EL-Nikkor 63 mm lenses
Nikon EL-Nikkor 63 mm F/3.5
Nikon EL-Nikkor 63 mm f/2.8 N  

As far as I am aware, there are three models of EL-Nikkor lenses with a focal length of 63 mm. Two have a maximum aperture of f/3.5 and are mounted in a black metal barrel with scalloped and knurled aperture ring. This is typical of EL-Nikkor lenses manufactured roughly between the late 1960s and early 1980s. Specimens engraved "Nippon Kogaku Japan" are older, while those engraved "Nikon" were manufactured from the mid-1970s to the early 1980s, after the Nikon brand name was adopted, and quickly started making its way onto photographic products. These two models are otherwise identical, and in the following discussion I am treating them together. The third model is in the N series and has a partly plastic barrel. It was the latest to be produced.

EL-Nikkor 63 mm f/3.5

The EL-Nikkor 63 mm f/3.5 (above figures) has an aperture ranging from F/3.5 to f/16, and is specified for an enlargement (on photographic paper) ranging between 2X and 20x, with an optimum at 8x. It is built with 6 elements in 4 groups and a tweaked double-Gauss optical formula not unlike that of the 50 mm f/2.8 (albeit the rear doublet and rear single element switch places in the 63 mm). The maximum aperture of f/3.5, with this focal length, would require a front lens element with a diameter of 18 mm, i.e., approximately the same as the diameter of the 50 mm f/2.8 in the same series (theoretically 17.9 mm, 22 mm in the actual lens). However, the 63 mm has a substantially larger front element (31mm). This increases the cost of manufacturing, and is sometimes done in high-end lenses to reduce or eliminate vignetting with the aperture fully open. The maximum negative size is specified as 32x45 mm, in itself a rather unusual format.

The lens mount is the standard 39 mm thread. The lens has a filter mount with a diameter of 40.5 mm, like the EL-Nikkor 50 mm f/2.8. As a whole, this lens is slightly bulkier than the 50 mm f/2.8. The lens coating has a definite yellowish colour.

This lens is scarce, and usually commands a high price on the second-hand market. The reason for this is that several Internet sources describe it as especially suitable for photography in the near-UV range. I test this claim below, together with the general performance of this lens. The unusual negative format also contributed to scarce numbers of this lens being sold. As an alternative to the 50 mm for 24x36 mm negatives, it was slightly less luminous, slightly more awkward to use for large paper sizes, and significantly more expensive. Its image circle was also too small to be selected for 45x60 mm (i.e., the next relatively popular format), and far too small for the more common 60x60 mm negatives.

EL-Nikkor 63 mm f/2.8 N

The EL-Nikkor 63 mm f/2.8 is part of the N series, i.e., the last series of EL-Nikkors of short and medium focal lengths (up to and including 105 mm - higher focal lengths were placed in the "professional" A series). It has the external plastic barrel characteristic of this series (with metal internal barrel, lens mount and filter mount). Like all lenses of this series, the aperture scale is illuminated through a window in the base of the lens. If this lens is reversed for photomacrography, the mask covering this window can be disassembled and put back rotated by 120°. This covers the window and prevents ambient light from reducing the image contrast.

The optical formula has been redesigned, like in all N-series EL-Nikkors. The use of more modern glass types and design techniques allowed a reduction in thickness of the elements, and perhaps an overall, moderate improvement in performance (which was already quite good). However, the 63 mm was the only lens in the N series to gain half a stop in maximum aperture, with respect to its predecessors. In spite of this, the size of the front element (31.5 mm) hardly increased at all. The one of the 50 mm f/2.8 in the N series (31 mm) did instead increase significantly from the preceding series, in spite of the maximum aperture remaining the same. As a result, the two lenses in the N series look virtually identical. The lens and filter mounts remain the same as in the preceding series. Like all lenses in the N series, this is multi-coated. This gives the front and rear elements a neutral or slightly bluish colour, and multicolour reflections from lens surfaces indicate that different types of coating are used on different surfaces.

Lens performance in UV photography

It seems that the fame of the EL-Nikkor 63 mm f/3.5 as a UV lens started when someone noticed that a Nikon brochure specifies that this lens is colour-corrected in the interval between 350 and 700 nm. Since the visible range ends at 400 nm, this seems to indicate a relatively good performance in the near-UV. Actually, there are two independent properties that make a lens usable in near-UV photography. The one mentioned above is focus shift, which is related to colour-correction. The second is discussed later on this page.

The spectral UV transmission of this lens is discussed and compared with other lenses here.

Focus shift

If the lens exhibits a substantial longitudinal colour aberration between the near-UV and visible ranges, an image focused in visible light will be out-of-focus in the near-UV (and vice versa). This is known as focus shift, and means that it will be necessary to determine, by trial-and-error, an offset to the focusing helicoid or bellows that must be set manually each time between focusing and exposing. In addition, this offset may change at different magnifications. A focus shift exists also in normal camera lenses, when used in near-IR photography. In the past, many of these lenses had a red dot or index mark, corresponding to the focus shift in the near-IR range. One had to focus in visible light, then mount an IR filter and manually shift the focus ring to the red index before exposing. Most modern lenses do not have this index mark.

My own research suggests that the "myth" of the 63 mm f/3.5 as a UV lens is likely to be based on a printing error - or an exaggeration by a sales representative - in a Nikon brochure (most likely, this brochure, which cites the 350-700 nm range for the 63 mm and a 370-700 nm for a few other EL-Nikkors, in addition to 380-700 nm for the remaining models). This other EL-Nikkor brochure mentions in its introductory text that all EL-Nikkors are colour-corrected between 350 and 700 nm. However, in the data sheets that follow in the very same brochure, only the 380-700 nm range is repeated - a total of 26 times! - while the 350 nm figure makes no further appearance. Of course, it must be said that there is no sharp cutoff at the end of the wavelength range that is colour corrected. Instead, colour aberration sets in gradually, and whether one specifies the end of the range at 350, 370 or 380 nm depends on how much aberration one is prepared to accept. Therefore, a sales representative may have decided to provide a slightly more optimistic figure, missing the fact that a different figure was reported many times in the following text. Other brochures (here, here, here and here) do quote only the 380-700 nm range for all EL-Nikkors, including the 63 mm f/3.5 where mentioned. So does all other Nikon literature I have seen.

In conclusion, I see no evidence that the 63 mm f/3.5 should be designed for a broader colour-correction range than other EL-Nikkors. This does not automatically exclude that, by a small accident, it could be slightly better than other models in this respect. However, the same could apply to one or more of the other focal lengths in the EL-Nikkor range, in particular the 50, 75 and 80 mm, which are all potential alternatives to the 63 mm. It may also be asked whether the N series, which is specified for the same colour correction, could constitute an alternative. I tested these possibilities here.

Incidentally, the reason why EL-Nikkors are designed to be colour-corrected slightly into the near-UV range is explained by Nikon with the fact that photographic films and papers are very photosensitive in the near-UV range. Since small amounts of near-UV are indeed present even in incandescent lighting, an extended correction range may give enlarger lenses a small performance edge (and also an additional selling point for salespeople to hammer customers with). However, normal glass is largely opaque to UV radiation, and tends to attenuate even near-UV. Thus, the thick condenser lenses used in most enlargers are likely to substantially absorb any UV present in the light source. The white paint used to coat the inside of enlarger bulbs may also be an effective UV barrier.

Transmission

This brings us to the second factor that affects the UV performance of a lens. If its optical elements are made of normal glass, they will absorb UV. In addition, most types of cement used to glue together lens elements do strongly absorb UV. Finally, lens coatings improve light transmission in a given range of wavelengths, but actually reflect abundant light outside this range. Any lens coating designed only for optimal visible transmission is likely to reflect UV. In normal photography, this is actually a desirable property, because UV in landscapes causes flare and a degradation of contrast. Thus, the coating of the front lens element is usually designed specifically to reflect UV. Therefore, unless a lens is specifically designed for use in the UV range (e.g., made of quartz and coated with layers specifically tuned to transmit UV), it is very likely to absorb and/or reflect strongly in this range. This does not conflict with a colour-correction range that includes also part of the near-UV. A lens may be corrected in this range, and at the same time allow very little of it to pass.

Virtually all camera lenses do transmit very well in the near-IR range, although several do exhibit unacceptable amounts of focus-shift or internal flare in this range. This is one of the factors that make photography in the near-IR without special equipment much easier.

Based on the above discussion, an acceptable lens for UV photography should (1) transmit usable amounts of UV, and (2) be colour-corrected between visible and UV, to avoid focus-shift. It does not need to be colour-corrected in the entire visible range. In fact, as a rule, it will be corrected only between the green - or blue - and UV. Lenses designed for photography in multiple spectra (e.g., near-UV to near-IR), on the other hand, should be colour-corrected throughout these ranges.

UV test

If you make a web search for EL-Nikkor 63 mm and UV photography, you will find the older version cited quite frequently in this context, and in positive terms. Not so the modern version, which sometimes is dismissed as unsuitable because multi-coated. This stimulated my curiosity, and I decided to find out for myself whether the two versions actually differ in UV performance.

As a comparison, I also used a UV Rodagon 60 mm f/5.6 (designed for UV-A close-up photography, and exhibiting no focus-shift between the blue and near-UV bands). As a filter to block visible light and pass only UV-A (and a little indigo, to leave some colour information in the picture), I placed a Schuler UV pass filter on the lenses. As a light source, I used a xenon HID rated at 11,000 K. I was at a loss for a suitable subject, since flowers are difficult to come by right now (20 cm of snow outside, sub-zero temperatures, and flower shops are closed). The subject of the following pictures is the best that my garden could provide.

The UV Rodagon is designed to give no focus-shift between blue and near-UV light. This produces very sharp pictures in the near-UV (in addition, diffraction is lesser in this band than in the visible range). Not surprisingly, this lens gives the best results among those tested here.

The EL-Nikkor 63 mm f/3.5 provides less brilliant and contrasted results. Compared with the preceding pictures, the one taken with this lens lacks the white highlights on the leaves evident with the UV Rodagon. This is possibly due to a poorer transmission of the shorter near-UV wavelengths. The El-Nikkor also exhibits a slight "fogginess" in the central region of the flower. The hairs around the flower and the lone petal, however, are quite sharp, so the resolution is quite high.

The EL-Nikkor 63 mm f/2.8 provides quite similar results to the preceding lens. The only differences I can seen are a slightly lower contrast and a slightly higher "fogginess" of the central region of the flower. I thought that the central "fogginess" could be caused by internal lens flare in the near-IR. Although the light source I used produces little near-IR and the UV-pass filter further cuts it, the latter still has a small peak in the near-IR. Unfortunately, I don't have a filter that further cuts near-IR without cutting also near-UV, so I cannot test this hypothesis directly. However, a simple test shows that the fogginess stays in the center of the flower even when the latter is moved off the center of the picture. This rules out internal flare, which always stays in the same place regardless of the subject.

In addition, I could remove the red and green component from the test picture, because near-IR is recorded by my camera largely in the red channel. In fact, this eliminates the central fogginess. Unfortunately, the resolution is also affected negatively (quite naturally, because the green channel information comes from twice the number of photosites than the red or blue information alone). By eliminating only the red channel, the result is much more detailed (last picture above), in spite of the fact that very little light is recorded in the green channel. Unfortunately, also the green channel has a fogginess at the center of the flower, albeit less evident than the red one. I now believe that that the fogginess may be caused by a strong near-IR reflection of the center of the flower. This is hardly a fault of the lens.

To get a better idea of the performance of the EL-Nikkor 63 mm f/2.8, I took additional test pictures (above), using a 40 W incandescent lamp. Exposure was manual and determined by looking at the histogram on the camera's LCD. All pictures used the same preset white-balance, and were not post-processed, except for modest adjustments in luminosity as necessary to produce uniform results. The first is without a filter. The red-purple tinge is caused by abundant near-IR in the light source. The second was taken with a B+W 486 filter, which is designed to cut both UV and IR. Illumination is a bit too "warm", but not bad considering the light source. This is approximately the spectral response of a normal, unmodified DSLR. The third was taken with a Hoya R72 filter, which is an IR-pass with a cutoff of 720 nm (i.e., just a little bit below visible light). The last was taken with a no-brand IR-pass filter with a cutoff of 820 nm (i.e., well within the near-IR).

It is interesting to notice that there is an obvious difference between the two IR filters. No visible light can be detected by looking through either filter. The recorded difference may be due either to a very slight red leakage, or to the red photosites of the camera sensor being more sensitive to short-wave IR than green and blue sites. All sites are equally sensitive, instead, to deeper IR. This should account for the reddish cast of the picture with the Hoya R72. The picture with the 820 nm filter is practically neutral in tone. The picture with the Hoya R72 also shows some slight tonal detail on the leaf at the top-left of center (a discoloration caused by disease, also seen in visible light), which is largely absent in the 820 nm picture (this is as expected, since most pigments that are detectable in the visible and near-IR become transparent in deeper IR).

It was only after performing the last test, in which I used as an IR source the incandescent lamp that had always been on throughout the preceding tests (to allow me to see what I was doing), that the obvious finally hit me. This lamp could well be the source of IR contamination that I was suspecting! Thus, I repeated the tests, but switched off the incandescent light during the actual exposure. The rest of the room was practically dark, so there was no ambient source of near-IR.

This simple precaution dramatically improved the performance of all three lenses and, especially, made the results more similar among all lenses. The central fogginess was much lesser, and eliminating the red channel completely removed it .Although the differences among the three lenses are still visible, they are much easier to compensate for in post-processing. Further reducing the IR component with an IR-cut filter that does not affect UV, mounted on the light source, and/or using a camera filter with better IR rejection (e.g., the Baader UV) may improve the results even more.

A "real" UV lens like the UV Rodagon does perform somewhat better. However, this is likely to be seen mostly in poor conditions, e.g., in the presence of IR contamination, or deeper into the UV range. When precautions are taken, the two EL-Nikkor 63 mm give remarkably clear near-UV results. The older 63 mm seems to transmit less UV than either the UV Rodagon or the newer 63 mm (the leaves are less brilliant and overall tone is darker). The modern EL-Nikkor transmits UV as well as the UV Rodagon, but it also transmits more near-IR than the other lenses (probably thanks to its multicoating). Focus shift does not seem to be a detectable problem if the aperture is stopped down. Again, this must be seen in the context of what range of UV is recorded by my camera (likely, not past 360 nm). The UV Rodagon probably performs better than the EL-Nikkors in deeper UV, but this is irrelevant if my camera cannot record it.

In conclusion, the two EL-Nikkors do not perform much differently from each other in the UV. A modest difference from a "real" UV lens is present, but not substantial. The older 63 mm gives a slightly better contrast in the presence of contaminating IR, but the newer model transmits slightly more UV. Small amounts of near-IR have a large impact in these tests, and care should be taken to eliminate this factor.

Photomacrography test

These lenses were tested reversed on bellows at a magnification of approximately 4x. Optically, this test is equivalent to printing a negative at an enlargement (on paper) of 4x (or to taking a close-up photograph at 1:4). However, it is more demanding than a close-up in terms of resolution, because the image projected on the camera sensor is four times larger than the subject, thus allowing the detection of small optical aberrations. As a comparison, I used the Micro Nikkor 70 mm f/5, which in other tests performed even better than the Zeiss Luminar 63 mm f/3.5. Because of the different focal length, I was forced to use a slightly different setup with this lens, which changed the illumination somewhat. This prevents an accurate comparison of the contrast with this lens.

The tests show that the Micro Nikkor 70 mm has a slight but visible edge in resolution. As mentioned above, the difference in contrast between the Micro Nikkor and the two EL-Nikkors may be due in part to a slightly different illumination setup. I find the picture from the Micro Nikkor better also in colour rendition of details, modeling of the relief and overall "roundness", but this comparison may not be entirely fair. In any case, this lens is the very best I own for this magnification range. It must be remembered also that the difference with the EL-Nikkors becomes visible only once individual pixels can be seen, and that photomacrography places extreme demands on resolution. When these lenses are used to take close-ups, results are virtually identical.

The two EL-Nikkors perform almost identically, but the modern model (from the N series) may have a very slightly better resolution and contrast, and a more neutral tone, with the older model being slightly warmer. Both models are quite good for photomacrography around 4x. In practice, in photomacrography, the modern (and usually cheaper) N model performs very slightly better than its older counterpart.

Conclusions

As a whole, the difference in UV performance between the two models of EL-Nikkor 63 mm is not very remarkable. If I could not afford a real UV lens, I would rather buy a modern EL-Nikkor 63 mm f/2.8 N at a much lower price, and get a good photomacrographic lens at the same time (slightly better for this application, in fact, than the older model). I definitely would not buy an EL-Nikkor 63 mm f/3.5 at a price that could buy me a "real" UV lens (i.e., around 1,000 US$).

Now that the secret is out, we will see whether the price of the EL-Nikkor 63 mm f/2.8 N will also increase out of proportion, or the price of the EL-Nikkor 63 mm f/3.5 will become more reasonable.

Finally, an eye opener: The 63 mm f/3.5 and f/2.8 N are not the only EL-Nikkors suitable for UV photography - quite the contrary! You do want to read also this page.