Best High ISO? Canon R5ii vs R3 vs R5 in-depth Test

[@darinl848]

with most thigs, there are trade-offs. stacked sensors are known to have faster read-out speeds and less rolling shutter and the trade-off is they have more noise and less dynamic range. some people are saying the Mk2 has better high ISO performance for some reason.

[@Makta972]

The R3 is so underrated. Best Canon Camera imo

[@the_camera_life]

Part of me was hoping the R5ii would be my new main camera but after this test I dont think the R3 is going anywhere

[@RayValdezPhotography]

Cant wait to see what the R1 can do. I already heard that in video it cleans up at 12800 like the a7s iii fx3

[@malcolmbullock4297]

@@the_camera_life So I have that dilemma. I have one R3 which I bought for sports - no issues. However, I'm starting professional weddings and portraits so I am/was thinking of an R5 ii as my second body. no video stills only. Reading the reviews I'm not sure what the mkii will give me over another used R3 apart from the larger image size/MP's which I don't need as I can frame shots and shouldn't need a massive crop. Any thoughts? Thanks VERY much for your video. Hugely helpful.

[@davidpruitt3255]

@@malcolmbullock4297 I would recommend renting to try out the cameras and see what works for you! You may find yourself better served with an original R5 if you like the increased megapixels or an R6 Mark II if you just want another camera with similar autofocus to the R3.

[@hrw1973]

thanks for the vid... saving me money for the moment... staying with my 1dxIII

[@the_camera_life]

The 1dxiii is still an epic camera! Are you thinking of checking out the R1 when it comes out?

[@hrw1973]

@@the_camera_life R1 is on my x-mas list.. yes

[@AdamBotond]

Great test! What you concluded is one of the reasons why I will go with the R3 instead of R5II. The other is the R3's 4K60 quality, which is oversampled. Thanks for sharing!

[@the_camera_life]

Awesome! I do still love the R3, that oversampled 4k is great out of that camera.

[@MilmidStudios]

Proud R3 owner!!!❤🎉

[@howardholtzman1331]

Excellent tests. I'm going to wait for the R1 test before I buy.

[@the_camera_life]

Thanks so much! I'm super interested to try the R1, I'm hoping my finances allow for it when it is available. Are you a sports shooter?

[@nip.pon172]

Would it also be possible to compare the ISO with the Canon R6 Mark II compared to the Canon R5 Mark II??

[@ThePhotographerGuy]

Great comparison. Love the real-world shooting.

[@andycollins5983]

It would be interesting to include the R6 MkII in this test.

[@the_camera_life]

For sure! I might redo it soon and include the R6 mark ii

[@mikede2464]

Wow...surprised the R3 did so well against the downsampled R5ii. I hope the R1 I preordered does as well as the R3 did in this video. Thanks for putting it together.

[@the_camera_life]

Yeah I was too, that’s one of the reasons I wanted to leave the whole process in the video. I really thought that down sampling would make things closer to even. I might do another test where I process with basic noise reduction then down sample to see if the extra resolution allows for more effective noise reduction but based on this test the lower MP sensors still have a big advantage. VERY keen to try the R1 when it’s available!!!

[@evilZardoz]

About to preorder the R1. Was going to go for the R5 II, but I need better high ISO performance, and right now, my 1DX II is giving me better results, and the R3 I borrowed from Canon for a weekend ran circles around the R5 for the ISO 12800+ use cases shooting events at dimly lit bars. I'm worried that it has come to "I hope the new camera that costs more than three year old camera, is just as good as the three year old camera". I would be overjoyed if it did better. If I was seeing a third to half a stop better low light performance than the R3, I'd be quite happy with forking over RRP for the R1 - especially if the colour improvements are the same as what we see in the R5 II.

[@mikede2464]

@@evilZardoz I personally don't think it will be much better than the R3. These days seems like-at best-incremental sensor performance in terms of low light/DNR/Noise. I'm buying into a different system (had a terrible experience with the Sony A1) and figure if I'm going to spend that much money on the R3 I might as well get the R1. If I already had the R3, from what I hear about it....don't think I would upgrade to the R1.

[@evilZardoz]

​@@mikede2464 I, too, had an atrocious experience with Sony, product quality and customer service with them refusing to honour warranty on a failed 35GM. It's my first experience with a non-Canon system, so I'm firmly sticking with Canon moving forward. Lesson learned! The R3 can be had secondhand for around half the new price of an R1, so that's very tempting. As long as the R1 isn't worse than the R3 in terms of low light performance, and if it has some of the colour rendition improvements of the R5 II, then I'll go for it. I need those cross-type AF sensors; the R5's AF system is hot garbage in low light and I can't wait to rid myself of it.

[@evilZardoz]

Awesome! I've got the R5 and I've been disappointed with the high ISO performance for events (even when delivering low res exports from LR) vs. my 1DX II.... Had an R3 on loan for a weekend and shot an event with it and was blown away with how useful ISO 12800 and even 25600 was again. Was hoping for better results from the R5 II - I do notice that the high ISO images appear to retain more detail, and - the way they expose blue LEDs as shown from your images suggests they tend to be less purple than the R5 - a huge problem I've had in the past has been when shooting live events with blue LED lights that tend to expose as purple. I felt my photo skills went down the toilet ever since I moved to the R5... I've been on the fence as to whether I pick up the R5 II (gripped is around 7200) vs an R1 at 10.5k - I'm already most of the way there, so I'm tempted to just go for the R1 at this point. After my recent experience with the R3, I would not be disappointed by moving to the R3 from the R5. My keeper rate and results for that event were back to where I was expecting them to be, and the camera was far more enjoyable to use. Still haven't heard much about the R1 vs R3 in terms of high ISO, but if there is any improvement at all on the R3, then I'm sold. A secondhand R3 looks like a great deal right now for those low light situations such as bars and clubs. I just wish I had better blue colours vs. the R5 II (or the Sony cameras, which do blue very well). Thanks for putting this together!

[@Jla-z9x]

I had a 1dx2 also and you may be interested to know just an original R6 is a stop cleaner at high iso for low light events so you may not need to invest much if that's priority. I got useable iso 51200 results at an event recently it was crazy

[@mareklesniak8768]

Thanks for comparison. Any chance to get those RAW files?

[@TheBigBlueMarble]

As post-processing software becomes more and more powerful, and AI gets better and better at removing noise, sensor noise becomes less of a concern. An interesting offshoot of this is that more and more photographers are moving to shooting manual with auto ISO.

[@TheBigBlueMarble]

Would it be a reasonable comparison to shoot these three cameras in absolute darkness and then compare the noise?

[@JGZphotography]

Next time, please conduct the test without any Lightroom (LR) or Adobe Camera Raw (ACR) adjustments, and shoot in RAW format. This will allow us to make a true comparison. I have both the R5 Mark II and R3 cameras, and at 12800 ISO, I would estimate that the R3 performs about 1-stop better in terms of noise. However, when using a noise reduction application like Topaz Photo AI, the noise results are much closer together, and the detail is better on the R5 Mark II.

[@WesternSydFooty]

r5 looks cleaner than R5II IMO

[@dscottstoness2436]

How do you know that it's not that R3 LR converter is better than R5's ? I have done lots of testing and found R5 better than R5ii at 25k iso. But after applying denoise they are very close. Maybe the converter for R3 adds some denoise? I am also wondering whether the LR raw converter is tuned yet.

[@the_camera_life]

Very good point! It’s possible things like that might be refined by adobe, I only tested in Lightroom as that’s what I use for processing Raws. Interesting that you also found the R5 a little better than the ii and you are right it’s very small and processing can make the difference tiny. Do you use the regular Denoise tools or the new AI denoise when processing 25k?

[@RayValdezPhotography]

r5 ii doesnt even have a real profile yet. the lens correction is generic when you pic a lens on the last tab

[@artofwomenphotography]

just wondering when the limited strap will be available to those who preorder the R5 MK II. Still waiting to get mine

[@rokg12]

did you use mechanical shutter on all or electronic shutter?

[@Jla-z9x]

Why does Norhhrup always say all that matters is total light and if 2 sensors gather the same total light they'll have the same iso performance regardless of megapixels when the higher resolution sensor is scaled down to match the size of the lower resolution one. You clearly proved that's bs here. Resolution does affect

[@johnykw]

I suggest don’t move very fast action when compare the photos, maybe freeze the screen around 2 second still shot compare the photos with 3 camera

[@the_camera_life]

Thanks so much for the feedback, this is the first video I have made like this. Next time I will take more time to setup the comparisons then be sure to hold them still for a longer time.

[@nevvanclarke9225]

It is always a compromise for the newer sensors that they are bringing out that with faster speeds. You are going to get some trade-off and that trade-off is the noise ratio which I'm not a scientist mathematician kind of guy but I do understand it but I don't think it's that bad to be honest and a lot of noise these days he can clean up with either Program separately that you can run or Lightroom denoise.... Other thing I do is just run a gradient over the areas that are dark noise tends to show up in dark areas more than it does light and coloured areas but I don't think it's that bad. there's also other gradients that I run as well like in Lightroom I reduce grain and reduce texture and clarity in certain areas that are just darker that don't have anything in the photo like a background area and there are so many tools we can use now that do take a little bit of time but you can get images that have been shot at 10,000 ISO to be quite usable if you prepared to be patient I remember shooting a funeral which was a really challenging thing to do because of the lowlight. It's funny isn't it? Because I think editing used to be seen as a bit of a bad word now it's out thing that can rescue it. Our images from noise and the noise reduction tools are amazing.. thanks for your video even though I'm not in the Canon ecosystem I still enjoyed it. The other thing that I always remember about lowlight images is to get our exposure right that's critical. I know that there's a tendency for some people and I know you didn't do that in this test but they try to under expose in certain situations and then try to rescue it impost and that's when it becomes a problem. Correct exposure will always help in this instance.. I remember watching the review for the Nikon cameras that were released recently and they had similar levels of noise as well.

[@cristirenault]

Iso 4000 is recomended for R5II...

[@the_camera_life]

Interesting, I’ll do some more testing at 4000 specifically and see what I can find. Any chance you have a link to that info so I can do some research?

[@MilmidStudios]

The R5 looked better than the R5mkii at 3200

[@rüzgarınadı]

thanks I need this video

[@the_camera_life]

So glad it helped! Sorry it was so long

[@etiennededurocher]

breathe my friend, breathe

[@RayValdezPhotography]

Lightroom is dumb you cant save the zoom for individual files. SO its annoying when you compare different MP files

[@prosunsport1]

Wow the R5 is far better sharper and less noise. 4 YEARS LATER AND Canon moved sideways

[@thelensmanphotography]

They stepped forward many steps with all of the capabilities of the MkII!

[@RayValdezPhotography]

seems like they focused on convenience and video. The video is way better. the original r5 was disgusting in terms of noise and DR in log. The only penalty is more noise in photo.

[@user-wl4df3yq5r]

This should of been a 10 min video max. Lots of unnecessary talking, quiet annoying actually. Skipped a bunch.

[@the_camera_life]

Thanks for the feedback and I completely understand. I only had 3-4 hours to make this video so I didn’t spend as much time editing as I should have. Will definitely try to cut the down in the future.

[@eliadallari]

it makes no sense to resize the photos you have to keep 45 megapixels vs 24 megapixels

[@the_camera_life]

I’m not sure what you mean? I resized all images to be the same 24mp so I could compare the noise quality side by side at the same viewing size. Do you think I should have upscaled the R3 to 45mp instead?

[@eliadallari]

@@the_camera_life yes but you have cut the pictures, keep original size !!!

[@kpopfanphotos]

@@eliadallari he didn't cut the pictures though. He scaled it down. This is how you properly do this comparison.

[@eliadallari]

@@kpopfanphotos it makes no sense!! the images must be left original, if one crops the electronic noise increases and loses detail, the r5 mark ii surpasses the r3 if you go to crop

[@AdamBotond]

​@@eliadallariHe did not crop them, just downsized them to make them comparable. It makes perfect sense. You should compare apple to apple, not apple to pear. Downsizeing 45 MP to 24 MP should actually hide some of the noise and create a more detailed 24 MP picture. In the this test, it has been proved that there is too much cost at high iso, when it comes to noise.