3
\$\begingroup\$

Given that Canon APS-C cameras differ in sensor size (e.g. 70D has a 22,5 x 15,0 mm and 700D has 22,3 x 14,9) is it possible for future models to switch from a 1,6 to 1,5 factor like Nikon crop sensors to gain some image quality? In other words, is there anything like stated maximum possible sensor size for EF-S lenses?

Improve this question
\$\endgroup\$

2 Answers 2

Reset to default
4
\$\begingroup\$

The diagonal of the 70D 22.5x15mm sensor is 27.04mm.
The diagonal of the 700D 22.3x14.9mm sensor is 26.82mm.
That's a difference of less than 1%.

The smallest APS-C sensor Canon has released is 22.2x14.7mm (EOS 1100D/Rebel T3).
The largest is 22.7x15.1mm.
This sensor first appeared in the 10D which was introduced before the EF-S lens system. EF-S lenses will not mount on the 10D, only EF lenses will. The EOS 300D/Digital Rebel, which is EF-S compatible, later had the same sensor.

The difference between the diagonal of the smallest and largest APS-C sensors Canon has released to date is only 2.3% (26.63mm vs. 27.26mm).

The diagonal of a 24x16mm sensor that yields a 1.5X crop factor is 28.85mm That's 5.6% wider than the largest APS-C sensor Canon has ever released.

I would be very surprised if Canon's current EF-S lenses would cover such a diagonal without significant light drop off in the corners.

Beyond that, though, your question may be making more of a 5% difference in the linear sizes of sensor than what is actually there. The differences in image quality between sensors has a lot more to do with other factors than a 5% difference in linear size. Just look at the improved performance at low ISO of the newly released 80D, with a 22.5x15.0mm sensor, compared to its predecessors.

Further, the perceived difference in the marketplace between sensors measured by testers such as DxO Mark and compared at ISO 100 but often used at ISO 1600 or 3200 is often more than the actual difference in real life situations. Especially when some cameras are tested using their manufacturer's own demosaicing algorithms and others are not tested using the manufacturer's own algorithms, but rather algorithms developed by the testing organization.

Improve this answer
\$\endgroup\$
3
  • \$\begingroup\$ Interesting though, DxOMark does not do any demosaicing or colour conversion. It means that cameras without low-pass filter and cameras requring less aggressive profiling will perform better in field - and Canon cameras generally do not match either criteria. \$\endgroup\$
    – Euri Pinhollow
    May 26, 2016 at 9:24
  • \$\begingroup\$ Sure DxO does demosaicing. Otherwise they could not measure things such as color sensitivity. The values output by a sensor are monochromatic luminance values with no color information. Color values are interpolated by demosaicing. Just go to DxO Mark and read their own methodology descriptions. \$\endgroup\$
    – Michael C
    May 26, 2016 at 9:33
  • \$\begingroup\$ let's discuss it in chat. \$\endgroup\$
    – Euri Pinhollow
    May 26, 2016 at 9:53
2
\$\begingroup\$

The answer will vary from lens to lens, and probably on other factors as well, such as whether the image stabilizer is in the center. :-)

The biggest problem with Canon's EF-S design has nothing to do with the image circle and everything to do with the short backfocus. Because the back of the lens sticks farther into the camera with EF, moving from a 1.6x crop to 1.5x might result in the mirror crashing into the back of the lens. I'm not sure what the limit is.

That's not to say that Canon couldn't get around this by making the mirror slide, slide and flip, etc., or by moving to a mirrorless design. With any of those approaches they/you could configure the device in software for a given lens to crop more or less depending on the corner fall-off.

Personally, I want to see a full-frame camera with a slide-flip mirror so that it can accommodate smaller, lighter EF-S lenses in a crop mode, or full-frame lenses in a full-frame mode as Nikon does. With that said, that would take a decent bit of engineering. Either way, it likely wouldn't result in noticeably better image quality for the reasons that Michael Clark already mentioned.

Improve this answer
\$\endgroup\$
1
  • 1
    \$\begingroup\$ The back of the lens can stick further into the camera with EF-S lenses on APS-C cameras with smaller mirrors than EF lenses on FF bodies with larger mirrors, but very few of them actually do. Only the very wide angle lenses do, and then only at the widest angle settings and the shortest focus distances. \$\endgroup\$
    – Michael C
    Jun 22, 2016 at 23:56

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.