...
Сергей-4030> PS На самом деле, меня вот какие вопросы занимают. Сколько, интересно, Кэноновых менеджеров уже повесилось в промежутке от D800 до сего дня, а сколько еше успеют повеситься до 13 сентября перед тем, как выживших сожрут с говном акционеры?
If you can't beat them, join them:
Правда, если они пихнули 40+МП сенсор в тушку с интегрированной батарейной ручкой, то и стоимость будет под 1Д (раза в полтора-два больше, чем у Д800)...
А вообще, поглядим...
Тут Хоган начал оправдываться, почему он таких как я поклонников обманул  [показать]Sensor Economics
August 27 updated (commentary)--Back when the D3 came out I made a number of comments about sensor costs, some of which have been embellished over the years into near Internet myth.
Basically, I wrote that I thought that a DX sensor was probably about US$50 in actual cost at the time, and an FX sensor, as far as I could tell, was about 10x more, or US$500. The part that got picked up more and embellished into mythdom, however, was my point that the rule of thumb in manufacturing is that one dollar of parts cost often ends up being about three dollars of consumer cost. In other words, a US$500 sensor cost implied US$1500 worth of price to the consumer. In other words, don't expect a US$1500 full frame camera (because all the other costs, when also multiplied by three would quickly push the consumer price higher).
Of course, Sony eventually had a US$2000 full frame camera (the now discontinued A850) and we now have rumors of Nikon (and not long after, Canon) getting ready to introduce entry-level full frame cameras at prices far lower than the current models, perhaps even below Sony's ending A850 price. So did something change?
Several things have changed. But I doubt that the DX/FX cost ratio has changed quite as much as some think it has, or that large sensor costs have reduced as much as people think.
One of the things that changed is the use of larger wafer sizes. Another is improvements in actually laying down the silicon. Improvements garnered by making smaller sensors with extremely small line size have had an impact, too, as a 24mp FX sensor doesn't have nearly as small features as state-of-the-art 24mp DX sensors; making changes to improve yield with smaller features should have implications on sensors with larger features, too.
A more interesting thing to contemplate is "what else is being made on the wafer?" Here's a challenge for you: take a 12" round area and in it place as many 24x36mm rectangles as you can. Notice something? Yep, you've got all these areas around the edges that aren't being used. Could you fit in some small chip designs in that space? You bet. It'd be nice if they were other smaller imaging sensors using a similar process. Hmm. Nikon is now creating its own sensors at CX, DX, and FX sizes. Have they figured out a yield benefit to mixing those intelligently? I'm going to alter this paragraph after talking to some sensor experts. One thing that some reading the original didn't understand was that most FX sensors require multiple lithography steps per sensor (DX sensors require a single step). The D3 sensor appears to require four, for example. It's actually this lithography step action I was trying to speculate about, though my original wording got in the way. Let's just reduce this to a simpler thing: can you produce what's on the wafer more efficiently?
I don't know the answer to that question, though after a few new conversations with those in the semiconductor business the answer is probably yes. Which leads me to believe that you could produce FX sensors less expensively now than you could five years ago. Not amazingly so, but a step forward in cost structures that would change product costs.
So let's back into the cost discussion a different way.
The D600 is rumored to be made in Thailand, the plant where all the DX cameras are made. It's rumored to basically use D7000 parts content where possible. Getting the sense that the overall product margins might be similar? Let's run the D7000 at introduction versus a US$2000 D600 at intro and see what happens:
D7000. US$1200 list. US$960 to Nikon. US$580 cost at 40% margin.
D600. US$2000 list. US$1600 to Nikon. US$960 cost at 40% margin.
A US$360 difference in cost mostly associated with the sensor. Make the DX sensor 1/10th the price (the factor I reported in 2007) and we have US$400 for an FX sensor, or about 80% of the US$500 price I guessed five years ago. Simply put: we don't need a drastic drop in sensor price to get a US$2000 FX body.
My guess is that Nikon thought that the hyped up market energy that would come from introducing a remarkable leading edge 36mp camera could generate really strong demand for something lower priced that maybe didn't quite go as far (e.g., the rumored 24mp D600). Introduce the D800 first, grab the "gotta have the best" customers first and get them to talk about how great it was, and then follow up with a lower priced option for those on more of a budget.
Which makes the next story all the more remarkable.
А, вот уже слух об ответе на ожидаемый по слухам Д600
: