6 hours ago, DreiFish said:

If anything, it's harder to compose your shot to hide/minimize the pincushion distortion at the edges with video than it is with stills.

I've shot video extensively with the 14-35 L and have used it frequently at 14mm.  What I really run into isn't the problem with pincushion distortion but depth of field.  I want the back ground in focus and that means I either need to use a really small aperture or be slightly farther away from my main subject.  Often that works out OK but then I find the corners distracting as they appear blurry and in a way that looks more like a mistake than an artistic choice.  Zooming in just a weee bit (like to 16mm) means that I find it easier to compose and not find the depth of field distracting from the image I like to create.  Most of this was shot at 14mm:

The exception to the 14mm was the ctenaphone (which was shot with a gopro...).  You can see what I mean with the shot of the baby fish hanging out on the plates of ice.  It looks like my focus was way off, but really the bottom of the frame was super close and out of the depth of field meaning I was not happy with the shot.  I've tried dialing this in using manual focus and focus peaking but it just never comes together perfectly in my opinion (hence me starting the fisheye thread) whereas it is a bit easier at 16mm because the subjects at the bottom are just farther away from the dome.  (This was shot through a 230mm dome, FYI)

On 5/31/2024 at 10:13 AM, Adventurer said:

F8

Samyang 14mm positition with my computed 35mm extension looks pretty nice.

The chess fields seem same size to me above and below water, what do you think?

 

Unfortunately I have slight vignetting in the corners with that Dome and port-opening.

I have to wait until next week, when I get my MARELUX 30mm Extension to see if it performs just as good. The vignette is also there outside of the water and it is purely an extension ring issue.

 

In the second test picture the I feel the letters and chess board underwater look smaller.

I am not sure this is my fault not being able to keep the chessboard exactly vertical or not.

If it is not my fault, the picture exhibits the lens misalignment backwards which gives me hope for the 30mm ring.

 

This shot is @ F8

How I build a simple garden test ground to verify theoretical lens positions.

Another test shot more far away @ F11 exhibiting smaller letters underwater:

Illustration on what should be observable.

 

Reviving these old findings, as I have done some Fisheye +1.4TC and +2.0TC tests today.

Quint essence: if your NPP aligment is not 100% prudent you will struggle hard to get a decent thin water line.

In fact you might constantly fail to get a decent over under split shot, just because of the NPP mispositioning.

 

We previsouly concluded that if you go wider than 20mm focal length the pincussion distortion get's pretty annoying and is hard to deal with as most of us tend to tilt the camera up and down for their underwater compositions.

 

Another thing to consider is the hyper-focal distance, to maximize sharpness of above water and below water scenes. On full frame sensors this will quickly force you to ditch diffraction concerns and set F22. On 20mm focal length you might even want more.  In the Canon RF lineup there are actually not many wide angle lenses which will allow this. Most of them have F22 as a maximum. Basically you have two roads to go: add a TC 1.4x or 2.0x to the Canon EF 8-15mm Fisheye and enjoy beeing able to set your lens to 21mm focal + enjoy F32. OR stay native in the RF mount and enjoy 5.5 stops OIS + IBIS and an ultra light weight and affordable lens, the: Canon RF 15-30mm F4.5-6.3 IS STM which has also proven to be sharper behind domes than it's more heavy and pricy siblings from Canon's L line. The latter ones however will not allow you more than F22 -> so here comes the real charming fact about the grey STM wide angle zoom: it does F32 @ 30mm focal length. At the highly desired 20mm focal length it will give you F25.

 

As I was also asked about the alignment tests with the charts above, I may have a more easy simplyfied wording for you all: If the stuff underwater looks bigger than what's on the surface, you lens is very likely to be positioned too close to the dome and you will need to buy more extension rings from your housing maker dealer. Very often these may be as many lenghty extensions as you can afford before you hit tunnel vision from the port system.