The pixel war

Such sensors are already out there, and being used in cameras like the canon S95. Pentax just don't use them (except for the Q).

It's a future aspiration I think... it will take time before Nikon start rolling their own sensors.

nikon use the own sensors in ther full frame cams d700 d3 d3s d3x.

http://nikonrumors.com/2011/01/26/the-sensors-of-the-d3-d3s-d700-and-d3100-are-m...

Okay.

High ISO and dynamic range are much more important than pixel count. Which is why most camera makers have been making better sensors rather than cramming in more pixels.

And yes, 10 MP is more than enough!

I agree on all counts.

I deliberately chose a 10MP version for my Sony compact, instead of the 16mp versions.

I wish the K-5 was 10MP. I think it would have even better IQ

So all round what would you say would be the ideal number of pixels to produce the best quality photos in a digital slr. A freind of mine who does weddings and studio portraits once told me if his fuji S5 pro was a little more advanced he would not purchase another camera I think I am correct in saying the same sensor was used in the Nikon D 200.
Ian

Assuming you're going to occasionally want to print A3-A2, I think 10mp is about the minimum for best results. Anything over that will be fine, the pixel count is actually not that relevant. For example the K5 manages to have better IQ than previous 10, 12, and 14mp models, despite having extra megapixels. The number of megapixels isn't going to reduce IQ per se, because the increase in pixel count comes hand in hand with other technical advancements.

Do you think the change of sensor was a step back over ?. Im sure I have read before of members prefering the old style sensor.
Ian

Dont do much printing to be honest however I can not see me needing any more than an A3.
Ian

While the body may be 95 percent Nikon, the sensor and image processing are 100 percent Fujifilm. The S5 Pro is equipped with the latest Super CCD SR Pro sensor, purpose built for this camera. It is an APS-C format sensor measuring 23.0mmx 15.5mm, slightly smaller than the DX sensor in the D200. Fuji’s SR sensors are different from conventional CCDs in several important respects. First, the main photocells are arranged not in a chequerboard pattern, but in a diagonal grid. Second, the main photocells are larger than the cells in conventional sensors. Third and most importantly, they have two photocells at each site, the larger ‘S’ cell tuned to capture the main image information, as well as the smaller ‘R’ cell tuned to capture brighter highlights. The theory is that this gives SR sensors greater dynamic range than conventional CCDs, so they can capture a wider range of shadow and highlight detail.

Which is the same sensor used in the K-5 I believe.

I've got into trouble for saying this before but the detail produced by my *istD (6Mp) was surprising. My K-7 (14.6Mp) needs more sharpening than the *istD to extract the detail.

The K7 resolves far more detail than the istD. I'm guessing your impression comes from looking at 100% crops of completely different shots, where the istD might look 'better' but you do not have to 'zoom in' to the shot to get to 100%, so it's a false impression. Anyway a picture tells a thousand words... guess which one is the istD and which is the K7. Just for fun I included the K5 as well. Each shot is a 100% crop of dpreview's sample picture.

From my limited understanding there's another way to get some indication of why the megapixel race is suspect. That is to use pixel density - how many pixels per square mm for eg - (as done in past reviews on another forum) as an indicator of why the race can't go on forever.

Pixel density is also a good way to better understand the limitations of compact camera sensors, and they were reached some time ago. APSC-sized sensors will soon come up against the same limitations, if that is seen in terms of comparable pixel densities.

Once that has been reached, the only way out is with FF - by providing a larger sensor area over which to spread the pixels (and thus reducing the pixel density that we started with here).

So the bottom line is that in a given area, if you wish to squash more pixels in, the pixels must be reduced in size (and so must the microlenses that service each pixel site). That approach has physical and practical limitations which will impact on camera performance & IQ - and there's the problem.