p.5 #1 · D800 spanked the 5DIII in every DXO category!
Besting the 5D3 in shadow detail at low ISO isn't actually that hard, There's quite a lot of cheap(er) cameras that do this without even breaking a sweat.
This a workably accurate comparison I made with the D7000 and the 5D3. Lens on the 5D3 was the 24-105L (@~70mm and F4.5), and I used the 50/1.8 (@F4.5) on the D7000 - to get the same field of view. Shutterspeed was set manually to 1/100s for both cameras, placing a white area outside the crop I show here just below clipping.
Caveats:
1) I've given the D7000 a 1Ev disadvantage by keeping aperture and shutterspeed the same
2) I've given the 5D3 a 1/3Ev disadvantage by using a lens that has a lower T-stop (the 50mm lens lets through 1/3Ev more light "per aperture stop")
3) I think I scaled the 5D3 down to 16MP, but for this comparison the scale doesnät really matter. The noise is there for both of them anyway.
p.5 #2 · D800 spanked the 5DIII in every DXO category!
theSuede wrote:
Besting the 5D3 in shadow detail at low ISO isn't actually that hard, There's quite a lot of cheap(er) cameras that do this without even breaking a sweat.
Cheaper cameras with something besides a Sony sensor? Are there any cheaper Canon's that can beat the shadow detail of their older FF brother?
p.5 #3 · D800 spanked the 5DIII in every DXO category!
This is no "extreme" lift, I think I used about +2 to +2.5 Ev locally in this part of the image. To give a feeling for the extent of the lift (no other local adjustments made!), here's one of the image frames in its' entirety.
p.5 #4 · D800 spanked the 5DIII in every DXO category!
A few years ago Canon had some really mediocre wides and were justifiably criticized for it. Now you can make an argument they have best wides in the world.
Canon invented high performance AF but somehow fell behind in the last half decade to Nikon. Now the new 61 point is without a doubt the most accurate AF I've ever used.
Hopefully they hear enough criticism about the lack of tonal/dynamic range in their sensors and like before, do something about it.
p.5 #6 · D800 spanked the 5DIII in every DXO category!
joeisayo wrote:
A few years ago Canon had some really mediocre wides and were justifiably criticized for it. Now you can make an argument they have best wides in the world.
Canon invented high performance AF but somehow fell behind in the last half decade to Nikon. Now the new 61 point is without a doubt the most accurate AF I've ever used.
Hopefully they hear enough criticism about the lack of tonal/dynamic range in their sensors and like before, do something about it.
Actually, no you can't make the argument that Canon has the best Wides. They've still got a unexceptional lineup aside from two brilliant TS-E wides. What Canon currently has (once the new 24 and 28 IS ship) is a lineup with only one glaring weakness, the 20/2.8. The rest of the Canon lineup good but not world class (aside from the 24L and 17L TS-E's)
Zeiss currently offers the best lineup of SLR wides in the world. And aside from the 17TS-E Nikon's lineup is generally stronger.
p.5 #8 · D800 spanked the 5DIII in every DXO category!
joeisayo wrote:
A few years ago Canon had some really mediocre wides and were justifiably criticized for it. Now you can make an argument they have best wides in the world.
Well, no, you can't. Only the 17 TS-E is world-beating. Nikon has better wides, as does Zeiss and Leica. Where Canon shines is the tele end but Nikon is very, very close there. And if you don't need AF/VR, Leica's R latest teles are better than both.
Canon invented high performance AF
Making something faster is not an invention. Leica invented AF, AFAIK, and everyone since then has added little twists and made it faster.
Now the new 61 point is without a doubt the most accurate AF I've ever used.
Let's wait for the labs to test it, shall we?
Hopefully they hear enough criticism about the lack of tonal/dynamic range in their sensors and like before, do something about it.
I wouldn't hold my breath. Canon seems to have other priorities than making great DSLR sensors these days, like movies.
p.5 #9 · D800 spanked the 5DIII in every DXO category!
theSuede wrote:
Besting the 5D3 in shadow detail at low ISO isn't actually that hard, There's quite a lot of cheap(er) cameras that do this without even breaking a sweat.
Wow... Numbers are one thing, but it's great to actually see an example. Thanks!
Canon will still sell a zillion of these 5D3s. I'd never advocate switching to Nikon just for one camera, but the argument is still interesting to ponder. Canon users are of course locked in by their lens mount, but there's also an element of simply continuing to buy what you know. I spoke with one photographer the other day who told me I was the first Nikon shooter he's ever met. I was shocked.
Apr 22, 2012 at 03:01 AM
Guest
Guest
p.5 #10 · D800 spanked the 5DIII in every DXO category!
Sorry for my blindness but which is which? And what raw-converter was used?
p.5 #11 · D800 spanked the 5DIII in every DXO category!
That's in ACR6.7. And the 5D3 is the right crop, D7000 on the left. The banding noise is less pronounced in DPP, but DPP also kills all available detail in the darker areas. This means no gain, according to me. If I want to blur the image, I'll do so myself, than you.
Tariq Gibran wrote:
Cheaper cameras with something besides a Sony sensor? Are there any cheaper Canon's that can beat the shadow detail of their older FF brother?
No, but there are cheaper APS-size Canon cameras that can at least equal their larger cousin at base ISO due to their smaller amount of banding noise.
Except for this: I know the Samsung NX200 will at least equal the 5D3 at base ISO. That's a Samsung APS sensor. The wee little Nikon V1/J1 also kills the 5D3 sensor (when you adjust for the almost three-stop disadvantage due to the more than seven times smaller sensor area). That's an Aptina sensor.
I actually brought the V1 out too, since the main reason for the excursion was to test the 5D3 video capabilities. Here's a V1 vs 5D3 still image shadow comparison (same shutterspeed and aperture, same angle of view)
Do note that this means the V1 has an almost 3Ev disadvantage due to sensor/pixel area - it's a 2.7-crop sensor! The difference between the 5D3 and the V1 "should" be almost 3Ev (like setting the 5D3 to 1/640s and ISO1250)
overview:
I can't find 3 Ev's of difference here. And you can clearly see that the 5D3 gains a lot of sharpness due to the 22>10MP downsampling. I'd say that the 5D3 is about 1.5 stops better - when it should be three stops better.
p.5 #12 · D800 spanked the 5DIII in every DXO category!
itai195 wrote:
Wow... Numbers are one thing, but it's great to actually see an example. Thanks!
Canon will still sell a zillion of these 5D3s. I'd never advocate switching to Nikon just for one camera, but the argument is still interesting to ponder. Canon users are of course locked in by their lens mount, but there's also an element of simply continuing to buy what you know. I spoke with one photographer the other day who told me I was the first Nikon shooter he's ever met. I was shocked.
Well, the main difference that gives the D800 an edge only applies to photographers that very often work in good light, low ISO conditions. From ISO1600 and up to 6400 I'd say the cameras are about equal (but the 5D3 gives more manageable files - most of the D800 file "weight" and all of the resolution advantage is lost due to noise at ISO6400. Above ISO6400 the Canon wins, in any practically viable situation. At 12800 you CAN make the D800 file just as good - but it takes a lot more user involvement than the files from the 5D3. Not very practical for everyday work.
p.5 #14 · D800 spanked the 5DIII in every DXO category!
carstenw wrote:
Well, no, you can't. Only the 17 TS-E is world-beating. Nikon has better wides, as does Zeiss and Leica. Where Canon shines is the tele end but Nikon is very, very close there. And if you don't need AF/VR, Leica's R latest teles are better than both.
Making something faster is not an invention. Leica invented AF, AFAIK, and everyone since then has added little twists and made it faster.
Let's wait for the labs to test it, shall we?
I wouldn't hold my breath. Canon seems to have other priorities than making great DSLR sensors these days, like movies.
I have the Canon 17 T&S, Zeiss ZE 21, and 24 1.4L II. I stand by what I said. The 24 1.4L is sharper than the Zeiss at F/2.8 all the way to F/5.6. The Zeiss has nicer colors (subjective) and slightly better borders. The Canon does much better with distortion.
The 17 T&S and 24 II T&S (i rent it) are sharper and have significantly less distortion than the zeiss. I do however prefer the color from the Zeiss 21 over the two Canon.
As to AF accuracy, the new 61 point AF on the 5D3 is more accurate of any Canon camera I've ever owned including the MK IV, 1Ds 3, and Nikon D3S.
p.5 #15 · D800 spanked the 5DIII in every DXO category!
joeisayo wrote:
As to AF accuracy, the new 61 point AF on the 5D3 is more accurate of any Canon camera I've ever owned including the MK IV, 1Ds 3, and Nikon D3S.
Also much better than D800 AF, if you believe in Lloyd Chamber's test (and he is a Nikon guy and he is in live with D800)... Lloyd finds d800 often backfocusing, in particular whit 3D objects, whereas he got terrific results with 5D3 AF and 85/1.2L
p.5 #16 · D800 spanked the 5DIII in every DXO category!
joeisayo wrote:
I have the Canon 17 T&S, Zeiss ZE 21, and 24 1.4L II. I stand by what I said. The 24 1.4L is sharper than the Zeiss at F/2.8 all the way to F/5.6. The Zeiss has nicer colors (subjective) and slightly better borders. The Canon does much better with distortion.
The 17 T&S and 24 II T&S (i rent it) are sharper and have significantly less distortion than the zeiss. I do however prefer the color from the Zeiss 21 over the two Canon.
This goes counter to my experience (apart from the distortion). Since you apparently have the samples, please post them so I can take a look.
p.5 #17 · D800 spanked the 5DIII in every DXO category!
fracas wrote:
Also much better than D800 AF, if you believe in Lloyd Chamber's test (and he is a Nikon guy and he is in live with D800)... Lloyd finds d800 often backfocusing, in particular whit 3D objects, whereas he got terrific results with 5D3 AF and 85/1.2L
The fact that the D800 has issues with the peripheral AF points when focusing f/1.4 lenses has been well documented on the net. According to Ming Thein, Nikon is aware of this issue. Word is that Nikon put D800 production on hold last week and it's reasonable to assume AF was the reason. Care needs to be taken when comparing Chambers' field tests of the two cameras' central AF points as he tested the D800 under different and more diverse conditions and for a greater number of trials in some cases in order to uncover AF misses.
p.5 #20 · D800 spanked the 5DIII in every DXO category!
joeisayo wrote:
Carsten I shoot for a living. My experience with the lenses comes from real world usage and not doing testing.
Wherever your experience comes from, there must be shots which demonstrate what you are saying. If you are trying to convince someone of something, or make a point, it is customary around here to be willing to back it up. A lot of claims vanish in the face of proof... and some don't. Which is this? That is why I am asking you to back up your claims.
....lol.....ok, the 17mm & 24mm TSE lenses...right?...........Canon has 2 great wide angle lenses, not the 'best' in the world.
