Canon EOS R Review (updated 12/19)

[Kyle Jones]

I'm not ready to review the camera, but I figured I'd create a thread to start sharing information as I gather it. This first installment is based on some high-ISO night shooting on the Northern California coast. This was a somewhat demanding shoot with some light pollution and moisture in the air.

My first impression was that the camera served me well. I purchased the control-ring adapter and took my first cut at setting up the controls. I assigned aperture to the control ring, left shutter speed on the top dial, and assigned ISO to the touch bar. I used the Rokinon SP 14mm f/2.4 for all photos, manually focused on a star that was easily visible in the viewfinder. I'll still play around with the controls over time, but this worked fine.

Before diving into details, here is the final processed image from the evening. The sky was shot at ISO 6400 f/2.8 for 30s. The ground was 5 minutes at ISO 1600 and f/4. Both images were processed in Lightroom and then blended in Photoshop. I did apply some extra noise reduction in Photoshop, particularly in the sky. I just received a 30 inch print from my lab which came out really nice. So from that standpoint, the camera did its job.

So lets take a look at ISO... Below are 3 images taken with the same effective exposure at ISOs 1600, 6400 and 12800. Each image was processed in exactly the same manner using Lightroom only. At web sizes they all look pretty good!

Here are 100% crops at the center. To my eye ISO 1600 is pretty darn clean, ISO 6400 is acceptable for my work, and ISO 12800 is noisy but may be manageable with some careful noise reduction.

Here are 100% crops at the top of the image. The shorter shutter speed of the ISO 12800 shot does a lot for reducing movement of the stars.

And here are 100% crops in the bottom right corner. The noise is noticeably worse, which is to be expected with less light in this part of the image. The ISO 12800 image has taken on an ugly magenta cast as well.

There have been reports of banding when shadows are pushed from the EOS R, so here is an image that I pushed hard. This is the 120s ISO 1600 shot shown above. The exposure is +3 and the shadow and black sliders are both pushed to 100%. It does look like there is a magenta vertical stripe near the left edge at the bottom, but overall the image held up better than I expected.

My next installment will likely be a comparison with the 5DSR at ISO 100 on a sunset shoot. Thoughts?

 

[JimFox]

Interesting review Kyle. I should do something similar with the D850 I suppose.

I like how you laid this out, and I like the 100% crops, those are nice. The ISO 1600 is super clean indeed. I only wish you had a ISO 3200 sample. My main ISO's at night would be ISO 1600, 3200 and 6400 as a max. Typically it's around ISO 3200, so for me at least, that's the ISO I am interested in too.

When you process these in ACR do you add any Luminance Noise Reduction? I have been finding that although it can soften the edges a bit, that a little Luminance NR in ACR seems to produce a cleaner image then if I do it later, where then sharpening seems to bring out those artifact halo's around the stars.

Good work here, and I look forward to seeing more. It would be awesome if we could get a couple of others helping to do gear reviews too here, that would be a great addition to FocalWorld.

 

[Ben Egbert]

Great start on this review. 30 inch print? That's impressive.

 

[Kyle Jones]

Interesting review Kyle. I should do something similar with the D850 I suppose.

I like how you laid this out, and I like the 100% crops, those are nice. The ISO 1600 is super clean indeed. I only wish you had a ISO 3200 sample. My main ISO's at night would be ISO 1600, 3200 and 6400 as a max. Typically it's around ISO 3200, so for me at least, that's the ISO I am interested in too.

When you process these in ACR do you add any Luminance Noise Reduction? I have been finding that although it can soften the edges a bit, that a little Luminance NR in ACR seems to produce a cleaner image then if I do it later, where then sharpening seems to bring out those artifact halo's around the stars.

Good work here, and I look forward to seeing more. It would be awesome if we could get a couple of others helping to do gear reviews too here, that would be a great addition to FocalWorld.

I do use some luminance noise reduction by default in my night images. I'll have to look it up. I wanted to base my impressions on how I normally do things rather than some more scientific/theoretical but unrealistic method. I didn't even think to try 3200. I do most of my night shots at 6400 and I wanted to see how well that worked vs my 6D. I also wanted to see how it held up at 12800. I'll see what I can add.

 

[Bob Denton]

A nice practical review.

 

[Kyle Jones]

Here is a 100% crop comparison at ISO 6400 against the Canon 6D. These were taken on different nights in different lighting (I sold the 6D to help fund this), but all settings and processing are identical and the same lens was used for both. The EOS R has a higher resolution, but I didn't want to scale it down for this comparison because I want the resolution increase to help produce larger images than I could get with the 6D.

And for the record, all files have luminance noise reduction set at 25.

 

[JimFox]

That's interesting comparison.

As to the ISO, my base is really ISO 3200. Rarely do I ever get up to ISO 6400 unless it's a totally pitch black situation. Usually ISO 5000 is max, but as I said I try to stay in the ISO 3200 range. Perhaps I am shooting mine darker then you are?

 

[Kyle Jones]

I'd gotten comfortable with the 6D at 6400 and even made big prints of files at 8000. There is some noise but it was always manageable if I exposed correctly. My 5DSR, however, struggles at anything over 3200. It all depends on the camera. I generally shoot at various ISO, but I've almost always been happiest with my files at 6400. This time I didn't even think to go lower. The problem for me is that I can't expose longer because the stars will move too much. That just makes the ISO 3200 files darker.

 

[Mike Lewis]

Kyle,

As a current 5DSr shooter who is looking very anxiously towards a high rez R camera body versus the next 5DSr (assuming they do not merge into the same camera) whenever they both come out, this is a very timely post which I will follow with much anticipation. As I like to shoot nightscapes a lot myself (although I have none yet that approach the imapct of this one you are using here for comparisons) I am keenly interested in the ISo noise performance of the exsiting R. I find my 5DSr a bit painful to use for night shooting because of the noise, which quickly gets to unpleasant levels to my eye above 800 ISO. As much as I LOVE the high resolution for large printing that I get from the 5DSr, if the next generation high MP bodies from Canon do not close the gap on high ISO noise performance significantly, I may decide that the 30MP of the EOS R you are showing off here is not too much of a resolution step down in trade for the better ISO performance I see here. The 1600 ISO stuff you show here looks particularly clean to me.

I will look for the dynamic range comparisons at whatever point you have the time to post them - those are of equal interest to me. Thanks so much for putting this stuff up!

ML

 

[Kyle Jones]

I will look for the dynamic range comparisons at whatever point you have the time to post them - those are of equal interest to me. Thanks so much for putting this stuff up!
ML

Thanks a lot Mike - here's the next batch!

This update includes another ISO comparison (this time from ISOP 800 through ISO 6400) as requested by Jim. Here's the full image at ISO 100:

I shot this at various ISO as shown below. It was getting darker as I shot these so I did push the ISO 6400 shot a little bit to keep the brightness about the same. Gives a better view of the noise anyway. Here are 100% crops at the center of the image. I'm still fine with ISO 6400 for night shots, but looking at this I'd probably try to stay at ISO 3200 or lower for anything else. All perfectly usable as far as I'm concerned.

Next up are some comparisons to the 5DSR. In these cases I shot the same scene with the same lens and settings with each camera at ISO 100. The light did change a little bit between shots. I shot with auto white balance (I like having the camera pick something so the "as shot" can be one of my choices in processing the RAW) and for some reason the EOS R files are bluer than those from the 5DSR. This even shows up in the "as shot" WB they picked. The 5DSR used something like 5100k (I can't remember exactly) and the EOS R used something like 6800k. At those settings, the colors look about the same (those are what I used below).

Here's the EOS R version:

And here's the 5DSR version:

100% crops of these files for comparison. The zoom difference is due to the extra resolution of the 5DSR. I didn't want to scale because I am interested in how much benefit in detail vs. noise I get from the higher resolution.

And since people love pushing shadows in reviews, here's a 100% crop comparison on the right edge (the darkest areas) with a 3 stop push. Note that the original image already used a 100% shadow recovery.

I ran a similar series at sunset with a reflected view of Half Dome, again at ISO 100. I did see the same WB behavior, which is something I need to explore further this weekend. These files were processed differently attempting to get a similar look. This suits what I'm after with this testing: what is the real performance difference when using my normal workflow? EOS R version:

And the 5DSR version:

100% crop at the center:

100% crop at the right edge with a 3 stop push (5DSR on left, EOS R on right)

 

[Ben Egbert]

Kyle, what was the iso of this last set? The EOSR appears sharper than the 5DSR, is this because of the difference in resolution? I also see that the last image the EOSR appears overexposed.

 

[Kyle Jones]

Kyle, what was the iso of this last set? The EOSR appears sharper than the 5DSR, is this because of the difference in resolution? I also see that the last image the EOSR appears overexposed.

Those are ISO 100. I'd chalk the sharpness difference up to resolution. The last image was a 3-stop push of the files, so yes it is over exposed.

 

[JimFox]

Hey Kyle, thanks for including the ISO 3200 in your testing this time. You sure were busy in Yosemite besides just shooting fall colors. To take the time to shoot these individual shots does take some time.

That EOS R is looking pretty good to my eye. For me I think that once you get to 24mp or above, that the IQ is more important then the resolution. Not that the extra resolution isn't nice to have.

 

[Ben Egbert]

Those are ISO 100. I'd chalk the sharpness difference up to resolution. The last image was a 3-stop push of the files, so yes it is over exposed.

I meant the difference between the cameras, the EOSR seems to have exposed brighter to start with if all else is equal. I agree with Jim, it's looking good, but I am probably going to wait for the high res version.

 

[Kyle Jones]

I meant the difference between the cameras, the EOSR seems to have exposed brighter to start with if all else is equal. I agree with Jim, it's looking good, but I am probably going to wait for the high res version.

I'd agree with that. My goals for this were: (1) Replace my 6D for night shooting at a higher resolution and as my backup camera; (2) Smaller/lighter than the 5DSR when I want something lighter; (3) A chance to get into the mirrorless platform; and (4) Check out the DR advantage of the 5DIV sensor. For me it hits on those goals. The 5DSR remains my primary landscape camera. I am still looking forward to the release of the CPL adapter - that will be really handy with the T/S 17, Tamron 15-30, Rokinon 14 f/2.4 that don't take filters.

 

[Mike Lewis]

Thanks a lot Mike - here's the next batch!

This update includes another ISO comparison (this time from ISOP 800 through ISO 6400) as requested by Jim. Here's the full image at ISO 100:
View attachment 13689

I shot this at various ISO as shown below. It was getting darker as I shot these so I did push the ISO 6400 shot a little bit to keep the brightness about the same. Gives a better view of the noise anyway. Here are 100% crops at the center of the image. I'm still fine with ISO 6400 for night shots, but looking at this I'd probably try to stay at ISO 3200 or lower for anything else. All perfectly usable as far as I'm concerned.
View attachment 13692

Next up are some comparisons to the 5DSR. In these cases I shot the same scene with the same lens and settings with each camera at ISO 100. The light did change a little bit between shots. I shot with auto white balance (I like having the camera pick something so the "as shot" can be one of my choices in processing the RAW) and for some reason the EOS R files are bluer than those from the 5DSR. This even shows up in the "as shot" WB they picked. The 5DSR used something like 5100k (I can't remember exactly) and the EOS R used something like 6800k. At those settings, the colors look about the same (those are what I used below).

Here's the EOS R version:
View attachment 13694

And here's the 5DSR version:
View attachment 13691

100% crops of these files for comparison. The zoom difference is due to the extra resolution of the 5DSR. I didn't want to scale because I am interested in how much benefit in detail vs. noise I get from the higher resolution.
View attachment 13702

And since people love pushing shadows in reviews, here's a 100% crop comparison on the right edge (the darkest areas) with a 3 stop push. Note that the original image already used a 100% shadow recovery.
View attachment 13701

I ran a similar series at sunset with a reflected view of Half Dome, again at ISO 100. I did see the same WB behavior, which is something I need to explore further this weekend. These files were processed differently attempting to get a similar look. This suits what I'm after with this testing: what is the real performance difference when using my normal workflow? EOS R version:
View attachment 13700

And the 5DSR version:
View attachment 13697

100% crop at the center:
View attachment 13698

100% crop at the right edge with a 3 stop push (5DSR on left, EOS R on right)
View attachment 13696

Kyle,

Wow, this is great info for me in particular. It looks like the improvement in noise with the R could be more significant than the loss in resolution, which really has been my big question in thinking about the new camera versus my 5DSr. For daytime scenic shooting, I think the 5DSr still makes more sense, but once it starts to get darker, or for shooting nightscapes, the R would seem to have an advantage.

I really see myself crossing over into the R system from the EF system at some point, but for right now I am with Ben, in the sense that I am still inclined to wait and see what next year brings as far as either the next generation 5DS family camera, or hopefully, a similar high resolution updated next body in the new R series. While realizing that the actual weight reduction realized in transitioning to the new system is not dramatic, every little bit helps I think, and perhaps as the RF lenses start to fill out more there might be some more impressive weight gains to be had going forward.

I did however manage to get one of my local shooting buddies to move up to the R from his 6D - so far he is loving it. So it may be harder to hold out the more time I spend with him. Spending your friend's money on cool stuff is always a double edged sword

In any case, all this work to shoot, organize, and write up these informative results is surely appreciated; it is definitely information I could not normally acquire without buying the camera myself.

If you do get out for any other opportunities at some nightscapes, it would be interesting to see some comparison shots between the R and the 5DSr on the same scene under those conditions, pretty much like you did in comparing to the 6D.

ML

 

[Ben Egbert]

By the way Kyle, I really do appreciate your work on this report and thanks for being a pioneer.

 

[Amy Nelson]

Excellent review Kyle!

I just spent 3 days at Arches and Bryce taking photos with both the 5D Mark IV and the new EOS-R. I found that the EOS-R over all produced a clearer crisper image during the day light hours, but I relied on the 5D more during early morning and late evenings. But this is my first experience with a mirrorless camera, this may change over time.

I look forward to more input from you.

Amy

 

[Kyle Jones]

Excellent review Kyle!

I just spent 3 days at Arches and Bryce taking photos with both the 5D Mark IV and the new EOS-R. I found that the EOS-R over all produced a clearer crisper image during the day light hours, but I relied on the 5D more during early morning and late evenings. But this is my first experience with a mirrorless camera, this may change over time.

I look forward to more input from you.

Amy

Did you see any weirdness with the white balance on the EOS-R? Mine (at least on that day in Yosemite) was behaving as if it had a blue filter installed.

 

[Mike Lewis]

Did you see any weirdness with the white balance on the EOS-R? Mine (at least on that day in Yosemite) was behaving as if it had a blue filter installed.

Kyle,

May not be the same thing, but did read this thread over on the DPReview R forum recently, sounds at least maybe related to what you are describing. It sounded maybe like a software and not a camera problem if I recall, although I am not sure it was really categorically figured out yet...

https://www.dpreview.com/forums/thread/4336401

ML