Patent: Canon EF 28-560mm f/2.8-5.6

[nvsravank]

this lens would be huge. same diameter as the 300 F2.8.

This lens would be F2.8 at what focal length - not 300? So does it have to be the size of a 300 2.8?

Jack

The diameter of the front element is (approximately) the focal length divided by the aperture.

Thus a 560 / 5.6 = 100mm. A 300 /2.8 ~ 100mm. Same diameter.

The definition of the F stop is the focal length of the lens divided by the maximum diameter of the aperture when opened fully.

Not exactly. The f stop is the focal length over the entrance pupil diameter, not the actual aperture diameter. The entrance pupil is the image of the aperture from the objective side; it is the image of the opening that can be magnified by the optics.

I'd argue the f stop is the focal length over the apparent size of the entrance pupil diameter. But even if we quibble over technicalities the optical fact remains that the aperture can't be larger than the front element because it can't appear to be bigger than that, ergo the focal length divided by aperture will give you a minimum front element size.

If you can point me to a lens that has a front element smaller than its focal length divided by its actual fastest aperture I'd like to see it.

The patent gave the dimensions of the front element. 1.9 mm thick with curvature radii of 114 mm on the front and 230 mm on the back. The front element should only be 60mm, no?

I'm trying to understand this phenomenon myself, but conceptually I understand it to mean that the elements in front of the aperture are causing the rays to converge towards the aperture so that the image of the aperture is larger than the front element; if you were trying to see the aperture you would have to angle the lens away from your eye to see the edge.

Wouldn't that mean a significant vignette ?
Also I have been searching on and off all morning to find a lens that does what you say. Can you point out actual production lens? Like private by design asked. I am unable to find it. Google is not being my friend.

But this topic has been good for me to learn some obscure/abstruse facts and came to read about front pupil and back Pupil.

Any way to set the aperture of electronic lenses to see this in action?

 

[mb66energy]

A lens with the radii and max thickness you mentioned has -- if both are convex -- a diameter of 34 mm and if the inner radius is concave, the outer one convex you will have the diameter of maximal 58mm -- but the thickness is then zero on the outside! Both lenses were positive.
If both radii describe concave surfaces 200mm diameter are possible (but pure nonsense I think) and the lens is negative.

Strange.

this lens would be huge. same diameter as the 300 F2.8.

This lens would be F2.8 at what focal length - not 300? So does it have to be the size of a 300 2.8?

Jack

The diameter of the front element is (approximately) the focal length divided by the aperture.

Thus a 560 / 5.6 = 100mm. A 300 /2.8 ~ 100mm. Same diameter.

The definition of the F stop is the focal length of the lens divided by the maximum diameter of the aperture when opened fully.

Not exactly. The f stop is the focal length over the entrance pupil diameter, not the actual aperture diameter. The entrance pupil is the image of the aperture from the objective side; it is the image of the opening that can be magnified by the optics.

I'd argue the f stop is the focal length over the apparent size of the entrance pupil diameter. But even if we quibble over technicalities the optical fact remains that the aperture can't be larger than the front element because it can't appear to be bigger than that, ergo the focal length divided by aperture will give you a minimum front element size.

If you can point me to a lens that has a front element smaller than its focal length divided by its actual fastest aperture I'd like to see it.

The patent gave the dimensions of the front element. 1.9 mm thick with curvature radii of 114 mm on the front and 230 mm on the back. The front element should only be 60mm, no?

I'm trying to understand this phenomenon myself, but conceptually I understand it to mean that the elements in front of the aperture are causing the rays to converge towards the aperture so that the image of the aperture is larger than the front element; if you were trying to see the aperture you would have to angle the lens away from your eye to see the edge.

 

[Wizardly]

this lens would be huge. same diameter as the 300 F2.8.

This lens would be F2.8 at what focal length - not 300? So does it have to be the size of a 300 2.8?

Jack

The diameter of the front element is (approximately) the focal length divided by the aperture.

Thus a 560 / 5.6 = 100mm. A 300 /2.8 ~ 100mm. Same diameter.

The definition of the F stop is the focal length of the lens divided by the maximum diameter of the aperture when opened fully.

Not exactly. The f stop is the focal length over the entrance pupil diameter, not the actual aperture diameter. The entrance pupil is the image of the aperture from the objective side; it is the image of the opening that can be magnified by the optics.

I'd argue the f stop is the focal length over the apparent size of the entrance pupil diameter. But even if we quibble over technicalities the optical fact remains that the aperture can't be larger than the front element because it can't appear to be bigger than that, ergo the focal length divided by aperture will give you a minimum front element size.

If you can point me to a lens that has a front element smaller than its focal length divided by its actual fastest aperture I'd like to see it.

The patent gave the dimensions of the front element. 1.9 mm thick with curvature radii of 114 mm on the front and 230 mm on the back. The front element should only be 60mm, no?

I'm trying to understand this phenomenon myself, but conceptually I understand it to mean that the elements in front of the aperture are causing the rays to converge towards the aperture so that the image of the aperture is larger than the front element; if you were trying to see the aperture you would have to angle the lens away from your eye to see the edge.

Wouldn't that mean a significant vignette ?
Also I have been searching on and off all morning to find a lens that does what you say. Can you point out actual production lens? Like private by design asked. I am unable to find it. Google is not being my friend.

But this topic has been good for me to learn some obscure/abstruse facts and came to read about front pupil and back Pupil.

Any way to set the aperture of electronic lenses to see this in action?

It just happens that I have a Quantaray (Tamron) 70-300 collecting dust in a closet that, when fully telephoto and focused all the way in has an entrance pupil that is larger than the front element.

 

[sjprg]

Whomever it was that called the 28-300 a dud, was a dunce. The lens does a great job and I hope the new 28-540 is an "L" lens that I can carry for another 10 years.

 

[Mt Spokane Photography]

Whomever it was that called the 28-300 a dud, was a dunce. The lens does a great job and I hope the new 28-540 is an "L" lens that I can carry for another 10 years.

The 28-300 is a great lens for what it does. But, carrying it everywhere is something I found to be impractical.

Personally, I expect this lens to be seriously considered by Canon. They seem to try and outdo the competition. It would be difficult to make a affordable lens better than the Sigma or Tamron super tele lenses, so they will go one step further. With Nikon now having a consumer grade super telephoto lens, Canon will try to outdo them all.

If such a lens were close focusing as they indicate, it might be a favorite for small birds too.

 

[Wizardly]

A lens with the radii and max thickness you mentioned has -- if both are convex -- a diameter of 34 mm and if the inner radius is concave, the outer one convex you will have the diameter of maximal 58mm -- but the thickness is then zero on the outside! Both lenses were positive.
If both radii describe concave surfaces 200mm diameter are possible (but pure nonsense I think) and the lens is negative.

Strange.

Outer surface is convex, inner surface is concave. The front group is a doublet.

 

[mb66energy]

A lens with the radii and max thickness you mentioned has -- if both are convex -- a diameter of 34 mm and if the inner radius is concave, the outer one convex you will have the diameter of maximal 58mm -- but the thickness is then zero on the outside! Both lenses were positive.
If both radii describe concave surfaces 200mm diameter are possible (but pure nonsense I think) and the lens is negative.

Strange.

Outer surface is convex, inner surface is concave. The front group is a doublet.

So we have a maximum of 58mm resulting in ca. f/10 - strange.

You mentioned a lens with larger entrance pupil than front lens element ...? If you "believe" in energy conservation this isn't possible: The front lens open diameter determines the area where light can pass into the lens and onto the sensor. This is THE LIMIT. Perhaps it is some visual effect which is more dramatic during zooming a lens?

 

[privatebydesign]

this lens would be huge. same diameter as the 300 F2.8.

This lens would be F2.8 at what focal length - not 300? So does it have to be the size of a 300 2.8?

Jack

The diameter of the front element is (approximately) the focal length divided by the aperture.

Thus a 560 / 5.6 = 100mm. A 300 /2.8 ~ 100mm. Same diameter.

The definition of the F stop is the focal length of the lens divided by the maximum diameter of the aperture when opened fully.

Not exactly. The f stop is the focal length over the entrance pupil diameter, not the actual aperture diameter. The entrance pupil is the image of the aperture from the objective side; it is the image of the opening that can be magnified by the optics.

I'd argue the f stop is the focal length over the apparent size of the entrance pupil diameter. But even if we quibble over technicalities the optical fact remains that the aperture can't be larger than the front element because it can't appear to be bigger than that, ergo the focal length divided by aperture will give you a minimum front element size.

If you can point me to a lens that has a front element smaller than its focal length divided by its actual fastest aperture I'd like to see it.

The patent gave the dimensions of the front element. 1.9 mm thick with curvature radii of 114 mm on the front and 230 mm on the back. The front element should only be 60mm, no?

I'm trying to understand this phenomenon myself, but conceptually I understand it to mean that the elements in front of the aperture are causing the rays to converge towards the aperture so that the image of the aperture is larger than the front element; if you were trying to see the aperture you would have to angle the lens away from your eye to see the edge.

Wouldn't that mean a significant vignette ?
Also I have been searching on and off all morning to find a lens that does what you say. Can you point out actual production lens? Like private by design asked. I am unable to find it. Google is not being my friend.

But this topic has been good for me to learn some obscure/abstruse facts and came to read about front pupil and back Pupil.

Any way to set the aperture of electronic lenses to see this in action?

It just happens that I have a Quantaray (Tamron) 70-300 collecting dust in a closet that, when fully telephoto and focused all the way in has an entrance pupil that is larger than the front element.

Care to elaborate on that? What lens, how did you determine the entrance pupil size etc etc?

 

[Wizardly]

So we have a maximum of 58mm resulting in ca. f/10 - strange.

You mentioned a lens with larger entrance pupil than front lens element ...? If you "believe" in energy conservation this isn't possible: The front lens open diameter determines the area where light can pass into the lens and onto the sensor. This is THE LIMIT. Perhaps it is some visual effect which is more dramatic during zooming a lens?

I'm having a hard time conceptualizing this myself given the geometry. Also, for specificity, the f number is 5.88, not 5.60; we're looking at 540.91mm/5.88 = 91.99 mm.

 

[Wizardly]

Care to elaborate on that? What lens, how did you determine the entrance pupil size etc etc?

I will absolutely defer to the smarter in this regard. If I'm misunderstanding the phenomenon please let me know. If you set a lens with a 62mm front element wide open and stand "infinitely" far away and cannot see the edge of the entrance pupil looking down optical axis unless you angle off from the optical axis, does that mean that the entrance pupil is larger than the front element itself?

I'm really struggling to understand how a lens with a 92mm entrance pupil can have a front element that is only around 60mm in diameter given the geometry specified in the patent.

 

[pierlux]

...but this isn't a 560mm...the patent is for a 540mm. Which brings the objective lens down a bit to 96mm. ... ...

The patent also indicates F-number = 5.88 at the long end, resulting in a 92mm front element by the same math. "Narrow" enough to accomodate 95mm filters. Or no filter thread at all, but a drop-in holder and built-in hood in which case this could be an "L" lens.

If this patent really represents the rumored supertelephoto zoom, the one I and others are waiting from Canon, the one which is intended to smack the Nikon/Sigma/Tamron offerings, then I don't know what to say.

A part of me, the rational one, is disappointed because I'd want something optically outstanding at the long end to the point that I'd prefer a supertelephoto prime over the zoom option and I really don't know how a 20x zoom could be optically outstanding.

Another part of me, the irrational and optimistic one, is excited because, as for the latest Canon releases, it can't be at least on par with, or superior to, its competitors IQ-wise and the extreme zoom range could help Canon sell more units and therefore help keep the price lower than most imagine. Maybe as low as $ 3-3.5K. Yes I know, lots of talk already about the price in past threads, but how is it otherwise possible to compete with $ 1.5-2K offerings?

 

[privatebydesign]

Care to elaborate on that? What lens, how did you determine the entrance pupil size etc etc?

I will absolutely defer to the smarter in this regard. If I'm misunderstanding the phenomenon please let me know. If you set a lens with a 62mm front element wide open and stand "infinitely" far away and cannot see the edge of the entrance pupil looking down optical axis unless you angle off from the optical axis, does that mean that the entrance pupil is larger than the front element itself?

I'm really struggling to understand how a lens with a 92mm entrance pupil can have a front element that is only around 60mm in diameter given the geometry specified in the patent.

You can't put your camera " "infinitely" far away ". You can only measure the apparent entrance pupil size from the point of focus, in this case where the sensor would be.

I haven't seen the block diagram/patent but it isn't possible to have a 92mm entrance pupil and a 60mm front element.

 

[privatebydesign]

I just looked at the patent.

It is an f5.88 at 540.91mm, so would require a minimum 92mm front element. The patent lists the front element diameter as 109.04mm ("The effective diameter of the front lens 109.04mm").

No mystery, no laws of optics and physics broken, just normal internet nonsense

Incidentally, if this became reality it would command an astronomic price, it is vastly more complex than a 300 f2.8 and has a similar sized front element. Current B&H 300mm price $6,099.

My guess would be if it covered a FF sensor that it would be a CN lens for video use and be aimed/positioned for 4k and 8k capture.

 

[Wizardly]

I just looked at the patent.

It is an f5.88 at 540.91mm, so would require a minimum 92mm front element. The patent lists the front element diameter as 109.04mm ("The effective diameter of the front lens 109.04mm").

*smacks forehead* must have gotten too used to the patents not explicitly stating the ED that I never even noticed. Also realize I had the radii on the wrong side and that convergence was never to occur. If you'll excuse me I'll be double checking all other equations for the last few days.....

 

[privatebydesign]

I just looked at the patent.

It is an f5.88 at 540.91mm, so would require a minimum 92mm front element. The patent lists the front element diameter as 109.04mm ("The effective diameter of the front lens 109.04mm").

*smacks forehead* must have gotten too used to the patents not explicitly stating the ED that I never even noticed. Also realize I had the radii on the wrong side and that convergence was never to occur. If you'll excuse me I'll be double checking all other equations for the last few days.....

No worries, we all have days like that! (I certainly do..........)

 

[neuroanatomist]

Well, it seems the laws of physics are safe again...at least for now.

 

[quiquae]

One reason why I suspect this lens is never going to hit the market: no IS.

Canon has not released a new lens more than 85mm long without IS since 2007 (800L).

 

[privatebydesign]

One reason why I suspect this lens is never going to hit the market: no IS.

Canon has not released a new lens more than 85mm long without IS since 2007 (800L).

Oh yes they have.

https://en.m.wikipedia.org/wiki/Canon_Cinema_EOS

Now if we consider the fact that Canon are pushing hard with the R&D on 8k, is it unreasonable to think they are going to want wider than Super 35 coverage of the current CN-E zoom lenses? How about the 4k DCI footage from the 1DX MkII and 1DC, both of which are wider than the CN-E zoom coverage so far.

 

[scyrene]

this lens would be huge. same diameter as the 300 F2.8.

This lens would be F2.8 at what focal length - not 300? So does it have to be the size of a 300 2.8?

Jack

The diameter of the front element is (approximately) the focal length divided by the aperture.

Thus a 560 / 5.6 = 100mm. A 300 /2.8 ~ 100mm. Same diameter.

The definition of the F stop is the focal length of the lens divided by the maximum diameter of the aperture when opened fully.

Not exactly. The f stop is the focal length over the entrance pupil diameter, not the actual aperture diameter. The entrance pupil is the image of the aperture from the objective side; it is the image of the opening that can be magnified by the optics.

I'd argue the f stop is the focal length over the apparent size of the entrance pupil diameter. But even if we quibble over technicalities the optical fact remains that the aperture can't be larger than the front element because it can't appear to be bigger than that, ergo the focal length divided by aperture will give you a minimum front element size.

If you can point me to a lens that has a front element smaller than its focal length divided by its actual fastest aperture I'd like to see it.

Playing devil's advocate... the MP-E maybe?

 

[privatebydesign]

this lens would be huge. same diameter as the 300 F2.8.

This lens would be F2.8 at what focal length - not 300? So does it have to be the size of a 300 2.8?

Jack

The diameter of the front element is (approximately) the focal length divided by the aperture.

Thus a 560 / 5.6 = 100mm. A 300 /2.8 ~ 100mm. Same diameter.

The definition of the F stop is the focal length of the lens divided by the maximum diameter of the aperture when opened fully.

Not exactly. The f stop is the focal length over the entrance pupil diameter, not the actual aperture diameter. The entrance pupil is the image of the aperture from the objective side; it is the image of the opening that can be magnified by the optics.

I'd argue the f stop is the focal length over the apparent size of the entrance pupil diameter. But even if we quibble over technicalities the optical fact remains that the aperture can't be larger than the front element because it can't appear to be bigger than that, ergo the focal length divided by aperture will give you a minimum front element size.

If you can point me to a lens that has a front element smaller than its focal length divided by its actual fastest aperture I'd like to see it.

Playing devil's advocate... the MP-E maybe?

You devil you!

But no, the MP-E 65 doesn't break any laws of optics either. This is the best image I could find of one from the front. If the outer two lines are 58mm apart, which they should be close to, then the inner lines are 24mm apart, 65mm / f2.8 = 23.2.