Equipment The 5D Mark III Thread

To further what motorsportcfd said, I had to sell my 50Ds specifically because the autofocus sucked so bad. I used to use my 50Ds as secondary bodies for weddings, and when I got back to review the photos I'd find that hardly any of them had nailed the focus. With the 1Ds2 using the same lens (Sigma 50mm) the focus was nailed virtually every time, even at f/1.4 which is quite something with a full frame body. I found that the peripheral focus points on the 50D were next to useless, whereas every single point on the 1Ds2 was virtually perfect.

Autofocus is really put to the test either when the subject is distant and moves very fast, or if you are working in a low light environment (like a church or dimly lit hotel, for example).

In these scenarios it really is night and day between the prosumer and pro grade bodies. I have still to use the 7D at a wedding and am excited to see how it will perform compared to it's little brother the 50D.

 

Interesting replies; thanks guys. Curiously, I note that motorsportcfd says that portrait photographers won't really notice the difference, but LennyRhys talks about autofocus performance for weddings. I presume you're talking about the more event-oriented stuff like receptions, rather than the static nature of the actual ceremony.

I'm curious what some of the old hands think about autofocus performance. I wonder if photographers were ever really hampered by fewer autofocus points, or if modern autofocus systems have become so capable that we tend to get lazy.

 

I'm not an old hand, but I can certainly chip in an answer to your questions.

The most common "solution" to the AF problem is focus-and-recompose, a technique that most photographers (in fast paced work) use if they want the main subject to be off center whilst using the center AF point. Some sticklers for textbook camera operation who are adept at changing AF points very quickly on the fly might prefer to make use of all 40-something points in their expensive cameras, which is fine; but exponents of the focus-and-recompose method find it an unnecessary faff and opt to use the center point because it is the strongest (cross type) AF point on all but the proest of pro cameras.

As for weddings, it's all to do with the light, not the movement of the subjects: I find that the ceremony can be even more of a challenge than the reception/dance because the light can be extremely poor in a church which has very heavy and dark stained-glass windows, and to make this even worse some churches won't allow photographers to use a flash during the ceremony...

With my 1Ds2 I used all the focus points - it was one of many things I loved about that camera, and I really miss it (but not the weight of it, lol). I think the 5D3 represents a very interesting compromise and it will probably be very popular with wedding photographers; at least the ones who don't jump on the medium format bandwagon.

 

Industry people I've met through contacts at work. Seems the general consensus is that manual Nikkor lenses are pretty much the weapon of choice for any professional using DSLR equipment. It's got nothing to do with optical quality either, as both Canon and Nikon lenses are both easily capable of outperforming the 1080P resolution.. it's about how they focus. There's less compression of the focus movement as you head towards infinity with manual Nikkors; they are more linear. This is essential when manually pulling focus.

Even some newer Nikkor lenses still have aperture rings too... which again, is very desirable in the all manual world of moving image.

There's also a recent fashion for old M42 screw lenses on 5D MKIIs as well for the same reasons.

Prior to the 5D MKII you could have got a used Noct-Nikkor 58mm f1.2 for around £400. You'll be lucky to get one for less than £1700 used now. That's no accident.

Modern digital cameras full stop are making people lazy. It's common now for degree photography students to actively resist learning how to use incident light meters, and practice exposure on film. They just don't see the point. Sad times. Even when I demonstrate a modern SLR failing to meter correctly, they still resist. Shoot, look at screen, delete if no likey... twiddle some dials... reshoot. "Well.. you can correct it afterwards!" LOL

Kids today.

I can remember shooting professionally for clients, and sometimes handing over un-processed E6, confident that everything was right. I mention that to students today and they look horrified. They still insist they don't need to do anything themselves of course.

These days... "good enough" is good enough.

 

I think you make an excellent and valid point Pookey, but I also think that it's not quite as clear cut as "all photographers who take advantage of correcting exposure after taking a shot are lazy" - if that's what you're saying, which I'm not entirely sure it is.

Photography has changed A LOT in the last ten years (you of all people will know this) and adaptation has taken place in professional circles - people don't treat digital cameras like film cameras, because that's not what they are. Light meters don't need to be bang on accurate because corrections can be made in post processing without extensive loss of quality, and that's not laziness - it's adaptation IMO... or, more to the point, for the true professionals who really know what they are doing and who are making the best use of the technology available to them, it's a good use of a resource.

Unfortunately, however, not even "good enough" is good enough. I lost a job last November - a day's conference photography - because some jumped-up schmuck with two Nikon D700s offered to do the conference for free, and the organiser told me this. A few weeks later, the conference organiser came to me with his head in his hands and said, "the photos are terrible... we should have hired you again."

C'est la vie.

 

Yep :/ Maybe it's the CF card that's pissing away power? It's a Kingston 133x though. I'm not sure.

 

It's the 5DII specifically - it doesn't have many AF points related to the size of the frame, basically leaving the edges as no-AF areas. There's also the fact that the outer AF points are less sensitive as well, leading to the central area being the only real 'usable' area if you're operating at speed.

My issues mainly stem from the lack of coverage, for which I compensate with the lock focus and shift composition technique. If I'm using a 1D however, I just select the AF point I need, as the coverage is basically global (hence the nickname 'ring of fire').

If Canon have fixed this in the 5D3, it's a winner.

 

I would be inclined to get your 5DII looked at, 20mins of video there is something seriously wrong, you should get a few hours, your CF card should fill up before your battery gives up.

For me the camera is what I wanted from the 5DIII and from all accounts it looks far more appealing then the D800, if I wanted that many MP I would get medium format, the 5DIII is a nice mid ground camera, they've improved on what was already the top selling entry level pro body and from all the reports I've heard from friends playing with it at Focus it just makes me want one even more.

I think yes canon could do with dropping the price a little bit but its still half the price of the 1DX and a hell of alot of camera, it does make the Broncolor Verso S2 I'm eyeing up look cheap though lol (damn you shopping list)

 

This is an interesting point. Some of my colleagues and I were talking with one of our video engineers yesterday during a digital cinema camera test. The conversation turned to some of the upcoming mid-to-high-range model camcorders that are starting to promise 4K resolution and beyond on 2/3" sensors. One point of the conversation centered on actual lens resolving capabilities, and the fact that Canon's best studio lens (approximately $70,000) can resolve 100 line pairs per millimeter over about the center 10% of the sensor. Outside that 10% area the resolution starts to fall off. For comparison, a 2/3" 1920x1080 sensor has 100 line pairs per mm; a Canon full-frame, 21.1 megapixel sensor equals about 78 line pairs per mm. What all this means is that mathematically in terms of lens/sensor resolving capability in the video world, Canon's flagship lens is not capable of full 1080 resolution on a full-frame DSLR sensor. If their $70,000 lens can't fully resolve 1080 video on a DSLR, how can another manufacturer claim 4K resolution in a $5,000 camcorder. The answer: the same way my iPhone records "720p" video.

I'm glad you brought up the desire to use Nikkor lenses for video. I'm going to have to keep an eye out for that, because it's not something I personally have seen a lot.

 

I've run into the same thing in my concert work. Some newbie college students that have been put in charge of categorizing all of my work I have turned in over the years have now started popping up to shoot these events. When I mention spot metering, manual exposure, and single servo continuous focusing, I lose their attention if not occasionally get rolling of eyes.

OK, Buddy, good luck with that matrix metering and zone focus on some coked up lead singer with strobes going on and off - when you are long gone, they will still be paying me to get the shot they need.

 

Actually that IS what I'm saying. Why? Because it's easy, that's why. So, by doing something easy in camera to get the exposures correct HAS to be better than any amount of post adjustment, no matter whether you CAN or not. There are loads of things that CAN be done, but the question is whether they SHOULD be done or not.

While being easy, it requires practice and skill to achieve, and students just don't want to do that when they can get "good enough" and CORRECT their MISTAKES afterwards. The fact that getting the exposures correct at the time of exposure would mean they wouldn't need to do any extra work doesn't seem to be of any importance to them.

Yes, you can correct MINOR exposure errors afterwards, and it's nice to have that headroom if you make a genuine mistake on a very important shot, but these are mitigation techniques, and shouldn't be daily working practice.

Then why do companies invest countless millions in R&D to make ever more accurate meters in their cameras? Surely a meter that's accurate to with a stop and a half will do if your argument is correct.

I think it is laziness. Just because you can make adjustments in RAW doesn't excuse a thing. What if you're taking a scene that fully utilises the histogram and wrings every last drop of dynamic range from the sensor? What then? You'll lose something, that's what, because there's no wiggle room left. To say that you have "adjustment" is a very misunderstood term. You only have adjustment possible if the scene you shoot is well within the dynamic range of the sensor. The more you approach that limit, the less tolerance you have. If your scene fully utilises the dynamic range, you have NO adjustment possible without compromising one end or the other of the Histogram... or changing the gamma response of the histogram. Fact.

If all you had to do to avoid this was spend some time learning a skill.. (outrageous I know).. then none of this would be an issue.

I never, ever take an incorrect exposure (unless I quickly snap the camera up to take quite literally a snap shot). This isn't because I'm super human, or have an excess of talent, it's because I understand the limitations of my equipment and fully understand exposure theory. I can look at a scene and evaluate whether my camera's meter will cope or not. If not, I'll ignore it, and reach for an incident meter. If I have none, I'll reach for a grey card. As BOTH methods are far more accurate than any reflective meter alone, no matter how many 3d metering cells and super duper witchcraft is purports to have. That also, is a fact. Incident meters merely measure the light falling ON an object, and ignore colour, contrast and tonal range... all the things that regularly fool reflective meters. My little Sekonic 308 is smaller than a pack of cigarettes, and roughly teh same weight. If you can carry an SLR around, and probably a bag of lenses... you can carry a meter. No real excuse.

No.. it's not. As I've said.. it's not foolproof, and if you are on the contrast range limit of the sensor.. you're screwed. It's a poor working methods that can be avoided if you just take the trouble to be a master of your craft. The thing is, that doesn't seem to be something people want to aspire to any longer. These days... if you can hide a mistake rather than spend time learning skills that avoid the mistake in the first place, they take the easier option. By definition, that's laziness.

No.. getting your exposure right is good use of your resource Getting it wrong and correcting it afterwards, is not. If there was some ADVANTAGE to not getting in right in camera and correcting it afterwards, I'd agree with you, but I challenge you to furnish me with one. (again, sports and press guys have a legitimate excuse for fast and free working... few other people have). The only perceivable advantage is that you can get away with doing less work, less study, less practice. In retrospect, that's never been, and will never be a good thing.

Sorry for the derailment. If you want to further the debate.... I think we should start a new thread.

 

Done.

 

I don't know why, but I'm so over the new gear...

Just like all companies they just got very good at these little updates. Apple, Adobe, Canon, Nikon...
They have a schedule... and with a 5DIII they just worked on a feature a YEAR.
A year for a audio jack...
a year for a sensor upgrade...
another year for autofocus..
and then a NEW MENU SETTINGS For Video!!!

If anyone has a 5DmkII I don't see a point in upgrade AT ALL.
If you're buying a new full frame from Canon then it's great!

At the end of the day when you look at a PHOTO and see what it took to get it.
The lighting, composition, knowledge of the equipment etc.
The few upgrades just disappear in the sea of variables and become almost irrelevant.
If you're doing sports photography, you should've been doing it with a 7D instead of a mkII instead

All are welcome improvements, But IT's SO NOT EXCITING. Just like iPad3 or Nikon D4 or the new version of Photoshop. I'm much more excited how the rebel from 4 years ago vs a rebel from now ARE ABSOLUTELY different beasts in a good way.

 

So the 5DIII is shipping, who's got one ?

 

I've had mine for a couple weeks now and, apart from the small form factor, it's one of (if not the) best cameras I've ever used. Overall the feel and build is much better than the MKII as is AF, storage capacity and VF. These were all qualms I had with the MKII and the very reasons I never upgraded from my 5D to the MKII. AF is on par with my 1DIII and seems slightly better in low light. While the camera does not allow AF with f/8 lenses it will with the pins on the TC taped. When taping the pins the camera will allow the use of 41pts, as opposed to one with all other current 1-series bodies, and all points seem to have the same sensitivity, speed and accuracy, which are actually pretty good. Apart from that having a huge point coverage within a FF body is nothing short of awesome...now I can finally select a point in the frame where I want to focus without having to focus and recompose the shot. Another major turn off in regards to 5-series bodies, having owned and used the 1DIII for years, was the lack of dual card slots...the MKIII addresses this issue allowing me to shoot on two cards in tandem. Another issue, albeit a moderately less important one, was the 5D's VF, especially when shooting macro and in low light conditions. The MKIII now offers a 100% VF which is noticeably brighter than the 5D's.

With that said there are a few snags that I find annoying enough to note. Overall, while the new AF system is worlds ahead of the 5D/5DII, the AF point illumination method is awful in comparison. With side VF illumination the camera initiates a red LED flash which bleeds across the VF, illuminating not only the points but also a large portion of the focus screen. Canon also got rid of the selected AF point pre-flash, which allows the user to see, before locking focus, which point is selected. Since the points are all black, seeing those points in low light can present an issue. Canon also decided, for what ever reason, to change the LCD review zoom button layout. The new process, while once learned is equally easy to use, doesn't offer any advantage over the prior method. It just seems like they changed it just to change it. Another drop of piss in the soup is the lack of any real IQ gains. Granted, the 5DII was already pretty good in this department, after several years of technological advancement one would expect at least some noticeable improvement in this area. While the 5DIII performs slightly better at extremely high ISOs it offers no gain in DR and overall little improvement in regards to high ISO performance.

Overall the camera is quite good and for me will be ever better once I receive the grip. I plan on adding another to my setup a bit later down the road.