How do i explain to someone its not about megapixels ?

Started Oct 4, 2013 | Questions thread
Great Bustard Forum Pro • Posts: 42,943
A little understanding just might help.

dholl wrote:

Mikael Risedal wrote:

you are wrong, and listen now to BOB

A stunning contribution from a valued cult member.

Mikael has contributed many valuable technical examples displaying the superiority of the Sony Exmor sensors over Canon sensors in terms of read noise (and thus DR) at base ISO.

By the way, as you know, I shoot Canon, and like my Canon equipment quite a bit.

But pack just 2mp on such a sensor, and you get Canon's revolutionary full-HD full-frame CMOS chip , which performs better in low-light than any other sensor (certainly better than those sensors with more (and smaller) pixels.

There is a very real rule-of-thumb here.

I'm disappointed that you would cite a single example as a "very real rule-of-thumb" when the preponderance of examples show the exact opposite.

However, let's discuss this exception to the rule. There are two primary sources of noise in a photo. The first is photon noise, which comes from the light itself -- more light, less noise (more specifically, 4x more light, half as much photon noise).

The second source of noise in a photo is read noise, which is the additional noise added by the sensor and the supporting hardware. The read noise becomes an issue when the signal (light) is very, very, very low.

Let's discuss just how low the light has to be before read noise becomes significant. At high ISOs, the 6D has read noise of about 2 electrons / pixel. With a QE (Quantum Efficiency -- the proportion of light falling on the sensor that is recorded) of 50%, this means that the read noise and photon noise are equal when 8 photons fall on a pixel. I think we can all agree that's pretty low light, indeed.

So, how would larger pixels help in low light noise? The only way this could happen is if larger pixels resulted in a higher QE (which they don't, since compacts and cell phones have QEs at least as good as DSLRs) or less read noise per area.

Now, it is by this last criteria that larger pixels may have an advantage. For example, it does appear as if all the latest sensors have about the same read noise / pixel. That is, regardless of the pixel size, at high ISOs, they seem to all lie at around 2-3 electrons / pixel, from compact to DSLR, despite very different pixel sizes.

Thus, if we measure the read noise over a given area of the photo (I like to use the µphoto -- millionth of a photo -- as the area), then, for example, we would compare not one pixel on the 6D to one pixel on the D800, but 16 pixels on the 6D to 36 pixels on the D800.  So, for a given read noise / pixel, then fewer pixels will result in less read noise / area.  Hence, the low noise capability of the 2 MP sensor you cite above could be nothing more than Canon making a sensor that has the same read noise / pixel as the 6D.

The question, of course, is why read noise would not be inversely proportional to the area of a pixel, which I am not qualified to answer (but Bob is).  I suspect that what is going on is that the read noise could scale, but the designers come up with a formula that works for a given tech and apply it across the board to all sensors of the same generation without scaling it as a function of the pixel size, for the reason that they don't want to take chances with a new tech.  Then, for the next generation of sensor, if they stay with the same tech, they do scale the design as they are now confident in the design from the previous generation, resulting in an improved efficiency.

However, this explanation strikes me as odd.  For example, a larger pixel should also have a higher saturation limit in proportion to its area, and, indeed, it does appear that's how things work out.  So why does this higher saturation limit of the larger pixel not result in greater read noise?  And given that it doesn't, why can't the smaller pixel have the same higher saturation limit as the larger pixel?

Well, like I said, I'm not qualified to answer these questions, but, hopefully, Bob can chime in.

The Lumia 1020 has not only a larger sensor than the F31, but also one that is significantly more modern, ie. has much lower read noise and higher quantum efficiency. Additionally the lens is faster. It has much higher performance potential in low light as well as other conditions than the F31 has.

Anyone fancy doing a F31fd vs 1020 ISO-1600 showdown?

Indeed, that would be great, but would only make sense with RAWs, since the in-camera jpg engine can do lots of things that will skew the results.  That said, the whole read noise issue discussed above will certainly put the Lumia at a disadvantage in low light unless its pixels have no more than 38% the read noise of the F31fd pixels (although, being newer tech, I should hope this requirement would be met).

Post (hide subjects) Posted by
(unknown member)
(unknown member)
(unknown member)
(unknown member)
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark MMy threads
Color scheme? Blue / Yellow