Canon EOS R1 Specifications [CR2]

[GoldWing]

I was expecting a high resolution studio type camera. While 30 mpix on my R has been great, I expected more resolution from the R1.

Edit: I don't know where I thought I'd read the R1 is going to be 30 mpix. My bad. 45-60 sounds great.

Why is Canon scared to let the user decide? Make it 50 MP and let us make the file size, based on our needs.

 

[frankchn]

Why is Canon scared to let the user decide? Make it 50 MP and let us make the file size, based on our needs.

Bandwidth between sensor and the DIGIC chip and processing speed of the DIGIC processing chip likely are limiting factors.

120 fps @ 30 MP 14-bit/pixel ~= 50 Gbits/second.

You can slice that data-rate however you want -- 120 fps @ 30 MP or 60 fps @ 60 MP or 40 fps @ 90 MP, and so on. For a sports/photojournalism focused camera I can see them choosing a lower MP and higher fps.

Note that Thunderbolt 4 maxes out at 40 Gbits/second, so 120 fps @ 30 MP is already higher bandwidth than that. 120 fps @ 60 MP or so might not be realistic.

We all want the 100 MP @ 1000 fps camera, but it is about as unrealistic as asking for the 20-2000mm f/2.8 lens.

 

[koenkooi]

Bandwidth between sensor and the DIGIC chip and processing speed of the DIGIC processing chip likely are limiting factors.

120 fps @ 30 MP 14-bit/pixel ~= 50 Gbits/second.

You can slice that data-rate however you want -- 120 fps @ 30 MP or 60 fps @ 60 MP or 40 fps @ 90 MP, and so on. For a sports/photojournalism focused camera I can see them choosing a lower MP and higher fps.

Note that Thunderbolt 4 maxes out at 40 Gbits/second, so 120 fps @ 30 MP is already higher bandwidth than that. 120 fps @ 60 MP or so might not be realistic.

We all want the 100 MP @ 1000 fps camera, but it is about as unrealistic as asking for the 20-2000mm f/2.8 lens.

DDR4 at 2166MT/s gets 54Gb/s easily in real life (had to measure that at work last week) and cfe 4.0 type B support 32gbit/s before overhead, so buffer clearing should be quite fast as well.

I suspect the digic has the sensor DMA its data straight to RAM, so the sensor speed is likely the limiting factor. The R3 can already do 195fps with clever deferred processing, so I think 120fps isn’t improbable.

 

[neuroanatomist]

We all want the 100 MP @ 1000 fps camera, but it is about as unrealistic as asking for the 20-2000mm f/2.8 lens.

Then we have nothing to worry about, @HarryFilm already made the latter so no doubt he can provide the former as well.

 

[EricN]

DDR4 at 2166MT/s gets 54Gb/s easily in real life (had to measure that at work last week) and cfe 4.0 type B support 32gbit/s before overhead, so buffer clearing should be quite fast as well.

I suspect the digic has the sensor DMA its data straight to RAM, so the sensor speed is likely the limiting factor. The R3 can already do 195fps with clever deferred processing, so I think 120fps isn’t improbable.

How long can that be sustainable without increasing heat or dropping voltage?

 

[koenkooi]

How long can that be sustainable without increasing heat or dropping voltage?

RAM doesn’t really have power management, it’s off, idle or busy, so the throughput or fill state doesn’t matter much.
Apart from cost, this is a big factor for bodies with LP-E17 batteries having such tiny buffers: less ram, less power drain.

The sensor and digic will likely use more power with faster fps, but I bet you’ll run into buffering issues before heat/power drain becomes an issue.

I haven’t heard people complaining about pre-capture draining their batteries with sony/nikon/om, I would think that would be a huge power drain!