I decided to start a new thread on this becasue my previous posts about this question (and many others') came under sidetracked forum topics.

I'll start with this image:

This is a shot taken recently on my first dives using my new Nauticam housing. ISO160, 1/250s, F11, 30mm (2x15mm). I chose this image as it has a lot of detail at the edges and is excellent cenrte focus.

Gear setup for reference:

• Sony A7rV, Metabones converter, Sony 2x TC, Canon 8-15mm Fisheye

• Nauticam Housing, 35.5 n100-n120, 60mm Extension and 140mm Nauticam dome port.

My question relates to ideal port extension, and image quality. I have seen others post about this specifically, but none of the images posted previously (IIRC) never really helped understand whole-of-image quality becasue they lacked detail in the edges. I think this photo does well becasue it does have a lot of this detail to evaluate how it's performing.

One of the previous, sidetracked posts (perhaps sidetracked by me), @Architeuthis had suggested 60mm extension was a good starting point for this setup. It's what i've used here.

My view is that the image quality is great for about 75%+ of the frame, but drops at the very far edges. Something like this range is where I think it starts dropping off:

A zoom-in on the areas that have a lot of detail..

Bottom right:

Bottom left:

Top left:

And for the sake of it, this is the centre:

Others might disagree and have better words to describe what I think I'm seeing. I'm sure theres a bit of out of plane blurring happening, but I think it's just slightly more than that.

My questions are:

1. Has anyone had any experience and good example images like this one that show off the performance of different extension lengths for this exact setup? We might be talking 5mm increments here.

2. Does anyone have any alternative views on what the ideal extension length might be (Ideally on the Sony and using the 140mm dome port)?

TIA!

Aren't the softer edges expected with the 140mm? I shoot a Zen DP100 with the Canon 8-15 (Nauticam NA-A7C + N100 ->N120 converter (no extension)) and run into the same, It's just the limitation of the dome size, thus the advantages of the larger domes. Can that be minimized with extensions?

21 minutes ago, foetusmachine said:

My questions are:

1. Has anyone had any experience and good example images like this one that show off the performance of different extension lengths for this exact setup? We might be talking 5mm increments here.

2. Does anyone have any alternative views on what the ideal extension length might be (Ideally on the Sony and using the 140mm dome port)?

TIA!

Welcome, Tia — two quick thoughts that may help:

1. There’s a must-read thread on corner sharpness with the Canon 8–15 fisheye. The author questions the Nauticam port chart and also ran extensive Nauticam port tests. It includes a simple “at home” position test you can do to confirm whether the lens is sitting at the correct position in the dome:
   https://waterpixels.net/forums/topic/1438-testing-nauticam-n120-port-extension-for-140mm-and-180mm-domes-with-wide-angle-lenses/

2. Even if everything is set up correctly, you can still run into depth-of-field limitations underwater (similar to what you’re seeing on land). Your sample image might be a good example of that. Former forum member Interceptor121 has a very good write-up on this, and he repeatedly notes that placing your focus point closer to the corners of the frame can help improve corner sharpness:
   https://interceptor121.com/2023/03/18/canon-8-15mm-with-kenko-1-4-teleconverter/

I don't really see an issue with some minor dropping of corner sharpness in an image like this. If you don't like the look, you can try a smaller aperture next time, or crop in a little tighter clipping off the edges. I personally like an 8x10 or 1x1 ratio for a lot of my photos. I do a lot of large format prints and find these to be the best options for display.

Also keep in mind that adding a teleconverter will always impact image quality to some degree by its nature. A 1.4x TC will have less of an impact than a 2x, but there is a degree of difference compared to native image quality. As @Dave_Hicks mentioned, there are ways to make the most of the situation.

In the end it’s a trade off of ultimate image quality vs having a little extra “reach” with the addition of the TC. Neither is better than the other, it’s just a trade off based on need and desired results.

In my opinion fisheyes are less sensitive to precise dome position. I took a great deal of time comparing the images at two different dome positions in the linked testing thread. I went as far as downloading the two comparison images and taking a 100% crop from the extreme corners and pasting them together in a single image to compare them. I squinted, compared different areas of the images but I really struggled to see an earth shattering difference in the two images; Here's the post, let me know if you can see anything significant.

As far as the the image presented by the OP, the corner crops where it is really blurry all appear to be foreground or background elements that are falling outside the depth of field. It seems like you must be really close as the BG behind the eel's head is quite out of focus, likewise the bottom right crop the foreground rock is quite out of focus it seems. With slightly less magnification you could look at focusing on an important foreground element rather than the subject , other wise crop out objectionable parts of the image or stop down a little more. Also looking at the composition before shooting to avoid large out of focus foreground areas which tend to look more objectionable than background being out of focus.

There is is also the fact that images viewed at 100% crop often don't look fantastic, but when presented in the final image and sharpened well, the overall image generally looks great.

As Chip said the 2x converter is softening things a little, it provides flexibility more so than the ultimate image sharpness. It's only really possible to do this because the bare 8-15 is such a sharp optic in the first place. You could consider also trying the 1.4x and give up on a little zoom range and trade it off for image sharpness.

While it is important to understand the various trade offs in dome ports and fussing over details, eventually you have to just get out and take some images. Work out how to best process and crop them and enjoy the overall image.

Congratulations to your new setup!

To me, the sharpness in the edges of your sample image look o.k. I believe this is what one can get with WA underwater - more or less...

Regarding extension, I remember that I once was using 5mm more as suggested by Dreifish (65mm instead of 60mm) and then I had got problems with vignetting at 16mm. Only with some photos, therefore I believe this was due to the IS. I also did not notice that IQ was recognizable different...

I personally am more concerned about the drop in IQ (sharpness, microcontrast) in the center when using a 2x Sony TC, the fisheye lens without TC performs clearly better. IQ is, however, still o.k. with the 2x TC and since the WACP-C/28-60mm combo does not provide perfect IQ either, I preferred the Sony 2x TC/8-15mm over the WACP-C/28-60mm on my last trips, since often I prefer the 180° diagonal over the 130°.

Also the travel weight is better (in case I take the WACP-C with me, I still take the Canon 8-15mm&140mm domeport in order to have the wider 180° angle in addition - my choice is to take the WACP-C OR the 2x TC (plus two extensions))...

Just a side note about using different TCs with Canon 8-15mm on Sony FF with the 140mm domeport that may be of interest (subjective judgement, no objective resolution testing):

I also have the Sony 1.4x TC and once tested it UW, as I was hoping to get better center IQ compared to the Sony 2x TC - at the cost of less zoom range. Surprisingly IQ was comparable, probably even worse, with the Sony 1.4x TC. I decided the Sony 1.4x TC is not a good combination with the Canon 8-15mm fisheye...

The situation is different with the Kenko Pro HD 1.4x TC. This TC performs well together with the Canon 8-15mm. Maybe IQ is a little better compared to the Sony 2x TC, but I am not sure about it...

Wolfgang

Edited yesterday at 11:11 AM1 day by Architeuthis

1 hour ago, Architeuthis said:

I remember that I once was using 5mm more as suggested by Dreifish (65mm instead of 60mm) and then I had got problems with vignetting at 16mm

If you look at the entrance pupil position, of the Canon EF 8-15 everybody in this forum who used and tested the small Nauticam and Marelux (approx 140 diameter) dome is suffering from a mis-positioned lens. Both these domes are not full spheres. They are missing a tiny bit and the ball is not exactly cut in half.

Due to that it‘s recommended to shoot them at minimum F16 or smaller aperture values.

If you want to max out the IQ and archive „superposition“ you have to die the death of vignetting and combine it with a 2.0x TC and in most cases you will also have to remove the sunshade of the dome underwater and use it from approx 18mm to 20mm focal length onwards. With my MARELUX and small dome this wil require 85mm of extension rings. Vignetting disappears from 18mm onwards.

Does anyone have any real world photos they've taken using this setup, but with a different extension?

33 minutes ago, foetusmachine said:

Does anyone have any real world photos they've taken using this setup, but with a different extension?

IMO, you are chasing a tiny improvement at best, there's thousands of great shots taken with the Canon 8-15 with the recommended dome dimensions. The test shots displayed on here of the banknotes certainly didn't set the world on fire.

I would be interested to see some shots also - test shots are generally preferred as they are a shot of a flat surface and you don't have to work out if the softness is due to optics or if the foreground is falling out of the depth of field.