Nikon D5100 vs. D5200

Started Jun 13, 2013 | Discussions thread
Herve5 Regular Member • Posts: 127
Re: Nikon D5100 vs. D5200

KennyXL wrote:

On this issue, is it correct to consider halving the D5200 resolution would improve its signal-to-noise ratio, ending with a 12 Mp image that would improve shadows "somehow" like a D5100 16 Mp?

All else being equal, if you have larger pixels on the same sensor surface area, then I imagine noise would be better handled. However, if you just take a D5200 sensor, but only use half the sensor (i.e. the other half is just sitting there idle) then you just get a 12 Mp image with the same noise performance as a regular D5200 sensor, but with half the image size.

I definitely don't think that's the way it works.

I believe the full sensor is used, and then the pixels are agregated almost on the fly.

This 'averaging' of the noise between adjacent pixels is the very reason noise should decrease IMHO. For instance, in conventional systems halving the resolution results in averaging 4 pixels, which means dividing noise by two. Indeed, with the camera in hand this should be almost testable with a proper image analysis software.

My question was rather to find if someone has more detailed insight about the way the pixels aggregation happen...

_sem_ wrote:

In principle yes, it is possible trade resolution for noise performance by proper downsampling, and the D5200 isn't bad in SNR scores normalized to a certain MP size at all. But here the problem is pattern noise which makes lifted images look worse than what the SNR score says. This may be partly recovered by banding.

Thanks for this reply!

I have difficulties understanding this issue of pattern noise. Could anyone point me to some references on this point?



 Herve5's gear list:Herve5's gear list
Leica Q Lemke Software GraphicConverter +5 more
Post (hide subjects) Posted by
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark MMy threads
Color scheme? Blue / Yellow