Join the discussion about work/life balance in relationships:

March 23, 2012

5D Mark III and D800: Thoughts on Camera Tests and Company Motivation

I've seen a lot of videos with footage from the 5D Mark III, but not as many with the D800. Most tests are either pretty outdoor scenes, or quick clips showing the improvements over the previous generation. These are certainly welcome, as they can help people decide on a pre-order or a purchase. For that reason, they are a necessary evil (or they can be a lot of fun for gear-heads). Part of writing in this community is having to wade through the dozens, if not hundreds of tests and videos, and it can obviously get a little tedious.

No Film School is going to have its own camera test, but instead of either shooting a short, or just a simple test, we're going to give you both. Not only that, but it's going to include both the Canon 5D Mark III and the Nikon D800. They are the two most interesting cameras at the moment because they are new, full frame, and reasonably affordable (that's debatable, I know).

I think one of the biggest things people are forgetting with these DSLRs, is that they are still cameras first, and video cameras second. Yes, it's the fault of Nikon and Canon for promising amazing video features and quality this time around. Neither company expected these cameras to take off for video people, because neither company saw the market for cinema-quality large-sensor cameras that were affordable and could use regular 35mm lenses (Japanese companies are historically slow to innovate, and clearly they weren't aware of the huge market for 35mm depth of field adapters). Nikon's offerings were rather pitiful until the D7000 this generation, and Canon didn't offer 24p on the 5D until a year and a half after the camera had been released.

But getting back to the point, Canon and Nikon have made tremendous still cameras. What so many in this community forget is that at this price range these are the best full-frame still cameras on the market - both of them huge improvements over their predecessors. The D800 is practically a medium format camera, and it's going up against cameras that cost $10,000 or more. The 5D Mark III can see in the dark, and it's got a professional autofocus system (unlike the poor autofocus ability on the 5D Mark II).

There's a whole group of people who will buy these cameras for the still imagery alone, because they make a living as photographers. Some people might be let down by what they've seen so far from the 5D Mark III and the D800, but they have to realize that those who will be strictly taking photographs with these cameras are a bigger market, and both companies attempted to listen to what those consumers wanted. This is a gross oversimplification, but Canon owners wanted the low-light ability of the D700, and Nikon users wanted the high-megapixel count of the 5D Mark II (you can't please everyone). Each company has a limited amount of resources that they can use for research and development, and up to this point, a lot of that R&D has been done by people who know a lot more about still images than moving images. That's why these cameras produce unbelievable still images, arguably better than any other brands out there. So from Nikon or Canon's perspective, getting someone to buy the camera just because of video features has been icing on the cake (that is starting to change though).

That doesn't mean we can't be critical of either Nikon or Canon. They are promoting the video features of these cameras as being complimentary to the photography features, rather than just an added bonus. We can't forget that these are still photography cameras. Many have compared these to the FS100, and in some cases just for video purposes, it might be a better deal. I'm not going to get into those comparisons right now, because I haven't used either camera yet - but that's going to change soon.

The other reason for this post is that I want to try to get a sense of what you guys in this community want to see in a camera test. The camera test portion is going to be as thorough as it can be - resolution, ISO, aliasing, rolling shutter - all the typical tests and more. A short film is rarely something you see right after a camera has been released, but that's the goal. I have some ideas about how I might try to use both cameras in the same short film, but what do you guys want to see? Do you have any ideas about how I can make it better?

We've covered some great tests and test films in the past, and if you haven't seen them before, they're embedded below.

These are just some of my favorites, but feel free to share some of your favorites below. Stay tuned for the 5D Mark III and D800 test coming in the next couple of weeks.

Your Comment

41 Comments

I am very impressed with this camera test. It is purely a test of the video features and stacks up DSLR`s against professional video cameras and it is thorough. I understand the production costs are great in this video but it educates and helps decipher what the results mean. These are key points in a camera test.

http://www.zacuto.com/the-great-camera-shootout-2011

March 23, 2012

0
Reply

Yeah, Zacuto's test is about as good as they get, but they spent thousands of dollars. Unfortunately I don't have that kind of money to spend on these tests.

March 24, 2012

0
Reply
avatar
Joe Marine
Editor-at-Large
Shooter/Writer/Director

This was a refreshing article - free of the pro-Canon or pro-Nikon jingoism (can I use jingoism in this context?) I usually come across. It helps to frame a good perspective from which to approach what are obviously two awesome cameras.
All that said, I am totally not a gearhead. In fact, I'm not even a photographer - but I am a filmmaker and my wife (who is a Nikon shooter) and I are in pre-production to shoot a film this summer. We have been considering shooting it on the D800 (have any features been shot on the Nikon?) because she already has glass and familiarity and we have a very (very!) low budget.
I'm not concerned that the film needs to look like it was shot on a RED or 35 or 16mm. We know it won't. But I would like to know some visible comparisons between the D5 MkIII and the D800 - resolution, moire, rolling shutting, aliasing artefacts, the blacks, etc. Like, what will an audience of laypersons see or notice in terms of a difference between the two. TBH, I'm not even that concerned about ISO - I don't believe that video shot at 12,500 or 6,400 will look good blown up on a 46" HDTV, let alone a movie screen (unless you can convince me otherwise). And almost anyone shooting narrative will have a lighting kit of some sort.
Hope that's not too pedestrian...

March 23, 2012

0
Reply
Shawn Whitney

I'd like to see the camera test have scenes/shots that are obviously testing the various metrics we're looking for (i.e. resolution, ISO, aliasing, rolling shutter). That way the film is a two-in-one.

Also, doing the same short film with both cameras would be cool, like John Brawley's Available Light Test would be cool, especially if there were a follow up of side by sides, or crop comparisons, etc.

March 23, 2012

0
Reply

Great insight! I've always liked reviews from Nofilmschool since you are mature enough to see the matters in wide angle.

I think it'd also be very helpful and great if you could illustrate changes in Codec. Mainly the issue of BT. 709 and how current NLEs have problem with properly reading 5dmk iii files.

Interesting is that now mk iii does proper 16-235 whereas mk ii did 0-255 though it was in Rec.709 format.

I'm curious how this change affects custom Picture Styles we use. As far as I know, CineStyle only uses 16-235 so I guess CineStyle must be safe. However, I wonder how other great custom picture styles(Marvel cine style comes to mind) would act on mk iii.

March 23, 2012

0
Reply
Tom

Personally I think the answer's simple, combine the specific tests you want to run into the short film's various scenes - and then let the camera handle each situation (alaising, moire, rolling shutter) as it would normally - don't try to cut around or mask the image defects, showing them to their full extent in-situ will give us the clearest idea of how the cameras perform in each test criteria.

March 23, 2012

0
Reply
Mark

D800 external recorder footage.

March 23, 2012

0
Reply
moebius22

Yes I agree. I think 100% of the D800 footage should be recorded to an external recorder. Nikon went out of their way to give us this feature, which was requested by nearly everyone. Canon did not include external recording capability, even though everyone begged for it for years, yet they raised their price significantly on the MkIII.

It would only be fair to use an external recorder as that would be the main reason to buy the D800 over the Canon. If the D800 footage is better than the MkIII with this advantage then so be it. If the MkIII fares well against an external device then kudos to Canon!

March 23, 2012

0
Reply
Lance B

Good points! +1

March 23, 2012

0
Reply
Shenan

in that same regard, HAVING to use an external recorder is the precise reason i would choose the 5D over the D800. I have no need for a dslr that is permanently attached to an external recorder.

March 23, 2012

0
Reply
Chris Hubbard

The thought of shooting 4:2:0 highly compressed 8 bit ever again is precisely why I want an external recorder lol!

Long before the c300 we used to talk about how great it would be to have the 50Mbit 4:2:2 Canon codec in a DSLR! But Canon thinks everyone using a DSLR can easily upgrade to the c300 - at 5 times the price! Obviously they want to protect a higher end c300 market and disregard the DSLR users who put them on the map in the first place.

March 23, 2012

0
Reply
Lance B

perhaps they both have their place then!

shooting with both the 5d AND the D800! heaven forbid! :p

March 24, 2012

0
Reply
Chris Hubbard

The Atomos Ninja is very small, and is just like having an HDMI field monitor mounted on your hot-shoe. When I read about this capability of the D800, I immediately sold my 5D II and bought a D800.

March 29, 2012

0
Reply
Markus

Marks what are your toughts on the D800? Was it a good idea to make the change?

March 30, 2012

0
Reply
Guilherme

Push the codecs with some intense grading.

March 23, 2012

0
Reply

Personally, I'd love to see some seriously low-light video, shot smartly, with whatever the flattest profile is for each camera, and then graded. Basically I want to see what the bleeding edge of these camera's capabilities are. I think it's a given that almost any camera looks good in the middle of its range with low contrast.

March 23, 2012

0
Reply
Geoff

Different people have different needs in a camera system. I tend to shoot at f/3.5, so Phillip Bloom's all WFO all the time, shows me nothing I want/need to know. If you shoot WFO, then his tests are valid for you.

On March 21st I rented a Canon 5D Mk III body and ran my own tests. $200.00 well spent. We shot from 3:00pm to 9:00pm and learned all we needed to know for video. Also shot some mount tests using a Sony NEX 5n at the same time, and now know how they will inter-cut with the 5D Mk III.

The next morning on the way back to the rental house, I snapped off a couple of dozen stills at f/1.8 to f/16 to check diffraction, etc.

March 23, 2012

0
Reply
c.d.embrey

What were your findings? Do you feel the MKIII is soft due to the low pass filter? Did you modify the camera styles? How did it hold up when pushed grading?

March 23, 2012

0
Reply

Sorry to disappoint you, but neither myself or the other three pros involved with the test found the 5D3 soft.

The best way to find the Real truth, as opposed to the Internet truth, is to test it yourself. I think it works as advertised, obviously there are Internet Gurus who disagree with me.

March 24, 2012

0
Reply
c.d.embrey

To me the worst image is in my opinion the 16mm...I liked the A and C from the beginning not by my artistic choice but purely in sensor response to highlights and color rendition. I would easily shoot with the E and F cameras although they are DSLRs. Excellent test IMHO.
Thank you

March 23, 2012

0
Reply
Augusto Alves d...

I did a test to compare two picture styles (http://vimeo.com/24439040).
If I could have the mkIII and D800 in hand, I think I would do something very similar to compare the two cameras.

March 23, 2012

0
Reply

The mk. 3 is an excellent and even affordable video camera. As is the FS100. I'm not sure why people expected canon to satisfy all their red-level dreams when the literally just rolled out their red platform at $15k.

March 23, 2012

0
Reply
jchilds

I would prefer not to have a short film as it's too difficult to compare shots if they're not super similar. I would prefer say four scenes/shots that are typical of footage shot with the cameras, that are relatively controlled and that would show their weaknesses and strengths.E.g.
A low light, high iso shot that's slightly underexposed and raised in post.
A high dynamic range shot.
A shaky shoulder stabilised fast moving shot.
A detailed close up.
A really detailed wide shot.
A moire inducing shot.
And some heavy grading examples and a good file to download. As someone mentioned I would also love if all the D800 stuff was HDMI out.

March 23, 2012

0
Reply

Agreed 100%. All I need is a few very short clips/scenes to see what I need to see. I love stories… but this is a tech story, not a narrative story. I'd get way more value seeing something like SCENE 1: shot once with each so I can see them back to back. SCENE 2 shot back to back, etc. This is, after all a comparison thing, yes?

March 29, 2012

0
Reply
Erik Stenbakken

That C300=awesome video is probably one of the smartest test/short hybrid I have ever seen. It seemelessly untegrated all the geekey features your jnterested in while presenting it in a comedic and interesting flow . Id like to see something like that from you guys...thts bad ass. And please externally record the D800, it deserves it.

March 23, 2012

0
Reply
Quobetah

The MK3 will have autofocus for video?

March 23, 2012

0
Reply
Baron

No, it's functions the same as the 5D II regarding AF.

March 29, 2012

0
Reply
Markus

Not so. I ran mine the other night and my 5DII would not AF during recording and the 5DIII did it. I never noticed it till I switched from the MkIII back to the MkII and I noticed that the MkII would not AF. Maybe I had the custom functions set up differently? I used the back of camera for AF operation and had the shutter one disabled in custom settings.

March 29, 2012

0
Reply
Erik Stenbakken

The 5D3 does not do continuous AF like a proper video camera however it is somewhat more useful than the AF on the 5D2. WIth the 5D2 you have to choose between either the excellent Quick AF (mirror flips down AF hits focus & mirror flips up again) that can only be used before recording starts & Live AF which is much slower but can be operated during recording. With the 5D3 you can use Quick AF before & Live AF during. Live AF does mess up recording for a second or two & has a tendency to hunt for focus if there isn't enough light &/or contrast but still can be useful as it's faster than stopping recording hitting Quick AF then restarting recording just so long as you know you will need a cutaway shot while it is finding focus. It is usually better to manually focus but Live AF can be useful in some situations e.g. a present there is no loupe that fits properly on the 5D3's screen so using Live AF may be more accurate than trying to focus on the screen held at arm's length.

March 30, 2012

0
Reply

I would prefer to see a kickstarter funded carnival where you use the funds to hire two hapless pro cinematographers that put the cameras through a gauntlet of ridiculous tests. Skin tones of arm wrestling sectogenarians dressed as superheroes sort of stuff. Can check moire on their plaid outer undies. Turn them into glamour models to check out how the cameras perform in that sort of lighting. If they aren't modeling well, shake the camera (need a controlled shake at same rate for each camera- electro mechanically pivoting) indicating dissatisfaction and giving a controlled sense of rolling shutter (and maybe shudder). Would that break the famous wall? I don't even know what else. At least it could be entertaining-so long as all the participants were game. Am I off base here?

March 23, 2012

0
Reply
Abersouth

I appreciate a hilarious and crazy idea stated with such elegance.

March 24, 2012

0
Reply
avatar
Joe Marine
Editor-at-Large
Shooter/Writer/Director

I do too. Who gave one?

March 24, 2012

0
Reply
Abersouth

Some guy who shares your name....strangest thing.

March 24, 2012

0
Reply
avatar
Joe Marine
Editor-at-Large
Shooter/Writer/Director

Thanks for playing.

March 24, 2012

0
Reply
Abersouth

The essential thing is that you use the optimal settings for each camera in a pro context. What an all-knowing master of the camera would use. It took a while to learn what those were for the 5d2 and that may also be true for this generation. If you are shooting one with the settings we wouldn't use in hindsight it's just unfair.

The C300 and C-line in general is what Canon is telling filmmakers to use if they care about features and quality. The 5D3 as you say is accidentally a video camera. But it will not ruin your script.

March 23, 2012

0
Reply
peederj

"will not ruin your script" = best comment I've seen in years on a tech-oriented string. we're shooting a verite doc (no scripting, thank you very much!) with F3, c300, and 7D; audio and lighting are much greater concerns than camera brands or codecs. in this biz the personalities are the greatest challenge, from running a network to making a sundance doc. decades ago I worked with top network editors who insisted that DPs sit in on an edit session or two . . . still great advice. keep up this solid thread

March 29, 2012

0
Reply
inbizallday

I made a little test based on the one from Dan Chung on how the footage of the 5Dm3 can look after applying sharpening filters in post. Check it out:
http://vimeo.com/39111785

March 24, 2012

0
Reply

has anyone seen this?
http://www.youtube.com/watch?v=6ZeM9ofnm4I
awful moire the d800 has...

March 25, 2012

0
Reply
ben

Yeah, how about a Spy Vs. Spy kind of thing? Two secret agent photogs, armed with their new cameras, in a shootout. You could do a mission briefing and set goals like:
1. Shoot an apple cart
2. Find a bald man in bright sunlight
3. Late night streetlight

Each test should represent one type of thing you want to know about the cams (i.e. color, highlights, night noise in the list above.) Then just add some knock-off Mission Impossible music and it's a go.
-Olaf

March 29, 2012

0
Reply

"Japanese companies are historically slow to innovate..."

Whaaaaat? Name a single DSLR or video camera that wasn't designed and conceived in Japan-- aside from some boutique manufactures, which don't matter much in the context of this article.

It's a lot like saying the U.S. is a slow innovator when it comes to putting humans on the moon.

March 29, 2012

0
Reply

Arri Alexa and RED come to mind - pretty big players in the market if I recall. RED specifically was doing 4K on a mass scale before any of the Japanese companies. They are very conservative with their products, I'm not saying they don't eventually release high-end tech, but sometimes it takes awhile. Japanese companies prefer incremental updates, that's just how they work.

March 29, 2012

0
Reply
avatar
Joe Marine
Editor-at-Large
Shooter/Writer/Director