Why can't Canon make a lower end EOS-1 body, 36-MP, 12 FPS, sports camera for $4-5,000, and then make another 40-50MP EOS-1 Body with 4-5 FPS for the higher end crowd for $8-9,000?
Nikon clearly has shown they can do it and made two very well priced FF cameras. (D800 and -E models) The technology is there and both would sell well.
If they wanted to save on costs, leave the video out.
First off, Nikon most definitely has NOT produced a 36mp camera capable of 12fps readout!! The D800 is 4FPS, period, when reading the full FF sensor! It is only capable of 5fps in 1.2x crop mode, and only capable of 6fps in 1.2x crop mode when an additional battery grip is attached. FOUR FRAMES PER SECOND is all Nikon could do with a 36mp sensor. Not 12fps. You are off by a factor of three.
You want a 36mp sensor that can be read out at 12fps? You are effectively asking for DOUBLE the processing power of the 1D X. At 18mp (which is really over 19mp in actual pixel count, as Canon masks off the border of their sensors for calibration purposes, but those pixels ARE read for each frame), the 1D X dual DIGIC 5+ processors pump about 250mb/s each, or 500mb/s total. To process 36mp, you would need either four DIGIC 5+ chips, or something along the lines of a DIGIC 6+ capable of handling about 500mb/s per chip, for a total processing power around one gigabyte per second. That is a hell of a lot of data processing. There is also not a CF card on the planet that could keep up, either. You would either have to pack in an absolutely monstrous frame buffer capable of capturing 100 frames before pausing, or simply tell your customers to suck it up and deal with a frame buffer only 15 deep and excessive lag time when it gets full as your huge frames are slowly written out.
High MP and high frame rate don't really go well together. For one, big pixels actually benefit IQ for high speed action photography, either on a tripod or hand held. You don't experience as much softening due to camera shake as you do with small pixels. Pixels 1/4 the area of the 1D X would produce less viable pixel-level IQ, so the benefit of having all those extra pixels ends up not really being worth the $20,000 or more you would have to pay for a 36mp camera that was basically capable of all the things the 1D X does now, with twice the resolving power.
True, Nikon does only have the 4FPS with the D800, but they do have the MP. Point made.
Conversely, is that truely "ALL Nikon could do"? I'm also inclined to think there are some designs that are incremental improvements - a "75% fix" so they can release another model in less time.
To simplify:
If they release a fantastic, "100% bells and whistles" capable camera, the market might not want the next version for 4-5 years. If they offered the 75% version, then the market might well desire the next version in 2-3 years.
Actually, I believe Nikon's problem is their frame buffer and how they handle writeout. Canon systems don't "lock up" when the frame buffer is full. You can continue writing new images to it as space is freed up when older images complete writing to the CF card. Your frame rate drops to a few per second, rather than 10-12 per second, but you can pretty much keep shooting indefinitely (or until your CF card fills up). The D800/E locks up when the buffer is full, and it seems you have to wait for the buffer to clear before you can take more shots. That, more than anything, is a real FPS killer, regardless of how fast the base rate is.
I would actually offer that, even at 4fps, Canon could to a hell of a lot better with their approach to handling the frame buffer.
I also posit that Canon CAN get close to a large MP and faster FPS camera by the time they get it to market in 2014.
1) The EOS-1 body style is established.
2) DIGIC 6 processors are going into cameras this year. By Mid-2014, they should be common place.
3) They should be able to have the power for 6-7 FPS and I am fine with a 15-shot buffer. Anything faster is almost video. (I shoot in bursts of three regardless. I try not to "spray and pray".)
4) Dual CF cards would be appreciated.
5) 5-6K on the price.
The 1D X currently has a 38 frame buffer and supports 12fps. (Actually, it supports 14fps, which is why I believe the DIGIC5+ supports an input rate of 250mb/s, and why the 1D X has two of them.)
If we run the numbers, today, with a pair of DIGIC 5+, Canon could make a 36mp camera with an 18 frame buffer and 6fps. Since the DICIG 5 is an established DSP, they wouldn't really need to invest additional research creating a faster one. Dual CF cards is a no brainer.
I figure Canon will probably go for ~40mp body at least. Even with a 40mp sensor, I think Canon could still get 6fps out of it with dual DIGIC 5+ (just barely.) If Canon releases a 50mp+ body, then I think the frame rate would have to drop to 4fps with a pair of DIGIC 5+. I don't foresee Canon making the same kinds of gains with DIGIC 6 over DIGIC 5 as they did with DIGIC 5 over DIGIC 4...they had over five years before to improve things. It's only been a year at this point. I think the gains would be marginal, although they might be able to eek out 5fps @ 50mp.
That said, if Canon really does go for the higher end of the range with a 50mp+ body, I think a slower frame rate would benefit I more. High frame rate dictates high frequency components, which introduces more read noise. I think it would be more likely to see 3-4fps in a 40-50mp body, as it would benefit IQ...potential lower read noise, probably better dynamic range, etc. I find high mp to be at odds with frame rate. While I can see and understand the desire of action shooters like bird and wildlife photographers wanting it, as smaller pixels means greater reach...at the same time this camera will be a powerhouse studio and landscape body. Both of those endeavors, particularly landscape photography, demand top notch IQ, of which lower ISO dynamic range is a critical factor.
