Are 5d3 owners happy with the video, ALL i?

[Capnbooboo]

I don't know about everyone else but a friend aka stringer sent me screenshots of videos and picture and the video is very noisy, looks terrible,

i tried the to compare my videos with photo taken at lower res as well and no matter what i did the video paused looks nice and large but the quality and noise can be compared to my ipad 3, total joke.

look at these taken by the stringer, first is a photo and then a video, totally different under the same day light and settings, he says they where taken with a 25/105mm lens, 125/f8, the video was shot at ALL-i

 

[Nishi Drew]

I don't know about everyone else but a friend aka stringer sent me screenshots of videos and picture and the video is very noisy, looks terrible,

i tried the to compare my videos with photo taken at lower res as well and no matter what i did the video paused looks nice and large but the quality and noise can be compared to my ipad 3, total joke.

look at these taken by the stringer, first is a photo and then a video, totally different under the same day light and settings, he says they where taken with a 25/105mm lens, 125/f8, the video was shot at ALL-i

Well video is video it's not meant to substitute taking photos. Pausing/taking a screenshot will never beat photos, while I've pulled stills from my 5DII that look as if they were photos (can post if request maybe), just less resolution and little information to edit. Also, the Mk3 is softer in video than the Mk2, while the Mk2 is sharper is suffers from moire.
I haven't heard about the iPad being production ready in quality... but of course in this case only the 5D can get nice and shallow when needed... the shots are obviously edited there, was that after receiving the stills and editing yourself or was it that way when you saw them? Because all that loss in quality/noise might as well come from whatever editing took place, this doesn't look like the rawest data to judge with

 

[Rienzphotoz]

: : : :

 

[Policar]

The video from the Mark III is pretty soft with some chunky noise but that second shot is unusually bad and was either shot or processed wrong to look like that. At best, Mark III footage won't look quite as good as the first frame grab, no matter what you do in post or on set. But it can get reasonably close. I don't know about the iPad but under ideal conditions the iPhone is sharper with better resolution than any Canon dSLR, but with less manual control and a worse "look" obviously. But for exterior video in low contrast light in wide shots in some cases it might produce the better image. I wouldn't count on it, though.

 

[Mt Spokane Photography]

When you look at a 2mp frame grab, its going to look bad. That's true for all digital video cameras operating at 2K.

4K is much better, and with 8K, single frames are very nice.

Show a frame grab from video of your iphone.

 

[Policar]

When you look at a 2mp frame grab, its going to look bad. That's true for all digital video cameras operating at 2K.

4K is much better, and with 8K, single frames are very nice.

Show a frame grab from video of your iphone.

Not true at all. Most cinemas project at 2k. The Alexa produces amazing 2k footage... This K stuff is nonsense.

If I find the time I'll shoot the same subject with my Mark III, C100, and iPhone 5 at similar focal lengths. But I already know what the results will be.

 

[Midphase]

I've shot my latest film on the 5D3, IMHO the video quality is noticeably better than the 5D2, particularly the noise when pushing the ISO past 800.

Judging video from a still is the wrong approach, also judging video that was shot incompetently vs. video which was properly lit and exposed is also not a way to judge a camera's capabilities.

Lastly, noise in video is generally much more accepted by the public than noise in still photography. I watch so many DVD and Blu Ray big budget films which look pretty sad in the shadows and darker scenes noise-wise...yet I don't hear many people complaining about how bad the film looked.

 

[risc32]

When you look at a 2mp frame grab, its going to look bad. That's true for all digital video cameras operating at 2K.

4K is much better, and with 8K, single frames are very nice.

Show a frame grab from video of your iphone.

Not true at all. Most cinemas project at 2k. The Alexa produces amazing 2k footage... This K stuff is nonsense.

If I find the time I'll shoot the same subject with my Mark III, C100, and iPhone 5 at similar focal lengths. But I already know what the results will be.

i think you're talking nonsense.

 

[Policar]

I've shot my latest film on the 5D3, IMHO the video quality is noticeably better than the 5D2, particularly the noise when pushing the ISO past 800.

Judging video from a still is the wrong approach, also judging video that was shot incompetently vs. video which was properly lit and exposed is also not a way to judge a camera's capabilities.

Lastly, noise in video is generally much more accepted by the public than noise in still photography. I watch so many DVD and Blu Ray big budget films which look pretty sad in the shadows and darker scenes noise-wise...yet I don't hear many people complaining about how bad the film looked.

Absolutely, but the Mark II and Mark III produce very similar footage (excepting aliasing issues) at low ISOs, with the Mark II arguably having the edge slightly for sharpness (due to false detail) and having a bit less read noise. At high ISOs the Mark III is far nicer, but this isn't that.

Again, the footage posted above is WAY worse than what the Mark III is capable of, and it's also under the conditions that are least impressive with the camera. But the crop taken from a still is better than the camera will ever produce for video.

 

[Axilrod]

I don't know about everyone else but a friend aka stringer sent me screenshots of videos and picture and the video is very noisy, looks terrible,

i tried the to compare my videos with photo taken at lower res as well and no matter what i did the video paused looks nice and large but the quality and noise can be compared to my ipad 3, total joke.

look at these taken by the stringer, first is a photo and then a video, totally different under the same day light and settings, he says they where taken with a 25/105mm lens, 125/f8, the video was shot at ALL-i

Are you playing back the footage inside of a NLE or just the original files? If it's in an NLE the quality will degrade when you press play but when you pause it goes back to full resolution.

As for ALL-I vs. IPB, pretty much all the tests say that there is no difference in image quality and the few people that have claimed to see a difference usually say it's the IPB that looks better. Pretty much the only advantage of shooting ALL-I is that your NLE will run much faster since it's already optimized for editing. Other than that it's a waste. I've shot hundred of videos in IPB and have never had any problems.

The first still looks about how the Mark III video should look, but something is definitely wrong with the 2nd one.

 

[Axilrod]

I don't know about the iPad but under ideal conditions the iPhone is sharper with better resolution than any Canon dSLR, but with less manual control and a worse "look" obviously. But for exterior video in low contrast light in wide shots in some cases it might produce the better image. I wouldn't count on it, though.

LOL I guess you're including the 1DX/1DC in "any Canon DSLR?" Look the iPhone is great for what it is, but I don't know how you can say that it's better than "any Canon DSLR" with a straight face In the Zacuto shootout it was painfully obvious which one was the iPhone, I mean the 7D didn't look that much better but definitely not as bad.

 

[Policar]

I don't know about the iPad but under ideal conditions the iPhone is sharper with better resolution than any Canon dSLR, but with less manual control and a worse "look" obviously. But for exterior video in low contrast light in wide shots in some cases it might produce the better image. I wouldn't count on it, though.

LOL I guess you're including the 1DX/1DC in "any Canon DSLR?" Look the iPhone is great for what it is, but I don't know how you can say that it's better than "any Canon DSLR" with a straight face In the Zacuto shootout it was painfully obvious which one was the iPhone, I mean the 7D didn't look that much better but definitely not as bad.

You got me on the 1DC. Never used the 1DX so I don't know about that one, either, but I'm guessing it's close based on footage I've seen.

All I said was it's sharper. Nothing about it being better overall... I implied the opposite pretty strongly.

 

[skitron]

My 2 cents (probably about all it's worth...) is the video grab looks like it was taken with a stock picture style and has not much apparent DR as a result. Plus looks like no sharpening in post. To me, these are the more noticable issues than noise.

I've not done much video with 5D3 but did do some testing to gear up for doing some in the future. My findings are that there's a pretty substantial difference to the look when using a picture style optimized for video like FLAAT or similar and then applying a RGB S curve in post as a basis for further grading refinments (if desired), as opposed to using a stock Canon still shot picture style.

For noise reduction, the Red Giant NR plug is pretty good, as is Neat Video NR plug (I own both) and I'd probably say the Red Giant product is the more flexible and easier to use.

Lastly, the softness is pretty easily handled to at least some degree with Adobe's native sharpening tools.

But yes, I'd agree what comes straight out of the camera could stand improvement.

As for all i-frame, I did see a test of moving cars at a distance that were very tiny within the frame and the all i frame looked smoother in the motion of the cars. Again, I've not spent much time on video with my 5D3 yet but that test leads me to believe all i frame is maybe helpful under certain conditions, and of course as already mentioned, if storage is not too much of an issue, it may help when you dump it into your NLE.

Jmo, ymmv, yadda, yadda.

 

[Policar]

Straight out of camera, unprocessed:

5D Mark III, neutral, low contrast, 1 sharpening, 17-40mm L at f5.6.

 

[Policar]

iPhone 5 video.

 

[cocopop05]

I love the video produced by my 5D Mark III. I use the 24-105mm f/4 L lens and in Premier Pro CS6 add 28-35 sharpening. In low light, even with that lens, the footage looks fab. No complaints from me.

 

[Capnbooboo]

I love the video produced by my 5D Mark III. I use the 24-105mm f/4 L lens and in Premier Pro CS6 add 28-35 sharpening. In low light, even with that lens, the footage looks fab. No complaints from me.

although i use my 5d3 in manual 90 percent of the time i do also see a quality difference, but do you guys think the problem is that the 5d3 processes at 4:2:0, until April (a weeks time) output at 4:2:2 will finally be available, but then we all need a ninja or other brand capture device via HDMi.

ive been testing my 5d3 to see whether i can pull of video and photos that look close enough to not tell the difference but i find its the lighting and i never get the same results with the same exact settings, it seems i have to tweek a custom setting for video and then switch to it from my typical manual or AV mode for photo rather then just go into live mode for video on the 5d3, you guys run into the same thing?