Nikkor AF-S 200mm f/2.8E PF (speculation)

[bobfriedman]

This is an interview to the Nikon engineers in charge of PF300:

https://dc.watch.impress.co.jp/docs/news/interview/688994.html

If you drop it into translate.google.com  does quite a good job

[Roland Vink]

I don't really understand the optophysical theory, but the element used in PF300 seems to be called "Blazed Grating" which is a special refractive optics utilizing the diffraction effect: [...]

Thanks Akira. I would need someone to explain it in more simple terms for me to understand, but it is obvious these are more than just simple Fresnel lenses.

I suspect besides the obvious Fresnel steps in the element, there must be a substructure which is the size of the wavelength of light, which is responsible for the diffraction effect. The wikipedia page says the blazed grating is optimised for one wavelength only, so I wonder how the lens works with white light (has anyone tried it for UV or IR photography?). Very interesting technology.

[Akira]

If you drop it into translate.google.com  does quite a good job

So, the interpreter will soon be filed to the redbook...

[Akira]

Thanks Akira. I would need someone to explain it in more simple terms for me to understand, but it is obvious these are more than just simple Fresnel lenses.

I suspect besides the obvious Fresnel steps in the element, there must be a substructure which is the size of the wavelength of light, which is responsible for the diffraction effect. The wikipedia page says the blazed grating is optimised for one wavelength only, so I wonder how the lens works with white light (has anyone tried it for UV or IR photography?). Very interesting technology.

Indeed it is an interesting technology.  I wonder when or if people would find the way to suppress the PF flare...

[Øivind Tøien]

Thanks Akira. I would need someone to explain it in more simple terms for me to understand, but it is obvious these are more than just simple Fresnel lenses.

I suspect besides the obvious Fresnel steps in the element, there must be a substructure which is the size of the wavelength of light, which is responsible for the diffraction effect. The wikipedia page says the blazed grating is optimised for one wavelength only, so I wonder how the lens works with white light (has anyone tried it for UV or IR photography?). Very interesting technology.

The 300PF works quite well with IR, as exemplified by muliple images posted here before. Here is one, wide open on D40x. Notice the lack of flare on the shiny chimneys.



Backlight hitting direclty can create some good flare and ghosting in IR, as I previously  exemplified here: http://nikongear.net/revival/index.php/topic,7350.msg118637.html#msg118637 , repeated here for convenience. With some effort one can show off the ghost of the fresnel pattern:



I doubt there are micro patterns overlayed on the Fresnel pattern. The DC watch article indicates otherwise. Perhaps the increasing frequency of the pattern towards the edges is at play?


I do not think there is much hope for UV transmission with the modern coatings of the lens.

Indeed it is an interesting technology.  I wonder when or if people would find the way to suppress the PF flare...

Akira, the PF flare is usually not a problem in practical visible light photography, and I believe the design already suppresses most of the effect compared to the original designs that came out (for instance Canon's first use of this technology). Other lenses will also flare in those situations: To make the PF effect show up one needs severely blown light points. Then instead of a monochromatic halo in regular lenses one can get a little of the rainbow pattern. I believe I posted a comparison to the AF 300/4 in the original 300PF thread.  I have also used it for solar imaging and astro photography without any problems.

A much more important property of the lens is the almost complete lack of longitudinal/highlight fringing; a little rainbow showing up in the most extreme flare cases is a very small cost to bear. For astro photography it is also worth noting that coma/astigmatism is much better with the AFS 300/4 PF ED VR E than the AFS 300/4 ED. The latter will give pronounced "winged" stars not very far away from the center (on a DX sensor) unless stopped down below f/5 or so. My AF 300mm f/4 ED has pretty bad purple/green fringing. (I almost always had to work on removal in reflections of sunlit eyes). This in spite of that all 3 lenses are highly regarded resolution wise.

[Jack Dahlgren]

Thanks Akira. I would need someone to explain it in more simple terms for me to understand, but it is obvious these are more than just simple Fresnel lenses.

I suspect besides the obvious Fresnel steps in the element, there must be a substructure which is the size of the wavelength of light, which is responsible for the diffraction effect. The wikipedia page says the blazed grating is optimised for one wavelength only, so I wonder how the lens works with white light (has anyone tried it for UV or IR photography?). Very interesting technology.

In the simplest terms red light through a prism refracts less than blue light. Red light through a diffraction grating refracts more than blue light. The diffractive lens element is used to counteract the refraction in the other elements of the lens, resulting in an achromatic lens without as many thick lenses which would otherwise be required to achieve the same thing. If you look at the diagrams in the article Akira posted you can see how the phase fresnel element reverses the order of wavelengths.

[Akira]

Øivind and Jack, thank you for the explanations.

I doubt there are micro patterns overlayed on the Fresnel pattern. The DC watch article indicates otherwise. Perhaps the increasing frequency of the pattern towards the edges is at play?

I think that the illustration in the DC Watch is simplified just to explain the concept of the blaze grating in general.

[chambeshi]

Here is Nikon's patent registered earlier this year, for a 400, 500 and 600. These bear close similarities to the 300 f4E PF, but differ in the rear groups. Note what appears to be a protective element (FL coated?) in the posterior of the proposed 400mm

https://nikonrumors.com/2018/02/01/the-latest-nikon-patents-400mm-500mm-and-600mm-f-5-6-phase-fresnel-pf-lenses.aspx/

https://www.j-platpat.inpit.go.jp/web/PU/JPA_H30017857/TKBS_GM302_Detailed.action

schematic of the 300 f4E for comparison
https://www.the-digital-picture.com/News/News-Post.aspx?News=14218

[Roland Vink]

In the simplest terms red light through a prism refracts less than blue light. Red light through a diffraction grating refracts more than blue light. The diffractive lens element is used to counteract the refraction in the other elements of the lens, resulting in an achromatic lens without as many thick lenses which would otherwise be required to achieve the same thing. If you look at the diagrams in the article Akira posted you can see how the phase fresnel element reverses the order of wavelengths.

Yes I saw that, but I don't understand the principles that cause it, the diagrams shown seem to be rather simplified and don't really show how it works. For example, it's not clear (to me at least) if the steps shown on some PF lens diagrams represent the large steps visible to the eye, or whether there is some microscopic structure ... I don't see how the large steps could cause diffraction effects. A normal Fresnel lens works by refraction only, which would have dispersion the same as a conventional lens, not reversed as with the FP lens. That's why I wonder if there is a microscopic structure which causes diffraction, similar to the tracks on a CD or DVD which give rainbow reflections due to diffraction. But if there is a sub-structure, then wouldn't it be easier to apply it to a non Fresnel lens? Or do you need both?

[Jack Dahlgren]

Yes I saw that, but I don't understand the principles that cause it, the diagrams shown seem to be rather simplified and don't really show how it works. For example, it's not clear (to me at least) if the steps shown on some PF lens diagrams represent the large steps visible to the eye, or whether there is some microscopic structure ... I don't see how the large steps could cause diffraction effects. A normal Fresnel lens works by refraction only, which would have dispersion the same as a conventional lens, not reversed as with the FP lens. That's why I wonder if there is a microscopic structure which causes diffraction, similar to the tracks on a CD or DVD which give rainbow reflections due to diffraction. But if there is a sub-structure, then wouldn't it be easier to apply it to a non Fresnel lens? Or do you need both?

You need both, and that fact is in the very name “phase Fresnel”. I am not certain exactly how it is constructed, but the diffraction angles through the transmission grating are fairly high so a flat grating won’t give you what you are looking for unless the angle it makes with the light path is carefully controlled. Fresnels  only save weight and don’t control against chromatic aberration so it alone would tend to degrade the image with only slight benefit to weight. It is the combination of effects here which makes it possible to control the diffraction to produce a useful image.