Nikon, being far wiser, has stayed at 12 MP, and has instead focused on improving the quality of their pixels (dynamic range, colour rendition, ISO performance etc.).

As for canon’s 1080p video… well it looks like its actually only resolves to 720… so again canon chases marketing numbers above image quality… in effect they have taken 720 video and wraped it in a 1080p format… great for marketing…

If the D3s delivers true 720 video format aside from the marketing hype the D3s will prove to be a better camera then 1Dmk4.

http://reduser.net/forum/showthread.php?t=37975

Everyone in this thread is arguing over “brand new” versus “tweaked” as though these terms mean something. Exactly how small of a design change qualifies as “tweaked?”

ANY change to ANY mask gets a new part number and can rightfully be described as “brand new.” And if we count the whole package as the “sensor,” the same could even be said of changes made to the microlenses or color filters.

In other words, if you can’t plug a D3 sensor into a D3s and get identical results, the D3s sensor can be said to be “brand new.”

Now certainly raw megapixel count (or video res) is a factor for many buyers – especially those easily bought by advertising or a salesman’s pitch. Canon certainly has an advantage there.

But, the numbers don’t lie. This is the division of DSLR (i.e. not including point-and-shoot) market share in Japan for 2008:

http://hierophyte.blogspot.com/2009/01/canon-and-nikon-tops-market-share.html

A lot has changed since then – new cameras, most of another year has gone by – and of course that only shows sales in Japan. But, considering the importance of Japan in camera sales (and Thom Hogan’s musings on market share), the overall numbers worldwide are similar.

2009 doesn’t look to be nearly as good a year for any camera maker, and Nikon has probably lost a little market share as a result of the minimal changes to its lineup. The new lower end Nikon bodies appear to be a bit weak to me, but so do the Canon offerings. Both companies use older cameras in the entry price segment as the lowest cost option, although it may just be old stock that is sold until all the cameras are out of their warehouses. Both companies have a decent hierarchy of bodies, with every level competing fairly effectively with each other. I would say Canon has a slight advantage on paper (specs vs. price as the consumer is concerned), but the Nikon rebates going on right now probably bring things a bit closer.

Above entry-level, Nikon and Canon diverge a little with their target markets so they’re not nearly as comparable. Those don’t really matter nearly as much though, because as you can see from the individual camera market share division, almost 60% of total DSLR sales are just Nikon and Canon’s entry level bodies. Given a 39% market share each, that means that over 75% of DSLR sales from each company are at the entry level! Now, higher level bodies are still vitally important in terms of overall business strategy for both companies. People buy in to both systems with the express possibility of moving up to bigger and better cameras, even if they don’t think they will right away. The “halo” effect of having professional level bodies and lenses goes a long way towards increasing entry level bodies.

Well, I could go on for forever but I think I’ll leave it at that. Remember, Nikon is a dedicated lens and camera manufacturer rather than a large electronics company like Canon. Performance of their camera division is essential for their survival as a business, and they know it.

Gains in the visible light spectrum are much smaller. I have not done the math myself, but some folk claim that we are very close to the theoretical limit for ISO response. But then again, theoretical models tend to be based on assumptions that are not necessarily absolute constraints. Instead they identify the net part that needs a design around.

When I was in college the latest result showed that 64Kb ram chips were the absolute maximum possible due to induced soft alpha particle errors. That is not the case of course, we have much larger RAM chips today because they use techniques to compensate for alpha particle induced errors.

At the end of the day a DSLR lens is a heck of a lot larger than your eyeball and that has very little difficult seeing well under artificial light. So I think that there is a way to go yet.

But resolution and ISO are not the only things I think we should care about. I would much prefer an electronic shutter, and the ability to sync flash at any speed and to do pan shots in video to an extra ISO stop.

For even double-page spreads on magazines, 12 MP is way more than any photographer needs for anything but major cropping – which means a different lens or framing should have been used anyway. That’s the biggest you’ll see photos in publication, so for the majority of the target market (photojournalists and even photographers) it is enough. Low noise however, is critical to remaining competitive in photojournalism. A camera like the D3s actually allows photographers to take photos where they couldn’t before – an enormous advantage in a highly competitive field.

And for those who really, really need more, the D3X is there as rkas mentioned. Jobs that demand that kind of resolution are going to justify the cost of the camera – and ones that don’t, won’t.

Of course, there’s certainly all sorts of other changes as well. Micro-lens adjustments, filter optimization, more efficient amplifiers and ADCs, better noise processing, etc. are likely all done as well. That’s all off-sensor, of course. I suspect that the biggest gains are to be had at the CMOS sensor though.

And I think, you will be the very few rare breeds who complains bout the MP, many photographers are pretty darn happy with the ISO improvements and accepts that Nikon keep the MP count the same. Besides, if MP is so freaking important, there is the D3x.

So stop whining just cause you cant own a D3x.