I think they are pretty good. Not amazing, but decent

For me it is a question of which lens performs better for true life colours. Light behaves in very different ways; if the camera sensor captures such shades and shadows. Sony A7 high scorer whereas Fuji tends to be more fussy; different but not worse. Most times it's easy to spot any Canon image by their REDS.

When I switched to R6 from 5d2, I noticed that the default color is different. I also noticed a pretty strong base correction in the green/magenta slider in lightroom/camera raw default settings. Overal "color logic" is pretty different compare to "good old canon". But I'm used to it. I'm a RAW shooter, and i always see colors in my way, i always make pretty strong manipulations in camera calibration and HSL, so main thing for me is that there are no incurable problems, no artifacts or defects, and I really don't see any fundamental problems. I see pretty good amount of data in all channels, all colors behaves well under corrections, all things works. Maybe situation is different for jpeg shooters, or for RAW shooters, who used to process colors not so heavy, I don't know, but i am happy.

@D135ima, good for you if you see it that way. For me the less I process the colors and images the best I feel. I want to have the closest to what I see in general, but not for some special photos. You know that in Jepgs you can also chose your colors before shooting, I don't rember the name though but you have four axes for colours.
Cheers.

Post processing is of course always possible (particularly for RAW shooters). But everyone is different. Some are more talented at PP. some enjoy PP more than actually capturing the images etc.

But IMHO where “color science” is important are three things. The camera that starts you off closest to your preferred look saves you time. The manufacturer that has their HSL such that the need to adjust these o
Are minimal makes things much easier. And to me the most important thing is that you don’t always have the luxury of waiting to do PP.

While on vacation, wanting to post things immediately to social media for example are times you would like the camera to have decent to great looking colors right if the bat. And then there is paid work. Providing some photos from a wedding to the bride and groom… or showing clients images on the back of the camera during a shoot.

R5 and 6 are trash for video because of the asinine micro-HDMI port. Yes, it's that big a deal. Nobody else uses this junk.

I think in many ways it's EF mount. in 1987 they made a very successful and far-sighted decision, to make fully electronic mount and still benefit from it.
But in the long run Canon may lose to Nikon if they doesn’t start making balanced no-nonsense lenses. You can't rely on EF forever.
I still don't have a single RF lens. And not because EF are very good, But because the RF is not just lens line, it's a field for ambition and experimentation. Except maybe 24-105 and 35 1.8, all other lenses are kinda strange and funny, artificial more expensive for features no one asked for, or full of compromises, or even defective. (Yes, I'm talking about these 4 (!!!!) lenses that do not cover the full frame)

What four lenses? This article's about a camera.

"even with noise reduction the Raw files look just amazing,sharp,clean etc"

THAT'S THE PROBLEM. Talk about not getting it...

Francis, I’m not sure what the problem is. If you don’t have any smearing or softening of noise reduction, then what’s the problem? I have the camera and it’s amazing! You probably don’t even have the camera to even know. I’ve had a ton of cameras and the R5/6 are my favorites. Show proof you have the camera, so we can see what you are complaining about. Maybe we can help.

Our scoring system was designed 10 years ago, so there are limited metrics for video. The S1 has less rolling shutter but can only do 60p from an APS-C crop, the R6 has more rolling shutter in 30p, less in 60p but can do high-quality 10-bit capture in all modes and captures from most of the width of its sensor, it has very good video AF and some of the best stills/video separation.

It's difficult to sum al this up in a single number, especially as everyone has their own priorities and ways of working (some people will find the R6's thermal limits a problem, other people will likely never encounter them, with the new f/w). For some people the R6 will be as good, if not better, for video, for others it won't.

Thank you for reply and additional info.

I understand. In that case it might be worth reviewing categorization logic.

Just as an example, the Above $2k guide lists the Nikon Z7, which is same price as the R6 but is not in same guide. OK, it does not target the same audience, but neither do the Sony A7Siii and A9, which however are in the same guide. Category is not price then, maybe a mix of that and audience.

I agree it's very difficult to categorize. There might be other ways of breaking down guides which can make easier for you to put them together and for readers to follow the logic and use them. Cheers.

There are downsides, whichever way you choose to break it down. If you make named categories, there'll be arguments over the cameras near the borders between categories, if you do it based on price, you have the problem that older cameras sell for less than list price or get re-positioned to make way for newer models.

The Z7 is a great example: Nikon lists the Z7 body only as $2799, but then says it's $2599 with a $200 rebate, so which price should we use? How long will that rebate last?

And that's before you hit the problem that prices vary at different rates in different countries (Non-US prices tend to start higher but drop faster, for instance), and that '$2000' is meaningless if you live outside the US and live in a country where tax is stated in the price...

But we are constantly trying to adjust and react to what's happening.

At this high price, 20MP (and no top plate LCD) does seem very mean, but where else could Canon have positioned it? A 24MP sensor would have looked better on paper, but slowed its other abilities slightly. Giving it a 32MP sensor would have eaten into R5 sales, as few people want 8K video or need 45MP. Canon are being seriously greedy with their pricing for the R6, providing an awkward range of mirrorless FF bodies/prices ranging from lacklustre (R/RP), expensive (R6), or stupid expensive (R5) with nothing in the price/spec/features sweetspot.

It seems from the review and DXO that the R6 has better DR than the R/MkIV? Even my X-T30 has better DR than the EOS R according to photonstophotos. The lack of a top plate LCD is bad form, though. Terrible.

@ashbashbeard
1) I would like to hear more from photographers, not vloggers, preferring a flip out screen. How that does not get in the way when in a hurry in flipping it out and framing not in line. As an action photographer I absolutely dread this, as a fact I often keep the monitor inactive and facing inward for how useless it is, at least it stays protected.
2) Mirrroless thinking does not necessarily mean small. I want a dedicated exposure compensation dial. I always shoot manual, frequently switching from auto to a set ISO. I want 4 direct dials for shutter, aperture, iso, and exposure compensation. Also, because the flip screen is more awkward to use in the common low and high angles, it makes it more unpractical to switch frequently from EVF to monitor ending up shooting all the time with the VF like a DSLR.
I want full Q menu and button customization, on my R5 I do not seem to be able to map specific functions buried in menus to a custom, more accessible location.

@ashbashbeard
Sony cameras have 2 dials side by side for aperture and exposure compensation, far more comfortable and intuitive than dealing with the quick dial on the back, which, btw, is not clickable in the Canons in the 4 directions dramatically limiting its usability. I find it funny that Sony bodies have been so criticized and here we have Canon with quite an inferior ability to directly access your functions. I do welcome the full separation of video settings, that is something that Sony should have too and I really do not get why it doesn't. Also, I think the lock button is a nice touch and I like it the way is designed flush on the R5.

It's cute that you are demanding to know DPR's internal publishing schedule.

Dpreview staff is endlessly debating between giving it a 94% or 95% score. They are not able to decide. LOL. Maybe they are even creating a new platinum award...

maybe they have been asked to hold off by Canon firmware upgrade 1.2 coming up... or something. Then again, that'd be unfair, the company should release finished products.
Either way, it's just weird and makes me think they either didn't test the R6 properly or they practice active click-bait, knowing people interested in the R5 keep visiting the website more until they've read that review, then don't visit as often anymore.

Alright, I should have specified, in video the colors are vastly different from each other. Both camera conspiracies and potato jet pointed that out.

@Breakfastphotographer, your tool shows the bald guy a bit yellowish actually.

Yea! Until the Sony A7rv comes along 😂😂🤣

24mp is only a very marginal increase in linear resolution over 20mp. Even the 32mp of the EOS R is only a small increase in practice.

That is true, but where do you draw the line? What if it was 18MP?, 16MP?
20MP is not a show stopper, but it is a bit disappointing.

do you all do huge prints of 1.5+ meters width? or when/where else do you guys need more than 20 MP?
Yes, yes cropping, but Ideally not too much of it...

Well the 5DMkII released 12 years ago had 21MP....20MP isn't the end of the world, but a little thin resolutionwise.

There has been 40mp+ medium format cameras back in 2007, doesn't mean the likes of the R5 are outdated as a result.

It's obviously not just about dem megapickles. Sensors have since improved dramatically in noise and dynamic range since then, which I would argue is much more difficult to do compared to slapping a few more mp into it.

@ashbashbeard: There's a lot of truth in that - we tend to overly focus on things that are easily verifiable, like pixel resolution, frame rate, maximum aperture. We focus much less on things harder to verify, like dynamic range, buffer performance, or accurate alignment.

At 90%, receiving a gold medal, I don't expect to see a camera that's so fundamentally flawed. The Panasonic S1, which outperforms this camera in nearly every way except autofocus and burst shooting, at the same price yet released 2 years early, only scores 88%. Same for the Z6, heck you can even throw the a7III in there. The R6 outscored its peers despite performing worst across the board with dynamic range, high ISO performance, resolution, coming out years afterwards, having cooked raws, and overheating issues that severly limit its capability. It's not like it's cheaper than its competition either.

@Less Channel, it's ISO and dynamic range and ISO performance is slightly less than Z6, A7 III and S1. This is what DPR wrote, "It's a little noisier than the best of its peers, but not to a degree that matters."

Look I don't even use Canon cameras but let's not pretend this is a terrible camera.

Yet it outscored all of those cameras. I didn't say that it is a terrible camera, but it has series drawbacks that get discounted, like the baked in noise reduction which is a deal breaker for some photographers, and the overheating which is a serious deal breaker for anyone looking to create videos. Yet R6 is rated higher than cameras that don't have these issues, which are all under 90% and none received Gold. All of which are also higher resolution, have better dynamic range, and better low light performance.
DPR staff may say that the scoring wasn't based on the videos recording capabilities, yet the camera is marketed as a hybrid camera. The first lines on their product page:

"The EOS R6 full-frame mirrorless camera is designed to meet the demands of both photographers and video content creators who want a strong and versatile camera to keep up with their active lives. It can go from photographing high-speed action sports to filming 4K cinematic video with ease..."

Hope you can fit a lot of action in that 5 minutes left for video after the sports photography with this Gold medal camera. Baffling. How does the camera score better while performing worse at a higher price while releasing later, with camera features that render your camera useless for extended periods of time if you employ them?

It's really hard to keep track of DPR's biases. One week they are Sony shils, the next week its all pro-Canon. There really should be a spreadsheet available.

It's simple, you can have a bit less resolution, a bit limited video record time, but you can't have an AF system that's unreliable and hunts everywhere and is Contrast-detection based. That's why it scores more than a Panny S1, its one or two flaws are too major while the r6 more flaws are much, much less important in a do-it-all gerenal 2K camera. That's why I don't recommend using the scores or awards for anything, what if you're a guy like me who only shoots in manual focus? Then the S1 is a much much higher score as a video camera with a focus puller and a film crew, and if you're a vlogger, the r6 is a much higher scorer in video, etc.

Just read the full review, know the flaws and strengths, choose accordingly

Then maybe you should think about getting a dedicated video camera rather than a stills camera that also does video.

Canon still have the most measurably accurate colours. And always up there with subjectively best also. Especially for natural lit portraits.

Not sure if your post is a poem or a conversation?

@SteveAnderson, you wrote "Canon still have the most measurably accurate colours.." Any evidence of this? The colour accuracy test using colour checkers that I saw doesn't show this. The top positions were dominated by Fuji and another competing camera manufacturer. I can't say about R5/R6 as they probably didn't measure them.

Handsome, Yes, Canon and Fuji consistently score most accurate. 7Dm2 with an avg de of 3.45.

This site linked and my own testing have always been in line. 20yrs worth of consistent testing here.

https://www.imaging-resource.com/PRODS/canon-7d-mark-ii/canon-7d-mark-iiA5.HTM

= Top notch and shows in photos.

https://www.imaging-resource.com/PRODS/sony-a6500/sony-a6500A5.HTM

= Terrible and shows in photos.

"Terrible and shows in photos."

May be but the link that you posted shows that A6500 is not far behind the canon you cherry picked. In fact at certain points the gap between ideal and camera was as far as it was in A6500 tests.

Post a better cherry picked camera for canon and worse than A6500 for sony, then may be you have some point.

@SteveAnderson, thanks for sharing your link. Canon looks excellent there. Sony A9 also looks good there too.

Would you say measuring more colours is better? Here they measure using a wide gamut colour checker (X-Rite ColorChecker SG color target, which has 96 different colours). If we compare your Canon R (which is one of the latest Canon cameras they tested), 6 colours exceed delta 20 and 1 colour even exceeded delta 30.8. For A7 III only 1 colour exceeded delta 20 and for A7R III only 2 colours exceeded delta 20. Obviously that's not the whole story. you may have a few colours that exceed delta 20 whilst the rest of the colours can still be incredibly accurate. This is what we see with A6400. So
you'll have to also look at the rest of the colours also

https://pdnonline.com/gear/cameras/the-best-cameras-for-color-reproduction-ranked/

Thanks for sharing your experience.

@SteveAnderson

"xzaar, I think you dont know how to read those plots well. They are worlds apart."

Keep deluding yourself. Your comments show that you do not understand the numbers at all.

"The Sony is off in the wrong direction by too much in the wrong plots."

Which direction is the correct direction??
There is no such thing as a direction when a scalar value is involved.

In this case there is a accurate value and then there is a value captured by the camera. There is error which is how much camera is off from ideal.

"It is not remotely close and no Sony plot is better than the Canon."

Not in the tests you posted.

Here is the thing, the maximum difference in canon and sony is very similar in their tests.

If x% difference in canon makes canon accurate and great than x% or less difference does the same for sony too.

You can delude yourself but if you want us to believe, bring proof.

I'd rather have this sensor than more MP and lower dynamic range. The 30MP r or 5d4 are similarly good.

With regards to being a noise purist, I don't have the same opinion. All the manufacturers have to process files, even raw files use an algorithm to process the colour information from the sensor.

I'd rather skip the noise processing too, but I accept it's there and won't be a disaster. I don't care about overheating video either, 1080p is more than enough for non pros and even Hollywood uprezzes 1080p for major 4k releases.

For me, coming from dslr and with good eyesight the evfs are stil not up to par with optical viewfinders, especially for enjoyment. I'm most interested in the 5r's betterviewfinder, but the price there really is too high for me at the moment.

i don't want cooked raws. Noise reduction on raws needs to be switchable by user, especially "full raws". canon can do whatever they want to jpgs, HEIFs, cRaws, mRaws etc. - they created so many filetypes i don't need or use.

But i want full choice and freedom to post-process them my way with regular raw files, starting with uncooked .CR3s out of camera, if user wants it that way.

up to now uncooked raws were a clear advantage with Canon, compared to other brands. Now we have less and less choice. Not good for Canon, because i won't buy their cameras, might as well get a Sony.

“I'd rather have this sensor than more MP and lower dynamic range.”

That’s a false dichotomy. The R5 has more MP and higher dynamic range.

Yes, the R5 has noise reduction too, up to (but not including) ISO 800.

The R6 has it with the mechanical shutter until ISO 1600 instead of 800, but not at all with the electronic shutter (unlike the R5 which has it in both modes).

demosaicing, embedded jpgs, addition/inclusion of various metadata entries for potential use in post-processing do not count as "cooking". Cooking means, additional measures are taken to "manipulate image" beyond these steps already in camera and cannot be avoided or excluded by user. Application of non-optional, forced de-noising is a prime example of cooking raws. It is something I do not want at all, unless i consciously select settings to enable it.

To my knowledge there is no also compression of Canon full raws. I only use .CR2 and .CR3 files.
Don't care for cRAW, mRAW and whatever other cooked, bastardized raw files cameras may be able to prioduce *if user consciously sets them to do so*.
Do you have any evidence that Canon is compressing .CR2 and/or .CR3 files?

CR2 and CR3 files use lossless compression. There is plenty of discussion on this, so look it up.

Colour response is a mix of the bayer filter and software calibration from the manufacturer itself, so this is also "cooked" into the RAW file. This is why colour matching profiles are tricky to 100% pull off between different camera manufacturers.

Everything else is still cooking, because the RAW file is processed to have these features added, and like everything else, cannot be removed.

Again, I don't think the noise reduction is there to simply decrease overall noise, as it only applies to ISOs up to 800 and there is no loss of detail. Even without it, the DR is still competitive too. Likely there is a different reason for it.

I would definitely count demosaicing as cooking unless there is metadata indicating how to get the original samples.

@all How did DPR prove that NR is applied to raw?
Since they use ACR, we don't know whether ACR applied it when converting raw to sth we can look at a computer screen (e.g. JPG). See also
https://www.dpreview.com/reviews/canon-eos-r6-review?comment=7943661741

@b823s7: it’s not just DPR, Bill Claff of PhotonsToPhotos.net found evidence of noise reduction by looking at spectral data: https://www.dpreview.com/forums/thread/4521740

The RAW files were mine, and I can certify that I have the format set to “RAW” and not “CRAW”.

Well, Bill writes "Unlike the R5 I did not detect NR using electronic shutter on the R6.". That sounds different from "Some noise reduction is always applied to low ISO Raws ...". Yes, it might be annoying that there is some NR in case of mechnical shutter. Now the intereseting question is why NR depends on the shutter, mechnical or electronic?

Why is it a problem, when literally no detail is lost? It's not like ISO 100-400 are high in noise to begin with, in fact they're not. You still get a decent 14 stop range of DR, with the added benefit of not having horrible Sony colours. Even if there is no NR, it's still about 13.5 stops of DR - quite competitive and still better than the 12-13 stops of the EOS R.

It just seems like you're trying to make nothing look like a problem. No one knows why Canon added it, but it's definitely not to make already extremely low noise ISO RAW files look better.

@ashbashbeard I agree
@all others: I think this is not a real problem. You are discussing a thing that has no relevance in reality if NR is applied to raw files starting at ISO 1600. It might be inacceptable at ISO 100 - 400. Landscapers might pixelpeeing and edit their images to death taken at ISO 100. If I'm forced to shoot at ISO 1600 then some NR is not my urgent problem. If I print @ 10" or 12" or publish on the web, nobody will see a difference. If I print at 6 ft I shouldn't shoot at ISO 1600.

@emfor: NR is applied *below* ISO 1600, not above.

@emfor NR is applied to ISO 100-400.

That range has virtually no noise to begin with on typical FF sensors (even some APS-C sensors), so the NR must be for a different reason. I speculate it's to remove slight sensor banding when shadows are pushed. The EOS R is criticised for this (even though it was barely noticable), especially prior to firmware updates which fixed most of it.

for RAWs it should be optional. maybe even as "default" setting - but definitely overridable when user sets NR to zero.

for jpgs, HEIC, cRaws, mRaws, etc. they can apply in camera whatever "beautifying measures" they want. but not non-optional in "real", full raws.

it is a real issue, even when this specific NR measure should turn out to be for an acceptable reason and (mostly) useful. full raws are the holy grail. they are digital negatives. i want to have the option to get them as untampered with as possible. up to know Canon was the only brand you could rely on in this regard. not any longer.

It isn't an issue when you literally get no artifacts or loss in detail when you push shadows. The fact it only really shows up in specific spectral tests shows this - typical NR in comparison is easy to spot by eye. It's also obviously not the same as NR in the likes of Lightroom, you do not see the characteristic mushy, painterly like look with a loss in fine detail.

I would understand if it did result in a loss in image quality in any noticeable way, like as seen with Sony's RAW files (eg. star eater problem and weird artifacts), but it's simply not the case here outside possibly removing slight sensor banding, although no one really knows why it's there.

Again, you're fussing over nothing like a typical spec sheet ogler. Camera gear just seems to be vessels of measurbatory numbers to you. If no one did the test, you simply would have never known or cared, instead nitpicking on some other pointless detail.

I think it is too early to conclude with any confidence that there are no negative effects associated with Canon applying mandatory, non-optional NR at low ISOs.

And even if it had "no negative side efffects" non-optional processing of raw files is an issue. Canon should continue to give users who want it the option to get "clean, untainted, as pure as possible" raw files - without any NR applied.

There is no lack of choice for non-optionally pre-processed files out of camera: jpg, HEIC, mRaw, sRaw, cRaw, whatever.

By taking away the only existing possibility to get clean, pure, lossless raw files Canon has lost one of their competitive advantages for me. They have put themselves on the same lower level as Nikon and Sony. To me it means 1 reason less to consider Canon gear for purchase in the future.