Wide-Angle Face-Off: Asahi 20/3.5 vs. Tamron 21/4.5

[OlDoinyo]

Good wide-angle candidate optics for ultraviolet, especially at the shorter focal lengths, are not always easy to find; many lenses in this range have already been shown to have such wretched bandpass that they are scarcely usable at all for UV. One notable exception is the Tamron 21/4.5 (and its rebadged versions,) which has been shown to have at least marginally acceptable sharpness at f/16 and enough bandpass depth to engage in many of the typical kinds of UV photography. But one must be on the lookout for others. One possibility that attracted my attention recently is the Asahi 20/4.5 Super Takumar, dating from the 1970s. First, it seemed that no one had evaluated it as a UV lens to date. Second, it is old enough that it may not contain crippling anti-UV countermeasures sometimes found in later optics. Third, some longer-focal-length Takumars not only have respectable bandpass, but are exceptionally sharp with very little chromatic aberration in UV. Fourth, vintage enthusiasts who have used this lens for visible photography have raved about how vivid the blue colors in their images were; one blogger said, "this lens loves blue." This last clue led me to think I might be onto something, so I obtained a specimen via Ebay to test--and the test did indeed prove somewhat interesting.

 

The lens as received proved somewhat more compact and considerably lighter than the Tamron, despite the similar focal length. Unlike with the Tamron, there is no provision for rear-mounting filters; there is a front filter thread of 58mm, but there is so little clearance without corner-clipping at the front that it is hardly likely that any filter at all could be mounted there; it is not even clear that one could use a step-up ring without encroaching on the image circle. However, this did not prove an obstacle to testing the lens.

 

The first test was a comparative pinhole bandpass test with the Baader U2 filter on the pinhole, presented in split-image format. A scrubbed aluminum cookie sheet was used as background in daylight; the rear disk was removed from the Tamron for this purpose:

 

 

In absolute terms, neither of these lenses is a bandpass champion, but we expect this with 20-21mm lenses. What is more interesting is the comparison. The color shown by the Tamron may be ever so slightly less vivid than that shown by the Asahi, but the difference looks quite small to me (and I had to adjust the two backgrounds to match due to changing light, so this test is not quantitatively precise.) These lenses have similar enough bandpass by this test that I am not sure that I can tell them apart on this basis, and it is questionable whether any such difference would be significant in real-world use. So far, so good. Next are the actual image-quality tests. For the Tamron, I used the rear-mounted UG11/S8612 disk, the only UV filter I have which fits this lens in any way. For the Asahi, I bodged the Baader U2 onto the front with step-down rings--not ideal, as it truncated much of the image rectangle, and it is a dichroic filter, which I would normally never use on a mid-to-short wide-angle optic; but one uses what one has. As subject matter, I chose three vehicles parked in a driveway, which provided both an abundance of chromatic surfaces and a convenient reference surface for white balancing (the grey driveway gravel.) All images are displayed BGR. White balancing was done individually on images.

 

First, I shot the middle of the scene with the pinhole itself, just as a color reference:

 

 

Next is the scene through the Tamron. At maximum aperture and infinity focus. the image is too blurry to be usable; it is not even as sharp as the pinhole image:

 

 

Stopped down to f/16, the image quality is better, and this is what I have done when I have used the lens for actual photography:

 

 

Zooming in on the automobile in the center, we see that there is in fact still quite a bit of blurring:

 

 

Examining some tree branches closer to the image edge, we see yet more blurring and quite a bit of chromatic aberration:

 

 

Now to the Asahi. Here is the image shot wide open, at f/4.5:

 

 

There is a lot of veiling and smudginess to this image, but it is clearly not completely unusable. Color resolution seems comparable to that with the Tamron. Zooming in on the center shows that the image is in fact a lot sharper than the Tamron could muster even at f/16:

 

 

Examining the same tree branches as before shows some blurring but considerably less chromatic aberration than with the Tamron:

 

 

Finally, we evaluate the Asahi stopped down to f/16. This lens seems less contrasty than the Tamron (although this could be due to the different filters used.)

 

 

The center now looks tack-sharp, and one can make out fine details:

 

 

Examining the tree branches again, we see that this image is the sharpest and that there is only a hint of chromatic aberration:

 

 

In conclusion, the Asahi 20/4.5 is a worthy competitor to the Tamron for wide-angle UV photography if a suitable filter can be devised and if the filter mounting problem can be solved. It is possible that the poorer sharpness of the Tamron images is partly a result of being unable to focus to infinity, this problem being exacerbated by the use of a rear filter disk much thicker than the factory-issue skylight disk furnished with the lens; indeed, some who have used this lens on UV-converted cameras with no rear filter in place have reported better sharpness than that shown here. But it is less clear that the difference in chromatic aberration can be explained this way; that is more likely a genuine difference in the properties of the lenses themselves. I can find no significant difference in the chromaticity of the images from the two lenses, although both lag noticeably behind the pinhole in that department.

 

As a footnote to this comparison, both these lenses are IR-capable as well, without hotspotting, though the Asahi may again be a bit sharper in this department.

 

Edit: I have now determined that the Asahi lens will accept a step-up ring without corner clipping. Experiments have shown that 62mm is not large enough but a 67mm step-up ring seems to accept a good-sized filter without encroachment.

 

Edit2: The maximum aperture of the Asahi lens was erroneously stated as f/3.5. It is in fact f/4.5.

[nfoto]

An interesting comparison - thanks for the heads-up.

 

In general, my 21/4.5 Tamron (without rear filter) on my Nikon D3200 (internal Baader U) is more than sharp enough to make good use of the camera's 24 MPix sensor. Focus shift in UV is very small and mostly accommodated by the larger depth of field brought about by the short focal length. Thus I focus it visually and don't bother using Live View.

 

The Tamron does well in IR, so confirms your observation there.

[Alex H]

I can attest to the fact that placing very thick stack of glass behind the lens, especially wide-angle lens, will cause multiple aberrations.

 

Asahi 20/3.5 seems like a very nice lens, but I see its is priced higher than I would want to spend...

[Andrea B.]

An excellent write-up, Clark.

Thank you!