Will there be an APS-C EOS R-series camera?

AlanF

Desperately seeking birds
CR Pro
Aug 16, 2012
12,355
22,529
A STORY

A man took his 12 year old daughter out to take some photographs. They had two cameras, one with what is called a Full-Frame sensor, and the other had what is called a Crop sensor. The stood behind a fence, looking at a house in the distance. The man gave his daughter the Full-Frame camera and had her look through the 300mm lens at the house. "It's still kinda small," she said.

The man put the same lens on the Crop camera. "Here, try this one," he said.

She looked through the camera and said, "Hey. the house looks bigger with this camera! It's like I'm closer; like my arms were reaching farther out in front of me!"

"Yes," said the man, "it's like you have greater reach."

"Oh, I understand," said the 12 year old.

"That's because you are still too young to be stupid," said the man, grinning.

THE END
There's a nice discussion about "reach" 10 years ago on dpr: "Reach" is a colloquial term for the ability of a lens to "pull in" detail from a distance. It doesn't have an official photographic definition. https://www.dpreview.com/forums/thread/2944342
Your story illustrates well the confusion between reach and field of view. If the viewfinder of the crop has the same magnification as the FF, then the image will appear larger and closer, which is nice for the 12 year old or someone using the camera as telescope because it is looking at a narrower field of view. But, for the photographer who wants to see the detail in the image, then the resolution of the sensor is crucial. A 20 Mpx 7DII will give a larger image in the viewfinder than for the same lens on a 50 Mpx 5DS, but they give as near as dammit the same detail when you process the image and crop the FF. A 32 Mpx Canon 90D and M6II with its 1.6x crop factor outresolves and gives more reach than a 20 Mpx Olympus micro 4/3 with its 2x crop factor with the same focal length lens. But, the Olympus will have a narrower field of view and bigger image in the evf.
 
Upvote 0
Jul 21, 2010
31,099
12,863
If the crop sensor has greater pixel density, then there is some rationale for “reach.”
Agreed. But Canon’s marketing statement that the R5’s 1.6x crop mode provides more reach is, to put it politely, a steaming pile of bovine scat.

Sure, one could argue that the 1.6x magnified view in the EVF (which I’m assuming happens with crop mode, but I don’t actually know) is a form of ‘more reach’. But if that’s true, that 12 year old girl could push the magnifying glass button and get so much reach she might be injured by bumping up against the wall of that house.
 
  • Like
  • Haha
Reactions: 1 users
Upvote 0

SteveC

R5
CR Pro
Sep 3, 2019
2,678
2,592
Here's how an APS-C crop mode on a FF sensor provides more reach in terms of pixels on duck:
View attachment 200370


Here's how an APS-C crop mode on a FF sensor provides more reach in terms of both being printed on 6x4 paper:
View attachment 200371

Yep, that's my understanding too. (If you think I'm contradicting myself, there's likely an issue of semantics between us, despite my efforts.)

If all you do is crop [either manually or with a crop mode], no difference in pixels on duck, but you MAY (if the print's pixel density after cropping is good enough for the medium you're printing (or displaying) on--which is not the situation in your example) believe you have more reach with the crop mode, when really all it did was save you a step in post-processing (and, likely a step in chimping).

Example for illustration: If you're displaying on a 1920x1080 monitor, as long as your crop mode gives you at least that many pixels, it gives you what I'm going to call "media reach" on that monitor. But I put that in quotes there because it's not real reach, it's just "digital zoom" (ugh! I could rant about point and shoot advertising practice!) but not so much as to start looking pixelated.
 
Last edited:
  • Like
Reactions: 1 user
Upvote 0

AlanF

Desperately seeking birds
CR Pro
Aug 16, 2012
12,355
22,529
If the crop sensor has greater pixel density, then there is some rationale for “reach.”
Pixel density is indeed crucial. Higher pixel density in crop cameras was the reason why many of us reached for our 7D to get more reach. But, cropping a FF to APS-C in our mirrorless in the menu doesn’t change the pixel density. What promoted this discussion were comments, including from Canon, that it increases reach.
 
  • Like
Reactions: 1 user
Upvote 0

AlanF

Desperately seeking birds
CR Pro
Aug 16, 2012
12,355
22,529
Yep, that's my understanding too. (If you think I'm contradicting myself, there's likely an issue of semantics between us, despite my efforts.)

If all you do is crop [either manually or with a crop mode], no difference in pixels on duck, but you MAY (if the print's pixel density after cropping is good enough for the medium you're printing (or displaying) on--which is not the situation in your example) believe you have more reach with the crop mode, when really all it did was save you a step in post-processing (and, likely a step in chimping).

Example for illustration: If you're displaying on a 1920x1080 monitor, as long as your crop mode gives you at least that many pixels, it gives you what I'm going to call "media reach" on that monitor. But I put that in quotes there because it's not real reach, it's just "digital zoom" (ugh! I could rant about point and shoot advertising practice!) but not so much as to start looking pixelated.
It's worthwhile using crop mode on the R5 (or elsewhere) if you are in situations where you would be cropping in post anyway, which is much of the time for me. Saving of RAW file space is certainly helpful, and there are reports that eyeAF on birds works at longer distances in crop mode - extenders certainly increase the range of bird eyeAF. I should really get into the habit of using crop mode more. But, when doing doing BIF and DIF I need the wider field of view as also for close-ups.
 
Upvote 0
"AlanF: A 32 Mpx Canon 90D and M6II with its 1.6x crop factor outresolves and gives more reach than a 20 Mpx Olympus micro 4/3 with its 2x crop factor with the same focal length lens."
An interesting comparison but if my math is right these are closer than one might think. The 90D has 21 Mpx within the same sensor area that the Olympus has 20. Since the Olympus lacks an AA filter, it is potentially sharper though.
 
Last edited:
Upvote 0

AlanF

Desperately seeking birds
CR Pro
Aug 16, 2012
12,355
22,529
"AlanF: A 32 Mpx Canon 90D and M6II with its 1.6x crop factor outresolves and gives more reach than a 20.4 Mpx Olympus micro 4/3 with its 2x crop factor with the same focal length lens."
An interesting comparison but if my math is right these are closer than one might think. The 90D has 21 Mpx within the same sensor area that the Olympus has 20. The tough part is how to deal with the different aspect ratios. If a 4:3 image is what is useful then the 20 Mpx Olympus would have more "useful pixels on duck" than the 90D once the 90D is cropped to 4:3. Obviously the opposite is true if a 3:2 ratio image is most useful. Hence, if printing 6x4 the edge goes to the Canon 90D; however, if printing for a US letter page or 8x10 the edge goes to the Oly. Since the Olympus lacks an AA filter, it is potentially sharper though.
According to the Nyquist theorem, the absolute limiting resolution of a sensor is half the sampling frequency, that is the number of pixels per mm. You don't need the aspect ratios to calculate the resolution, you just need to know the pixel sizes: the Canon 90D has 3.19µ (313.5px/mm) and the Olympus 3.32µ pixels (301.2px/mm). Thus, the absolute Nyquist limit of the Canon sensor is 4.08% higher resolution than the Olympus. Of course, the resolutions will be lowered by Bayer matrix, low pass filters, noise, lens etc etc. But, the important take home message is that just because the M4/3 has a 2x crop factor and the APS-C only 1.6x it doesn't mean the M4/3 outresolves (has greater reach than) the APS-C. Pixel density (pixel size) is a crucial factor in resolution/reach. The relevant Canon and Olympus pixel sizes are close, but I had done these calculations before I made my comment.

Your argument that you can get more "useful pixels on duck" because pixels could be lost on cropping APS-C to a 4/3 ratio is flawed. An Olympus M4/3 sensor is 17.3mmx13mm, and the Canon APS-C 22.3mmx14.9mm. So, the Olympus sensor fits inside the Canon APS-C sensor, whose image can be cropped to the same size as the Olympus without losing any pixels that are in the Olympus. The same focal length lens on the Canon 90D APS-C will always put 1.0408x1.0408 times the number of pixels on the duck than does the Olympus M4/3, 8% more.
 
Upvote 0
According to the Nyquist theorem, the absolute limiting resolution of a sensor is half the sampling frequency, that is the number of pixels per mm. You don't need the aspect ratios to calculate the resolution, you just need to know the pixel sizes: the Canon 90D has 3.19µ (313.5px/mm) and the Olympus 3.32µ pixels (301.2px/mm). Thus, the absolute Nyquist limit of the Canon sensor is 4.08% higher resolution than the Olympus. Of course, the resolutions will be lowered by Bayer matrix, low pass filters, noise, lens etc etc. But, the important take home message is that just because the M4/3 has a 2x crop factor and the APS-C only 1.6x it doesn't mean the M4/3 outresolves (has greater reach than) the APS-C. Pixel density (pixel size) is a crucial factor in resolution/reach. The relevant Canon and Olympus pixel sizes are close, but I had done these calculations before I made my comment.

Your argument that you can get more "useful pixels on duck" because pixels could be lost on cropping APS-C to a 4/3 ratio is flawed. An Olympus M4/3 sensor is 17.3mmx13mm, and the Canon APS-C 22.3mmx14.9mm. So, the Olympus sensor fits inside the Canon APS-C sensor, whose image can be cropped to the same size as the Olympus without losing any pixels that are in the Olympus. The same focal length lens on the Canon 90D APS-C will always put 1.0408x1.0408 times the number of pixels on the duck than does the Olympus M4/3, 8% more.
You are right. My math and yours agree, my extrapolation to print size was in error and has been edited out from above comment. Published spec. numbers I used don't have as many significant digits as what you list, nevertheless, I agree the 90D will have slightly more 'pixels on duck' than the Oly (between 5-10% depending on rounding of input numbers). I still think that the Oly without an AA filter may have the edge in apparent sharpness. I don't have a 90D to test however. comparison on https://www.imaging-resource.com seems to support that hypothesis.
 
  • Like
Reactions: 1 user
Upvote 0
Jul 21, 2010
31,099
12,863
I still think that the Oly without an AA filter may have the edge in apparent sharpness.
Since the purpose of an AA filter is to blur an image slightly, it certainly makes a difference. It’s not just presence or absence, they come in different strengths. The R3 has a lower strength AA filter, which is the basis for Canon‘s statement that its 24MP sensor delivers better resolution then the 30 MP sensor in the 5DIV.
 
  • Like
Reactions: 1 user
Upvote 0

Jack Douglas

CR for the Humour
Apr 10, 2013
6,980
2,602
Alberta, Canada
I'm not really qualified to comment but, not unlike the arguments about extenders, there was one thing that stuck out for me and Alan touched on it. When you can see a distant subject larger you have a better chance of getting a spot AF point where you want it.

I have no idea what the spot AF areas were for crop cameras so maybe the focus area was simply larger and so there was no advantage. I also have no idea how the spot AF behaves switching to crop mode with the R5 but it seems maybe there is some difference. Nevertheless, seeing a subject better still is somewhat advantageous.

With my 6D and 300 2.8 in the early days I tried shooting the little birdies in the far away trees with just the 1.4X because supposedly the IQ would be better but quickly abandoned that for 2X based on what I could see as I was trying to get AF on subject.

Jack
 
  • Like
Reactions: 3 users
Upvote 0

AlanF

Desperately seeking birds
CR Pro
Aug 16, 2012
12,355
22,529
You are right. My math and yours agree, my extrapolation to print size was in error and has been edited out from above comment. Published spec. numbers I used don't have as many significant digits as what you list, nevertheless, I agree the 90D will have slightly more 'pixels on duck' than the Oly (between 5-10% depending on rounding of input numbers). I still think that the Oly without an AA filter may have the edge in apparent sharpness. I don't have a 90D to test however. comparison on https://www.imaging-resource.com seems to support that hypothesis.
Thanks for concurring with my calculations. You are right that the AA-filter could make a significant difference, which is why I covered myself by writing that the Bayer matrix, low pass filters, noise, lens etc etc will lower resolution, and Olympus is rather good at getting the best out of their sensors. There is one additional factor, you get about 15% more pixels in height and 30% in width with the APS-C over the M4/3, so if you can fill the frame, the APS-C could gain another 15-30%. Look, I am not running down the Olympus, they make great cameras with some really neat features, like having a buffer of shots before you click the shutter, but I am just pushing the analysis that crop factor does not necessarily equate to to reach.
 
  • Like
Reactions: 1 user
Upvote 0

AlanF

Desperately seeking birds
CR Pro
Aug 16, 2012
12,355
22,529
Since the purpose of an AA filter is to blur an image slightly, it certainly makes a difference. It’s not just presence or absence, they come in different strengths. The R3 has a lower strength AA filter, which is the basis for Canon‘s statement that its 24MP sensor delivers better resolution then the 30 MP sensor in the 5DIV.
I find slightly better resolution with the 45 Mpx R5, with its novel Canon AA-filter, than the 50 Mpx 5DSR, consistent with Canon's claims and measurements I've seen.
 
  • Like
Reactions: 1 user
Upvote 0

dcm

Enjoy the gear you have!
CR Pro
Apr 18, 2013
1,088
846
Colorado, USA
The confound has been that until relatively recently, crop sensors had higher MP counts than FF sensors, and sadly in the minds of many people correlation does mean causation.

Don’t normally see confound used as a noun. Many dictionaries no longer list that form.
 
Upvote 0