-flash 2 or 3 times to illuminate all parts of a subject. I hand hold my UV-flash and usually fire it 3 times around a flower subject. This technique has a learning curve: how close to hold the flash, how long to set the exposure time, and more.

[JMC]

PDL1, have you considered cross polarised visible light photograhy for skin to look at hyper and hypopigmented areas?

 

The issues with using UV is the background melanin in the skin becomes more and more prominent as wavelength decreases, making it appear darker and darker. This could be an issue depending on what you trying to image.  Also, surface features such as the dermatoglyphic lines become more visible in the UV, making it harder to see differences in pigmentation. 

 

Cross polarised visible light photography gets rid of the shine from the surface of the skin and hides lines and texture. It therefore makes it much easier to see pigmentary changes and differences.

 

Cross polarised UV invivo imaging can be done, but it requires an awful lot of light, and perhaps might not be worth the effort compared to using cross polarised visible light photography.

 

Drop me a message or an email and I can send you some background on the skin imaging work I have done if you've not seen it already.

[PDL1]

@Andy Perrin, good thing to note on the ring light! Thankfully, what we're looking to do shouldn't require a ton of color. Perhaps having it being exceptionally narrowband with the LEDs will be of greater utility to prevent a lot of extra pigmentation or noise coming through that wouldn't be clinically useful. On the flip side, perhaps we may be missing some clinical information as well. We'll see and post updates as we can!

 

Thank you so much @Andrea B.! That's all really helpful! We're trying the torch lights right now, but the biggest problem is mounting so that they're stable. We may just see if either the nurse or the patient can hold the torch while we grab the picture.

 

@JMC, I'll shoot you over a message!

[PDL1]

@Andrea B., for focus stacking, my only experience with that was photogrammetry in visible light of small objects like amulets. When I had seen it, it was done with an automatic focus lens. With the Yongnuo, I haven't been able to get autofocus to work, so I've been using it primarily in manual mode. The Rayfact is a manual lens, so that can't be changed. I was reading up on doing focus stacking with a manual lens, but if you have any tips from your botanical experience, I'd greatly appreciate!

[Andy Perrin]

Focus stacking is usually done with manual focus I thought? You need very precise control of camera position so usually a rail of some kind. Photogrammetry is totally different, I’ve done that myself. For photogrammetry you want the whole object in focus as much as possible, with the camera at a distance from the object, but for macro focus stacking it’s about changing the focal plane in very small increments with narrow depth of field. The techniques are not related although both produce depth maps.

[dabateman]

Wow I learned something new. The Canon RP camera was the first camera Canon implemented focus bracketing on.  However the list of lenses that can be used with this feature is extremely limited:

Canon RF 24-105mm f/4L IS USM Lens

Canon RF 28-70mm f/2L USM Lens

Canon RF 35mm f/1.8mm f/1.8 IS STM Macro Lens

Canon RF 50mm f/1.2L USM Lens

Canon EF 16-35mm f/4L IS USM Lens

Canon EF 24-70mm f/4L IS USM Lens

Canon RF 35mm f/1.8 IS STM Macro Lens

Canon EF-S 60mm f/2.8 Macro USM Lens

Canon EF 100mm f/2.8L IS USM Macro Lens

 

 

For the R5 and R6 its a bit more:

Canon RF 35mm f/1.8mm f/1.8 IS STM Macro Lens

Canon RF 50mm f/1.2 L USM Lens

Canon RF 85mm F1.2 L USM Lens

Canon RF 85mm F1.2 L USM DS Lens

Canon RF 600mm F11 IS STM Lens

Canon RF 800mm F11 IS STM Lens

Canon RF 15-35mm F2.8 L IS USM Lens

Canon RF 24-70mm F2.8 L IS USM Lens

Canon RF 24-105mm f/4L IS USM Lens

Canon RF 24-105mm F4-7.1 IS STM Lens

Canon RF 24-240mm F4-6.3 IS USM Lens

Canon RF 28-70mm f/2L USM Lens

Canon RF 70-200mm F2.8 L IS USM Lens

Canon RF 100-500mm F4.5-7.1 L IS USM Lens

Canon EF 100mm f/2.8L IS USM Macro Lens

Canon EF 180mm f/3.5L USM Macro Lens

Canon EF 16-35mm f/4L IS USM Lens

Canon EF 24-70mm f/4L IS USM Lens

Canon EF-S 35mm f/2.8 Macro IS STM Lens

Canon EF-S 60mm f/2.8 Macro USM Lens

Canon EF-S 18-135mm f/3.5-5.6 IS USM Lens

 

 

So for your camera you will need to move the focus ring on the lens or move the whole camera and lens forward and back to get your stack. There are some devices you can buy that do either of these things automatically, a clip that snaps around the lens focus ring or a rail that moves the whole camera. 

Or if you have enough light and steady hands, you can just tilt you feet ever so slightly forward and back between your shots. 

You will need software like Helicon or Zerene stacker to merge the shots.

 

An alternative light might be to use a blb fluorescent bulb.  You can get handheld units that are 4W, used to check id's. They aren't horrible and broad range UVA. Also very cheap to play with, around $5 -$15. They take 4 AA batteries. 

This is the one I have, when I ordered 4, only 3 worked. But the seller shipped me a replacement that also failed and then refunded me for the failed light. So not a bad seller: 

https://www.ebay.com/itm/Portable-Handheld-UV-Light-Torch-Blacklight-Counterfeit-Bill-Money-Detector-2in1-/262048004002?mkcid=16&mkevt=1&_trksid=p2349624.m46890.l49286&mkrid=711-127632-2357-0

[dabateman]

Attached are two photos of the rig I made using these lights with my monochrome pi HQ camera. 

I have them with Halex #1 hanging brackets, bolted to 2 L brackets on an Olympus macro bracket.  

I use a Halex #2 hanging bracket to hold the Nemo flashlights to a tripod and Halex #1's to hold the Convoy flashlights to a tripod. 

[Andrea B.]

PDL1:  Focus stacking needs a focus rail - either manual or automated. Example:  Stackshot

 

Flowers are large enough so that I can usually make a nice focus stack with less than 10 frames. But if you use an automated rail the edges will be cleaner and the noise less because the shifts are precisely measured.

*********

 

 

We're trying the torch lights right now, but the biggest problem is mounting so that they're stable. 

 

If you Google "flashlight clamp" or "torch clamp", a lot of solutions will pop up. Here is just one example of many. The base of the gadget might be magnetic (as shown in this example) or it might be some kind of clothespin-like clamp or vice-like clamp. There are lots of variations.

 

 

 

Next Google "Gorilla Pod with flashlight clamp" and you can see more possibilities. I liked this one. If the flashlight has a diameter smaller than this clamp, it is easy to pad it with foam pieces, duct tape, a clean kitchen sponge or whatever works.  The Gorilla Pod legs are bendable. Different sizes are available.

 

[Andrea B.]

David, that is an impressive set up!! Deserves its own topic together with some examples. 

[Andrea B.]

I forgot to link to a manual focusing rail. There are so many available. I'm going to randomly pick just one for purposes of generic illustration. 

LINK

That link contains a video if you scroll down.

 

Later:  I picked that focusing rail at random, but it looks kind cool !

 

 

[The Usual Disclaimer:  Ultravioletphotography.com is non-monetized and has no affiliation with any product or vendor.]

 

[nfoto]

I use Manfrotto Studio Clamps and Magical Arms in my setups. Rather expensive, but immensely versatile.

 

Manual focusing rails are too coarse in stepping for high-magnification work.

[ulf]

Here are parts similar to the magical arms  and clamps that might cost less:

https://www.aliexpress.com/item/1005002824294118.html?spm=a2g0s.8937460.0.0.79e92e0eX7eV3a

 

[PDL1]

Sorry I've been away everyone! I'm still looking into this focus stacking but just wanted to update you all with some finds!

 

TLDR: The LDP LLC UV Ring Light is great! I wouldn't bother combining it with a UV flash. It can be hard to manually focus the Yongnuo lens with the ring light on, but it would be a lot easier with a tripod.

 

To get semi-decent non-macro indoor photos, I was having to use settings like 1/60, F ~2-4, ISO 3200. I wasn't getting as clear of a picture as I wanted, and I was getting pretty noticeable chromatic aberration on some. This was with the Kolari UV Flash as well as the Kolari torch, as well as a separate ~100 LED UV flashlight. Without the ~100 LED flashlight, I couldn't even find the subject in the viewfinder--it was all just dark. The flashlight helped find them so I could focus; I assume the flash then took over most of the illumination.

 

I was taking some handheld shots with the following equipment:

• Canon EOS T100 by Kolari
• Yongnuo 50mm f/1.8 EF Lens
• Kolari Vision 52mm UV Bandpass Filter
• Macro Adapter Canon EF Lens Extension Tube x 1
• LDP LLC UV Ring Light

 

I don't really have any good spots to photograph, but I was able to find one using the camera. You can't see this spot at all without the camera, and it's pretty faint even trying to capture it with my set-up. Other than being compressed down to JPEGs and adding the settings to each photograph, I haven't adjusted them in any way.

 

In any case--I had just gotten the UV Ring Light, and wow. It felt like a game changer when I put it on the lens. I could immediately see everything, even with only a weak fluorescent kitchen bulb on the other side of the room. I didn't have to flashlight at all. When I used the Kolari UV Flash, it didn't seem to provide a whole lot of increased lighting and instead just threw off my white balance with the ring light. I could adjust my settings closer to what I have been taught is more ideal. 

 

I was playing with settings, specifically trying to find a good spot for the F-number with the Yongnuo. Even with the various F-numbers, I never had a noticeable change in depth-of-field. It was all on the order of millimeters, so I think I'm just going to try to find a good F-stop that balances adequate illumination with appropriate sharpness. Based on my trial run/pictures, it looks like I can get by with 4.5 to 6.3.

 

With the Yongnuo, it was a little hard to focus the lens while the ring light was on, but still doable. Would definitely be a lot easier with a tripod. Even though the light doesn't have any soft feet, it didn't scratch up my lens body at all. That being said, I definitely wouldn't recommend jostling it too much while it's on the lens. It is powered via a power source, so you'll need to be within ~8 feet of a power outlet.

 

 

 

 

 

 

 

 

 

 

[PDL1]

On 3/8/2022 at 12:25 PM, Andy Perrin said:

Focus stacking is usually done with manual focus I thought? You need very precise control of camera position so usually a rail of some kind. Photogrammetry is totally different, I’ve done that myself. For photogrammetry you want the whole object in focus as much as possible, with the camera at a distance from the object, but for macro focus stacking it’s about changing the focal plane in very small increments with narrow depth of field. The techniques are not related although both produce depth maps.

 

We didn't normally use it for even our smaller/handheld artifacts. Just the amulets that were about the size of a quarter? Some of which had inscriptions that we were trying to get in high detail. These were items not usually amendable to typical photogrammetry. Those who did the amulets were part of a different team, I had only seen their set-up once, and it was something like 5-ish years ago. At the time, I had thought the camera would automatically focus stack ~20 different shots then synth it into one mega picture, but I could very well be misremembering.

 

@dabateman, you have a very impressive set-up! I agree with @Andrea B. that it's worthy of a further discussion post! I don't know how much I could pull off from that on my own with the remaining budget, but I will definitely look into it! You'll have to pardon my ignorance as I've not yet mastered how to tell which lenses are UV-permitting and don't have any coatings on them--but for the Canon lenses you listed, do any/all of them work with UV photography? For my own future reference, if anyone could point out how they find out this information, I would greatly appreciate 

 

@Andrea B. Now that I have the ring light, I won't need a torch since the ring light itself serves as one and has it's own mount. But I really appreciate your suggestions and guidance! Thank you so much for all that you do!

 

On 3/10/2022 at 2:27 AM, nfoto said:

I use Manfrotto Studio Clamps and Magical Arms in my setups. Rather expensive, but immensely versatile.

 

Manual focusing rails are too coarse in stepping for high-magnification work.

 

Could you elaborate on the coarseness? It looks like the Manfrotto would require manual adjustment too, but I absolutely acknowledge that I'm likely missing something here. Is the problem with coarseness because it's a 2D focusing rail, whereas the 3D Manfrotto setup can give you more ways to adjust?

[nfoto]

The Manfotto items are just used to keep the light(s) in the desired position. Focud stacking is done using a Stackshot macro rail and its programmable controller. The stack then is processed by Zerene.

[Andy Perrin]

11 hours ago, PDL1 said:

We didn't normally use it for even our smaller/handheld artifacts. Just the amulets that were about the size of a quarter? Some of which had inscriptions that we were trying to get in high detail. These were items not usually amendable to typical photogrammetry. Those who did the amulets were part of a different team, I had only seen their set-up once, and it was something like 5-ish years ago. At the time, I had thought the camera would automatically focus stack ~20 different shots then synth it into one mega picture, but I could very well be misremembering.

Photogrammetry is not focus stacking -- the objective is not to get an image with photogrammetry, it's to make a 3D model. The object is imaged from different locations or different angles and the computer reconstructs the 3D locations of each point on the object, along with the color at that location. In photogrammetry, the whole object is supposed to be in focus for each picture so that the coordinates of each point can be identified. 

 

As an example, I did this grave from the Miles Standish cemetery in Duxbury, Massachusetts.

 

First I took photos from many different angles and masked the background in Metashape (the photogrammetry software):

etc. (I took many more photos for accuracy).

 

The pictures are then combined by Metashape into one 3D model:

[dabateman]

Those are the only Canon lenses that will work with Canon's focus bracketing. I don't think any of them would be great for UV.

Sadly the 40mm f2.8 stm lens isn't listed and its probably the best Canon uv capable lens.

[PDL1]

16 hours ago, Andy Perrin said:

Photogrammetry is not focus stacking -- the objective is not to get an image with photogrammetry, it's to make a 3D model. The object is imaged from different locations or different angles and the computer reconstructs the 3D locations of each point on the object, along with the color at that location. In photogrammetry, the whole object is supposed to be in focus for each picture so that the coordinates of each point can be identified. 

 

Correct, but what I was trying to get at is that they were using focus-stacked images for the data input into making the photogrammetric models. The subject matter was very similar to this. Other than being the size of a coin, I can't remember why the models would turn out so poorly using regular photography without the focus-stacked pictures. Somewhere in the process they found that focus-stacked macro images made for significantly better models on these small artifacts. It was several years ago and involving an adjunctive team, so I'm very fuzzy on all of the details/rationale. But that was previously my only experience even seeing focus-stacking in action.

[Andy Perrin]

Hm. Well I guess without more details we will never know why they did that then.