The 32mp R7 gives me room to crop for birds and skyscape, a bit of astrophotography and for sports when I cannot get close to the action. The buffer is better than the R10 but not by a big margin. I use faster SD cards and I find the R7 can keep firing and taking shots when the R10 would "pause" more often. Both produce great image quality so that part I'd say is a draw. AF of R7 is slightly better, which says a lot because the R10 AF is amazing.
14 ms, just below the R5 on the chart. To me, that’s definitely in the “good for normal shooting” range.do we know where R6II reads out?
Thanks, that is really good for $250014 ms, just below the R5 on the chart. To me, that’s definitely in the “good for normal shooting” range.do we know where R6II reads out?
Thanks for sharing this. In what ways are you finding the AF better on the R7 than the R10?
Using the EVF, I can see better in the R7 than on the R10, which enables me to pick focus points faster. The R7 can also lock in focus more effectively when light starts to be scarce. I find the R10 starts to struggle shortly after sunset or when blue light kicks in. The R7 can focus on less contrasty sceneries better too. This is using various lenses such as UWAs RF16mm, RF24mm up to EF100-400mm Mark II.
Ties in with Duade Paton's findings on AF - with the dark f/11 combination of the RF 100-400 + 1.4X TC he tested the R10 on. And Canon spec that the R7 is more sensitive AF to very low light.
This is definitely something I have experienced
This I haven’t noticed so far
This I haven’t noticed so far
Maybe it’s something with the lenses, I’m mostly using long focal lengths and typically relatively fast lenses.
Variability can be driven by lenses indeed. In terms of fast lenses, I've used the EF50mm f/1.2, the RF24mm f/1.8, the RF35mm f/1.8 aside from the telephoto I referenced above.
Given that the R10 and R7 share as far as I can tell the same generation of sensor technology (very similar readout speed) and the same Digic X it seems odd that there would be a difference in AF performance between the two cameras. If anything I’d suspect that the R10 could actually have an advantage. I happen to own both cameras and the original post here about AF performance being different really got me curious. Last night I did some comparisons between the two cameras with my EF 500mm f/4L IS USM and RF 100-400 and so far I haven’t been able to detect a difference. Not saying that there isn’t one, just I haven’t been able to figure out how to detect it yet but I am very curious.
Not really, when the R7 has a sensor with substantially faster readout. That means more AF calculations per second.
Nope, not when the denser sensor reads out faster.
That's another matter altogether. Maybe Canon's current FW can't leverage the difference. Maybe it's only measurable in edge conditions. Maybe optimal technique can turn the difference into a negligible one. It's exceedingly hard to say. But, with everything else constant, a faster sensor readout should deliver an improvement in AF effectiveness.
would suggest based on what published - and the rolling shutter issues on the R7 - not a substantial distance in read out a factor. Neither is good from rolling shutter perspective.
Where did you get this information, from my admittedly crude tests the readout speeds are the same. If there are more accurate/competent tests available I’d love to see.
I’m not so sure that’s actually true, doesn’t the whole sensor need to be read out?
I guess the question is, is the R7 sensor readout actually faster than the R10 sensor. Also, if you can’t actually detect the difference in actual use does it actually mean anything.
The only hard numbers I could find are from Duade Pattons youtube channel, although I don't know how those numbers were derived/measured. There is a comparison of the R10/R7/R5/R6 showing drone props, which does show the relative performance between those 4 cameras.
I can see a difference in the prop shots above, but I can't say one is noticeably better than the other. So I agree with your "If you can't see if, it doesn't matter"!
Thanks for thisThe only hard numbers I could find are from Duade Pattons youtube channel, although I don't know how those numbers were derived/measured. There is a comparison of the R10/R7/R5/R6 showing drone props, which does show the relative performance between those 4 cameras.
The shot of the drone props shows that the R7 and R10 are reasonably close.
I can see a difference in the prop shots above, but I can't say one is noticeably better than the other. So I agree with your "If you can't see if, it doesn't matter"!
IMHO a poor comparison (you have to count the "ghost" blades here). Better to show the bend in a VERTICAL blade.The only hard numbers I could find are from Duade Pattons youtube channel, although I don't know how those numbers were derived/measured. There is a comparison of the R10/R7/R5/R6 showing drone props, which does show the relative performance between those 4 cameras.
The shot of the drone props shows that the R7 and R10 are reasonably close.
I can see a difference in the prop shots above, but I can't say one is noticeably better than the other. So I agree with your "If you can't see if, it doesn't matter"!
It’s been tested by more than one person, and there’s a chart making the rounds. To be perfectly honest, I can’t be bothered to fish it out now
Sensor readout speed depends far more on the silicon substrate technology than it does on photosite count. Naturally, if you have two chips, one with 8MP and one with 24MP, made using the same process, the less dense one will have a faster readout. But this doesn’t scale linearly, and different processes can have vastly different readout speeds, enough to offset any losses from extra photosites. Like it happens with the Canon 24MP and 32MP chips.
It does. Canon sensors today use similar strategies to Sony chips, meaning that they drop down to a much lower bitrate when in “EVF mode”, which is also how they find focus. Ten milliseconds don’t sound like much, but that means a couple of extra frames for the processor to scan during AF tracking each second. Over time, that definitely adds up.
