[GUB]

The plan was to compare projections and centre sharpness.
Wet day here in NZ so looking out the door is as good as it gets. But does nicely for comparing projections.
These are on the K-1 and the Samyang is saved as apsc. So for simplicity I resized it 150% in the details.


The Fujinon 16mm 2.8 is a M42 and seems to be very uncommon. A lovely big lens with internal filters. I will open a review on it when I have got some experience with it.
Has anyone else here got one?
The Fujinon seems to have a wider field of view than the Zenitar but maybe softer on the corners. Needs a better subject to compare that.

The Samyang seemed noticeably faster than the others - guessing a t stop difference.
Don't compare the exposures here - it wasn't a t stop comparison. The shutter speed was varied in M mode to roughly match the change of aperture.

Attached Images

     

[ZombieArmy]

I think the Fujinon and the Tokina are best in terms of center sharpness at 5.6 but the Tokina reeaally sucks wide open. The Zenitar isn't far behind the Fujinon wide open or stopped down.

The samyang is a completely different focal length and apsc so that's not exactly fair.

In terms of projection I think the Samyang is the clear winner in my eyes. The Zenitar and Fujinon projections are almost identical to me.

[GUB]

The samyang is a completely different focal length and apsc so that's not exactly fair.

Agree. This is a comparison for it not a competition.
I read somewhere lost online that the Fujinon threw a 180 degree FOV on FF which is what the Samyang does on apsc.
And when you compare the extreme corners of the two then yes there is not much in it. And warning - I moved the pink shoes.
But the 8mm cropped should be more like 12mm on FF.
I guess it is all about projection.
But it does clarify something for me - that the Samyang doesn't really have a purpose on the K-1 now that I have these 16mm fisheyes. Unless it's newer coatings can bring something in high contrast shots. I have always been impressed with it's ability into the sun.

[stevebrot]

But the 8mm cropped should be more like 12mm on FF.

That is where things get strange. The FOV along the field diagonal for a 180° fisheye should be the same for all three lenses* when paired with their intended image frame, regardless of their focal length. Magnification, however, is another matter; it follows the focal length.


Steve

* The Zenitar is also 180° FOV on the diagonal for 24x36mm FF.

[GUB]

That is where things get strange. The FOV along the field diagonal for a 180° fisheye should be the same for all three lenses* when paired with their intended image frame, regardless of their focal length. Magnification, however, is another matter; it follows the focal length.


Steve

* The Zenitar is also 180° FOV on the diagonal for 24x36mm FF.

It is to do with the projection isn't it?
Given that the Samyang head in the detail is obviously coming from a wider angle lens (it is smaller) yet the over all framing is about the same.

[stevebrot]

It is to do with the projection isn't it?
Given that the Samyang head in the detail is obviously coming from a wider angle lens (it is smaller) yet the over all framing is about the same.

I have been thinking about this and you are probably correct that the projection may defeat the preservation of area or volume in different portions of the frame and comparison must be done for like against like. Samyang makes a 12mm FF fisheye, also with stereographic projection. I suspect that it will provide almost identical framing and proportions to the 8mm.

FWIW, I have been shooting with both the Zenitar 16mm and the Rokinon 8mm for several years and have often been put off by the Rokinon's projection. The Zenitar is quite predictable while the Rokinon is not. The Rokinon handles curved contours quite well, but fails badly when presented with something like a flat wall receding left to right as in the photo below:



The Zenitar would not have flattened the perspective at the left. The article linked below may also be pertinent in that both lenses use equisolid projection.

https://www.ephotozine.com/article/nikon-10-5mm-zenitar-16mm-fisheye-lens-co...s-review-14651


Steve

[GUB]

I find the Samyang 8mm erratic in sharpness - often smeary like above but every now and again a good shot pops out. It seems to do coastal well. Contrast perhaps or perhaps because colours are a little less saturated.

I have been trying to "visualise" projection in a practical way. To distort the image as such it is like it is changing FOV (focal length) from wide at the centre of the image to narrow on the outside.
Seems impossible with the front elements because the light going through any one point them is very unfocused and destined for all parts of the sensor. And also for the rear elements of standard 50mm type lenses because the light is almost equally unsorted as it leaves the lens.
But I suspect on very retrofocus design the light passing through the last element is already close to focused so the last element(s) could be the driver for projection. (This is why the exit element on a wide is much smaller the the front element.) The outer edges of this element could be ground to a different focal length compared to the centre
Have I thought it through even remotely correct?