Canon RF 70-400mm f/4.5-5.6L IS USM to be announced this year [CR2]

Joules

EOS 7D MK II
Jul 16, 2017
619
576
Hamburg, Germany
Maybe... But maybe the shorter flange distance (same mount diameter) is a gain that balances out any compromise. Guess we will see.
From the initial material that Canon released, it seems that a major advantage of the RF mount is the ability to design the lenses with far greater rear elements. The assumption that Tele RF glass can't be made smaller just because of the flange distance may or may not be true. But there's definitely more to it that just the distance.
 

usern4cr

EOS M50
Sep 2, 2018
33
29
Canon is coming out with amazing lenses. I'm a non-professional, more into still-photography than movies, and want to move up from MFT to FF. I hope their upcoming R body(s) will compete with the A7... line and exceed it in well designed ergonomics & (especially)menu system so that I can happily buy into the R system.

I want high quality lenses that I can hand-hold comfortably, and so the largest front element for me needs to stay between 70 and 75mm to keep the size and weight reasonable. That means f/2.8 could be up to 210mm, f/4 up to 300mm, f/5.6 up to 420mm, and f/8 up to 600mm. This 70-400mm f/4.5-5.6 should be a wonderful lens for f/5.6, and I hope they keep the same high image quality with this 5.7x zoom. Now could they do the same quality 5.7x to get something close to a 24-135mm f/2.2-2.8, 35-200mm f/2.2-2.8, 55-300mm f/3-4, and 105-600mm f/6-8 ?

If they had a high quality 105-600mm f/6-8 then I could see myself buying 2 of their new bodies with that lens for all telephoto work on one, and the 24-70(or 100)mm f/2.8 for wide angle to portraits on the other. That would probably cost $8-12K in total, but still well within my reach and I would happily buy it.

Canon has an opportunity to have cellphone, MFT, APS-C and DSLR users jump into FF with their R system instead of the current leader A7... system. I hope they succeed in this.
 
Last edited:

Danglin52

Wildlife Shooter
Aug 8, 2018
145
114
Canon is coming out with amazing lenses. I'm a non-professional, more into still-photography than movies, and want to move up from MFT to FF. I hope their upcoming R body(s) will compete with the A7... line and exceed it in well designed ergonomics & (especially)menu system so that I can happily buy into the R system.

I want high quality lenses that I can hand-hold comfortably, and so the largest front element for me needs to stay between 70 and 75mm to keep the size and weight reasonable. That means f/2.8 could be up to 210mm, f/4 up to 300mm, f/5.6 up to 420mm, and f/8 up to 600mm. This 70-400mm f/4.5-5.6 should be a wonderful lens for f/5.6, and I hope they keep the same high image quality with this 5.7x zoom. Now could they do the same quality 5.7x to get something close to a 24-135mm f/2.2-2.8, 35-200mm f/2.2-2.8, 55-300mm f/3-4, and 105-600mm f/6-8 ?

If they had a high quality 105-600mm f/6-8 then I could see myself buying 2 of their new bodies with that lens for all telephoto work on one, and the 24-100(or 135)mm f/2.8 for wide angle to portraits on the other. That would probably cost $8-12K in total, but still well within my reach and I would happily buy it.

Canon has an opportunity to have cellphone, MFT, APS-C and DSLR users jump into FF with their R system instead of the current leader A7... system. I hope they succeed in this.
I think you need a bigger element if you want to get to a useable 150-600 in low light. The f6-f8 would be a non-starter for me because it would not provide enough light for wildlife photographers in early / late light without driving the shutter too low or the ISO too hight (noise). My preference would be a f5.6, but no more than an f6.3, which gets into a similar range as the Sigma/Tamron. I just want better IQ and faster AF than the Sigma/Tamron. The other issues will be how far into the zoom range the lens requires f5.6 or f6.3
 

usern4cr

EOS M50
Sep 2, 2018
33
29
I think you need a bigger element if you want to get to a useable 150-600 in low light. The f6-f8 would be a non-starter for me because it would not provide enough light for wildlife photographers in early / late light without driving the shutter too low or the ISO too hight (noise). My preference would be a f5.6, but no more than an f6.3, which gets into a similar range as the Sigma/Tamron. I just want better IQ and faster AF than the Sigma/Tamron. The other issues will be how far into the zoom range the lens requires f5.6 or f6.3
I understand your worry about f/8 being too dim for low light use. But at 600mm you would have to get a much bigger front element (> 107mm) to get f/5.6 and it would be much bigger & heavier - which might be ok for you and many others, but it would just be too big for me to carry. If I had to keep the lens at or below f/5.6 I would rather use the 70-400mm f/4.5-5.6, and crop(1.5x) when I want 600mm (which would result in a smaller MP image at f/8). In fact, even if they offered a 105-600mm f/6-8 I might prefer the 70-400mm f/4.5-5.6, relying on crop as needed, and hope the smaller MP image was still detailed enough.

I can tell you that I have the Olympus 300mm f/4 pro lens which at 2x crop is equivalent to a FF 600mm f/8 lens. It is about as big & heavy as I can comfortably (enough) carry. I have taken stunning hand-held photos with it (thanks to it's 6.5 stop dual IS). My only complaint is that it is not a zoom and so it is hard to do good BIF(emphasis on flight) shots on it and having a zoom would fix that. I have not been bothered by low light issues with it.

And you could even consider a possible RF 55(or so)-300mm f/3-4 and use a 2x crop of it to get 110mm(or so)-600mm f/6-8. That would reduce the MP size by 4x, so you'd need a hi-res image sensor (80MP) to result in a 20MP image which would be the same as with the (2x crop) Oly 300mm f/4. The main difference is that you could now also use that RF lens for 55(or so)-300mm shots at full MP size at f/3-4 which would be great! I might even choose that best as a 2nd zoom with the RF 24-70(or 100) f/2.8 lens.
 
Last edited:

Danglin52

Wildlife Shooter
Aug 8, 2018
145
114
I understand your worry about f/8 being too dim for low light use. But at 600mm you would have to get a much bigger front element (> 107mm) to get f/5.6 and it would be much bigger & heavier - which might be ok for you and many others, but it would just be too big for me to carry.

I can tell you that I have the Oly 300mm f/4 pro lens which at 2x crop is equivalent to a 600mm f/8 lens. It is about as big & heavy as I can comfortably (enough) carry. I have taken stunning hand-held photos with it (thanks to it's 6.5 stop dual IS). My only complaint is that it is not a zoom and so it is hard to impossible to do good BIF(emphasis on flight) shots on it and having a zoom would fix that. I have not been bothered by low light issues with it.
The problem is that IS doesn't help you with subject movement. I have had many situations where I am pushing up into high ISO (12,800) when using f5.6 at the minimum shutter speed required to capture the animal.
 
  • Like
Reactions: SecureGSM

usern4cr

EOS M50
Sep 2, 2018
33
29
The problem is that IS doesn't help you with subject movement. I have had many situations where I am pushing up into high ISO (12,800) when using f5.6 at the minimum shutter speed required to capture the animal.
You're correct - current IS doesn't help with subject movement (AFAIK). If you're happy with a bigger & heavier lens then I hope Canon offers you the choice (I commend Canon for offering so many great lens choices). I do think that future cameras might use IS movement to track objects *during* the exposure which would help with that (so that the moving subject stays in focus even if it smears the background). But then again, that would still give a bad photo that would be fixed by shorter exposure needing bigger aperture - sigh!
 
Last edited:

AlanF

Canon 5DSR II
Aug 16, 2012
6,341
4,432
I understand your worry about f/8 being too dim for low light use. But at 600mm you would have to get a much bigger front element (> 107mm) to get f/5.6 and it would be much bigger & heavier - which might be ok for you and many others, but it would just be too big for me to carry. If I had to keep the lens at or below f/5.6 I would rather use the 70-400mm f/4.5-5.6, and crop(1.5x) when I want 600mm (which would result in a smaller MP image at f/8). In fact, even if they offered a 105-600mm f/6-8 I might prefer the 70-400mm f/4.5-5.6, relying on crop as needed, and hope the smaller MP image was still detailed enough.

I can tell you that I have the Olympus 300mm f/4 pro lens which at 2x crop is equivalent to a FF 600mm f/8 lens. It is about as big & heavy as I can comfortably (enough) carry. I have taken stunning hand-held photos with it (thanks to it's 6.5 stop dual IS). My only complaint is that it is not a zoom and so it is hard to do good BIF(emphasis on flight) shots on it and having a zoom would fix that. I have not been bothered by low light issues with it.

And you could even consider a possible RF 55(or so)-300mm f/3-4 and use a 2x crop of it to get 110mm(or so)-600mm f/6-8. That would reduce the MP size by 4x, so you'd need a hi-res image sensor (80MP) to result in a 20MP image which would be the same as with the (2x crop) Oly 300mm f/4. The main difference is that you could now also use that RF lens for 55(or so)-300mm shots at full MP size at f/3-4 which would be great! I might even choose that best as a 2nd zoom with the RF 24-100 f/2.8 lens.
The current 90D and M6 II have smaller pixels than the highest density Olympus sensor, 3.19 vs 3.32 micron pixels. The current 100-400mm II on them gives you as much reach as the 300mm Pro + 1.4xTC on an Olympus, and that is a nice light Canon combination. A 70-400mm on an 83 Mpx R would be quite a useful BIF and perched bird combo.
 
  • Like
Reactions: pj1974 and Joules

usern4cr

EOS M50
Sep 2, 2018
33
29
The current 90D and M6 II have smaller pixels than the highest density Olympus sensor, 3.19 vs 3.32 micron pixels. The current 100-400mm II on them gives you as much reach as the 300mm Pro + 1.4xTC on an Olympus, and that is a nice light Canon combination. A 70-400mm on an 83 Mpx R would be quite a useful BIF and perched bird combo.
Canon APS-C has 1.6x crop factor (correct me if I'm wrong). So a 400mm on them is 640mm on FF. Oly has 2.0x crop factor. So a 300 on them with a 1.4TC gives 840mm on FF. They would be closer to each other if you dropped the TC from the Oly so you'd get 600mm on FF.

I agree that a RF 70-400mm f/4.5-5.6 would be extremely useful for BIF and perched bird combo (and so many other great things!). It might be the best extra-wide-range tele-zoom for non-professionals there is for awhile. If you have a RF 24-70 f/2.8 and want a zoom for good portrait use (f/2.8) then you need a RF 70-200mm f/2.8, so adding a 70-400mm may be too redundant (and you'd instead want a 200-600 etc.). But if Canon came out with the RF 24-100mm f/2.8 they patented recently then it would cover the portrait range enough (at f/2.8) so you might consider just getting the RF 70-400mm f/4.5-5.6 for non-portrait telephoto use (and only 2 zooms would allow you to buy just 2 bodies if you don't like swapping lenses, and I don't like swapping them if possible)

Oh, I hope Canon comes out with some great R bodies soon!
 
Last edited:

AlanF

Canon 5DSR II
Aug 16, 2012
6,341
4,432
Canon APS-C has 1.6x crop factor (correct me if I'm wrong). So a 400mm on them is 640mm on FF. Oly has 2.0x crop factor. So a 300 on them with a 1.4TC gives 840mm on FF. They would be closer to each other if you dropped the TC from the Oly so you'd get 600mm on FF.

I agree that a RF 70-400mm f/4.5-5.6 would be extremely useful for BIF and perched bird combo (and so many other great things!). It might be the best extra-wide-range tele-zoom for non-professionals there is for awhile. I'd certainly love to have it - oh, I hope Canon comes out with some great R bodies soon!
I used the term "reach", which means resolution. The calculation you present is for "field of view", which is different from reach. For example, a 300mm lens on an 80 Mpx FF sensor has the same field of view as a 300mm lens on a 20 Mpx FF sensor, but the 80 MPx has twice the resolution or "reach". This is because pixel density comes into the calculation of reach or resolution. A 32.5 Mpx Canon APS-C sensor (pixel size equivalent to a 83 Mpx FF) has slightly higher resolution than a 20 Mpx MFT sensor (equivalent to an 80 Mpx FF), 3.19 vs 3.32 micron pixels. A 300mm lens on a 32.5 Mpx Canon APS-C sensor or an 83 Mpx FF sensor will just outresolve a 300mm on a 20 Mpx MFT, that is they all have very similar "reach" but very different fields of view.
 
  • Like
Reactions: pj1974 and Joules

usern4cr

EOS M50
Sep 2, 2018
33
29
I used the term "reach", which means resolution. The calculation you present is for "field of view", which is different from reach. For example, a 300mm lens on an 80 Mpx FF sensor has the same field of view as a 300mm lens on a 20 Mpx FF sensor, but the 80 MPx has twice the resolution or "reach". This is because pixel density comes into the calculation of reach or resolution. A 32.5 Mpx Canon APS-C sensor (pixel size equivalent to a 83 Mpx FF) has slightly higher resolution than a 20 Mpx MFT sensor (equivalent to an 80 Mpx FF), 3.19 vs 3.32 micron pixels. A 300mm lens on a 32.5 Mpx Canon APS-C sensor or an 83 Mpx FF sensor will just outresolve a 300mm on a 20 Mpx MFT, that is they all have very similar "reach" but very different fields of view.
Thanks for the explanation. Could you post what the equation(s) is? I'd be interested in looking at it.
 

Joules

EOS 7D MK II
Jul 16, 2017
619
576
Hamburg, Germany
Thanks for the explanation. Could you post what the equation(s) is? I'd be interested in looking at it.
What Alan is describing is the ability to crop beyond the crop factor of your format.

The common formats with their dimensions and following area and crop factor play a role here:

FF
36 mm * 24 mm = 864 mm^2
Diagonal of 43.27 mm
Crop factor 1 (Reference)

Canon APS-C
22.2 mm * 14.8 mm = 328.56 mm^2
Diagonal of 26.68 mm
Crop factor 1.62 (43.27/26.68)

Mirco Four Thirds
17.3 mm * 13.0 mm = 224.9 mm^2
Diagonal of 21.64 mm
Crop factor 2 (43.27/21.64)

The different areas mean that for a given lens, each sensor format will show a different field of view, regardless of resolution. But on the larger formats, given enough resolution you can crop down so that you are left with an area that's the size of one of the smaller formats.

The term resolution is ambiguous. It can mean the absolute number of pixel in an image, for instance. Or it can be a measure of actual detail if it is given as a density.

Cropping an APS-C image down to a MFT format reduces the total resolution down to 224.9 / 328.56 = 0.685 ~ 68.5 %. Meaning if your total APS-C resolution is 32.5 MP (EOS M6 II, EOS 90D) you'd end up with 22.3 MP. Which is more than MFT has to offer and why more megapixels on an identical sensor size can deliver more detail should be intuitive.

An alternative way to get to this right away is to simply look at the pixel density or like Alan did, the size of each pixel. For a 32.5 MP Canon APS-C that would be 328,56 * 10^-6 m / (32.5 * 10^6) = 10.109 Micrometer^2, and taking the square root it get's you 3.18 Micrometer which is the width and height of a pixel that Alan gave. There's some rounding error involved.

So square root of (sensor area divided by number of pixels).
 
  • Like
Reactions: AlanF

uri.raz

EOS RP
Jan 5, 2016
213
134
Cropping an APS-C image down to a MFT format reduces the total resolution down to 224.9 / 328.56 = 0.685 ~ 68.5 %. Meaning if your total APS-C resolution is 32.5 MP (EOS M6 II, EOS 90D) you'd end up with 22.3 MP. Which is more than MFT has to offer and why more megapixels on an identical sensor size can deliver more detail should be intuitive.
That assumes the lens' resolving power is good enough to get a real gain from the additional resolution.

IIRC, past threads indicated there's always some gain, but it drops like 1/X^2, so in the example above it could be somewhere from a little less than the 10% sensor resolution difference (reasonable for good tele lens) to nearly nothing.
 

SteveC

M6 mk II
Sep 3, 2019
540
375
An alternative way to get to this right away is to simply look at the pixel density or like Alan did, the size of each pixel. For a 32.5 MP Canon APS-C that would be 328,56 * 10^-6 m / (32.5 * 10^6) = 10.109 Micrometer^2, and taking the square root it get's you 3.18 Micrometer which is the width and height of a pixel that Alan gave. There's some rounding error involved.

So square root of (sensor area divided by number of pixels).
This question may just serve to show my ignorance, but isn't there something else going on too? I'm going to have to describe what I mean when I really need to draw a picture, so please bear with me:

Take a 50 millimeter lens, for some (notional) Full Frame mount X. It has to project an image of the subject onto a sensor, and that image has to be of a certain precise width by the time it hits the focal plane. If it's too big, then the camera will behave as if it's cropping, too small and it will need cropping after the fact.

A different mount, Full Frame Mount Y, with its 50 millimeter lens presumably must project the same subject to the SAME width, or you will have two different "sizes" of image on the two different full frame sensors, and 50mm won't be behaving identically on the two mounts. But note, the X and Y lenses might, due to the mount design, be at very different distances from the sensor.

So it's not just the physical size of the pixels, it's the angle they subtend on the projected image. Now one would HOPE that the full frame sensor in every full frame camera subtends the same angle on the image (given the same focal length of lens); and therefore that comparing, say, 24MP sensors on the two cameras actually is comparing apples to apples instead of apples to oranges...but I don't know. Is this standardized so that not just the size of full frame sensors is identical, but also their field of view?
 

Joules

EOS 7D MK II
Jul 16, 2017
619
576
Hamburg, Germany
IIRC, past threads indicated there's always some gain, but it drops like 1/X^2, so in the example above it could be somewhere from a little less than the 10% sensor resolution difference (reasonable for good tele lens) to nearly nothing.
A difference of 20 vs 22 MP is basically nearly nothing. I don't think anybody here claims differently. But more is more, even if it is a negligible difference. This was just about some theory, not making any statement on the pros and cons of sensor formats, I think.

one would HOPE that the full frame sensor in every full frame camera subtends the same angle on the image (given the same focal length of lens); and therefore that comparing, say, 24MP sensors on the two cameras actually is comparing apples to apples instead of apples to oranges...but I don't know. Is this standardized
I'm not sure if I understand the question or can really answer it. I believe you're asking if the field of view depends on more than the focal length? Basically, asking if the lens has to properly scale the image so that some technical distance within the lens actually correlates to the field of view we know as 50mm?

Reading the Wiki on the matter, it appears that the focal length as we use it is an effective value anyway and is directly related to the FOV, with the only other variable affecting it being the sensor diagonal:

"Focal length (f) and field of view (FOV) of a lens are inversely proportional. For a standard rectilinear lens, FOV = 2 arctan (x/(2f)), where x is the diagonal of the film"

From: https://en.m.wikipedia.org/wiki/Focal_length

So if it's labeled 50mm, that's what you get - in combination with your sensor diagonal. And that is a variable that you can shrink to you liking using cropping. Was that what you were going after?
 

usern4cr

EOS M50
Sep 2, 2018
33
29
That assumes the lens' resolving power is good enough to get a real gain from the additional resolution.

IIRC, past threads indicated there's always some gain, but it drops like 1/X^2, so in the example above it could be somewhere from a little less than the 10% sensor resolution difference (reasonable for good tele lens) to nearly nothing.
I see that this is proportional to 1 / pixel area. So as the pixels shrink down to almost nothing this number goes through the roof. Personally, I don't see much value in this resultant value. But I do thank you all for explaining it to me.

One thing that's not being considered here is that a FF picture which is stored as a FF picture and later cropped down 2x to get a MFT size image is not really the same as taking a MFT image. When you crop after the fact, you can adjust where you want to crop from within the entire FF image. I often find I want some off-center portion of the image that would not exist if initially stored as the smaller image, and so the freedom to crop anywhere from a larger image into the smaller size is much more important than just storing the smaller image in the first place.
 
Last edited:

SteveC

M6 mk II
Sep 3, 2019
540
375
So if it's labeled 50mm, that's what you get - in combination with your sensor diagonal. And that is a variable that you can shrink to you liking using cropping. Was that what you were going after?
I think it is, thanks! it tells me they compensate for it, and a (say) Sony 50mm on a Sony full frame will have the same FOV as a Canon 50 mm on a Canon EF body, or a Canon R 50 on a Canon R body, all three pics taken from the same place. Since the sensors are of identical size and presuably identical aspect ratio (right?), the SOLE determinant of how much a pixel sees compared to the others should therefore be the size of the pixel on that sensor.

Needless to say the Canon picture will look better than the Sony one. :D