Fujinon or Hasselblad 30mm for Xpan

Is there any reason not to get the Fujinon version of the 30mm lens for the Xpan, rather than the Hasselblad branded version? The Fujinon seems to be less expensive. Also, is the center filter for the 45mm lens the same as for the 30mm? My guess is not.

..you seem to know that the Fuji and Hasselblad 30mm lenses are the same. The centre filters for the 45mm and 30mm are NOT the same. Neither diameter nor 'strength'. Neither lens is really useable in 'pan' mode without filter. The 30mm lens filter has a factor of x3 - so not an auwful lot of light reaches the film even at full aperture of f/5.6. There is no vignetting when it's in place and the photos show perfect illumnation across the frame. 30mm lens was sold as a kit with the centre filter and viewfinder and was very very expensive.

I'm curious about the above statement that the centre ND filter for the 30mm lens is only 3x maximum density. -1.6 EV (3x) isn't really a lot of vignetting for any wideangle lens. Plus vignetting at wide apertures is mainly mechanical and due to oblique rays being blocked by the apparent cat's-eye shape of the iris. Stop down and vignetting reduces considerably.

 

So it's difficult to see how a centre ND filter can work perfectly at all apertures.

 

Depending on the film you're using, you might well get equally good results by using an image editor to correct the vignetting.

 

It might be cheaper to make a little 'windmill' such as was used on the Goerz Hypergon lens.:)

RJ, Rodenstock and Schneider recommend stopping down two stops when using their center filters.

 

When shooting reversal film with a very wide angle lens, a CF is essential. 1.6 stops down center to corner is quite a lot. There's a limit to what an image editor can recover.

 

For a fuller discussion, please see

 

6x9 lens by Dan Fromm (1) (this is really a link to an article on CFs. I don't know why the PN editor substitutes irrelevant text for it, but it still works. I know, it isn't a bug, it is a feature.)

 

 

You might also want to take a look at http://www.galerie-photo.com/horseman-4x5-exposure-meter.html, which prompted me to write the CF piece.

Edited November 24, 2017 by dan_fromm|2

I'm curious about the above statement that the centre ND filter for the 30mm lens is only 3x maximum density. -1.6 EV (3x) isn't really a lot of vignetting for any wideangle lens. Plus vignetting at wide apertures is mainly mechanical and due to oblique rays being blocked by the apparent cat's-eye shape of the iris. Stop down and vignetting reduces considerably.

 

So it's difficult to see how a centre ND filter can work perfectly at all apertures.

 

Depending on the film you're using, you might well get equally good results by using an image editor to correct the vignetting.

 

It might be cheaper to make a little 'windmill' such as was used on the Goerz Hypergon lens.:)

...appears to give perfect coverage. I don't use it all all apertures, probably mostly at f/8. Not my favourite lens for XPAN - that would be a Nikon 135mm. Pity a proper one was never made.

.... Plus vignetting at wide apertures is mainly mechanical and due to oblique rays being blocked by the apparent cat's-eye shape of the iris. Stop down and vignetting reduces considerably.

 

So it's difficult to see how a centre ND filter can work perfectly at all apertures....

 

Geometric cat-eye vignetting due to the central iris is the same for all stops. So a center filter placed a reasonable distance from the aperture will work similarly for the aperture range. Lenses where vignetting decreases with smaller apertures is due to undersized front or rear elements.

"Geometric cat-eye vignetting due to the central iris is the same for all stops."

 

- Sorry, but it most certainly is not. Otherwise vignetting wouldn't reduce upon stopping down. A wide open lens shows an oblique 'mandorla' or almond shaped intersection of the aperture with the edge of the lens. This becomes less elliptical as the aperture is closed down, and becomes a (distorted) full circle at very small iris diameters.

 

Simply viewing the aperture/iris at an angle through the rear of a lens while stopping down will clearly illustrate the mechanics of vignetting.

 

The commonly blamed Cos^4 law accounts for very little of the vignetting effect seen with most practical lenses.

 

"There's a limit to what an image editor can recover."

- That's why I opened by saying 'Depending on the film you're using...'

 

I still think 1.6EV is quite a low degree of vignetting compared to that reported on other lenses with a narrower angle of view.

The commonly blamed Cos^4 law accounts for very little of the vignetting effect seen with most practical lenses.

 

Are you sure? It certainly bites, and hard, for angles of view as large as 90 degrees.

 

To put it another way, do you mean to say that extreme wide angle lenses aren't practical? I have a weakness for them, can't agree with that proposition. If that's what you meant.

 

Rodenstock and Schneider recommend using center filters for lenses that cover 100 degrees or more. The consensus of users on the US LF forum is that the shortest focal length that can be used with no movements on a 4x5 camera without a CF is around 90 mm. The angle of view is ~ 80 degrees. Add decentering movements or go shorter and there's no avoiding a CF. The corresponding minimum FL for 2x3, which I shoot, is around 65 mm. I sometimes shoot a 47/5.6 SA on 2x3 without a CF, whether I can get away with it depends on the subject.

"Geometric cat-eye vignetting due to the central iris is the same for all stops."

 

- Sorry, but it most certainly is not. Otherwise vignetting wouldn't reduce upon stopping down. A wide open lens shows an oblique 'mandorla' or almond shaped intersection of the aperture with the edge of the lens. This becomes less elliptical as the aperture is closed down, and becomes a (distorted) full circle at very small iris diameters.

 

Simply viewing the aperture/iris at an angle through the rear of a lens while stopping down will clearly illustrate the mechanics of vignetting.

 

The commonly blamed Cos^4 law accounts for very little of the vignetting effect seen with most practical lenses....

<snip>

 

The 30mm lens has a ~100 deg field of view. The cos^4 geometric vignette for that is a little over 2 stops. The center filter provided for that lens is 1.5 stops, optimized at f11.

 

The oblique view of any stop, regardless of size, is ovalized.

 

The change in vignette with aperture is due to undersized front (or rear) elements, which result in the cateye bokeh with the sharp corners. When the lens element is large enough, you get a smooth oval.

 

The $3k 30mm Hasselblad lens does not have an undersized front element (not significantly, anyways). It also does not use any optical trickery to reduce geometric vignette (like retro focal lenses offer do).

If the cos^4 law held true, then a 180 degree fisheye lens would never be a possibility, would it? No; but such lenses do exist, because there are optical design 'tricks' that can effectively magnify and elongate the exit pupil away from the lens axis.

 

The cos^4 law is just another theoretical optical law that has little relevance to practical photography with real lenses. We don't use virtual thin-lens models in front of our actual cameras.

 

Tom, may I suggest you just look through the rear of a wideangle lens at an oblique angle sometime?

If the cos^4 law held true, then a 180 degree fisheye lens would never be a possibility, would it? No; but such lenses do exist, because there are optical design 'tricks' that can effectively magnify and elongate the exit pupil away from the lens axis.

 

The cos^4 law is just another theoretical optical law that has little relevance to practical photography with real lenses. We don't use virtual thin-lens models in front of our actual cameras...

 

Cos^4 is a geometric property, which is present in every lens, even in your fish-eye. The optical trickery used (for retro focal lenses as well) is distortion. You compress the image at the periphery to increase the intensity, which shows up as barrel distortion. A fish-eye is barrel distortion taken to a limit. But you cannot use this trickery for the back focus, as you need to project the image onto a flat plane.

 

The back focus of retor-focal FF SLR lenses is around 55~60mm, and a 60mm lens has a cos^4 of ~1.2 stops. This is close to the residual vignette in most SLR wide angle lenses (for a reason). When you are forced to retro-focus a lens to clear a mirror, then you may as well use some optics to take advantage of the low vignette back focus cos^4. This is usually done by distortion in the front focus, and un-distorting in the back focus (or as people do digitally, fish and de-fish).

 

Tom, may I suggest you just look through the rear of a wide-angle lens at an oblique angle sometime?

 

If you look obliquely through the back of most wide-angle lens, you will clearly see the edge of the front element occluding the aperture. This is the cause of vignetting at large apertures. When you reduce the aperture until it is free of the occlusion, any further changes in aperture does not significantly change the vignetting. Which is what I said in my first post - "Geometric cat-eye vignetting due to the central iris is the same for all stops." This is also true if optical trickery mentioned above is used to reduce vignetting. For curvilinear lenses, the residual vignette is dominated by the back focus cos^4 (or about 1.2 stops for more SLR lenses).

However, this thread is about the Hasselblad/Fuji 30mm f/5.6 lens. It has a symmetrical design with a back focus of ~30mm and a field of view of 100 deg (ie: not retro-focused). The design minimizes distortion over vignetting, which means the vignette is dominated by cos^4. This is true for most pre-digital large format wide angle lenses, where the use of a center filter is common. In these cases, the centre filter is optimized for the non-occluded aperture.