Hi,

I have a particular question.

For a scientific project I need to photograph fish habitats. Photos will then be converted to simulate a fish viewing perspective (correction for acuity and photoreceptor sensitivity). The studied fish are small benthic fishes (like gobies). In the past I have used a large dome but I am not close to the bottom enough to simulate a fish viewing perspective. A flat port would be better but the field of view would be quite limited. Another issue with, for exemple with a 28-60 (I have now a Sony A7RV) is the minimal focusing distance: 30cm at 28mm, meaning that the foreground is blurry.

How would you solve this trade-off between having a large field of view, close focusing distance and sitting on the bottom ?

Thanks for any advice.

Fred.

I do not have it by myself (too much bulk as well as too expensive), but I imagine that the Nauticam EMWL with Angled relay lens may be perfect for this issue (CFWA and perspective from the bottom).

(different "objective lenses" from 60° to 160° are available)

Does it have to be rectilinear? What FOV are you targeting?

What field are you looking at? A fisheye (or add a teleconverter) would allow this and can be used in a small dome if required, or you could look at the MWL-1 it gives a 150 deg diagonal field, it's used with a Sony 50mm macro, downside is it needs to be shot at f14-16 to give its best. It can get close to the bottom as it's only 100mm diameter. It also works with a Sony 28-70 f3.5-5.6 lens between 60 and 70mm to give a 150-110° diagonal field.

You could probably use a Canon 8-15 with sony 2x combo with a zen 100mm dome (at the cost of a little corner performance) to get fields between full 180° diagonal fisheye and something approaching a 28mm FF rectilinear in terms of horizontal field and it will focus down to the dome.

Thanks, these are useful suggestions. I should clarify that I am trying to sample the spatial statistics of the fish’s visual environment, rather than reproduce its entire visual field in each photograph.

I do not yet have a biologically justified FOV target. The priorities are a viewpoint very close to the substrate, close focusing, and enough usable coverage to sample the habitat efficiently. I can take photographs in several directions.

It does not necessarily have to be rectilinear at capture. A fisheye with a small dome could work if I can calibrate the underwater setup and extract rectified views for analysis. The angular extent of those views needs to be known so that I can simulate the fish’s visual acuity correctly.

I'd definitely play around with some fisheye images and correction before deciding that is the route. If you are just going to use the center of a fisheye, then it is less important (but I would hate to try and get quantitative data from it). I can say the canon fish eye correction looks bad and the middle isn't anything like rectilinear. But if you have the sony - the WWL1 (or WACP-1) seems to be the logical choice for close focus on Sony. For Canon I would just use the 24mm Macro behind the small dome (140/180). Super sharp, easy to use just not that wide, very easy to light smoothly.

I imagine that the only way you can get close enough to the bottom is a very small camera... i.e a maximum size of a mobile phone in a housing, an action camera, or a custom solution like a modified security, or spy camera, or some sort of a probe camera with a separate monitor. It would be interesting to see how low a mobile phone in a housing can go with the lens closest to the substrate. I would investigate the mobile phone option and how it fits with macro-photogrammetry/TagLab type workflow. Looks like you can get an intervalometer app for a mobile phone and other options for 3-D photogrammetry apps with modern phones.