16MP sensor improvements?

Started Apr 24, 2013 | Questions thread
Najinsky Veteran Member • Posts: 5,739
Re: GF-3 / OM-D Shoot-out

grumpyolderman wrote:

Najinsky wrote:

Now it was OM-Ds turn to have the additional recovery process as well. Wavy lines mostly recovered but vertical lines in parasol now lost on the OM-D too. Wavy lines on the GF-3 shot now lost too, although when you know they are there you can get a hint of them in the grain.

So it seems pretty clear the OM-D keeps a clear lead up to F16. At F22 some of the lead has been erroded by diffraction but is still picking up more detail.

At least, that's how I read it, what's your take?

I think you are spot on, and thanks for this nice comparison, better to show than to argue endlessly...:-)

The arguing is a sport entirely independent of the photography so can happen with or without the images. Arguing what the images show, arguing how the testing was done, etc.

But at least an image help add a context. The original sunflower image, taken with a GF-3 and PL45 @ F22 is severely diffracted. This is predicted quite clearly from the DPR's MTF test.

But while the tests might lead one to think the results would therefore be unusable. At slideshow resolutions or even as an image for a good sized calendar, it would still work well with some interesting detail to explore:

Sunflower Bud

of course the PL45/2.8 is a pretty sharp lens, I guess with something like the 14-42 the results would not have been as clear cut, so I wonder when we will get the first 24mP m43 sensors to take full advantage of all these great new lenses coming our way?

Less clear cut, but there's still some usable detail down at that yellow 750lpph line and the extra resolution/efficiency of the 16MP sensors v the older 12MP are still better able to discern it. Even many kit lenses have fairly sharp centres, it's really away from the centre where they suffer most.

I'm not really chomping for a 24MP sensor just yet. I think a nice 18MP sensor, 14bit raws, but a really quantifiable boost in sensor efficiency (it's ability to make to most of the light it receives) would be make for an IQ leap as significant as from the 12MP technology to the OM-D.

It you take DXOs rating of the Olympus 75/1.8 tested on a GH2, it scores 11pMP (perceptual megapixels). This means 5pMP are being lost and this is primarily due to the sensor efficiency.

If we had a significant boost in sensor efficiency, and 14 bits to carry that extra data, (even if sticking with only 16MP) we would see substantial improvements in IQ.

A lens like the 75/1.8 might well then test at 14 or 15pMP, which is well into the mid-territory of some FF Nikons or Canons with premium glass (although the best FFs and glass are now scoring in the low twenties).

Edit: Thanks for commenting. It took a long time processing and preparing the results, nice to know at least one person looked!

-Najinsky

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