Just a couple of questions about the new R6 II

koenkooi wrote:

Ephemeris wrote:

Alastair Norcross wrote:

Justme wrote:

Alastair Norcross wrote:

Seems like a pretty good one camera solution. The R6II in crop mode will lose the high ISO noise advantage, of course. I think the R7 performs a bit better than the R6II in crop mode, and has almost four times the pixels, but if you do a lot of shooting that doesn’t require reach as well as the bird shots that you like, the R6II looks to be a good compromise.

Alastair, you just threw a monkey wrench into my plans. Why would the R6 MK2 in crop mode lose the high ISO advantage? Probably because I don't know how crop mode works on the R6 MK2.

Have a look at this chart, from photonstophotons (the guy who runs the site posts quite a lot here). It gives you readings both for the full R6II sensor, and the sensor in crop mode, compared with the R7. It's not exactly the same as high ISO noise performance, but it's a pretty good proxy. As you can see, the R6II in crop mode is pretty much the same as the R7, but achieves that through built-in noise reduction:

https://www.photonstophotos.net/Charts/PDR.htm#Canon%20EOS%20R6%20Mark%20II,Canon%20EOS%20R6%20Mark%20II(APS-C),Canon%20EOS%20R7

As I said, if you want one camera to do everything, and focal length limited applications (like birding) are only a part of what you shoot, the R6II looks like a good all-rounder. But for the specifically birding part of your shooting, the R7 will be significantly better (because of the much higher pixel density). Shooting the R6II in crop mode will be like using a cross between the 30D and 40D, but with much better AF, and slightly better high ISO noise performance.

Bills site is super helpful and interesting.

I wonder if the question, and for Justme please forgive me if this isn't the case, but I wonder if you were asking if we have a camera (day the R6Ii) with some high ISO performance, why that may be reduced if we crop? It's the way I read your query and your response but I may have misunderstood.

The simplified explanation that made sense to me is the following:

Noise performance is done by printing (or pretending to print) the image at a standard size (say 8x10) and looking at it at a fixed distance (say arms length).

Since noise has a physical size on the image/sensor, cropping the image will enlarge the size of that noise, since that cropped portion of the full image will get blown up to 8x10.

This is why looking at an image at 100% magnification for noise isn't indicative for "image quality", since you're looking at pixel level noise, not image level noise. I still find it useful to look at 100%, but I'm an engineer, not an artist

Sounds helpful.  I suppose another way is that if we wanted to look more closely at the image to see more detail then we may magnify and find we don't have more detail. Noise, or some unwanted information often looks unpleasant (not entirely sure why).