Agreed, thanks for that note Richard. I personally try not to shoot anything over 1000 or so...even iso 800 produces crunchy skin and skies in some cases... For RAWS, as I dont shoot jpeg.

I pay a lot of attention to these small bodies because my M(1) has produced such wonderful images over the years. I have been desperately trying to have a true upgrade to this camera, however there's always seems to be caveat. When i begin to travel again, i want to be ready for all that sweet landscape and portrait action. But its been a struggle to say the least:

It took 3 m62 bodies to get one from canon that had acceptable sharpness from the sensor, and that's not even discussing the...different color response compared to the older crop sensors. On the other hand, the m50's AF has been very impressive, even when dark! So for me, the issue's have been threefold- reliable AF/softness (m62), noise (m50), lack of ibis (all), and of course portability.

Interestingly, TDP noted that the image quality is actually better than previous sensor versions. The iso100 color checker looked good, compared to other reviewed non FF cameras. (you really would need to compare the charts directly to see the difference/upgrade) Next time i find a camera that gets the job done (like the M1), im buying two or three! Getting a reliable replacement has been the absolute worst experience.

Sorry. Even in JPG mode the results are exactly what I could expect

But now if you add Panasonic Lumix S1 into the mix ..

https://www.dpreview.com/reviews/image-comparison?attr18=lowlight&attr13_0=canon_eosr7&attr13_1=canon_eosr6&attr13_2=panasonic_dcs1&attr13_3=sony_a1&attr15_0=raw&attr15_1=raw&attr15_2=raw&attr15_3=raw&attr16_0=25600&attr16_1=25600&attr16_2=25600&attr16_3=25600&attr126_1=1&attr126_2=1&attr126_3=1&normalization=compare&widget=1&x=-0.2073375511198577&y=-0.6951009465630864

@Richard Butler, the reason I compared in Jpeg is because I use only it in low light, I find that Jpeg are not only less noisy but faster to edit and the inbox Jpeg engine is better than we what can do manuelly. Thanks for the Jpegs though, for me it is 1) Canon R6 2) Sony a7IV (with a bigger margin vs the R6) , 3) Sony a1 4) Canon R7.
Thanks for the new comparison.

@JPM29

Try DXO PureRaw and you'll never shoot JPG again.

@thx1138
As I referred to image processing, so related to .jpeg, R7 provide more details than 5DSR, mainly at darker arias (scene set to bulb), where 5DIV is way behind both.
And yes, to be fair, we need to add vastly superior glass on R7 into equation.

Well, I wouldn't say the results at ISO 6400 are bad. If you change from Full to Comparison, the field becomes more even (because the R7 has more pixels).

Of course, the XT-4 is still cleaner, but it's not a huge difference. If you look closely, the R7 is noisier but sharper. I bet if one denoises the R7 image one can get very close to the level of detail and noise of the XT-4. See the comparison in the low light version.

Now try this one: insert the 700D (my current camera). That's what I call an advancement! At high ISOs, the R7 is more than 1 stop cleaner.

@shutterbugnx
Since this is a traditional FSI sensor and not stack, I'm not so sure if the read out speed has anything to do with the sensor and not the electronics off-sensor?

@RecklessCoding
Yes, a camera is more than it's sensor. Features and ergo play a big part too. But the studio scene is purely aim at looking at the sensor and the processing electronics.

@Jon555
COMP button shows no improvement at all over the noise at ISO6400. The brush has more details, but that could be a variable of the lens used and also the AA filter strength (or lack of).

Overall I do hope Canon has a higher end APS-c camera in the works that's 32MP BSI sensor (non stack) and better noise/DR as this FSI one.

Frankly when I look at fellow members R7 shots, I don't quite like the jello effect (rolling shutter). Surely a big room for improvement.

I was doing the same thing, buying f/1.4 lens for crop camera, what a fail. f/1.8 lenses on ff are much better price and performance wise. Imagine 14/1.4 lens on crop camera you'd have to buy to match performance of 20/1.8 lens on ff. What a monstrosity that would be, if physics of this universe even allow such construction.

@Sacred. Seems like someone missed the release of Fujifilm Viltrox 13mm F1.4. Yes it is a bit heavier than an Fullframe equiv, but the image quality is one of the best currently in Fuji system. It was also on sale for 360$, pretty reasonable.

rawdinal be prepared to carry that damned thing! hahaha!

Alam12 RP does it right if yo do not need speed or lots of physical controls.
I'd like RP + RF 24-105 F4 IS L

No job for the RAW processor to do. All NR is turned off.

Yeah the skin tones look very different

Perhaps more care is needed in quality control of the prints or an explanation when things go this wrong:

https://www.dpreview.com/reviews/image-comparison/fullscreen?attr18=daylight&attr13_0=canon_eosr7&attr13_1=canon_eos90d&attr13_2=sony_a6600&attr13_3=fujifilm_xt4&attr15_0=jpeg&attr15_1=jpeg&attr15_2=jpeg&attr15_3=jpeg&attr16_0=100&attr16_1=100&attr16_2=100&attr16_3=160&normalization=compare&widget=852&x=0.8614538545877983&y=0.1579134576030639

Or is this what we should expect from this camera compared to the others?

Okay keep comparing the colors on the fading prints with captures taken at very different times in the print’s lifetime and draw the wrong conclusion.

Richard. You don’t have to replace them. You’d be better off just removing them and leaving white space.

Exactly. Its worse than pointless to use faded prints.

But this only happens rarely. The R review, and not any others that I can remember. The prints are usually comparable to the differences i see in my testing of these cameras. Not talking about any conspiracies either.

And then once in a while this happens. You have to bring it up. Either print them for the scene like DPR seems to normally do or remove them from the test.

Notice i never said anything bad about the skin colours, just the prints appear horribly off, but how do we know its the prints? Some form of explanation would be needed.

The portraits are going to be replaced, but the challenge is what to replace them with. That's not something I can join at do off-the-cuff that decision will affect all future test shots.

The existing prints have deteriorated faster than expected, but I haven't have the time to update them (constantly thinking the replacement is imminent).

Not having been able to visit the office for nearly two weeks isn't helping.

It's nice that DPR use the same exposure. In this way, you can really compare ISO performance.
While comparing the R7 to the 90 D, the more obvious difference is sharpness : the RF 50 1.2 is much sharper even at F5.6 than the 50 mm 1.4. I d'ont think that the difference is related to a weaker AA filter

rawdinal - the original 5D was launched in 2005, before the ISO standard changed. The standard at the time specified a fixed gap between clipping and middle grey (in JPEG), comparable to the one that DxO uses to normalize its Raw results.

The ISO standard was revised in 2006 in such a way that it gave manufacturers more control over how they thought the sensors should be used.

Modern cameras don't 'fail' on 'real ISO,' they are intentionally configured to capture more highlight information than the pre-2006 standard dictated. The aim was to give better JPEGs with better highlight roll-off and make better use of improvements in sensor technology.

No manufacturer attempts to comply with the saturation approach DxO's uses for normalization (which the latest version of the ISO standard appears to explicitly distance itself from). DxO's charts can indicate how much extra highlight information the manufacturer is attempting to capture. This is not a measure of ISO accuracy.

It is not correct to say that the EOS 5D II was 'less sensitive than the iso on the camera may suggest.'

In our studio scene, the Mark II has received 1/3EV more exposure and has delivered an image approximately 1/6EV brighter (the 1/3EV increments meant we couldn't perfectly match the two camera).

The two cameras produce JPEGs within around 1/6EV of one another, when given the same exposure. That's what the ISO standard demands. It does not state that the Raw files should place middle grey the same distance from clipping, which is what DxOmark measures.

I accept that the more useful part isn’t simple at all. But these images are definitely misleading!! (That much is simple!)

Anyway glad you guys have some thoughts about this and plan to do something about it.

:)

@Richard Butler -- what kind of prints are you using that the pictures are changing their colors between the cameras tested within the same year or even the same month? And how come fading prints are suddenly turning green?
I have large prints on my walls and they still look pretty good after a decade.

All these cameras, if I remember correctly were tested this year, yet only R7 shows the green skin tones. I have hard time believing the prints changed so dramatically in such a short time.
https://www.dpreview.com/reviews/image-comparison/fullscreen?attr18=daylight&attr13_0=canon_eosr7&attr13_1=sony_a7iv&attr13_2=omsystem_om1&attr13_3=nikon_z9&attr15_0=raw&attr15_1=raw&attr15_2=raw&attr15_3=raw&attr16_0=100&attr16_1=100&attr16_2=200&attr16_3=200&attr126_2=1&normalization=print&widget=852&x=0.8695994694630135&y=-0.22295447680236602

Read his comments above. There’s another comment thread about it.

@MikeRan -- I haven't seen any satisfactory explanation neither here nor anywhere on DPR site.
First, it's clear they already changed the prints (at least once), you can easily see it by different glue marks. Second, I haven't seen even the worst quality prints that fade into green in a matter of 3-6 months as you can see from different cameras tested at approximately same time.
I don't doubt the prints fade, though some of my photos are 40 years old still looking like new as well as all my 10 y.o. inkjet prints, so DPR inexplicably must be using some very quickly degrading print process. But even that doesn't explain what we see here with cameras used at the same time frame.
Btw, another indication that fading is not the (only) explanation is the ColorChecker. The colors have differences, but of course, you can't judge skin tones from the ColorChecker chart.

Perhaps not an explanation but a clear admission and advisory which you may choose to ignore if you don’t like it. Anyway draw your own wrong conclusions if you like. Also Richard’s comments are as official as they get too…. But since it seems that’s not good enough for you, well I guess this won’t be either.

https://m.dpreview.com/articles/9306199729/sony-a7r-iv-added-to-studio-test-scene-comparison

“Please note that since it was originally printed, the portraits on our test chart have faded. As such, they should not be used to draw comparisons between the color output of different cameras. We are aware of the issue and we are working on a solution.”

@MikeRan -- you bet I don't find all these explanations holding water. Why is that nobody wants to address the fact that the same "faded" images rendered differently with different cameras? And also, are they faded into green? Would that be quite unusual fading effect? Why then no other camera got that?
Also, I remember in film age there were standardized reference images for different skin tones. I still remember Kodak cards. What happened to those, why suddenly DPR has a problem that didn't exist in the past?

@Foskito -- oh, I didn't know flickr has a set of skin tones reference images for different cameras. Could you provide a link?

Did you see the link? That link is literally from the past. Haha. Deny all you like. Misjudge all you like. I don’t care.

You mean how the camera looks or the image quality ?

Both actually. The difference is that the camera looks like it is 10 years old, whereas the image quality seems like it's from 3 years ago.

Actually, my camera is almost 10 years old. But in terms of dynamic range, noise, and other relevant factors, it still beats the R7. Check https://www.photonstophotos.net/Charts/PDR.htm

However, that is not the point. The looks of the camera are my subjective opinion. And the sensor design is simply at least 3 years old, that is a fact. You cannot dispute either of that.

@mr predator - Do you really not see the benefit of lenses designed for just for cropsensor? Cheaper, lighter, smaller.

@hfolkertsma
Show me. It's mostly due to being slower in comparison or more expensive except sometimes in very short focal lengths. Means most of the time it wouldn't have much of a benefit otherwise there would be an aps-c "holy trinity" as these three sell well.