Wet lenses and modern water contact optics - how do they work?

[Guest]

This is a piece I have been asked to write for a long time but I had resisted so far...

 

https://interceptor121.com/2024/05/11/underwater-wet-wide-angle-lenses/

 

The answers are all in the article and in most cases in the exif of your images if this contains focus distance....

 

Have fun!

[TimG]

Really interesting stuff, Massimo. Thanks

 

Could you just explain your final paragraph a little more? I don’t quite follow….. Thanks!

[Guest]

28 minutes ago, TimG said:

Really interesting stuff, Massimo. Thanks

 

Could you just explain your final paragraph a little more? I don’t quite follow….. Thanks!

The water contact optic doesn’t provide any stops improvement edges or not because it runs out of depth of field as it goes all the way to infinity compared to a dome solution

so you end up f/13 or even smaller at close range. Which I guess most people have found out

 

instead if you are shooting schools with nothing in front it looks great

 

unless you shoot a pool wall or something flat this type or solution doesn’t have necessarily any benefit over traditional ones due to the way it works

 

At close range it also looses field of view so there are several considerations

Edited May 11 by Interceptor121

[Guest]

This post shows some shot I took in Italy last year where the issue of depth of field is very apparent

I was shooting f.8 to maximise sharpness of the lens but I had not considered the depth of field so many shots are blurry in front when the subject is not the nearest thing to the lens

 

 

https://interceptor121.com/2023/05/23/sony-28mm-prime-vs-28-60mm-zoom-with-the-nautical-wwl-1/

 

The shots with the grouper are the ones where the issue is clearer those should have been taken at f/13 and smaller apertures

[Eric Hanauer]

On a recent trip to Cozumel I shot video on the entire trip with one lens: the LUMIX 14-40mm behind a Nauticam WWL-1, Panasonic GH5II. I don’t know to what degree 4/3 optics differ from full frame, but my imagery was sharp corner to corner. And it focused virtually on the lens. I was so impressed that I sold my big dome port. 

[Guest]

17 minutes ago, Eric Hanauer said:

On a recent trip to Cozumel I shot video on the entire trip with one lens: the LUMIX 14-40mm behind a Nauticam WWL-1, Panasonic GH5II. I don’t know to what degree 4/3 optics differ from full frame, but my imagery was sharp corner to corner. And it focused virtually on the lens. I was so impressed that I sold my big dome port. 

Wet lenses work much better on cropped sensors and compacts because of the increased depth of field

f/8 is in reality f/16 full frame. When you look at compacts the increase is even more

In fact on 2x crop MFT system a wet lens works perfectly

 

However on a full frame camera the situation is different due to the larger sensor

 

The depth of field of the wet lens is still higher than the naked lens of a factor of (1/m)^2 however when you compare it with a dome that reaches infinity at 4x from the entrance pupil of a shorter focal lenght half in my example the effect is less especially in front of the focus point the near part

On the same account though if you use the Canon 8-15mm behind a dome the IQ is indeed even better

Edited May 11 by Interceptor121

[Guest]

3 hours ago, TimG said:

Really interesting stuff, Massimo. Thanks

 

Could you just explain your final paragraph a little more? I don’t quite follow….. Thanks!

I have clarified the statement at the end

In essence compared to a fisheye with a teleconverter on a smaller dome (140mm) at equal field of view a wet lens system provides a minimal to no improvement at close range also because it looses some of its demagnification effect

Compared to a shorter optic say 14mm so half the focal length in a 9" dome instead the wet lens system looses straight on depth of field but recovers overall due to the fact you need to stop the lens behind a dome to fight the aberrations

Yet the dome optic may have over 1 stop advantage and more resolution depeding on the lens used

 

In practical terms...

 

If you want to shoot very close those systems do not offer really a benefit assuming you have a performing fisheye zoom at equal field of view

If you want to shoot fish schools the indeed have a penalty on shorter rectilinear optics and as normally the background is behind the benefit of a wet solution may be for example size unless you have a massive one and reduced drag 

 

The wet lens also offers flexibility in the zoom range which is very nice to have however as you need to stop down more and more to insane numbers when you zoom at the end most of the resolution gets eaten away by diffraction on a full frame system

 

Those optics are generally more suited to the smaller sensor they were originally designed for

 

For what I am concerned if I need to shoot schooling fish in 3 knots current I will take my WWL-1 because it has less drag  but not a 230mm dome port 

 

If I need to shoot in calm conditions the big dome may actually offer a better solution (but so far I have not been diving in such calm conditions!)

 

If I am shooting very close I would definitely not take a wet optic nor a rectilinear lens in a big dome but a fisheye with or not teleconverter

 

At the end all those solution are complementary and there is no silver bullet that fixes it all

 

Edited May 11 by Interceptor121

[Adventurer]

Bravo 👏 Massimo!

Amazing ✍️ write up and thank you for your efforts and time you put in this. This is the point where I would like to know what you do for living?

I can imagine you have a physics / engineering background or are in involved in optical design.

[Guest]

11 minutes ago, Adventurer said:

Bravo 👏 Massimo!

Amazing ✍️ write up and thank you for your efforts and time you put in this. This is the point where I would like to know what you do for living?

I can imagine you have a physics / engineering background or are in involved in optical design.

Firstly the article is conceptual there are of course data points however the exact way certain things work in real life I think are not even fully known by the people that design the products.

I think anybody with engineering or physics background can work out things you don't need to be doing this as a job. In this specific the details are in the exif of certain cameras although not necessarily 100% exact brand by brand help understanding the concept.

I have also limited the knowledge article to equipment I have. I do believe that is broadly extendible however anyone should draw their own conclusions when making decisions. For example in certain system in effect there are no good lenses that focus close and indeed certain choices are preferrable to others.
When you look closer into problems you can see that there is a lot of generalisation that goes on and in addition things move with time as camera and lenses develop. You should continuosly review performance as working on old assumptions is not effective

Edited May 13 by Interceptor121

[Adventurer]

Article Feedback:

 

“When we look at wet lens specifications, we are given the magnification of the lens alone for example 0.36x or 0.57x. This does not mean that if your lens is 28mm it becomes 28×0.36=10.08mm during the process.

 

What it means is that if something is at 1 metre from the wet lens this will indeed look as if it was 2.8 metres and therefore the field of view will be larger.”

 

When reading it, I have a hard time to understand how you derive the 2.8m distance from the factor 0.36x or 0.57x of the water contact optic. Maybe you want to refine this without leaving the reader two choices? Also adding a small line with short computation example would be nic. If this is not computed data at all it’s worth noting that it comes from the EXIF and is empirical from a specific rig combination.

[Architeuthis]

Big thanks for writing this together Massimo... 👍

 

As I understand your findings, there is a big difference between WWL1/WACP and a domeport: The domeport produces a small virtual image that is close to the domeport and is photographed by the (rectilinear) WA lens at short object distance. In contrast WWL1/WACP add additional optical elements to the front of a standard lens, the entire assembly (standard lens + WWL1/WACP) being an UW WA lens with fisheye characteristics. This is indicated by the long working distance that you read from the EXIF data (the distance is, of course, incorrect, as the specification in m is for the pure lens without WWL1/WACP and over the water)?

 

 

Wolfgang

 

[Guest]

1 hour ago, Adventurer said:

Article Feedback:

 

 

“When we look at wet lens specifications, we are given the magnification of the lens alone for example 0.36x or 0.57x. This does not mean that if your lens is 28mm it becomes 28×0.36=10.08mm during the process.

 

What it means is that if something is at 1 metre from the wet lens this will indeed look as if it was 2.8 metres and therefore the field of view will be larger.”

 

When reading it, I have a hard time to understand how you derive the 2.8m distance from the factor 0.36x or 0.57x of the water contact optic. Maybe you want to refine this without leaving the reader two choices? Also adding a small line with short computation example would be nic. If this is not computed data at all it’s worth noting that it comes from the EXIF and is empirical from a specific rig combination.

1/0.36x=2.77 meters however this is not true at all shooting distances as proven by the example at close range

 

Empirical means backing up a theory with actual data which is what this article does. As I wrote correctly

 

When you combine the wet lens with your camera lens the demagnification effect may be affected  by  the distance between the two lenses, the length of the optical path inside the camera lens and body, and inside the wet lens.

 

When you read a wet lens specification on inon webpage you are finding something like this

 

Maximum incident view angle (underwater) (*9) 

(9) When using XXX system . actual view angle may vary depending on camera / housing.

 

Now one would like to know if the stated field of view on Nauticam site are accurate.

Let's have a look at one example Sony 28mm vs Sony 28-60mm both are reported to have 130 degrees field of view however based on my EMPIRICAL observation I can tell that is not the case and the 28mm is slightly narrower and therefore those specifications have to be validated combination by combination.

 

If you want to do your own test you can proceed by installing a tripod in a pool at specified distances from a shooting target and then reading what the camera thought it was focussing at. You will find that the relationship is not always linear and therefore the magnification is not either and also that depending on which part of the frame you focus the system reports different distances due to the fact the add on lens is not corrected for distortion. I am satisfied the model I have has the degree of accuracy to make decisions and therefore personally I am not going to do that.

 

Ultimately the exact precision does not even matter is the principle that matters as it allows to compare with other optical system specifically with dome ports

[Guest]

40 minutes ago, Architeuthis said:

Big thanks for writing this together Massimo... 👍

 

As I understand your findings, there is a big difference between WWL1/WACP and a domeport: The domeport produces a small virtual image that is close to the domeport and is photographed by the (rectilinear) WA lens at short object distance. In contrast WWL1/WACP add additional optical elements to the front of a standard lens, the entire assembly (standard lens + WWL1/WACP) being an UW WA lens with fisheye characteristics. This is indicated by the long working distance that you read from the EXIF data (the distance is, of course, incorrect, as the specification in m is for the pure lens without WWL1/WACP and over the water)?

 

 

Wolfgang

 

Correct. The water contact optic pushes the image further away not only the focus point the entire image becomes smaller as you were shooting from far.

The dome instead focuses at virtual distance that is very close however the field of view is unchanged and remains within the variation of the lens focus breathing.

The systems are very different indeed

[Architeuthis]

How do you get the subject/object focus distance from the EXIF data? I guess you have a special trick or program?

 

=> I just had a look at raw files from Sony A7R5 in LRc, obtained with Canon 8-15mm fisheye, Sony 28-60mm and Sony 20-70mm, but "Focus Distance" or "Subject Distance" values are missing in my case...

 

Wolfgang

[Guest]

15 minutes ago, Architeuthis said:

How do you get the subject/object focus distance from the EXIF data? I guess you have a special trick or program?

 

=> I just had a look at raw files from Sony A7R5 in LRc, obtained with Canon 8-15mm fisheye, Sony 28-60mm and Sony 20-70mm, but "Focus Distance" or "Subject Distance" values are missing in my case...

 

Wolfgang

You need to use exiftool or other advanced viewer that can decode the more obscure tags

[Adventurer]

5 hours ago, Interceptor121 said:

You need to use exiftool or other advanced viewer that can decode the more obscure tags

On Mac AppolloOne is a good tool to browse and look at your image data: https://www.apollooneapp.com

Edited May 13 by Adventurer

[Adventurer]

I would like to take Massimo’s instructive article and apply it to two slightly unusual underwater contact optics to better understand what’s going on and maybe derive a practical application.

 

Have a look at the two for 60mm full frame focal length designed:

 

Nauticam MWL-1 and

Kraken / Weefine WFL-09s

 

Both lenses claim interesting 150deg FOV @ 60mm focal length when used in optimum. For Nauticam I am unable to find the lens factor but Weefine published 0.32x magnification. So we can assume the factors are similar.

 

Nauticam recommends to shoot their lens at least at F16 aperture for good sharpness. The Nauticam lens has 7 Elements in 5 groups and the Weefine/Kraken 7 Elements in 6 groups. So both factories take on the problem looks very similar.

 

According to Interceptor this means that an on object in one meter distance will appear at 3,13m when viewed through the lens underwater.

 

You can apply classical topside depth of field calculator to a 60mm lens @ F16 and learn that softness of that lens reported might be solely blamed to

a DOF issue. There is also one Wetpixel Review by Jack Connick and one on divephotoguide by Mathew Sullivan, where both authors report that you have to significantly stop down to F29 or F32 if you use this lens with a 100mm or 105mm macro.

 

Especially the images by Mathew exhibit shallow DOF in wide angle shots, I think.

 

Some underwater photographers might have mistaken these shallow DOF with missing corner sharpness on dome ports, I think, and did not consider these lenses as an improvement.

 

I however see these two as highly underestimated because they bring flexibility to your dive and highly optimize remote travel size and weight.

 

The downer is however that not many mirrorless Systems have native 60mm Macro lenses. Nikonians being in the supreme position, followed by Micro4thirds Users which have a 30mm Macro turning into 60mm FF equivalent. Sony FF Users need a Nikon adapter and Canon Users would need to twist an EF-S 60mm macro and live with slight vigneting on FF if used on the wrong focusing distance and aperture.

 

So,… what does this give of any use to us?

Well.. how about if we utilize classical topside DOF calculators and look at hyperfocal distances and apertures. 
 

What we then can learn from these, is that we simply have to use these lenses at higher working distances.

Sadly though both lenses were marketed heavily with the claim that you can focus right on the front element. The point where you would suffer most from shallow DOF issues.
 

When shooting these two lenses with more subject distance and your strobes brought forward towards the subject you will buy back lens sharpness.

examples: 

 

The recommended Nauticam example at F16 on a full frame 60mm lens renders as follows:

 

Hyperfocal distance focussin point on Land = 7.55 m

underwater Equivalent 0.32x HF AF distance = 2.42 m

On Land everything sharp from 3,77m to Infinty 

UW (0.32x) sharp from approx 1.2m

 

Changing these to a 100mm lens this significantly drops:

 

The NOT recommended example at F16 on a full frame 100mm lens renders as follows:

 

Hyperfocal distance focusing point on Land = 20.9 m

underwater Equivalent 0.32x HF AF distance = 6.7 m

On Land everything sharp from 10.45 m to Infinty 

UW (0.32x) sharp from approx 3.35 m

 

Having worked through the above these lens-rapers go heavily against the average underwater photographers intuition and could make a small exception of the steadily heard “get close”.