Opticallimits
new Nikon gear - Printable Version

+- Opticallimits (https://forum.opticallimits.com)
+-- Forum: Forums (https://forum.opticallimits.com/forumdisplay.php?fid=4)
+--- Forum: Nikon (https://forum.opticallimits.com/forumdisplay.php?fid=12)
+--- Thread: new Nikon gear (/showthread.php?tid=1291)

Pages: 1 2 3 4 5 6 7 8 9 10


new Nikon gear - Scythels - 01-07-2015

@Joju, you will lose approximately one stop of light with a 1.4x teleconverter.  f/4 will become f/5.6.

 

@Brightcolors

Quote: 

Scynthels, you seem to not be aware that the reason for the short length is due to the special nature of the PF element, and that such an element does its work without increased aberrations (and so without the need for extra elements)? same as with the DO elements from Canon.
 

That's not true at all. 

 

A fresnel lens allows you to focus light with less material than a spherical lens.  It's based on the theory that light is bent when it hits the material and does not get further bent inside the material.  A fresnel lens has many small prism-like structures on it, to focus light you just increase the angle of inclination continuously as you move away from the center of the lens.  You can reduce the necessary amount of material in this way by producing a nearly perfectly flat lens that still focuses light (and thus can be very thin), or you may apply it to an already curved surface to simulate a much higher refractive index and thus focusing the light much more strongly. The fresnel element will not contribute third order aberrations to the system, but higher order aberrations must be calculated as they will be present in sufficient quantities so as to impact the system.  This is also already a strongly curved surface, so it imparts rather large aberrations already.

 

To drive the telephoto ratio further from one you must increase the power of the positive and negative elements.  To first order (that is, arrangement of thin lenses with positive and negative power simulating the net result of the system) this allows you to place the positive and negative lenses closer together to achieve the same result (and thus, shorten the overall length of the lens).  The trouble with this is that spherical aberration can become quite immense when you bring thick lenses close together, astigmatism will rise sharply the closer they get together, and it becomes more and more sensitive to assembly tolerance as well as the presense of local minima within the optimization of the lens, as spherical will quickly invert from being over to under corrected.

 

The asymmetry of power about the aperture stop will also drive distortion up rather quickly.

 

Nikon has used a fresnel lens to massively increase the power of the "P" group in the telephoto arrangement and met it with a large increase in the power of the "N" group.  The fresnel's contribution to the power of the "P" group is free, diffractive effects aside, but the "N" group's power absolutely is not.  Many extra elements were clearly needed to correct the aberrations induced, and either coma or astigmatism has still been corrected imperfectly.




new Nikon gear - JJ_SO - 01-07-2015

Quote:@Joju, you will lose approximately one stop of light with a 1.4x teleconverter.  f/4 will become f/5.6.
 

Nikon gives for the TC-17EIII 1.5 f-stops and for the TC-20EIII 2 f-stops. They don't mention a decreasing aperture for the TC-14EIII, so either they "forgot it" or I have to believe my colleague who owns those converters (in version I and II, I believe) and told me that with the 1.4× converter the f-stop remains the same.

 

Anyway, 40% to me means f/4.8, so I don't think it's a full f-stop. And (my) alternative lens in question is a 150-600/5-6.3.

 

Which means the new Nikon ends up at f/8 with 2× converter. So, I got a 300/4 and a 600/8 at 755 + 330 grams, roughly 1kg against a 2.9 kg zoom lens which is larger, has M 105 filter (Nikon: M 77) and a ton of settings to adjust. Price difference from Sigma to the Nikon combination: I expect app. 700 $.

 

No need to decide now - the Nikon will hit the shelves beginning of february, I still have no confirmed delivery date from Sigma representative.



new Nikon gear - stoppingdown - 01-07-2015

Quote:Nikon gives for the TC-17EIII 1.5 f-stops and for the TC-20EIII 2 f-stops. They don't mention a decreasing aperture for the TC-14EIII, so either they "forgot it" or I have to believe my colleague who owns those converters (in version I and II, I believe) and told me that with the 1.4× converter the f-stop remains the same.
 

??? I own it (TC-14EII), and definitely the max aperture I can dial with it and the 300 f/4 is f/5.6... I wouldn't understand how a teleconverter could keep the same aperture.

 

In any case, this Nikon official page explicitly says there's 1 stop loss:

 

http://www.nikonusa.com/en/Nikon-Products/Product/Camera-Lenses/AF-S-Teleconverter-TC-14E-III.html



new Nikon gear - JJ_SO - 01-07-2015

I checked the Swiss and German sites of Nikon, they don't mention a loss of light (what appears not logical to me).

 

After your post I was looking for the manual of the TC-14IIIE and there was a sentence like "The teleconverter decreases the effective aperture by one stop." So, my information was not correct, sorry.




new Nikon gear - Scythels - 01-07-2015

Think of it logically.  The teleconverter is behind the aperture stop, it cannot alter the aperture stop as seen from the front.  A 300mm f/4 lens requires a 75mm entrance pupil diameter, the teleconverter increases the focal length 40%.  420/75 = 5.6.  Precisely one stop lost.




new Nikon gear - JJ_SO - 01-07-2015

The flare is pretty weird, although sharpness at this sample looks very good. Kind of disappointing, Nikon asia was announced on Nikon Rumors to deliver full res pictures and what I got are all cropped or worse, downscaled Sad




new Nikon gear - stoppingdown - 01-07-2015

Quote:The flare is pretty weird, although sharpness at this sample looks very good. Kind of disappointing, Nikon asia was announced on Nikon Rumors to deliver full res pictures and what I got are all cropped or worse, downscaled Sad
 

I see. It's quite funny that people in charge of ads don't realise that it's absolutely useless to post a downscaled picture when a new lens is released. Anyway full-size samples will follow soon, I think.



new Nikon gear - Brightcolours - 01-08-2015

You can say it is not true at all, but it is.

 

The PF lens, like the DO lens configurations used in Canon DO lenses, make the lens all on their own shorter, without causing CA because it is combined by a single refractive lens and without the need of many more elements.

 

Unless Nikon is unclear about their own PF lens:

[Image: Nikon-Phase-Fresnel-PF-lens-explained-550x319.png]

And Canon does similar things, but with multiple DO elements bonded togethet without gap in the 400mm DO II to prevent the flaring the Nikon lens shows.

http://www.canon.co.uk/For_Home/Product_Finder/Cameras/EF_Lenses/DO_Lenses.aspx

http://cpn.canon-europe.com/content/product/lenses/ef400mm_f_4_do_is_ii_usm.do

 

Like I pointed out before, the simple fact is that the 300mm f4E PF has the same number of elements as the 400mm f2.8E FL.... And the latter has no PF element.




new Nikon gear - Scythels - 01-08-2015

Quote: 

The PF lens, like the DO lens configurations used in Canon DO lenses, make the lens all on their own shorter
No, it doesn't.  What a fresnel lens does is use so-called microoptical structures to bend the light in lieu of a continuous curve as you would see with a normal lens.  The formula for the apex angle of each prism is given by A(ρ)=ρ/(f(n2-n1)).  In the case the fresnel surface lies along a curve so we must complicate the math mildly by converting what we want to optical power, as the curve itself gives power to the lens.  Optical power, is the measure of how strongly something focuses light. 

 

For any lens surface we have ϕ=(1/R)(n2-n1) where R is the radius of curvature, n2 is the refractive index of the medium to the right of the surface, and and n1 is the refractive index of the medium to the left of the surface.  For any optical system, uppercase phi (browsers do not draw phi correctly...) Φ=1/efl - the effective focal length.  In laymans' terms this is the focal length. 

 

Let's have a little thought experiment, shall we? 

 

The 300mm f/4 is a telephoto design.  This you must surely agree with.  It has a phenominal telephoto ratio of just .466, a ratio which i believe may be a world record of some kind. 

 

in its simplest form a telephoto lens is a positive lens followed by a negative lens.  There are two basic ways to increase the power of a lens, to make its radii of curvature shorter and thus steepen the slope of its surface, or to use a higher index material for the glass.  There are other methods such as using a fresnel lens or using a GRadient INdex (GRIN) material but they are (ironically) outside the scope of this bit here.

 

The fourth lens in the new 300/4 has a net negative power.  The first surface has positive power but it likely has been curved in the way it has simply to adjust the angle of incidence on the rear surface without a goal of having positive power there, or it is a concession to allow a thinner lens overall. 

 

Let us imagine that we want the rear surface of the fourth lens, considered in isolation, to have a focal length of -50mm.  I do not believe that is an unreasonable guess given how strongly curved it is and the fact that it has a fresnel lens. 

 

Now take a moment of basic theory with snell's law.  If a lens is curved, the angle of incidence with respect to the surface normal is not constant along the surface.  Because for non-GRIN lenses the refractive index is constant, the angle light is bent at is not constant.  Unfortunately, for basic spherical surfaces Δθ for the angle of refraction does not exactly mirror the angle of incidence with respect to the optical axis, so you get defocus when you utilize a lens.  This is the origin of all aberrations, but spherical is easiest to imagine. 

 

The stronger your curvature, the more you "aggrivate" the Δθ disagreement.  Ideally, you want very weak curvatures. 

 

Because the radius of curvature necessary to create a long focal length is very large, the slope of the lens surface is small.  This creates only weak geometrical aberrations and is why long focal length lenses tend to perform better than short focal length lenses.

 

Nikon has shown us in their diagram that they have combined a high dispersion lens with with whatever the fresnel "element" is cut into, it's probably NSF2 or a "neighbor" glass.  The material the fresnel element is cut into is most likely a crown.  Given that there will be manufacturing difficulty a nice cheap crown was likely chosen, either NBK7 or NSK4 or one of their neighbors.  I'll use NBK7 for an example, but in this thought experiment it doesn't truly matter.

 

NBK7 has a refractive index of 1.517.  Air has a refractive index of 1.  The radius of curvature necessary to achieve a focal length of 30mm can be calculated with -1/50=.517(1/x) = -25.85mm.  That is an extremely strong curvature and would induce absolutely monstrous aberrations.  But, if you add a fresnel component and say that the fresnel componenet will contribute 1/2 the power, now your radius of curvature must only be -51.7mm, which is still strong but nowhere near as bad as the value without a fresnel componenet.

 

Why do you want a short focal length - isn't this a telephoto lens?

 

Yes, it is.  But the adjustment of the telephoto ratio requires an adjustment of the Positive and Negative lens arrangement.  There is a "good zone" to be in regarding how close these two elements are to each others' centers of curvature that is developed in nodal aberration theory.  If you want to shrink the overall length you must get them close together, so they must have roughly complementary short focal lengths (for example 50 and -50mm). 

 

----

 

There is a great deal of positive optical power in front of the fresnel element, the job of that surface is to reduce the power.  The very strong radius of curvature has unfortunately induced many aberrations which must be undone.  To undo them nikon has chosen to use additional elements, though they could have done it with aspheric surface(s) if they had wished to. 

 

-------------------------

 

Quote: 

 without causing CA because it is combined by a single refractive lens and without the need of many more elements.
 

Elements are used to correct much more than polychromatic aberrations.  There are two basic techniques to address problem elements and improve the performance of a lens - splitting and compounding.  By splitting a lens you distribute its power across more surfaces and thus reduce the radii of of curvature, "piggybacking" off of the refractive index of the elements to provide the power.  This allows you to reduce geometrical aberrations.  One may also compound a lens into a doublet.  This allows the creation of an achromatic doublet if one wishes to correct axial and lateral CA, or the simulation of an element with very high refractive index if one wishes to reduce the curvatures at play. 

 

There are fifteen other elements in this lens.  There are other chromatic aberrations that must be addressed, and the monstrous geometrical aberrations created to achieve the exceptional telephoto ratio are being corrected through the addition of more elements.

 

Quote: 

And Canon does similar things, but with multiple DO elements bonded togethet without gap in the 400mm DO II to prevent the flaring the Nikon lens shows.
The only similarity between the third generation DO technology and nikon's new PF technology is the use of a fresnel surface.  They aren't even the same type of fresnel surface.  Flare from fresnel surfaces is caused by reflections off of the legs of the "triangles".  Canon has eliminated that by very precisely manufacturing two equal fresnel surfaces and cementing them together.

 

Quote: 

Like I pointed out before, the simple fact is that the 300mm f4E PF has the same number of elements as the 400mm f2.8E FL.... And the latter has no PF element.
I'm not sure what your point is.  Element count only directly affects cost, manufacturability, size, and weight.  It says nothing about anything else directly.  The 400mm's design is made easier by having a longer focal length.  It is made harder by having a wider aperture.  These two "fight" regarding design complexity, but the 400mm pursues a less aggressive telephoto ratio.

 

Now please, brightcolors.  Stop pretending you know so much about lenses.  You know far less than you think you do.



new Nikon gear - JJ_SO - 01-08-2015

Scythels, the new 300/4 is exactly 75 mm shorter than the old one. I'm sorry, I don't want to defend BC's often weird posts but sometimes he happens to hit the nail even with false assumptions  Rolleyes

 

As far as I understand this fresnel stuff, it's the same geometry that makes the optical systems of a lighthouse work or the old fresnel screens in analog cameras or those flat loupes or "wide angle" lenses on some rear screens of buses work. All those systems have in common "lightweight" and "short lens length" and, if saving money was the reason to make this reading loupes, the quality usually is better than a plastic loupe with spherical elements and worse than one with grinded and polished glass. I think, the manufacturing process of this fresnel stuff is quite tricky. 

 

Apparently this PF element appears to make the lens shorter by quite a strong bit (⅓) and lighter by nearly 50%. You can write a scientific book about why this is not caused by the PF element. Alone. But it contributes to it. As for the weight reduction, this is also based on massive use of (high-tech) plastics (maybe the PF element is also made of plastics?) and the electronically driven aperture and focus drive.

 

Still, it's impressive enough.