High ISO comparo: 5DIII vs. D800

[DavidRiesenberg]

The DPR tests are hardly in good lighting at 3EV. If anything, shooting at ISO 25600 with shutter speed of 1/4000 is hardly a real life situation. In the end the results are very impressive in both tests.

 

[kenraw]

what is the point of constantly comparing these two cameras?
you either have a canon or Nikon and get on with it.
or if your that bothered buy one of each.

lets get out and shoot already.

 

[psolberg]

Well, as expected on 100% comparison 5dIII blows D800 out of the water. It's a t least 1 stop better, but we must not forget that scaled down images just might look pretty much the same, as the D800 has so much detail captured.

Wondering how do they compare scaled down???

What I'm afraid will happen to the d800 is that many wedding photographers won't need it as it is too slow and the files are too big, and many are already shooting Cannon, which has better lenses up it's line (don't argue). Fashion community might embrace it.

could not dissagree more.

The same could have been said of the 5DmkII. The D700 was faster and produced cleaner images and smaller files. did that stop the 5DII? How many weddings were shot with the SLOW 5DmkII? I'll tell you that more that I can count. It seems every wedding photog with a canon set had a 5DmkII and was ok with the speed. As you said, at 100% the 5DIII is clearly better and has ISO ranges that are not possible with the D800, althought the quality of these is questionable.

I suspect that downscaling a 36MP file to 22 will improve a the images about a stop or so just like scaling down the 5DmkII images to 12MP improves them as well. The argument was always made that the 5DmkII gave you the option to scale down when you needed it, and preserve high resolution when you didn't. The same argument now applies to the D800.

Ultimately, I think the biggest gains of the 5DIII are for sports and action photographers that need a lot of speed. Wedding and portrait photographers made livings with "slow" 5DmkIIs so why can't D800 photogs? With the cost per gigabyte always dropping down, 36MP files are not that much of an issue compared to 22MP files. It is just 1/3 more pixels.

So while I totally see that sports photographers will enjoy the new 5DIII over a D800, I don't think the new model has anything over the D800 and in fact, I think the lower MP for a market that is based on a lot of printing and cropping may actually make the D800 the better camera.

 

[K-amps]

Canon 5dMarkus the 3rd. ISO shoots:

http://www.dpreview.com/galleries/reviewsamples/photos/1779787/iso-50?inalbum=canon-eos-5d-mark-iii-low-light-iso-samples

+1: I am no PP expert but I took the ISO 102k image and did some very basic PP to it. Resize/ Nik's define for noise/ contrast adj and got this usable image. It is not something I would consider client worthy but considering I get images like this from my 5Dc at 800-1600 ISO, thats about 6 stops better ISO perf than my current gear ... impressed!

 

[K-amps]

Look at 5:36, that tells me what I needed to know...

Canon 5D Mark III Hands-on First Impression

 

[MikeHunt]

Canon have provided a sample video with the 5D3 coupled to various EF L lenses, some of which was shot at ISO 3200 and some at ISO 6400 (see below). Now compare that to any preview Nikon D800 HD video that is available ???

Canon 5D3 HD Sample Video

 

[cps_user]

So while I totally see that sports photographers will enjoy the new 5DIII over a D800, I don't think the new model has anything over the D800 and in fact, I think the lower MP for a market that is based on a lot of printing and cropping may actually make the D800 the better camera.

as a wedding photographer, I think I still prefer the 5d III (not just because it'll fit my EF lenses ;D) since at this point 36mp raw files would not be very great for my workflow: I would need more memory cards and more memory card swapping, file ingestion for making a same day slide show to be played in the evening would take much longer (I usually have only an hour for making this) and I work off the NAS so larger files would slow me down a bit there.

I don't crop a lot and what's left is almost every time more than large enough for making 11x14 inch wedding albums (largest size I do) and nice enlargements for clients.

in the future, yes maybe larger files would be great, but at this point as a wedding photog, I'm quite happy with this.

 

[RafaPolit]

...Ultimately, I think the biggest gains of the 5DIII are for sports and action photographers that need a lot of speed. Wedding and portrait photographers made livings with "slow" 5DmkIIs so why can't D800 photogs? With the cost per gigabyte always dropping down, 36MP files are not that much of an issue compared to 22MP files. It is just 1/3 more pixels...

I don't think this encompasses the whole picture... It's not the storage of the files, either on camera or inside a hard drive, that is the issue but actually working with those files. At 18MP, when you start adding layers of adjustments and corrections in photoshop, for instance, the resources on the computer start to see themselves heavily demanded. At 22MP things get much worse, but at 36MP, unless you have heaps and heaps of RAM, you will be struggling with any change you want made, and probably could go for a cup of coffee when applying a heavy filter such as lens blur or radial blur.

Lightroom, also RAM hungry, will really struggle loading and developing 36MP images to a degree I would consider beyond reasonable limits.

Noise wise, I'm impressed with what I see on the 5DmkIII, and it's certainly more than 1 stop better than the 5DII under any circumstance. But it's too early to draw final conclusions as very little real samples are out there.

Best regards,
Rafa.

 

[V8Beast]

In light of these D800 raws that I dug up, I might have to retract my worthless opinion. There's a link to download the raws and jpegs in a zip file.

http://www.arekgmurczyk.pl/

In comparing jpegs, the 5DIII files look like they have a clear edge, but compared to this set of raw files on Imaging Resource that have already been posted, both cameras appear equal in the noise department:

http://www.imaging-resource.com/PRODS/canon-5d-mkiii/canon-5d-mkiiiTHMB.HTM

It's an apples to oranges comparison since they aren't images of the same subject/scene, but I have to admit that this is a surprisingly good showing by the D800. I was expecting the 5DIII to beat it by a comfortable margin due to the disparity in megapixel count.

In comparing the D800's raws to its jpegs, the noise is pretty similar. I think this validates people's suspicions that Canon is apply some heavy in-camera noise reduction in its jpeg files.

 

[KeithR]

I was expecting the 5DIII to beat it by a comfortable margin due to the disparity in megapixel count.

...Which graphically demonstrates the complete nonsense of that line of thinking.

Just look at this comparison between the D700 and D800 - the 36 mp D800 is all over the 12 mp D700 at the image level, and that's after conversion in a Raw Therapee, which is well known not to be great at high ISOs, and using cobbled-together profiling for the D800...

 

[SpartanWarrior]

I think both will be great cameras, but I got to say dpreview is filled with fanboy galore they are already making it sound like the D800 is better than D4 in high ISO, they wish, I will also say the 5D III comes close to D4 at high ISO but better detail from 5D III.

 

[V8Beast]

...Which graphically demonstrates the complete nonsense of that line of thinking.

Allow me to elaborate. When looking at a 100 percent crop of a 36 mp image compared to a 22 mp image, I'd expect the odds to be stacked in favor of the 22 mp files unless the 36 mp image was downsized accordingly. In viewing the raws from both cameras at 100 percent, I can't see much a difference in noise.

I'm not complaining about the 5DIII's megapixel count, as I think 22 is plenty, but considering that Canon didn't increase it much at all over the MKII, I was hopeful of a substantial improvement in ISO. It looks like what we got was a substantial increase in in-camera noise reduction. I'm not going to complain about that either, but Canon's advertised 2 stop ISO advantage of the MKIII over the MKII is obviously for jpegs, now raws.

 

[PhilDrinkwater]

It's an apples to oranges comparison since they aren't images of the same subject/scene, but I have to admit that this is a surprisingly good showing by the D800. I was expecting the 5DIII to beat it by a comfortable margin due to the disparity in megapixel count.

I don't know the tech side of this, but I wasn't personally expecting this. The 5dii was only 1/3rd stop behind the d700 (although Nikon fanboys claimed a lot more) and it had twice as many MP.

When taking resolution out of the equation (ie. resizing to the lower resolution of the two) it seems noise is about the same. MP doesn't seem to affect noise all that much..

 

[Kernuak]

...Which graphically demonstrates the complete nonsense of that line of thinking.

Allow me to elaborate. When looking at a 100 percent crop of a 36 mp image compared to a 22 mp image, I'd expect the odds to be stacked in favor of the 22 mp files unless the 36 mp image was downsized accordingly. In viewing the raws from both cameras at 100 percent, I can't see much a difference in noise.

I'm not complaining about the 5DIII's megapixel count, as I think 22 is plenty, but considering that Canon didn't increase it much at all over the MKII, I was hopeful of a substantial improvement in ISO. It looks like what we got was a substantial increase in in-camera noise reduction. I'm not going to complain about that either, but Canon's advertised 2 stop ISO advantage of the MKIII over the MKII is obviously for jpegs, now raws.

From my quick tests last week, the impression I got was that the MkIII has around a 1.5 stop advantage over the MkII with the noise reduction switched off, but I didn't have any images to take home with me to look properly. I had to compare the images on the back of the screens, as the card slots were taped up.

 

[SpartanWarrior]

I don't know the tech side of this, but I wasn't personally expecting this. The 5dii was only 1/3rd stop behind the d700 (although Nikon fanboys claimed a lot more) and it had twice as many MP.

When taking resolution out of the equation (ie. resizing to the lower resolution of the two) it seems noise is about the same. MP doesn't seem to affect noise all that much..

They claim a lot of false stuff without proof, 5D II and D700 maybe not even 1/3 stop but less, here is my 5D II and my friends D700 that I tested and I have many more with same results,
shot in RAW same settings, cropped 100% and converted to Jpeg nothing else
5D II ISO 12800

D700 ISO 12800

5D II Left 12800

D700 12800

and here are 5D II mRAW ISO 12800

 

[Actionpix]

5DIII and D800 are completely different cameras. So why would you try to compare them? As long digital cameras are being sold there always has been the trade of between amount of pixels and amount of noise. Also you can not expect large files and high fps together. All this is getting better with every camera but the difference remains. Just get what you need most. (Being able to make the pick in one camera system would be nice though.)

 

[Cali_PH]

Canon have provided a sample video with the 5D3 coupled to various EF L lenses, some of which was shot at ISO 3200 and some at ISO 6400 (see below). Now compare that to any preview Nikon D800 HD video that is available ???

Funny timing, I was just reading about the Nikon D800 promo video debacle. Nikon's apparently admitted they used footage from at least a couple of people without permission for their video, one of which said he shot with a Canon 5DmkII (unless this is a fake story?). In the comments section I see people pointing out more clips taken from non-D800 sources. My apologies if this is old news for most.

http://wideopencamera.com/cameras/nikon-steals-canon-5d-mk-ii-footage-for-d800-promo-video/

 

[XanuFoto]

Good photographers will be able to work with either bodies and produce stellar images. Everyone else with be just comparing specs and complaining. Both cameras are good tools if you know what you are doing.

 

[LetTheRightLensIn]

5DIII and D800 are completely different cameras. So why would you try to compare them? As long digital cameras are being sold there always has been the trade of between amount of pixels and amount of noise.

But the point here is there hasn't always been that much of a trade. It now looks like the crazy MP D800 will not only have a lot more dynamic range at low ISO but do just about nearly as well at high ISO. And it's actually often been like that. Within anything not going to crazy extremes, more MP has usually offered more pluses than often trivially minor losses in high iso performance.

Also you can not expect large files and high fps together.

That is why the 1.3x crop and 1.5 crop modes of the Nikons are cool. 36MP 4fps, 25MP 5fps, 16MP 6fps.

Canon is getting close to large files and high fps at 22 and 6 (if it had been the 6.9-7.5 of the rumors it would've have fully attained that goal).

 

[justsomedude]

But the point here is there hasn't always been that much of a trade. It now looks like the crazy MP D800 will not only have a lot more dynamic range at low ISO but do just about nearly as well at high ISO. And it's actually often been like that. Within anything not going to crazy extremes, more MP has usually offered more pluses than often trivially minor losses in high iso performance.

+1

And, I really really really hate to say this, but it looks like Canon may (and will?) lose dominance to competition from the d800. I will be totally honest, I have been a 100% die-hard Canon fanboy through and through since my 40D days - but these sample images coming out are making me think twice. Toss in my own disasters with 7D AF (this pretty much sums up my 7D experience) and the 580exII PocketWizard fry-out issue I also had, and Canon's reputation in my mind has sustained some chinks in its armor.

I can't help but think that others around the CR forums are feeling the same way. When other die-hards start saying things like "well - you can't compare them [the d800 and 5D3]," "they're two different cameras," and "well - I'm sure both bodies will take great shots!" ...you know there is trouble in paradise. It's like we're already setting up excuses for the lack of improvements that one would expect in the 5D3 after 4 years of R&D. It seems everyone is trying to deny the fact that the d800 shots look, well - let's face it - f*cking incredible. And nothing trickling out from MKIII land is really that mind blowing.

It's discouraging. It's heart wrenching. It's making me wonder WTF Canon has been doing the last four years.

It's like they had a board meeting and said, "let's address the forum kids bitching about AF - so toss in the 1DX system - and maybe improve noise a smidge" - and called it a frickin' day (or 4 years). Don't get me wrong, the AF and stop improvements will be very welcomed features to most 5D users. But when Nikon has clearly been hard at work with Sony researching sensor tech - you gotta show a little more after 48 months of R&D. Seriously - how can anyone deny those d800 shots? They're just bad ass. ...and quite frankly, they make me want to cry. I WANT those images to be from the 5D. :'(

To boot - that $500 cost premium over Nikon isn't helping Canon much. Sure - they'll get stellar sales on the 5D3 (in full disclosure - I'm already one of the buyers) - but so far every single buyer is living off of 5D2 experiences, and their own hopes and dreams of what the 5D3 *should* be. If that doesn't materialize, you can bet your bottom dollar every single one of us will be looking at the d800 shooter in the corner wondering, "what if?"

I've never considered abandoning all of my Canon gear for Nikon. And I hate to sound trollish - and expect the flames to come - but the d800 samples are the first images that have made me seriously consider "the switch". I'll wait for my hands on with the 5D3 and real world reviews before I make such a huge business decision - but the bug is already there in the back of my mind.

And just the fact that the bug is there is what's pissing me off the most. I just wish Canon had done more this go 'round.

:-/