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

[jrista]

Welcome, Canon, to the year 2012!! ;D

Well, they may be four years late, but it looks like they have finally caught up to the rest of the pack. It looks like the D810 still has the edge with their ISO 64, however when you bump the D810 up to ISO 100 on DPR's comparison tool, it's a pretty level playing field...D810 still has an edge, but it's color is off:

I also prefer the more neutral blacks of the 5D IV, 1DX II and D5500 over the reddish hue of the D810. Good to finally be seeing Canon's DR up to modern standards. The 5D IV looks better to my eyes than even the 1DX II.

 

[StudentOfLight]

Last week I put together a speculative chart (http://www.canonrumors.com/forum/index.php?topic=30675.msg619797#msg619797) based on the beta images from DPR.

I think the DXO measures line up pretty well, so I'm comfortable with saying that the 5D-IV has delivered on my expectations in terms of image quality.

This too is ignoring the potential extra highlight recovery from dual-pixel-RAW, which is discussed on RAWgigger: http://www.rawdigger.com/howtouse/canon-dual-pixel-mode-highlights-are-there

I'd guess there is at least an extra 0.5Ev available in extreme highlights when shooting dpRAW and it might just be a software upgrade away...

 

[StudentOfLight]

Welcome, Canon, to the year 2012!! ;D

Well, they may be four years late, but it looks like they have finally caught up to the rest of the pack. It looks like the D810 still has the edge with their ISO 64, however when you bump the D810 up to ISO 100 on DPR's comparison tool, it's a pretty level playing field...D810 still has an edge, but it's color is off:

I also prefer the more neutral blacks of the 5D IV, 1DX II and D5500 over the reddish hue of the D810. Good to finally be seeing Canon's DR up to modern standards. The 5D IV looks better to my eyes than even the 1DX II.

I'm very happy with the color performance. So happy Canon fixed the red channel. I look forward to seeing astro images from the 5D-IV.

 

[willhuff.net]

What a weird place for the 5DS to be in. It was designed to be the top image quality camera for landscapes and studio work, but now the mk4 has better dynamic range. I wanted to get the 5DS but the sensor wasn't much improved from the 5Dmk3 so I held off. Now I have to decide if 20mpx or more dynamic range is more important. I wonder how long it will take for the 5DSmk2 to come out.

 

[3kramd5]

What a weird place for the 5DS to be in. It was designed to be the top image quality camera for landscapes and studio work, but now the mk4 has better dynamic range. I wanted to get the 5DS but the sensor wasn't much improved from the 5Dmk3 so I held off. Now I have to decide if 20mpx or more dynamic range is more important. I wonder how long it will take for the 5DSmk2 to come out.

Hence why some of us think that there will be a 5Ds update "sooner" rather than later - i.e. it would be on the cards for next year.

I don't think it's a weird place. If you want uber high resolution, you have the S/R. If you want more flexibility with your files (arguably more important in the low light/fast shutter scenarios), you go 5D4.

Gentleman's bet there will be no new 5DS/R in 2017? They will almost undoubtedly update it eventually, but next year? I think you're dreaming.

 

[ahsanford]

Gentleman's bet there will be no new 5DS/R in 2017? They will almost undoubtedly update it eventually, but next year? I think you're dreaming.

Agree. There's as much a chance for a 5DS2 in 2017 as there is for a 7D3 in 2017. (Which is to say: no chance at all)

Next year is (all but certainly) a 6D2 in the front half of the year and then a 'who knows' window in the schedule where Canon will innovate or refresh in unlikely ways: an SL2, a dedicated astro rig, a fixed lens FF mirrorless rig, perhaps a 50 MP 1D rig a la the 1Ds3, etc.

I just don't see them violating their long 4+ year routine with 5D bodies, and certainly not just to chase one stop of base ISO DR over the current model (the 5DS sits somewhere between the 5D3 and 5D4 on DR if I recall). The 5D4 would have to have unleashed some face-melting game changer of a sensor to warrant a 'trickle-down tech refresh' after only two years. I just don't see it happening.

- A

 

[3kramd5]

What a weird place for the 5DS to be in. It was designed to be the top image quality camera for landscapes and studio work, but now the mk4 has better dynamic range. I wanted to get the 5DS but the sensor wasn't much improved from the 5Dmk3 so I held off. Now I have to decide if 20mpx or more dynamic range is more important. I wonder how long it will take for the 5DSmk2 to come out.

Hence why some of us think that there will be a 5Ds update "sooner" rather than later - i.e. it would be on the cards for next year.

I don't think it's a weird place. If you want uber high resolution, you have the S/R. If you want more flexibility with your files (arguably more important in the low light/fast shutter scenarios), you go 5D4.

Gentleman's bet there will be no new 5DS/R in 2017? They will almost undoubtedly update it eventually, but next year? I think you're dreaming.

There was a rumor earlier in the year that it would be updated this year ...

So it's a bet!

 

[jrista]

What a weird place for the 5DS to be in. It was designed to be the top image quality camera for landscapes and studio work, but now the mk4 has better dynamic range. I wanted to get the 5DS but the sensor wasn't much improved from the 5Dmk3 so I held off. Now I have to decide if 20mpx or more dynamic range is more important. I wonder how long it will take for the 5DSmk2 to come out.

If you are controlling your light source, such as studio or portrait photography, then the 5Ds is still probably king. All that resolution, you could downsample by 20mp and average out the noise, and recover some DR if you really needed to. But I think the more important thing with that camera is the stellar resolution for controlled lighting scenes.

 

[ahsanford]

...
I just don't see them violating their long 4+ year routine with 5D bodies, and certainly not just to chase one stop of base ISO DR over the current model (the 5DS sits somewhere between the 5D3 and 5D4 on DR if I recall).
...

Sure if you are selective about "routine" as it applies to 5D3 -> 5D4.

Inside Canon I doubt very much that they have such "routines."

I think they do have predictable schedules, actually:

http://www.northlight-images.co.uk/cameras/Canon_rumours.html#canon_age_chart

The 1DX (recently) and 5D# line (throughout) has a 4-4.5 year refresh. All money would also be on a 4 year 6D refresh cycle, as it's now a cascade of 1DX# --> 5D# --> 6D# in roughly six month chunks just like the last time (with the 1DX --> 5D3 --> 6D).

I see the 5DS# probably following suit with the 5D# and going every four, but Nikon seems to be at 36 MP and isn't pushing Canon to climb higher just yet. Perhaps it's a longer refresh cycle for that. We'll see.

The 7D line had one 5 year refresh with a mid-cycle firmware/functionality upgrade, and they are doing just that again with the 7D2. I peg that as a 5 year cycle product.

The XXD line is the somewhat inconsistent animal, but 3 years seems about right.

- A

 

[Maximilian]

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

Nope!

Because I still give a s*** on DXO scores.
I am glad to see that Canon does improve their sensor tech and that even DXO recognizes it
But I can sleep quite well because real live differences don't justify that price explosion of 27% MRSP in Germany.
I can calmly wait for even better sensors or better lenses.
Canon won't see my money for a new body too soon.

 

[Aglet]

Welcome, Canon, to the year 2012!! ;D

Well, they may be four years late, but it looks like they have finally caught up to the rest of the pack. It looks like the D810 still has the edge with their ISO 64, however when you bump the D810 up to ISO 100 on DPR's comparison tool, it's a pretty level playing field...D810 still has an edge, but it's color is off:

I also prefer the more neutral blacks of the 5D IV, 1DX II and D5500 over the reddish hue of the D810. Good to finally be seeing Canon's DR up to modern standards. The 5D IV looks better to my eyes than even the 1DX II.

I'm very happy with the color performance. So happy Canon fixed the red channel. I look forward to seeing astro images from the 5D-IV.

the noise in those pushed dark areas still looks a bit vertical-bandy in the 5d4 but not nearly as much as the older tech bodies.
Still happier with my ABC bodies and non-Canon glass for the stuff I shoot but glad the big C gang finally has another seriously improved imaging machine out there.

 

[ahsanford]

...
The 1DX (recently) and 5D# line (throughout) has a 4-4.5 year refresh. All money would also be on a 4 year 6D refresh cycle, as it's now a cascade of 1DX# --> 5D# --> 6D# in roughly six month chunks just like the last time (with the 1DX --> 5D3 --> 6D).
...

Why are you ignoring all the history?

5D -> 5D2 = 3 years
5D2 -> 5D3 = 3.5 years (and rumors of natural disaster making the release later than it should have been)
5D3 -> 5D4 = 4.5 years

Seriously. Look at the entire graph.

There are exceptions, sure, but don't miss the bigger picture. Canon lets pricier products stay on the market for a number of years to maximize profits. Natural disasters and slight reshuffling of the schedule due to other releases are going to happen that alter the timing a bit. It's not precise like a train schedule -- it's just a rough timing rule of thumb.

I also see them hitting a 'cross-brand rhythm' in which the FF product lines work a roughly 4 year schedule and each get their 6 months in the spotlight (leaks / announcement / marketing push / pre-orders / initial orders ship / reviews, etc.) for the 1DX, 5D and 6D lines. Obviously, three rigs X 6 months = only 18 months of a 4 year cycle, so yes, there is room for other releases. Those gaps in the calendar are when we get the fairly unpredictable updates (XXD), altogether new product lines (5DS, EOS M), major L lens releases, video products, and odd flavors of Canon innovation:

• SL1
• XC10
• The first EOS M before it became a steady line
• Astro cameras

...and it ends up being a pretty busy 4 years. I am not setting a clock on that 4 years, but you get the idea.

Could they expedite the 5DS to drop it in behind the 6D release in the FF train of annoucements? It's certainly possible, but it seems both out of the ordinary for Canon to do with a FF line and it seems like Canon isn't (yet) being pressured in the ultra high resolution space. Nikon could drop a 50MP+ rig with a better sensor and push Canon to change their plans with the 5DS2, but my money's on Canon being Canon and sticking to their plan.

- A

 

[Deleted member 7828]

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).

Yeah, I'm beginning to sense that as well. A major breakthrough will be necessary apparently. Still, 1 stop may be workable. Curious to see the direction they take the next 6D - maybe that camera becomes the low-light master, staying in the ~22-24MP range?

 

[ahsanford]

Yeah, I'm beginning to sense that as well. A major breakthrough will be necessary apparently. Still, 1 stop may be workable. Curious to see the direction they take the next 6D - maybe that camera becomes the low-light master, staying in the ~22-24MP range?

Canon is not set up like Sony with a7S line being the lowest res + best for video + best low light. The 6D is supposed to be a consumer focused / affordable FF option a la the Nikon D610. It's not supposed to be the best at anything (other than FF value).

So I see the 6D outperforming the 5D3 sensor as a bit of a one-off that Canon will not repeat; people who sunk $3499 into a 5D3 were decidedly non-plused to see a 6D at $2099 (at launch) with a slightly better sensor and -3 EV center AF. So Canon needs the 5D4 and 5DS to completely outclass the the 6D2. Or if the 6D2 ends up anywhere near the 5D4 for sensor performance, Canon will punish the product elsewhere to protect 5D sales -- likely by nerfing key features on the 6D2 to dissuade those with 5D4 money from buying a 6D2 instead.

The betting man would expect a flip screen + DPAF with the 6D2, but I wouldn't be surprised to see Canon limit the 6D2 with a 1/4000 max shutter, no 4K, fewer AF points than the 5D line, lower burst rate, lock it out from having anti-flicker or Dual-Pixel RAW, etc.

- A

 

[Jack Douglas]

Yeah, I'm beginning to sense that as well. A major breakthrough will be necessary apparently. Still, 1 stop may be workable. Curious to see the direction they take the next 6D - maybe that camera becomes the low-light master, staying in the ~22-24MP range?

Canon is not set up like Sony with a7S line being the lowest res + best for video + best low light. The 6D is supposed to be a consumer focused / affordable FF option a la the Nikon D610. It's not supposed to be the best at anything (other than FF value).

So I see the 6D outperforming the 5D3 sensor as a bit of a one-off that Canon will not repeat; people who sunk $3499 into a 5D3 were decidedly non-plused to see a 6D at $2099 (at launch) with a slightly better sensor and -3 EV center AF. So Canon needs the 5D4 and 5DS to completely outclass the the 6D2. Or if the 6D2 ends up anywhere near the 5D4 for sensor performance, Canon will punish the product elsewhere to protect 5D sales -- likely by nerfing key features on the 6D2 to dissuade those with 5D4 money from buying a 6D2 instead.

The betting man would expect a flip screen + DPAF with the 6D2, but I wouldn't be surprised to see Canon limit the 6D2 with a 1/4000 max shutter, no 4K, fewer AF points than the 5D line, lower burst rate, lock it out from having anti-flicker or Dual-Pixel RAW, etc.

- A

When all the 6D bashing was going on I bought it because it seemed better value for a camera that would get upgraded to ....... looks like the 5D4 over the 1DX II given I'm not that committed to doing video and essentially no sports.

The 6D delivered close to 1DX IQ based on my friend and I shooting together. I knew nothing compared to now and so this feels really good because I feel much more confident in making this upgrade.

Provided these early reports continue to be positive, I'm committed to waiting until spring hoping for a price drop and then it'll be 5D4. Couple this with my 400 DO II purchase and I'm almost giddy.

Only negative has been the few weeks of enduring all the Canon bashing by whiners.

Jack

 

[ahsanford]

When all the 6D bashing was going on I bought it because it seemed better value for a camera that would get upgraded to ....... looks like the 5D4 over the 1DX II given I'm not that committed to doing video and essentially no sports.

Ah, that's the key distinction, though. With stills, the 1DX did a jillion things the 6D could not (burst, AF, metering, tracking, high ISO, etc.). But for a stills-only shooter, one could argue that other than AF (in fairness, a really big deal), having a 1/8000 shutter and +1.5 fps, a 5D3 was not that much better than a 6D, certainly not to the level a $3499 vs. $2099 rig ought to have been.

Further, the 5d line was the cheapest 'step-up-to a new FF rig' option available to crop shooters before the 6D came out, so some crop-upgrading stills-only folks in 2012 had painful buyers' remorse that they had to purchase an overkill $3499 professional rig (with video features they'd never use) to get tickets to the FF party and then ostensibly have the camera they probably would have purchased instead -- a solid starter FF rig -- show up a good 8 months later.

That's why 1DX people didn't give a damn about the 6D, while the 5D3 people were saying "Hey, wait a minute!" once it was announced, once reviews started rolling in, etc. The 6D is not better than the 5D3 on aggregate (at all), but if you don't shoot video or high fps / servo AF work, it's a far more compelling value.

- A

 

[PhotoCat]

Where have you been Canon for the past 6+ years? Glad to see you again! Yay!
Thumbs up on the new Canon sensors!

 

[3kramd5]

Where have you been Canon for the past 6+ years?

Developing, producing, and selling product.

 

[ejenner]

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

Exactly what I was thinking - about myself.

Looks like my 5DIII just turned into a POS . I'm happy with it and 1/2 stop DR wouldn't have swayed me, but I know what will happen now. I will spend 18 months trying to convince myself I don't need it to finally break down and buy it when I could have been using it for all that time. Sort of did that with the 5DII, then the 5DIII came out and did the same.

Maybe this time will be different. 5DIII IQ is good really, as is the AF which was my biggest issue with the 5DII.

But... if DR and high ISO is really that much better........

Bollocks, here we go again. :'(

 

[jrista]

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

Exactly what I was thinking - about myself.

Looks like my 5DIII just turned into a POS . I'm happy with it and 1/2 stop DR wouldn't have swayed me, but I know what will happen now. I will spend 18 months trying to convince myself I don't need it to finally break down and buy it when I could have been using it for all that time. Sort of did that with the 5DII, then the 5DIII came out and did the same.

Maybe this time will be different. 5DIII IQ is good really, as is the AF which was my biggest issue with the 5DII.

But... if DR and high ISO is really that much better........

Bollocks, here we go again. :'(

You might be better off waiting anyway. The current price is, IMHO, ridiculous. I purchased my 5D III a couple years after it's initial release for $2650. The price will hold until the initial furor about it fades, then it will start dropping. By 18 months, you should be able to find it for more reasonable prices.

I am quite happy Canon finally got into the DR game, a little sad that it looks like their time horizons for each new product release get larger and larger (we went from ~3 years between releases to 4.5-5 years between releases...), while the competitions time horizons have shrunk amidst the competitive environment. So, by the time Canon releases a 5D V, we might very well be looking at 16-bit cameras with 15 stops of DR from Sony, Nikon, etc.

In the higher end sensor space, read noise levels have dropped to around 1-1.5e-, in a few cases less, with forecasts for 0.75e- down to 0.5e- read noise on the horizon. These read noise levels are paired with 30ke- to 45ke- full well capacities. That is around 15 stops of DR now, with the possibility of 15.9 stops in the future with sub-electron read noise levels. I don't expect to see such sensors in consumer grade DSLRs for a while, but they are finding their way into other kinds of devices, including the kind of scientific cameras we use for astrophotography.

At Canon's telescoping snails pace, it will be 2021, maybe 2022 before we see the 5D V. Big question is...will it still have 13.6 stops of DR, will it still have some banding, will their cameras still exhibit amp glow, etc...?