Canon announces development of the EOS R5 full-frame mirrorless camera

[joestopper]

The magnification between the front of a lens and the physical diaphragm is the only reason an 18-55mm f/3.5-5.6 is not an 18-55mm f/3.5-11, a 70-300mm f/4-5.6 is not a 70-300mm f/4-17, or a 70-200mm f/2.8 is not a 70-200mm f/2.8-8.

The "effective aperture" is the entrance pupil as viewed by the subject (i.e. from the front of the lens). With narrow angle lenses (telephoto lenses), the front element must be at least as large as the entrance pupil because the light it is focusing is almost collimated. With wider angle lenses, the entrance pupil can actually be larger than the diameter of the front element when viewed from a point on the len's center axis, though that causes severe vignetting and "cat's eye" bokeh for objects in the periphery. To keep the entire entrance pupil visible from the entire field of view is why many wide angle lenses have those bulbous front elements that are much larger than the e.p.

When light is refracted by a converging lens, the field density increases in proportion to the magnification, just as the cross sectional area through which the light passes decreases in proportion to the magnification. With a simple thin lens, there is a point halfway between the lens and the point of focus where all light passing through the lens passes through a single point where the two cones of light meet at their tips. That's why the image projected on the film/sensor is inverted.

Mixing up several things:
The discussion started from a *prime* lens and not a zoom. And I stated, the front lens element must be at least the size of the aperture.in a meaningful design. And that is true (and not a contradiction to your statements regarding a zoom).
However, if you do magnify in front of the aperture then the t-stop is quite different from the f-stop value. While that can be the case for a zoom at some focal length, it makes no sense for a prime.
Proof: Check f-stop and t-stop values for various primes and you will see they are very close (and the reason why t-stop value is always a bit worse than f-stop value, is the loss of transmission through glass).

 

[joestopper]

She also is unsure if 8k will be cropped. Hmmm. How can the math that was told to me earlier be wrong??

Ask her this question: How much can an 8k be cropped on a 45mp sensor?
(Hint: it is far less then the claimed 2x that was posted earlier by someone in this forum)

 

[sanj]

I shoot bands in dark venues. On my 5D 3 I am often dancing between 3 to 10 thousand iso. I try very hard to avoid going so super high as post processing is a bitch. With auto iso the camera would want to go to those super high iso's regularly. I can't give the camera that option.

Then would your camera not go to slower shutter speed giving you blurred photos? What is worse - blurred photos of high iso noise?

 

[Kit.]

Ask her this question: How much can an 8k be cropped on a 45mp sensor?
(Hint: it is far less then the claimed 2x that was posted earlier by someone in this forum)

Technically, it can be cropped 444x and then upscaled.
Knowing Sony trolls, they might as well suggest that it's the case.

 

[Michael Clark]

Mixing up several things:
The discussion started from a *prime* lens and not a zoom. And I stated, the front lens element must be at least the size of the aperture.in a meaningful design. And that is true (and not a contradiction to your statements regarding a zoom).
However, if you do magnify in front of the aperture then the t-stop is quite different from the f-stop value. While that can be the case for a zoom at some focal length, it makes no sense for a prime.
Proof: Check f-stop and t-stop values for various primes and you will see they are very close (and the reason why t-stop value is always a bit worse than f-stop value, is the loss of transmission through glass).

Every prime I've ever seen magnifies (either negatively for a retrofocus design or positively otherwise) between the physical diaphragm and the front of the lens. Are you saying Zeiss 50mm Planar lenses are not meaningful designs? Their e.p.s are slightly larger than their front elements.



It's pretty easy to illustrate that wider angle lenses can have larger entrance pupils than front elements.

 

[yeahright]

I shoot bands in dark venues. On my 5D 3 I am often dancing between 3 to 10 thousand iso. I try very hard to avoid going so super high as post processing is a bitch. With auto iso the camera would want to go to those super high iso's regularly. I can't give the camera that option.

you could always limit the maximum allowable value in Auto ISO

 

[StoicalEtcher]

Hmmm, after reading all that the biggest surprise was you didn't use the term 'Maths' because I always assumed you were British. ...

What have we done to upset you this time??

 

[slclick]

What have we done to upset you this time??

What am I, the Larry David of the Forum?

 

[davo]

Ask her this question: How much can an 8k be cropped on a 45mp sensor?
(Hint: it is far less then the claimed 2x that was posted earlier by someone in this forum)

I also notice she specifically says IBIS works in conjunction with "RF" lens stabilization for up to 7-8 stops. I wonder if it will do the same for adapted EF stabilized lenses afterall? I have been lead to believe it will.....maybe not though.

 

[Michael Clark]

Then would your camera not go to slower shutter speed giving you blurred photos? What is worse - blurred photos of high iso noise?

One sometimes trades keeper ratio for better keepers. If timed properly (for the subjects strumming guitars and singing) and stabilized properly (for the camera), one can get usable images at much slower Tv than conventional wisdom dictates.

EOS 5D Mark III, ISO 5000, 1/80, f/2.2, EF 135mm f/2 L




EOS 7D Mark II, ISO 1600, 1/320, f/3.2, 100mm (EF 70-200mm f/2.8 L IS II)

 

[davo]

Then would your camera not go to slower shutter speed giving you blurred photos? What is worse - blurred photos of high iso noise?

Blurred photos are no good. Depending on if my subject is static or moving I am always tweaking my shutter speed as well as slow as I think I can get away with so I can avoid those extreme iso's.

 

[Michael Clark]

you could always limit the maximum allowable value in Auto ISO

It's still a pain in post for each shot to be at different ISOs, and manual exposure usually works better anyway.

 

[davo]

One trades keeper ratio for better keepers. If timed properly (for the subjects strumming guitars and singing) and stabilized properly (for the camera), one can get usable images at much slower Tv than conventional wisdom dictates.

ISO 5000, 1/80, f/2.2, 135mm

View attachment 188725


ISO 1600, 1/320, f/3.2, 100mm

View attachment 188726

Yea its often a frantic dance. If the venue and players are well lit its no issue, but if not you have to stay on your toes to avoid motion blur. What lens am I using at the moment? What is the lighting like at THIS moment on my subject? How much motion? Can I drop my shutter speed at all? No? Well up the iso then.....Oh the stage brightened!...turn down the iso and raise the shutter. Swap lens to tele? well now I have to raise the shutter more which means raise the iso....well now the stage went darker!....cant raise iso anymore and cant slow the shutter,,,so now I just have to wait for things to improve

 

[Michael Clark]

Mixing up several things:
The discussion started from a *prime* lens and not a zoom. And I stated, the front lens element must be at least the size of the aperture.in a meaningful design. And that is true (and not a contradiction to your statements regarding a zoom).
However, if you do magnify in front of the aperture then the t-stop is quite different from the f-stop value. While that can be the case for a zoom at some focal length, it makes no sense for a prime.
Proof: Check f-stop and t-stop values for various primes and you will see they are very close (and the reason why t-stop value is always a bit worse than f-stop value, is the loss of transmission through glass).

Where, exactly, are the T-stops going up in proportion to the magnification as these constant aperture zoom lenses increase magnification between the aperture diaphragm and the front of the lens?

 

[davo]

you could always limit the maximum allowable value in Auto ISO

I would BUT I STILL only want the iso the go up to the highest limitation as seldom as possible. Post processing at the higher iso is not ideal at all. You can get usable images but nothing like when 2-3 stops lower

 

[BillB]

I also notice she specifically says IBIS works in conjunction with "RF" lens stabilization for up to 7-8 stops. I wonder if it will do the same for adapted EF stabilized lenses afterall? I have been lead to believe it will.....maybe not though.

Well, if we don't know after 1100 CR posts to this thread whether Canon IBIS will interact with EF lenses, the chances are she doesn't either. There are still a lot of missing pieces to this puzzle. Even if the EF lenses do not support IBIS, there will still be 5 stops of stabilization.

 

[Michael Clark]

Yea its often a frantic dance. If the venue and players are well lit its no issue, but if not you have to stay on your toes to avoid motion blur. What lens am I using at the moment? What is the lighting like at THIS moment on my subject? How much motion? Can I drop my shutter speed at all? No? Well up the iso then.....Oh the stage brightened!...turn down the iso and raise the shutter. Swap lens to tele? well now I have to raise the shutter more which means raise the iso....well now the stage went darker!....cant raise iso anymore and cant slow the shutter,,,so now I just have to wait for things to improve

Getting the color right in post can go a long way to making it look like there was better/more light than there actually was.

EOS 5D Mark III, ISO 5000, 1/200, f/4 (24-105mm f/4 L IS - temporary outdoor stage vibrating with the music)




EOS 7D Mark II, ISO 5000, 1/250, f/2.8, 105mm (EF 70-200mm f/2.8 L IS II)

 

[Sharlin]

I also notice she specifically says IBIS works in conjunction with "RF" lens stabilization for up to 7-8 stops. I wonder if it will do the same for adapted EF stabilized lenses afterall? I have been lead to believe it will.....maybe not though.

EF lenses may not gain as much because they’re not designed to coordinate with the body on stabilization, and the EF interface is likely not fast enough to do that in any case. Whereas combining IBIS and lens IS was likely one of the ”killer apps” of the RF mount since the very beginning.

 

[Rule556]

The EF 135mm f/2 L is a harder lens the find on the used market than many others. Not too many who have them are willing to let them go unless they are a "below average" copy.

Time is a luxury I have.

 

[Ozarker]

Will EF image stabilization interact with R series IBIS? Of course it will.