Are smaller sensors more efficient...

Started Nov 2, 2013 | Discussions thread
Flat view
Great Bustard Forum Pro • Posts: 43,932
Are smaller sensors more efficient...

...than larger sensors, as a general rule?

Specifically, do smaller sensors for a given generation have higher QEs and/or lower read noise per area of the photo?

Let's consider the following cameras from smallest to largest sensor:

  • G15: 12 MP 4.7x, 1.8 micron pixels, 59% QE, min read noise ~ 1.6 electrons per pixel = 5.5 electrons per µphoto, max saturation = 7335 electrons per pixel = 88020 electrons per µphoto.
  • RX100: 20 MP 2.7x, 2.4 micron pixels, 52% QE, min read noise ~ 2.0 electrons per pixel = 8.9 electrons per µphoto, max saturation = 10040 electrons per pixel = 200800 electrons per µphoto.
  • EM5: 16 MP mFT, 3.7 micron pixels, 53% QE, min read noise ~ 2.5 electrons per pixel = 10 electrons per µphoto, max saturation = 25041 electrons per pixel = 400656 electrons per µphoto.
  • D5200: 24 MP 1.5x, 3.9 micron pixels, 53% QE, min read noise ~ 1.9 electrons per pixel = 9.3 electrons per µphoto, max saturation = 27517 electrons per pixel = 660408 electrons per µphoto.
  • 6D: 20 MP FF, 6.3 micron pixels, 50% QE, min read noise ~ 2.2 electrons per pixel = 9.8 electrons per µphoto, max saturation = 76606 electrons per pixel = 1532120 electrons per µphoto.
  • D800: 36 MP FF, 4.7 micron pixels, 56% QE, min read noise ~ 3.0 electrons per pixel = 18 electrons per µphoto, max saturation = 44972 electrons per pixel = 1618992 electrons per µphoto.

So, I'm thinking this is a very good range of sensor sizes and pixel sizes to compare for a generation of cameras for the discussion.

First off, I notice that there doesn't seem to be any significant difference in QE based either on sensor size or pixel size, unless there's cheating going on -- that is, someone is using a weaker CFA to increase QE. So, if anyone can demonstrate that larger sensors (or larger pixels), as a rule, use weaker CFAs, that would be worth noting.

However, when it comes to read noise per µphoto, we have an interesting situation. The G15 is, by far, the lowest, the D800 is, by far, the highest, and the rest are about the same.

But the G15 has, by far, the lowest pixel count, and the D800 has, by far, the largest pixel count, so it seems that it's the higher pixel count, not the larger sensor size, resulting in more read noise per area.

As far as max saturation per area goes, that seems to fall right in line as proportional to sensor area.

Now, at base ISO, the read noise is significantly higher for some of these sensors (especially the 6D), but my understanding is this is due to a deficient ADC and more a Canon issue than it is a sensor size or pixel size issue.

So, how say you all? Is it reasonable to claim that sensors of the same generation are, as a general rule, equally efficient, at least from 4.7x to FF? Or would a pattern have emerged if I had worked the numbers for more cameras and/or performed a more rigorous analysis?

Canon EOS 6D Canon PowerShot G15 Nikon D800
If you believe there are incorrect tags, please send us this post using our feedback form.
Flat view
Post (hide subjects) Posted by
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark MMy threads
Color scheme? Blue / Yellow