Canon EOS 5D Mark III low-light ISO series samples

Depends on your camera condition, and it's clicks usage, you should be able to sell it for around $1500, given some rumors the Mark ii price will further drop around the 2K range. Also, if you shoot in RAW, you may not get as much low light boost as you think, as the boost is for JPG only done via Digiic 5+ processor's internal noise filter.

Mar 6, 2012permalink

lensberg

You actually believe that the D800 is the most exciting DSLR of 2012...? The D800 is only revolutionary as far as resolution goes... everything else about it is borderline mediocre...

The 5D Mark III is almost guaranteed to outperform the D800 on every front... High ISO, autofocus, FPS, per pixel quality, video frame / bit rate... Uncompressed video is more of a gimmick on the D800... at least from the sample footage i've seen... recording at roughly 24 Mbit/s it is not satisfactory at all...

Nikon is going to rue the day it chose quantity over quality... Even the sample images provided at Nikons website are average at best... The days when Canon did it and got away with it successfully are long gone... Consumers are a lot more quality conscious now...

Whether its sporting events, weddings, concerts / parties or studio work... the 5D Mark III will be far more preferable for its IQ and overall handling...

Mar 6, 2012permalink

Zeis

Most of the people complaining here are terrible photographers. Hacks, hobbiest, debbie digital & uncle bob and some sad, sad pro's who can't take a decent images to save their life.

Who cares what the technical abilities of these new Nikon and Canon cameras are if you can't use them. Theses camera are better than the last round of new releases and people back in the day (so so long ago) and people were creating amazing images with such outdated gear.

My suggestion is to learn how to push the camera you already have and not worry about all this bull people are slinging here in these forums.

Mar 5, 2012permalink

chopsteeks

Yup, I am perhaps the world's poster boy for bad photographers. I need all the technical I need to make my images good enough to be able to put on a frame and hang at my apartment's wall.

Is there really anything wrong with this ?

Mar 5, 2012permalink

Except the point here is to justify if the Mark iii is worth spending another 3.5K for someone who's already using the Mark ii buddy! The marketing is misleading until someone realize it's not true 2 stop low light performance improvement but a hardware noise filter gimmick.

Mar 6, 2012permalink

I'm pretty sure discussing sample images is done so that people can compare to what they are currently using, to highlight what may be good, not so good, etc. Not to insult your peers because you happen to disagree with them. But whatever, stay classy.

Mar 6, 2012permalink

Chayus

Your comment can be applied to literally every i have read on this website

Mar 6, 2012permalink

Paullubbock

hey, hey, I resemble that remark!!! sir!

Mar 6, 2012permalink

What's the point of low-light improvement on JPGs? Most pros shoot in Raw. This is basically faked low-light improvements, as in theory, it's the same as applying a software noise reduction filter in PS, or Lightroom, except it's done internally by the Digit 5+ processor.

I want to see true low-light quality improvement by the Mark iii's sensor instead of hardware based noise reduction in JPGs only!

Mar 5, 2012permalink

Nick Criggar

I agree with your argument for most types of photography, but let's say you're a stringer for the local paper, and you have to get the shot to get paid versus being a studio or landscape professional. In this case, the Copy Chief doesn't care whether you've shot RAW or JPG. Your photo isn't going in a frame on some Newlyweds' bedroom wall at 12x18 inches, it's going on the center truck of the sports section at 4 columns, and it's for the local football game which ends 90 minutes before press time. In this scenario, you need that ISO 102800 performance (especially because you're shooting at very high shutter speeds), and you want the photo to be processed via in camera settings before you even give the files to your pre-press department. That's the value of low-light improvement on JPGs. Where deadlines matter, and other in-camera physics cannot be manipulated (e.g. shutter speed and f-stop) great low-light ISO performance and automatic image manipulation are paramount.

Mar 5, 2012permalink

Nick: I agree, you've listed a very good scenario where JPG is appropriate for certain professional jobs, and having hardware noise filter is beneficial. The issue I'm having is this need is for a very small percentage of the mass market, where as this feature is flaunted as if it's for the masses: a 2-stop low light performance boost, which it is truly not. Applying filter to reduce noise is not a performance boost, it's smart touch up.

Mar 6, 2012permalink

Frank Palm

Truly disappointed on this upgrade, in my perspective 5D Mark III is old even before it´s on market. AF should have been on mark II, resolution is the same as mark II and the viewfinder just sucks. Way to go Nikon, Sony and Panasonic.

Mar 5, 2012permalink

Bodhi Dharma Zen

full of Canon fanboys, this thread is

Mar 5, 2012permalink

arthurzhang

are you kiding me? This one abviousely is a nikon fanboy...

Mar 6, 2012permalink

SkyTamer

Just imagine how great the 7D II will be, for a fraction of the 5D MIII. Or at least that is what I keep telling myself every time I see the 3.5K price tag.

Mar 5, 2012permalink

d3xmeister

What a bunch of spoiled whiners. Louie CK was right when he said ,,today everything is fantastic and nobody's happy,,

Mar 4, 2012permalink

healer81

He also said, what contribution did you make to society you nothing. Companies like Canon, Nikon have been serving people for decades and capturing their most precious moments.

Mar 5, 2012permalink

PicOne

Couple clarifications desired.
EV3 equates to 8 seconds at F/8. Was lighting adjusted so that all exposures at each ISO were eg. 8 seconds at f/8, or by ISO 25600 shot was at eg. 1/30th at f/8 using the same lighting as provided for ISO100 shot? I'm assuming aperture remained constant?

Also.. to see noise and reduction effects, IMO kinda strange to choose a lot of "plastic" looking painted objects where 'watercolor' effects are already part of the subject. Only the stuffed animal has detail, but my opinion, this mostly falls outside the DOF of the scene.

Mar 4, 2012permalink

Andy Westlake

You could just look at the exposure data that's provided in the gallery for every single shot.

As for the suitability of the scene - it may not be 100% ideal, but given 30 minutes with the camera to put something together, it's what was on hand.

Mar 4, 2012permalink

ICPix

expensive yes
MP count ok
I think Ill shoot my next wedding with the 40 MP Nokia.

Seriously, I might not have the cheese to buy this body, or the D800 for that matter, but I can rent both and enjoy the quality. Personally Id rather spend 3k taking courses and having more original artistic thoughts in my own work.

Mar 4, 2012permalink

Wubslin

@JakeB - no, I use a Canon 5D. Not sure what you're on about.

Mar 4, 2012permalink

EdinburghDoug

I applaud Canon for not pushing the pixel density and concentrating on IQ. I am happy with my MkII IQ. More often than not I have the central AF point active and simply recompose before shooting, so 61 AF points are little use to me except in some fast moving scenarios and in this case I'd probably want more than 6FPS. The inclusion of an SD slot, that allows Eye-fi is welcome. But the extra 1.2MP make little difference and the increased ISO range, moderate shift to 6 FPS and marginally higher res of the rear LCD are not enough to tempt me. I'll stick with the MkII. I'll continue with the 7D for more dynamic shoots and enjoy its telephoto benefit and 8 FPS.

Mar 4, 2012permalink

qwertyasdf

sensible

Mar 5, 2012permalink

santhan

is good

Mar 4, 2012permalink

Wubslin

Yet another winner for Canon! I wonder where this leaves Nikon? Maybe they should just break up the company and give the money back to the shareholders.

Mar 4, 2012permalink

JakeB

Let me guess -- you use a P&S, right, sport?

Mar 4, 2012permalink

Jacques Cornell

I like what Canon's done to make the 5DIII a better low-light and action camera, but at $2,200, the D700 is still very compelling unless you really need more resolution, which, in my event photography business, I don't.

Mar 4, 2012permalink

AbrasiveReducer

It leaves Nikon making cameras and Canon making cameras. And people who use Nikon cameras will continue using Nikon cameras. And people who use Canon cameras will keep on using Canon cameras. This "battle" exists only in your mind. And, as this stuff gets more and more expensive, users are even more locked in to one brand or the other.

Mar 4, 2012*permalink

Marco 2k7

nikon's shareholders are canon's shareholders....and vice versa. they just aim to cover the most of camera sellings owning two companies

Mar 5, 2012permalink

Where this leaves Nikon is with the D800, the most talked about and exciting DSLR of 2012.

Mar 6, 2012permalink

cptrios

The high ISO shots are impressive enough, but we really need to see RAW conversions. The JPEG engine looks incredibly plasticy.

Mar 3, 2012permalink

Jacques Cornell

I'm very pleased with the direction Canon has taken with the 5dIII. I've been wishing for 1 Series AF and clean ISO 3200 performance in a compact body, and Canon seems to have finally delivered. As an event photographer who routinely shoots in low light, I was not impressed by the 5DII's ancient AF module or the 7D's high-ISO noise performance. I've long felt that the D700 was a nearly ideal do-everything camera for most pros - I know many who use it - and now it seems that Canon finally has a worthy challenger. Now if they'd just make a 7D equivalent with fewer pixels and cleaner files...

Mar 3, 2012permalink

Tape5

Photography can also be a sickness. When I talk about photography I am thinking about Blakemore's tulips, not a sweaty girl hitting a baseball bat at high speed. I understand high ISOs believe me. Photography will always be pulled between information and art. I am only saying that most photographers are knowingly or otherwise in need of artistic expression in their lives and are not earning their family income by taking 9 fps indoor shots of fast moving things. And if anyone takes a picture with ISO 12000, I love to see that too. We all have different priorities and different mortgages.

Mar 3, 2012permalink

Smartypants

Tape5,
Some of us here pay some or all of our mortgage through this “sickness” you claim to know about and need to deal with both the technical and artistic elements of capturing an image on a daily if not hourly basis. Sorry if I’m being rude, but I just wanted to let you know that photography for some of us is more than a hobby or a way to take picture of the kids it’s what we do to pay the bills.

Mar 4, 2012permalink

MGJA

@DPR:

First, thanks for moving towards and standardized high-iso test.

But that said, you probably should take a look at the scene while it's still new and you haven't shot dozens of tests with it. People have bene criticizing it, and some of that is unfair - it needs to be a tripod shot for comparability and lighting needs to be stable even if low for the same reason. The subject, though, are just about as bad as they could have been for this test - NR smearing doesn't really impact most of the items in the picture. The rat, which is affected, is out of focus. Please consider replacing some of the plastic figurines with hair, texture and feather samples, ideally in patterns that tend toward small. Those things in central focus. If possible, try to normalize the highest-iso shot at 1/250 or possibly even shorter, since high-iso is so often used to freeze action. It will not be a perfect test - what is? - but it will be better.

Mar 3, 2012permalink

Andy Westlake

This particular scene isn't supposed to be a standardised high ISO test, and it certainly won't be shot with all cameras. It's simply one that I have available in our London studio (as opposed to the main Seattle studio), which will sometimes appear in quick sample shots like these for cameras we first get hold of in the UK.

Mar 3, 2012permalink

cesaregal

Andy,
is it difficult for you test D700 and 5D Mark II in similar way?
And D800 when available.
Readers could see real improvements.

Mar 4, 2012permalink

Rubenski

The real pro's even hesitate to shoot with anything over ISO 100. But we're no pro's, if we were we wouldn't be posting here (come on, just admit it...)

Mar 3, 2012*permalink

lol pro concert or sports photographers are shooting a ISO 1600+ regularly.

Mar 3, 2012permalink

What in the world are you talking about? Maybe that might have been true in the film days, but these days we *regularly* shoot at 800 and higher. In fact, I hardly ever shoot ISO 100 anymore! My *standard* ISO is 400. With the higher quality of today's sensors, there's rarely any reason *not* to shoot over ISO 100.

Mar 3, 2012permalink

Gothmoth

rubenski = clueless

Mar 3, 2012*permalink

leomartinez

Photography is not about pixels or noise. Photography is about visual language: To comunicate and to convey moods, feelings, situations...

A pro is someone who make all his incomes from one activity. A real pro would shoot ISO 409600 or even with a stenopeic box, if has to.

You have to see some of the pictures in books (Sebastiao Salgado) or magazines, it´s no unusual to see some noise or grain. If you limit yourself to ISO 100-200 you will not be able to reach the extense variety of photographic situation. I refer you to the doble page illustration of the article "The sky is falling" (American Photo sep/oct 2011).

It seems that we tend to forget other working scenaries. It is impossible, or at least absurd, make super specific cameras, or cameras "a la carte". ISO 12800 is there because it exist a necessity. A real pro necessity.

And no, I am not a pro. But professionals do the same (comment and discuss), from his blogs or web sites. That is not a big difference to me.

Mar 3, 2012permalink

Carlos C

well said.

Mar 3, 2012permalink

JakeB

Why are clueless trolls who know nothing about DIGITAL photography allowed to contribute?

Ban these pests.

Mar 4, 2012permalink

Smartypants

Really Tape5,
Reviewing the thousands of photos cataloged in my Lightroom database I see many over ISA 400.
My Alpha 900 in auto ISO mode routinely sets ISO between 400-800 and I believe many other cameras do this also.
I use ISO to help create depth of field and adjust it as needed up to 1600 and down to 100 as part of my settings when I shoot in manual mode to help me hit the sweet aperture setting on my optics.
I don’t agree at all.
It would be nice if and when ISO could be made useable higher with good noise control but now at least for my camera an ISO of 3200 is about the limit

Mar 3, 2012permalink

Tape5

An entire history of photography is behind us with rarely any photographs using ISOs over 400. Where did all these pros come from who are suddenly craving ISOs over 25,000,000? If you have nothing to scratch, you have to buy the itch first I guess.
I wonder exactly how many photojournalists do we have in this world? Two per capita?

Mar 3, 2012permalink

You are forgetting sports photographers who need fast shutter speeds often in low light conditions. You simply can't shoot an Indoor football or hockey game at ISO 400.

Mar 3, 2012permalink

Skip M

And wedding photographers who have to, occasionally, shoot in dimly lit churches w/o flash...

Mar 3, 2012permalink

"An entire history of photography behind us..." was severely limited by the poor high ISO performance of film, and more initially, the poor high ISO performance of digital. However, as high ISO performance gets better and better, there's no reason why photographers need to limit themselves to lower ISO settings.

High quality high ISO = freedom! Freedom to shoot when, where and what you want, even at low light levels. It sucks when you have a great image ruined by motion blur because your shutter speed was too slow for the action or movement in the scene. The capacity to select a very high ISO is a huge help in addressing this issue.

Sorry, but there are plenty of situations that are dimly lit (namely during a period of time called "night"; ever hear of it?). And there is movement in the world that requires higher shutter speeds. That's where high ISO comes in.

Mar 3, 2012*permalink

Christo256

I can say as a pro for twelve years that what the high iso means is the difference between saying to the client the 'image will be too difficult without lights' to saying 'yes we can do that.'

Mar 3, 2012permalink

plenty of photographers were using ISO 800 - 1600 just before digital arrived, and they continued to do so after it did arrive.

many liked the ISO 800 -1600 range simply because at the time at least the ISO 800 film negatives were already offering finer grain matching ISO 200 of years before.

this was the time many more enthusiasts and pros alike ventured out to do a lot of 'ambient' available light photography and not even ever carrying an external flash or use of built-in flash at all.

it was a time of lots of exploration of non-flash photography in low light which in the past necessitate flash, which unfortunately, gave indoor lit shots that 'flash' look... a look i call 'club black' (where the people subjects are well lit, but the context you find them in is completely unknown because it all was 'blackened')

that 'club black' look is typical of film shots taken in the older era when higher ISOs were not available, and if they were, early versions were quite coarse grained.

Mar 4, 2012permalink

rootsup

raw with nr off, we all know this. also real world shots with plenty of different scenarios. not sure what happened here.

Mar 3, 2012permalink

Christo256

Firstly I agree with some of the comments about the limitations of this test. They are however slow shutter speeds so probably continuous lighting. I would think it would be a better test if the lighting was a little more difficult with some serious highlight and shadow areas to show the dynamic range. I would also like to know how strong the AA filter is in this camera as this has not been mentioned and that is another area where this camera could be improved.
22 megapixels is great as long as it isn't strangled by the AA filter.

Mar 3, 2012permalink

lensberg

Considering the fact that this image was taken at the extreme setting of ISO 102800 ... it is truly a fantastic achievement! Easily comparable or perhaps even a little better than the Nikon D3S ... The 5D Mark III's noise fluctuations are extremely well controlled in both its magnitude and spatial frequency... virtually no traces of chroma noise present & luminance noise - definitely a 2 stop advantage over the 5D Mark II ... I seriously doubt whether anybody would actually use this setting for anyting other than emergency purposes... but at least it produces an acceptable result for reduced resolutions....

Mar 3, 2012permalink

JanneFJH

Hi Again.
Since the Mark II is kind of ¨Canon very well known std ¨, the same pics from that camera would help evaluation a lot.
Best regards.
/Janne

Mar 3, 2012permalink

avgcitizen

I like the idea of a test with low level light to better map with real-world use, but this scene is far from useful. Too many similar smooth surfaces, no indicator of focus point, etc.. I'll bet most peepers are looking at the furry rat and trying to figure out what is noise reduction and what is just out of focus. Please, someone at DPReview, explain the use of this scene? Is parent company Amazon.com keeping you on a low budget diet?

Mar 3, 2012permalink

Really to make this scene better all they have to do is move the rat to the center and focus on it.

Mar 3, 2012*permalink

They shot it at f11 where DOF is 1.5 meter. With focus seeming to be on the samurai, the rat is well inside the depth-of-field range.

Mar 3, 2012permalink

Hynee

I just checked out the low ISO images to find what was in focus, troll.

Mar 3, 2012permalink

The back half of the the rat and the lunch box are clearly out of focus. plus most lenses are sharpest in the center. The whole point is the rats fur is really the only place in the test image to judge the amount of detail being lost to noise. So why not jsut make that the focus of the test?

Mar 3, 2012*permalink

DarkShift

@marike6 No the focus point is at 1.5m. DOF is much shorter with 100mm lens.

The lighting is too flat in this scene. It should also have some dark fine detailed objects like textiles in focus area.

Mar 3, 2012permalink

@DarkShift Oops, You're right, I just re-read that.

Mar 3, 2012permalink

ah... okay.

focus is about 1.5m... (just read that now too)

ok... so DOF is about... 13.9 cm deep (for 100mm f/11)

interesting
i did my own 'visual' analysis with only the objects shown as references and made an approximate subject distance (focal point) of about 1.15m to 1.29m... or avg of 1.22m

DOF for 1.22m would have been: ~9 cm deep

also hazarded a guess the wood figurine is 8" tall and 1.27m away...

if so, DOF for 1.27m would have been: ~9.8 cm deep

i guess i was off

off by ~4.1cm of DOF, that is, and 0.23m distance...

at least one object of known scalable length horizontally or vertically would have been helpful.

unsure of mfr objects in UK are imperial or metric...
:D

Mar 4, 2012*permalink

Jason1234

I like a lot about what i am reading - dont know if digic 5 is usefulto raw shooters....
just remember the 5d mark ii when it came up - the pre-order price that everyone paid was changed upon deliver, increasing by about $600 for everyone -

Canon and it's distributed extorted everyone with a $600 priceincrease upon delivery - i spoke to Canon exec's who really couldn' tjustify the bait and switch....even agreeing that Canon hedges their own orders from suppliers...and the deposits from canon distributers did not say the price could change upon delivery.....

I say make sure your pre order is a hedged price (gauranteed) like Canon does themselves to it's suppliers.

Mar 3, 2012*permalink

Alfonso Bresciani

I'm glad I pre ordered it yesterday from B&H then

Mar 3, 2012permalink

Christo256

I ordered two 5d mark 2's when they first came out late 2008 and Canon then did the sneaky price change. I have to say that Robert White was very decent and charged me the original quoted price. They actually took it on the chin!! Now I live in NZ so cant buy from them but think Canon very underhand to do this to loyal customers. They knock up the price when they say the pound or US dollar is weak but don't change it when currencies regain their strength. If I didn't own loads of Canon lenses I would switch camps just due to their unfair policies.

Mar 3, 2012permalink

US prices on intro for 5d2 (Sept 2008) was US$2699
in Canada, our 5d2 (Sept 2008) was C$2800 (it was too low for exchange rates)

so just as news popped up 2008-Nov-06, official Canadian prices were to go up to C$3100, i happen to catch a last minute pre-order with Black's Photo, they had a pre-order special price with C$100 off... so I was lucky I was getting not just the regular price of C$2800, but at a special C$100 off for the pre-order to total C$2700 instead.

the exchange rate on that day was ~ C$1.17/USD
so, in effect, i paid US$2393 in 2008

just added sales taxes, and they offered free shipping too...

whew

got it delivered in 2008-Dec-18 (while away on vacation)
:D

odd thing though, while Black's must have gotten a hold of 5d2's early, they only sold them online, and i had never seen them in their regular stores, so they definitely were giving local consumers a great opportunity for a purchase not normally available except in more pro shops or bigger electronics stores.

Mar 4, 2012*permalink

tbcass

Why take a camera that is meant to produce high quality, high resolution photos and only show samples crippled by poor lighting conditions? I want to see the best the camera can do.

Mar 3, 2012permalink

We really can't win, here, can we?

We are hoping to publish real-world shots as soon as possible and will shoot our standard tests under studio lighting.

Mar 3, 2012permalink

jeffoto

many people WANT to see what the camera can do in sub-optimal conditions. As a photojournalist of 25-years, these examples look absolutely amazing and can open up a whole new world film could never have.

Mar 3, 2012permalink

Poor lighting? This still-life is evenly illuminated by flash lighting from 1 or 2 monolights. How would you suggest they light this scene? Should they pull the sun from the sky?

Mar 3, 2012*permalink

Edmond Leung

Jeffoto, but the digital also could never have the taste of film. They are two different world and no need to compare with each other. Right?

Mar 3, 2012permalink

http://natanlorenzi.com/wp-content/uploads/2012/03/DSC3776.jpg

you didn't read the description... it was a test to see the 'best' images in the WORST of conditions... and DPR offered us that... this is far better than showing 'easy' lighting and parameter setting conditions, where IQ obviously would be much better.

this approach makes perfect sense, and i am glad DPR is doing just that. it allows us to see the limits in a very brief mini-test, and does the job we want. in a very short time span.

Mar 4, 2012*permalink

Natan Lorenzi

It isn't fair, with controlled light, and NR strong, it'll never show the real low light capacity of this camera. To prove what i'm saying, here is a sample shoot by me with a Nikon D90 + AF-S Nikkor 50mm f/1.8G @ iso 6400 f/10 and 1/3s. That is great result, clean with details and low noise, with controlled light.

what you think about? We need real life samples!

Mar 2, 2012permalink

I totally agree and said the same thing below. I like DPR, but to call this "Low-Light ISO Series Samples" is really misleading. Low light does not mean studio strobes, it means natural light. You are not testing the high ISO ability if you are adding artificial flash lighting to the scene.

Mar 3, 2012permalink

It is low light (around 3EV) and is more challenging than natural light.

It is not shot with studio strobes. NR is set to default, not strong.

Mar 3, 2012*permalink

@R Butler
If it is not shot with studio strobes how did you light it? Continuous lights?

You say this test is more challenging than natural light. Challenging to shoot or challenging for the sensor? I would question whether this scene is more challenging than natural light for the camera, as an ISO test. I don't think we'd see the ISO 100 and the ISO 6400 looking so similar under natural dim light situations. All the objects are evenly lit so there are few shadow areas where you'd typically see noise so images look great at most ISO settings.

I'm not complaining, just trying to understand the test. Thanks.

Mar 3, 2012permalink

Revenant

DPR do state that they used "low level halogen lighting" at about 3 EV. Take at look at the shutter speeds they used, eg. 25 sec at ISO 50. They wouldn't have needed those long exposures had the scene been well-lit. So yes, it's very dim continuous lighting.
They also state that "At a colour temperature of about 3000K and low light intensity (approx 3 EV), these images are also neutrally white-balanced. This requires extreme amplification of the blue channel (and to a lesser extent the green) relative to the red, accentuating any noise or banding that might be introduced by the camera. This should therefore be considered close to the worst-case scenario - under many situations you'll see less noise at high ISOs, for example when shooting indoors in daylight."

Mar 3, 2012*permalink

this is hardly a low light scenario

your lit 50mm shot at
iso 6400 f/10 and 1/3s (and processed in photoshop)
comes out to EV-ISO100 of 7.7EV
was it JPEG straight out of camera or not?
or a NEF or JPG pp w/ ps?
what was original ISO when captured, 100 or 200?
was it pp to ISO6400?, say, by more than 1/3 EV stops?, +2EV? +3EV, more?

DPR's lighting conditions quite a bit lower
iso 100 f/11 and 13s
DPR's shooting conditions are much lower at ~3.3EV
this is a jpg straight out of the camera i believe.

point is, we don't know if you are comparing like-for-like lighting, which makes a big difference

Mar 4, 2012*permalink

dmiller62

Is it possible to record video in the magnification modes (which can be done in the latest Digital Rebel, for example, but which isn't supported on the 5D Mark II?) I was hoping for this - you'd think they would support it - but I can't see anything in what's been written up so far about it.

Or is this not supported? (the language in the official product description is ambiguous. It talks about not being able to change magnification during video recording - fair enough, the Digital Rebel doesn't allow that either - but it doesn't say that you can't start recording video at all when you're in one of the magnified modes)

Mar 2, 2012permalink

fyngyrz

Andy:

Does the 5DmkIII preproduction unit DPR is in possession of have the ability to turn OFF the high ISO noise reduction data damage mechanism, uh, I mean, feature?

If so, can you turn it off for JPEG and/or RAW?

Thanks!

Mar 2, 2012*permalink

altenae

Good question...

Please DPREVIEW post some RAW files......

Thx

Mar 2, 2012permalink

I don't think we've ever been allowed to provid Raw files from a pre-production camera. We also don't post Raw files before we have Adobe Camera Raw support and can post converted versions.

Mar 2, 2012*permalink

daravuth lay

I see banding everywhere at iso 100 (lateral + vertical). I gonna switching to ... Sony!! Canon why u did this to me?! I hate u

Mar 2, 2012permalink

altenae

Also switching ??

Well .........Goodbye !!!

Mar 2, 2012permalink

daravuth lay

i just found strange that after 200 comments, nobody talked about banding. Which is nice, if there were banding, they should also be visible in jpeg aren't they? I ll just wait until price of MK II drop and upgrade from my 50D. I m no pro and cannot justify the cost of the MK III

Mar 2, 2012*permalink

daravuth lay

j/k

Mar 2, 2012permalink

Mescalamba

You should written "bazinga!". :D

Btw. banding can be fixed by SW.. (or camera FW:) ). I would wait for RAW, overNRized JPEGs dont tell much.

Mar 2, 2012permalink

most likely still TTL metering... if not... shooting ISO 100 in the dark... w/ RAW... and then 'fixing it' in pp (i already know Sony folks do the same as other Nikoners... haven't learned to use ExpSim LV of Canon, provided they even know what it is or know if they have it or not, or know how to use it...)

i did an impromptu deliberate altering of all JPEG with an extreme Negative 032/127 Contrast and extreme upping of Gamma from 1.00 to 2.00... (as a curious check)

zero banding on all images for all ISOs (as expected)

why zero banding? because all images were properly exposed in-camera to begin with for each ISO, including the lowest to highest.

Mar 4, 2012permalink

seta666

I have to apologise to the administrators because of using Dpreview images on other site to show the heavy NR processing.
Anyone can download fullsize images and see by themselfs that (Eg, the wood textures around the miniature camera dissapear at high ISO, anything over ISO 1600 is over processed) but I guess it is easier just to show it
http://g1.img-dpreview.com/2DECE5831124497BB23543B38ECD4389.jpg

Mar 2, 2012*permalink

for fine details and NR, you chose an ISO 3200 most rivals fare far worse too, but the 5DMkIII will fare better, than any of the following here, including Canon's own PowerShot G1X:

@ ISO 3200:
PowerShot G1X vs NIKON D3s vs SONY A900 vs NIKON D4:
http://tinyurl.com/724ctue

maybe, if you showed just how different ISO 1600 was, just before it jumped to ISO 3200, then you would have shown it was 'near' identical to ISO 100.

just to give everyone a hint of what to expect, we all know the 5D3 will surpass the humble PowerShot G1X (and just a bit less than the 1DX), here is an official DPR sample of ISO1600 for its rivals:

@ ISO 1600:
PowerShot G1X vs NIKON D3s vs SONY A900 vs NIKON D4:
http://tinyurl.com/7aue8zb

to which, we can see, rivals cannot even keep up with a humble PowerShot G1X, to which we know the 5D3 will surpass.

too bad DPR didn't show D3x

i am sure, the D800 will be fine at ISO 1600, given it will have higher Mp, but it most likely not match the 645D... but we'll see

Mar 4, 2012*permalink

JKP

Good catch. Seems like Canon has not been able to improve image quality nor pixel count significantly this time. I guess, the new auto focus system has eaten all the R&D resources of 5dMIII. The good thing is that I don't need to feel the lust to buy one. :) Maybe next time, 5dM4...

Mar 4, 2012permalink

seta666

Sydue, I do not understand what you say, are you comparing the sensor of a gx1 (worse than that on the 7D) with a nikon D3S? You either are flipping or you work for Canon and think we are stupid.

In order to compare sensor we need raw data, no magical JPG processing with lots of NR. The nikon D3s is well over two stops ahead of the G1X (Noise, DR, color depth, etc...)

I advice you to visit DXOmark so you can understand what a sensor needs in order to be good

Of course the 5D mkIII will surpass the G1X, and the 5D mkII too but I yet have to see it beats the nikon D800 and I do not mean only noise but also DR and color depth

And please, do not give me more links with over processed JPGs for comparisson
http://tinyurl.com/724ctue

Mar 4, 2012*permalink

Rob

Please post a few of the RAW image files. I want to know if the loss of detail is noise reduction coming off the chip, or noise reduction processing for jpg.

Mar 2, 2012permalink

HowaboutRAW

Rob:

Got any idea of what software you'd use to extract those proposed raws?

Mar 2, 2012permalink

Rob

DPP from Canon for now I guess, maybe it would require an update which Canon could provide. Soon an update from Adobe unless they pull the "you have to update us as well" like they did with the 5d2.

Mar 2, 2012permalink

HowaboutRAW

Rob:

Right and DPReview does post raws with the full reviews.

Hope Adobe updates ACR soon.

Mar 2, 2012permalink

First thing studio shots with flash lighting is not a good test of a camera for low light / high ISO performance. A camera will always look better, less noisy in artificial flash lighting.

Secondly, the only way to view these samples is by downloading the full-size images, and opening them on your computer. They look OK, but I'm hearing a ton of super positive comments, but upon viewing the full size images like ISO 6400, it's clear that NR is heavy, and it is smearing away fine details like the mouses fur. I don't remember my 5D Mk II having such heavy NR in JPEG, but then I hardly ever use JPEG, so...

Lastly, a comment below states these to be on par with the D3s. They are not. They may be an improvement of 1 stop over the 5D Mk II, but nowhere near the D3s. The D3s high ISO images are clean WITHOUT smearing fine details.

Mar 2, 2012permalink

Exactly. When I looked at the images I was like "why is everyone going gaga over these?" The only thing being demonstrated by theses images is canon's NR algorithms.

Mar 2, 2012*permalink

If you tried reading the accompanying text, you'd realise these are NOT shot under flash lighting. They're shot under 3000K lighting that more closely resembles Tungsten lighting, rather than our standard, daylight-balanced studio lighting.

Richard - dpreview.com

Mar 2, 2012permalink

Flash light or continuous light the effect is the same. You ADDED light to the scene.

Mar 3, 2012permalink

KerryBE

Then why did you emphasize that using flash lighting was a major problem?

Mar 3, 2012permalink

only if the artificial lighting is a bright luminance chosen intensity, one can choose intensity so it is just above candle light.

any daylight light would be much too intense even in shadow/indoor in an indirect context.

Mar 4, 2012permalink

Helmut_S

Congrats to Canon overall, for specifically addressing the Mark II's glaring deficit in AF. I left for Nikon then, and do not regret that, but would not have gone through such a big move had the Mark II been more than a video gimmick. Now Canon has two real great contenders with the 1 DX and the Mark III. Good times...

Mar 2, 2012permalink

Personally I would wait until people have a chance to actually use the new AF system before calling it a success.

Mar 2, 2012permalink

ppastoris

Helmut_S, did you actually use the 5D2? :-)

Having owned that camera since it came out, my experience is that the AF there is excellent for most things, with the exception of super low light (essentially candle light) and tracking extremely fast or off-center subjects.

Mar 3, 2012*permalink

lensberg

The ISO performance from this new beast really seems to be pretty much on par with the Nikon D3S ... Oh my god... ISO 25600 is extraordinarily good! No doubt there is noise reduction present at that setting... but the D3S isn't exactly devoid of NR either... Granted the 51200 & 102400 are definitely extreme steps... but quite frankly no better than what you would get from the D3S... How come there are no low light ISO test samples from the D800 available...???!!! Probably don't want to hurt their sales with the results... in the coming months...

Mar 2, 2012permalink

HowaboutRAW

You have RAWs from the Canon 5D III, what software did you used to extract the RAWS and where can I see full-sized tiffs of your results?

Mar 2, 2012permalink

Patrick H Lockwood

The only low light shots, because I'm a wedding photographer, are settings which I would use during low light in a reception, without flash, and that would be:

Shutter 1/125 on the 24 - 70 ( to freeze action and accommodate any or most focal lengths - and the fact that I don't have real steady hands, don't have time to think and make adjustments, most of the time.)

Shutter 1/250 on the 70 - 200

Aperture, F/2.8 the apertures on the 24 - 70 / 70 - 200

Aperture F/2 on the 50mm 1.4

And whatever high iso for the test.

The settings they use for the tests are useless, from my vantage point.

And, I wish they would make a 24 - 70 IS. But the high ISO capability probably negates the need for IS on that lens.

Mar 2, 2012permalink

jywang

Is the 24-105 f/4L lense worth getting as part of the bundle if I already have the original 24-70 F/2.8L and 70-200 F/2.8L IS?

Mar 2, 2012permalink

IMO no but your results may very.

Mar 2, 2012permalink

probert500

vary

Mar 2, 2012permalink

Thanks typo police the internet is safer because of you you!

Mar 2, 2012*permalink

Tom Melanson

I've used both and the 24-105 is very good, indeed. If I didn't need the extra reach and felt the extra speed were worth it, I'd spring for the 24-70 II. My wish would be for the 24-70 MK II to have IS, but then.."if wishes were horses, beggars would ride" :=-)

Mar 3, 2012permalink

snake_b

I love how people slam Canon and Nikon on the forums and slight them, using such names as "Canikon". However, how can people argue with these results, along with real, tangible improvements over the previous models?

Mar 2, 2012permalink

Jaimito

Pretty impressive. clean up to 6400 iso. very acceptable12800. 25600 is when noise becomes more distracting, but it is still quite usable. The H1 and H2 are for emergencies, kinda like it is for the MKII at 12800, 25600.

Mar 2, 2012permalink

M Jesper

Take that, people who said they saw no reason to upgrade ;)

Mar 2, 2012permalink

Eden S

What if those people shoot studio with controlled lighting or outdoor landscape?... Just saying your own needs don't always match those of others. I welcome high ISO capability, but not everyone needs it, so not everyone may care to upgrade.

Mar 2, 2012permalink

M Jesper

Hey man, i winked.

Mar 2, 2012permalink

samhain

Blows me away people whining about this cam. Personally, I'm impressed. Good job canon

Mar 2, 2012permalink

Looking at images with a critical eye is not the same a "whining". And Henry, how could you possibly know the finances of other DPR members?

Mar 2, 2012permalink

Dylthedog

I have the money but with the D800 a serious sum less I can live without the extra 2 FPS and at the moment, and as far as we know, that's the 5DIII's only advantage. Nikon are in the driving seat on price and features.

This camera is slightly less than TWICE what I paid for my 5DII (£1600 vs £3000) and while it really looks like the camera I always wanted, Canon aren't being realistic within the current market.

These shots do look great but as a RAW shooter, that's what I want to see. If they show this much advantage over my 5DII in RAW I'll be amazed.

Mar 2, 2012permalink

Alfonso Bresciani

yeah but were these shot in JPGs?? I wanna see shooting from raw!

Mar 2, 2012permalink

Jaimito

wait for formal test on a production copy.

Mar 2, 2012permalink

Alfonso Bresciani

I pre-ordered yesterday so I just hope at least 12800 from raw will be pretty clean

Mar 2, 2012permalink