7D Mark II Video Tested By Gizmodo

[Canon Rumors Guy]

<p>Gizmodo has published a comparative video test against the EOS 5D Mark III & EOS 70D. It basically solidifies what cinematographers and videographers have been saying since the launch of the EOS 7D Mark II; it’s not a camera for them.</p>
<p><iframe src="//player.vimeo.com/video/106524090" width="500" height="281" frameborder="0" title="Canon 7D Mark II Video Comparison" webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe></p>
<p>I’m starting to think that Canon is going to be expanding Cinema EOS to the prosumer level, and will not be developing the photographer DSLR for that market. That sort of approach could payoff, as they could focus ergonomics and software to the video segment. This could also lead to people buying multiple EOS mount products. If you can’t get new customers, at least sell more stuff to the people that are already your customer, I’m borrowing from <a href="http://www.bythom.com/" target="_blank">Thom Hogan</a> on that one.</p>
<p><strong>Canon EOS 7D Mark II $1799: <a style="color: #900000;" href="http://www.bhphotovideo.com/c/product/1081808-REG/canon_9128b002_eos_7d_mark_ii.html/bi/2466/kbid/3296" target="_blank">B&H Photo</a> | <a style="color: #900000;" href="http://www.adorama.com/ICA7DM2.html?KBID=64393" target="_blank">Adorama</a> | <a style="color: #900000;" href="http://www.amazon.com/gp/product/B00NEWZDRG/ref=as_li_tl?ie=UTF8&camp=1789&creative=390957&creativeASIN=B00NEWZDRG&linkCode=as2&tag=canorumo-20&linkId=4IHYPE3ZKJN5VL4X" target="_blank">Amazon</a> | <a style="color: #900000;" href="http://www.cameracanada.com/enet-cart/product.asp?pid=7dmarkii" target="_blank">Camera Canada</a></strong></p>
<p>Source: [<a href="http://reframe.gizmodo.com/comparing-7d-mark-ii-video-to-its-fellow-canon-dslrs-1636490542" target="_blank">GM</a>]</p>
<p><strong><span style="color: #ff0000;">c</span>r</strong></p>

 

[SoullessPolack]

What exactly are we supposed to see here? Or want to see? They all seem pretty identical, the only clue telling otherwise is the slight changes in perspective when switching out cameras. Is there something missing that a "videographer's camera" would have?

 

[Rowbear]

What exactly are we supposed to see here? Or want to see? They all seem pretty identical, the only clue telling otherwise is the slight changes in perspective when switching out cameras. Is there something missing that a "videographer's camera" would have?

+1. I must be dumb...but I don't get it either :/

 

[SpecialGregg]

I will say that it seemed like the 7D II was a stop better than the 70D once they got to ISO 6400.

 

[rs]

This is my take on it: That brick wall in the background of the first scene has some pretty heavy moire on the 70D; the 7D mk II and 5D mk III seem to be much better on that front. In terms of sharpness, in this test the 5D3 is better than the other two. When it comes to noise, the 7D2 appears to be between the other two, possibly closer to the 5D3.

 

[mkabi]

What exactly are we supposed to see here? Or want to see? They all seem pretty identical, the only clue telling otherwise is the slight changes in perspective when switching out cameras. Is there something missing that a "videographer's camera" would have?

+1. I must be dumb...but I don't get it either :/

To answer both Rowbear and SoullessPolack... not that I saw the video per se (well not yet, but I have an idea).
The sensors are identical, and even if the sensor is different its still downsampling from the bigger sensor, because its 1080p.

The 70D is cheaper and has swivel touch screen, which makes it more ideal for videographers (in my opinion).
The only thing the 7DII brings to the table in terms of video is the 1080/60p and the ALL-I/IPB/IPB light, which is not enough for someone to upgrade from a 70D. Moreover, there are other options like the G4H, A7s, and now the NX1 that has always had 1080/60p but also have 4K (they all have more to offer than the 7DII). In my opinion, having 4K does not make it a videographer's camera.

A "videographer's camera" is just much more than just 4K... and well 7DII even less than that...
If they had put all the functionalities of what Magic Lantern is offering, without Magic Lantern tampering with it. Then it would be a videographer's camera...

 

[Drizzt321]

My big question is, what were the recording settings? Was this the clean HDMI out to a quality recorder and then transcoded with a professional package? I assume (given I'm not a pro) that output at 4:2:2 via uncompressed HDMI will probably make a pretty big difference. Otherwise, probably not too big a surprise if they used the same codec to save out from 7d2 & 5d3. Canon probably didn't really change that around.

 

[ajfotofilmagem]

This is my take on it: That brick wall in the background of the first scene has some pretty heavy moire on the 70D; the 7D mk II and 5D mk III seem to be much better on that front. In terms of sharpness, in this test the 5D3 is better than the other two. When it comes to noise, the 7D2 appears to be between the other two, possibly closer to the 5D3.

Well observed. The brick wall in the center have terrible moire only in 70D. The bar code on the bottles also have moire only in 70D. In general, 7D Mark ii seems to have better video than 70D.

 

[LetTheRightLensIn]

Basically it looks like they stopped line skipping on the 7D2 so it won't get the nasty aliasing and moire of the 70D or 7D and it will have a good one stop or more SNR advantage over those (due to not line skipping), BUT it's still the same waxy, blurry looking mess. A 5D3 with ML RAW will utterly blow it away for color, detail, editing latitude and usability.

A VERY feeble video effort this late in the game.

It figures that the second Canon realized that they accidentally had something huge with DSLR video the bean counters went and ruined it and crippled it and tries to push people to 1DC and Cxx for anything solid. (thankfully ML saved the 5D3, and so long as you are willing to deal with the troubles of RAW, it makes that a very fine 1080p cam indeed).

They didn't even give it a zoomed video mode! Despite it seemingly reading the entire sensor and having dual digic 6 power! And it being the reach/wildlife body!

Panny GH4 video totally destroys this, so does internal A7S video (never even mind A7S to 4k Ninja Shogun), so does 5D3 with ML RAW, so does SONY 4k camcorder, so most likely will the new new Samsung and anything else arriving soon.

Other than for those for whom DPAF is key, I think they will be very disappointed in the video sales this brings in. A day late and two dollars short.

The AF should be awesome and the fps are great, so to the sports/wildlife shooters it should sell very well though, but they blew it for the video crowd.

 

[Ebrahim Saadawi]

Horaay finally a test. The video tells a lot by the way.

First, the 7D mk III is NOT line skipping like the 70D, it's using pixel binning like the 5D mk III, in short, they solved aliasing and moire issues.

Second, the low light performance is improved by 1.5 to 2 stops, approaching 5D III level.

Third, no improvent in detail over the 5D mk III. Very similar.

-First APS-C Canon without moire and aliasing, and with great low light performance, I really like that. The camera is also very nice in the audio features (headphone jack, ability to silently change levels whilst recording), Dual pixel AF with speed/sensitivity adjustment, 1080 slow motion, lack of distortion when using Canon lenses, Etc

it's just a bit disappointing in resolution just like the 5D mk III was, yet the 5D proved to produce very nice results with post-sharpening and it looked very natural.

I would have loved higher resoluton, but they seem to have reserved that specific feature for the C100.

 

[LetTheRightLensIn]

My big question is, what were the recording settings? Was this the clean HDMI out to a quality recorder and then transcoded with a professional package? I assume (given I'm not a pro) that output at 4:2:2 via uncompressed HDMI will probably make a pretty big difference. Otherwise, probably not too big a surprise if they used the same codec to save out from 7d2 & 5d3. Canon probably didn't really change that around.

No clean HDMI out recording makes barely any difference. I tried it on the 5D3 and it was the same mess, you could barely tell any difference whatsoever, only on frames where the entire frame changed and even then only compared to non-ALLI basically. ML RAW made a HUGE difference though. That's really good quality.

It's not the codec that does the damage, it's something in the de-bayer or later processing stage where they ruin it all. Either DIGIC utterly stinks at debayer and image processing or they chose to use really amateurish large radius sharpening with a ton of DR and use DIGIC badly, or marketing has them apply a blurring filter at the last stage. The proof is in the HDMI uncompressed recording looks almost the same, but the ML RAW recording looks infinitely better.

 

[IslanderMV]

I only checked the 7D M2 and 5D3. Look at the Empire State building in the distance and the roof "mushrooms" in the foreground. Some of the air vents look washed out in the 7D M2, the Empire State has no detail in the 5D3.

I could not find Waldo anywhere !

 

[transpo1]

There is barely a difference between the 3- that's what we're supposed to see here. Canon seems not want to compete in the video DSLR market anymore. Sad- it's the end of an era, really

If they're smart, they'll develop a second line of Cinema EOS DSLRs that are more affordable (and thus more competitive) for filmmakers and videographers.

 

[LetTheRightLensIn]

Although, on the plus side, since it doesn't line skip, in a year or so, when Magic Lantern comes out for it, it should be able to produce an excellent 1080p so long as you are willing to deal with RAW and Magic Lantern proves possible to write for it. So in a year or so you might get to notch 1080p RAW out of it for less money than a 5D3. So given time it might be an awesome 1080p camera (for those not scared off by RAW video).

 

[Lee Jay]

Let me be frank. Image quality matters much, much less for videos than for stills. Videos have many more frames, audio, and all of it in time sequence to provide more information than the viewer of a single still can get from a single still frame.

The reason I wanted 4k, even if it was low end 4k, was for stabilization in post, which can cost you 3/4 of your pixels in some cases. The reason I want digital zoom is so I can get more zoom range out of a cheaper and better narrower zoom lens (18-135 versus 18-300, for example).

But, for properly framed shots, this performance is more than acceptable.

 

[DanielW]

I will probably buy this camera because of the AF system alone, as by now I already have some respectable collection of almost-in-focus shots of my quick little man.
On the other hand, were I a videographer, a 7D mark-whatever would never be in my wish list. Horses for courses?
Cheers,
Daniel

 

[talicoa]

I'm sorry but to me it looks like they didn't stop down and they have some focus issues. Sure it is soft if you look at elements outside the DOF. Did they use a good lens? I don't know, it doesn't seem like sharpness wasn't maximized here.

 

[LetTheRightLensIn]

Let me be frank. Image quality matters much, much less for videos than for stills. Videos have many more frames, audio, and all of it in time sequence to provide more information than the viewer of a single still can get from a single still frame.

The reason I wanted 4k, even if it was low end 4k, was for stabilization in post, which can cost you 3/4 of your pixels in some cases. The reason I want digital zoom is so I can get more zoom range out of a cheaper and better narrower zoom lens (18-135 versus 18-300, for example).

But, for properly framed shots, this performance is more than acceptable.

That all depends. The difference in detail and quality between ML RAW 5D3 video and regular 5D3 video is pretty intense to me. And A7S 4k video, wow, it's like looking out a window. The difference is pretty darn easy to see. Easier to see than most web stills image differences.

Of course it also depends a bit what you are shooting. Nature/scenics footage it probably matters than for a scripted story.

 

[Lee Jay]

Watching a documentary in a true large format IMAX theater is impressive visually. For most everything else, whether it's on the big screen or on my SDTV makes essentially no difference to me.

I find the difference between ML raw video and native 5D3 video to be entirely immaterial.

 

[LetTheRightLensIn]

Watching a documentary in a true large format IMAX theater is impressive visually. For most everything else, whether it's on the big screen or on my SDTV makes essentially no difference to me.

I find the difference between ML raw video and native 5D3 video to be entirely immaterial.

Wow.

No difference seen in detail and texture between waxy 5D3 native video and richly textured ML RAW and no difference in blown highlights and crushed black detail between the two?

No difference between 4k and 1080i and 480i even for nature?

I guess people sure see things differently. To me the difference is almost unimagineably great.

In all seriousness are you sure you are not slightly near sighted and perhaps just not seeing any resolution and detail differences so the only thing that impresses you is raw screen size or when something is blown up soooo large that you can finally start to see fine details?