Canon EOS 5D Mark IV Achieves Best Canon Sensor Score at DXO

[sebasan]

The "bash" on DxO was always in their scores, not in their measures.
In this case, comparing with the scores of the A7RII and D810, the 5DIV scores and overall score seem short for me.

 

[3kramd5]

I think a number of 5D3 folks not planning on upgrading may have an uncomfortable morning...

... why? If they aren't planning on upgrading, it's because 4 doesn't offer anything compelling over 3. A convoluted, secret sauce sensor score posted at DXO probably change that for most people.

I ordered a 1Dx (used) after 5D4 specs came out. I don't regret that choice. The sensor is but one subassembly that makes up a camera. I think the marginal differences in two to three performance parameters between competitors' sensors gets an inordinate amount of play relative to the real world difference they make, and the amount they matter to most consumers.

 

[ahsanford]

I think a number of 5D3 folks not planning on upgrading may have an uncomfortable morning...

... why? If they aren't planning on upgrading, it's because 4 doesn't offer anything compelling over 3. A convoluted, secret sauce sensor score posted at DXO probably change that for most people.

I ordered a 1Dx (used) after 5D4 specs came out. I don't regret that choice. The sensor is but one subassembly that makes up a camera. I think the marginal differences in two to three performance parameters between competitors' sensors gets an inordinate amount of play relative to the real world difference they make, and the amount they matter to most consumers.

Understood, but remember that folks do (or do not) upgrade for a ton of reasons.

Some are in on day one strictly based on specs/features (Does it use the video format I prefer? How many f/8 AF points does it have for my birding? etc.), while others value real-world use and reviews to make sure the product delivers on its specs, while yet another group of people absolutely love what currently shoot with and just want to know how much better the sensor is.

If you are in that third group, DXO just stated (and I'm hoping others will corroborate) that this is a rig worth upgrading for. It appears to be a considerable upgrade over the 5D3 in (obviously) resolution, low light performance and dynamic range. If you care about those things or run into the limits of what the 5D3 can do in those metrics, this may a rewarding upgrade for you.

But if you're in the other camps -- let's say you're in the 'review' group and you want to know the AF hit rate, or if you're just a spec junky who wants a 30 MP FF rig with DPAF, this DXO report (and other sensor testers's work) may do nothing for you, I agree.

- A

 

[hajiaru]

Some of the posts in here might break the Internet. LOL Some people will never be happy with what Canon delivers...just continue worrying about your gear and taking awful images and blaming the gear. (Hint: it's not the gear)

mostly video guys are disappointed in the camera after waiting 5 years too much 4k crop they couldn't frame the subject and New DiGiC 6+ = 4K in MJPG no proper codec no faster memory cards ;D ;D ;D ;D Why do u present ur product in 1080p when it shoots 'crop' 4K

 

[Big_Ant_TV_Media]

so i think we all owe CANON a big TIME SORRY FOR BASHING YOU THESE PAST YEARS
"the 1s who bashed them" they have been improving just at there own pace, there sensors can now ONLY get better" dont u all agree but wow, so we can safely say we have as good if not better sensor in our 5D4 as the 1DX if not better?

I don't know if I would go that far, but the 5D4 sensor closely matching the 1DX2 sensor performance despite +10 MP is an impressive feat.

- A

i didnt say it was but it does closely match it??

 

[Refurb7]

I think a number of 5D3 folks not planning on upgrading may have an uncomfortable morning...

Not me. I never found the dynamic range or low light performance of the 5D3 to be lacking, so the 5D4 is nice but won't make me rush to upgrade. If and when I upgrade to the 5D4 I will enjoy the benefits, but until then the 5D3 is quite a fine camera. Does everything I need. The fact that the 5D4 is 2 ounces lighter is more interesting to me than the sensor.

 

[Refurb7]

On a side note, give it to dxo to think the 5D IV's predecessor is the 1D III... :

That is DxO logic and accuracy at its finest! Good to know that the 5DIV is a real improvement over its predecessor, the 1DIII.

 

[IglooEater]

What happened to the DxO bashing?

It's still here; you must have missed my post on page 1.

 

[scyrene]

I think a number of 5D3 folks not planning on upgrading may have an uncomfortable morning

Not this one, at least...

As a 5D3 owner, there is one area the 5D4 falls short for me and that is lack of noticeable improvement in resolving power and sharpness - despite the resolution bump. The other stuff is nice, but ultimately end IQ is what I'm after. Other than the 5DSR, I'm better off sticking with what I've got and upgrading lenses instead...

Tests are fun to discuss, and this one from DXO seems pretty close.
I have the 5DS and sold the 5D III. The 5D IV came late yesterday.
I took one jpg test shot last night and was blown away by the improvement. The biggest thing I noticed is how well the reds held together at ISO 12800. Contrast was better and edge detail without colors bleeding was impressive. The DR improvement is noticeable.
I am very happy with my decision.

High ISO jpeg processing seems one area where significant improvements are still being made.

 

[scyrene]

I think a number of 5D3 folks not planning on upgrading may have an uncomfortable morning

Not this one, at least...

As a 5D3 owner, there is one area the 5D4 falls short for me and that is lack of noticeable improvement in resolving power and sharpness - despite the resolution bump.

How are you determining that? It runs counter to common sense.

Simple. I presume Act444 wanted more pixels in the 5D4 than 30 MP.

- A

Honestly, I'd have preferred a more modest MP boost (say, 24) and a 1-2 stop improvement on high ISO. But, if you've (Canon) chosen to push me towards more MP instead (with the direction of 5D line), ok fine - but I want more sharpness and detail as well. At a minimum, the same if not more pixel-level crispness. From the RAW files I've looked at so far, this is where the 5D4 disappoints me somewhat. I'm sure it's a fine camera, but it's not worth a $3500 upgrade to me at this point.

2 stops at high ISO may not be possible. I've seen it repeated here fairly often (by people I pay attention to) that at the highest ISO settings, we're approaching the physical limits of what is possible (caveats: raw images, with Bayer-filtered silicon sensors).

 

[East Wind Photography]

lot of camera for your money, including a sensor that boasts the best DxOMark sensor score to date for a Canon DSLR

What a turn off. Like i would buy a camera just because dxomark scores the sensor high. They are a bit full of themselves to even say that in their article. So done with them.

 

[3kramd5]

I think a number of 5D3 folks not planning on upgrading may have an uncomfortable morning...

... why? If they aren't planning on upgrading, it's because 4 doesn't offer anything compelling over 3. A convoluted, secret sauce sensor score posted at DXO probably change that for most people.

I ordered a 1Dx (used) after 5D4 specs came out. I don't regret that choice. The sensor is but one subassembly that makes up a camera. I think the marginal differences in two to three performance parameters between competitors' sensors gets an inordinate amount of play relative to the real world difference they make, and the amount they matter to most consumers.

Understood, but remember that folks do (or do not) upgrade for a ton of reasons.

while yet another group of people absolutely love what currently shoot with and just want to know how much better the sensor is.

If you are in that third group, DXO just stated (and I'm hoping others will corroborate) that this is a rig worth upgrading for.

Sure, but those people would not have already made up their minds, so there is no internal struggle with "but I wasn't going to upgrade."

I may have misinterpreted what you wrote as suggesting this DXO post would alter people's decisions.

 

[3kramd5]

...
... why? If they aren't planning on upgrading, it's because 4 doesn't offer anything compelling over 3. A convoluted, secret sauce sensor score posted at DXO probably change that for most people.
...

Whilst the "one number" scores get lots of scorn, looking at the component scores is very telling.

For example..

5D4 - Low light: 2995
5D3 - Low light: 2293

5D4 - DR: 13.6
5D3 - DR: 11.7

So if you are a low light shooter or landscape shooter, the 5D4 is going to be better than the 5D3.

Interestingly, the gaps to the A7RII/D810 are somewhat closer now...:

5D4 - DR: 13.6
A7RII - DR: 13.9
D810 - DR: 14.8

5D4 - Low light: 2995
A7RII - Low light: 3434
D810 - Low light: 2853

Those would be covered in the statement from the second half of my post: "two to three performance parameters"

Will this camera provide slightly better photos in edge cases? Sure. In general use? I don't think those numbers are going to bust open barn doors.

 

[privatebydesign]

I think a number of 5D3 folks not planning on upgrading may have an uncomfortable morning

Not this one, at least...

As a 5D3 owner, there is one area the 5D4 falls short for me and that is lack of noticeable improvement in resolving power and sharpness - despite the resolution bump.

How are you determining that? It runs counter to common sense.

Simple. I presume Act444 wanted more pixels in the 5D4 than 30 MP.

- A

Honestly, I'd have preferred a more modest MP boost (say, 24) and a 1-2 stop improvement on high ISO. But, if you've (Canon) chosen to push me towards more MP instead (with the direction of 5D line), ok fine - but I want more sharpness and detail as well. At a minimum, the same if not more pixel-level crispness. From the RAW files I've looked at so far, this is where the 5D4 disappoints me somewhat. I'm sure it's a fine camera, but it's not worth a $3500 upgrade to me at this point.

If you don't want the camera for whatever reason that is fine by me, but to compare different sized pixels on a one to one basis is an entirely fallacious concept. The 5D MkIV sensor vastly out performs the 5D MkIII sensor.

Please post a screen shot of the comparison images that drove you to your conclusions.

 

[Cali Capture]

As a 5dMarkII user I obviously Pre-ordered the IV. All I can say is it's a HUGE upgrade!
Form factor alone makes it feel like a whole new generation of camera.
Touchscreen menus= much faster workflow, instant focus check by finger spread zoom on higher res. screen
Autofocus: was a center point and recompose shooter- now trusting the wider AF points much more- That's not east thing to change, but the IV is really allowing me to "let GO"
30 is better that 20mp- just ask someone with 50! My guess is that most images in the future will end up on a 4K TV, and those screens will continue to get larger ( 70" and 80" are common now) and the MP race is not over yet!
Nightvision Camera- I threw on the new 35mm f/1.4 II and shot around my house with the lights off, dim Nightlights allowed auto focus, I then switched off autofocus and could shoot handheld MF in the near dark, ans see where I was going! You don't need a lot more ISO than that!
HDR features are great, Stills and Video
Timelapse anyone?
I've been buying great canon glass since I got late into the Mark II, I now feel like I have a body to take advantage of those lenses and look forward to a new 50L and the 16-35mmIII. This is a great camera and if you make money with your photos you will probably get one (outside the DX and RS users). I can see the argument for waiting for a price drop if your on the MkIII. But one comment I heard makes sense, You get the bottom half of the fudge sunday by waiting. It's the "Cumulative" that makes the IV worth buying, and that has always been why the 5D series has been the best selling DSLR!

 

[Ozarker]

I think a number of 5D3 folks not planning on upgrading may have an uncomfortable morning

Not this one, at least...

As a 5D3 owner, there is one area the 5D4 falls short for me and that is lack of noticeable improvement in resolving power and sharpness - despite the resolution bump. The other stuff is nice, but ultimately end IQ is what I'm after. Other than the 5DSR, I'm better off sticking with what I've got and upgrading lenses instead...

Doesn't matter to me at all. I am extremely happy with my 5D Mark III. Extremely.

I'll be waiting until two years into the Mark V cycle. That doesn't bother me a bit either.

More great glass is on my list before a body upgrade.

I can't afford to be as big a Fanboy as I'd like to be.

 

[Deleted member 7828]

Please post a screen shot of the comparison images that drove you to your conclusions.

The following shots are from DPP. Both feature a comparison window featuring the studio scene pic from DPR. RAW files downloaded and loaded into DPP 4.5. 5D3 pic on the left, 5D4 pic on the right. Noise reduction turned off, sharpening set to 0 on both images (option unchecked). Focus is on the green foliage for reference. You can also see that the FL & exposure on both is the same (1/40, f5.6, ISO 100, 85mm). 1st shot shows 5D3 shot selected, 2nd shot shows 5D4 shot selected (to show that settings are indeed the same)

Now, at equivalent sizes the 5D4 will have the edge, and the MP difference is still enough to put the 5D4 ahead overall...but when I view both images at 1:1, it seems to me 5D4 has a softer appearance...I dunno, am I seeing things?? Have my eyes turned on me?

to compare different sized pixels on a one to one basis is an entirely fallacious concept.

...hmm, Is it though? Maybe I'm doing something wrong, but I will next include the 5D3 vs. the 5DS (not the R) to show what I'm getting at here. Unlike with the 5D4, when both are viewed at 1:1 you can CLEARLY see the 5DS resolving detail with similar crispness as the 5D3. This leads me to draw the conclusion that the 5D4 has a stronger AA filter than both the 5D3 AND 5DS...and DPR seems to more or less agree as well:

"It's a similar story if you compare the 5D Mark IV with the higher resolution EOS 5DS R or even the EOS 5DS, whose low pass filter appears to be weaker than the IV's." - DPR 5D4 Review, Image Quality Section. (Emphasis added)

 

[3kramd5]

...
... why? If they aren't planning on upgrading, it's because 4 doesn't offer anything compelling over 3. A convoluted, secret sauce sensor score posted at DXO probably change that for most people.
...

Whilst the "one number" scores get lots of scorn, looking at the component scores is very telling.

For example..

5D4 - Low light: 2995
5D3 - Low light: 2293

5D4 - DR: 13.6
5D3 - DR: 11.7

So if you are a low light shooter or landscape shooter, the 5D4 is going to be better than the 5D3.

Interestingly, the gaps to the A7RII/D810 are somewhat closer now...:

5D4 - DR: 13.6
A7RII - DR: 13.9
D810 - DR: 14.8

5D4 - Low light: 2995
A7RII - Low light: 3434
D810 - Low light: 2853

Those would be covered in the statement from the second half of my post: "two to three performance parameters"

Will this camera provide slightly better photos in edge cases? Sure. In general use? I don't think those numbers are going to bust open barn doors.

The low light score is a 30% improvement ... I think that's going to be noticeable.

How does a 30% increase in that score translate to pictures? Is a picture taken in low light going to be 30% better? I don't qualitatively think my photos taken with A7R2 in low light are 50% better than those taken with my 5D3. In fact, in my experience the A7R2 struggles significantly more with focus in low light than does the 5D3, so many of my alpha photos look qualitatively worse (which is sorta what I was getting at with my statement about undue attention given to a few parameters of a single subassy within a larger end item).


Is it saying that in low light, a photo taken with the 5D4 at ISO3000 will look equally as good as one taken with 5D3 at 2300?

 

[Deleted member 7828]

OK, now including comparisons of the 5D3 and 5DS (again, NOT the R version) to illustrate my above observation. Again, no sharpening or NR in either shot, and two screenshots to show this is the case.

How does a 30% increase in that score translate to pictures? Is a picture taken in low light going to be 30% better? Or is it saying that in low light, a photo taken with the 5D4 at ISO3000 will look equally as good as one taken with 5D3 at 2300?

My guess is it means you can expect roughly a 1/3 stop improvement on high ISO shots (in RAW, with no NR applied)?

 

[3kramd5]

How does a 30% increase in that score translate to pictures? Is a picture taken in low light going to be 30% better? Or is it saying that in low light, a photo taken with the 5D4 at ISO3000 will look equally as good as one taken with 5D3 at 2300?

My guess is it means you can expect roughly a 1/3 stop improvement on high ISO shots (in RAW, with no NR applied)?

Well their definition of the score includes:

An SNR value of 30dB means excellent image quality. Thus low-light ISO is the highest ISO setting for a camera that allows it to achieve an SNR of 30dB while keeping a good dynamic range of 9 EVs and a color depth of 18bits.

So I have 30dB SNR and 9EV DR and 18bit color depth at those ISOs. But again, that doesn't really tell me how it translates to the viewing of a photo.