New Rumor: The R1 will be everything you dreamed of and more

The rumor came out 4 weeks early

I really don't know where to start with this rumor, but since it came out somewhere else,I just have to pitch in with my thoughts here.  To be fair the original post of this does state to take it with a grain of salt, however, I suggest, forget the grain, just fill up the wheelbarrow.  Anyways onto the specifications;

• Approx: 85MP global shutter CMOS imaging sensor
• Approx: 85MP at 20fps, 21MP at 40fps unlimited continued shooting
• Full Sensor Large Quad Pixel Auto Focus
• 15.5 EV+ Dynamic Range
• ISO 160-1638400
• 5-axis IBIS up to 9 stops
• 3.5” 9.33MP 1280nit 10-bit WCG RGB-OLED touch screen with 120Hz AE-AF
• 9.44MP 120Hz SuperSpeed AE-AF electronic viewfinder
• Price $8,500USD

Is it possible?

Well, some things may be possible.  No really. 

Let's for starters assume this is running Dual DIGIC X. Believe it or not 85MP at 20 fps is indeed possible.  That would be around 1.7Giga-pixels per second, which is under the theoretical maximum of 1.8GigaPixels per second with 2 DIGIC X processors. So this could in theory happen especially if Canon drops the bit rate down to 12 bits. 

The 40 fps would be lens limited unless it's AE fixed - because apertures would start flying apart if AF/AE happened in between each exposure.

Canon never really mentions dynamic range so clearly, so I find it odd that this would specifically state 15.5EV dynamic range.  I would have expected "stacked" to have been mentioned because i don't think it's possible for Canon to do a global shutter sensor non-stacked and deliver stills dynamic range that high.  I also find it curious that Canon was able to do 15.5EV on an extremely fast sensor. If it's stacked it may be possible though.  Also, this is around 1EV better than the R5, so while it's twice as fast, has global shutter, Canon even managed to eke out an extra 1EV dynamic range.  I'm not buying the dynamic range specification at all. It seems implausible that Canon would manage all that in one sensor.

The touch screen? I'm not believing this specification.  There would there be little reason for it to be 10 bit.  1280 nit display would be difficult without going to HDR display, as even smartphones have struggled with that.  Samsung's S21 Ultra hits slightly over 1000 nits in normal mode and exceeds HDR at 1500 nits but only for HDR video playback and can't occur for the entire screen display. No other smartphone manufacturer (yet) exceeds 1000 nits.

The displays also have little to do with AE/AF which is mentioned in these points, and Canon has never once mentioned how fast they calculate out AE/AF.  For this to reach 120hz though would require nearly 600GB/sec data transfer from the sensor (assuming AF reads are done at 12 bit ADC bit depth and each side computed separately).  That's around 10 times the speed of PCIe version 4.  Can lenses even AF that quickly? I'm not sure.  Linear USM may be able to keep up, but it's hard to say if these would be exclusive to specific lenses.  9.44M dot (not MP, that's certainly wrong) EVF does indeed make sense since Sony already has that out on the A1 and A7S III.

All in all, it sounds like someone took the A1 specifications and then went crazy. I mean really crazy.

Then again, the R5 caught me by surprise and I didn't even think it was possible - so while it looks crazy - maybe Canon has pulled out all the stops, or even went with another company for the sensor. 

What do I really think of it? look at the subtitle ;)

• CanonWatch