Gear Talk > EOS Bodies - For Video
Philip Bloom Comparison 5d3 vs D800 vs D4
JasonATL:
--- Quote from: psolberg on May 23, 2012, 10:40:57 AM ---the problem is that the killer dslr doesn't even had 4:2:2 output and the codecs fall appart easily with motion, and additionally it's Full1080p mode doesn't resolve detail to be called true 1080. this makes the D800 the better "killer" IMO for a lot of applications.
--- End quote ---
psolberg - Do you have a link to where someone has tested or established the actual video resolution of the D800? I've done shot my own test charts with the 5D3, but I've not yet seen the same done with a D800.
psolberg:
--- Quote from: JasonATL on May 23, 2012, 04:30:46 PM ---
--- Quote from: psolberg on May 23, 2012, 10:40:57 AM ---the problem is that the killer dslr doesn't even had 4:2:2 output and the codecs fall appart easily with motion, and additionally it's Full1080p mode doesn't resolve detail to be called true 1080. this makes the D800 the better "killer" IMO for a lot of applications.
--- End quote ---
psolberg - Do you have a link to where someone has tested or established the actual video resolution of the D800? I've done shot my own test charts with the 5D3, but I've not yet seen the same done with a D800.
--- End quote ---
I have not seen it done with charts side by side comparison but you can just google it and you'll find several reviews having the same conclusion. Philip bloom even mentions it and his review is about as fair to all cameras as I have ever seen. The issue is that a lot of people are looking at re-compressed video or youtube and that's a poor evaluation of any camera.
The best evaluation of the sampling of the D800 comes from
http://falklumo.blogspot.de/2012/04/lumolabs-nikon-d800-video-function.html
as per his conclusions, the D800 is actually a 2240 x 1260p camera which then scales that down to 1080p. I think that is what accounts for the better image quality.
Their guess is that the canon's sampling method pixel bins to a lower resolution before encoding and then creates 1080p video which Andrew from ESO hd calls 720p upscaled to 1080p. The exact details will likely remain unknown but the footage of the D800 is definitively sharper regardless of how it arrives to it.
Andrew over at EOS HD mentioned he'll be doing a shootout of the D800 soon so stay tuned for his comparison since he posts still images of frame grabs which is the proper way to compare video short of the original mov file.
Meanwhile I would refer anybody curious about the uncompressed HDMI out (which is far more interesting than the resolution) to this video (download the original and not the vimeo embedded video)
https://vimeo.com/40788982
then compare STILLS
http://proofs.iamron.com/Personal/Atomos-Ninja-VS-D800-In-Camera/22567372_w9PxNn#!i=1806243743&k=fmHtjv9
there you can see the weakness of the built in codecs. the 5DIII's codec is equaly weak with motion leading to a blocky output. But you can also see how the D800 image can be sharpened even beyond its already good levels if you capture it at high bitrates.
I'm ordering the mosaic engineering filter as soon as they sell it. I'm hoping it will not soften the output too much. Philip bloom didn't say if it did but I'm wondering if the scrutinzed it.
awinphoto:
--- Quote from: psolberg on May 24, 2012, 07:37:42 AM ---there you can see the weakness of the built in codecs. the 5DIII's codec is equaly weak with motion leading to a blocky output. But you can also see how the D800 image can be sharpened even beyond its already good levels if you capture it at high bitrates.
I'm ordering the mosaic engineering filter as soon as they sell it. I'm hoping it will not soften the output too much. Philip bloom didn't say if it did but I
'm wondering if the scrutinzed it.
--- End quote ---
If i'm not mistaken, didn't bloom say the d800 COULDN'T be sharpened without introducing bad artifacts but the 5d3 COULD be sharpened to correct it? It may have been another video but i got from blooms video was the 5d3 had to be sharpened in post and the D800 needed a 3rd party filter and assuming you corrected for both, they both are great cameras, noise issues aside for the d800.
bp:
--- Quote from: awinphoto on May 24, 2012, 02:39:41 PM ---If i'm not mistaken, didn't bloom say the d800 COULDN'T be sharpened without introducing bad artifacts but the 5d3 COULD be sharpened to correct it? It may have been another video but i got from blooms video was the 5d3 had to be sharpened in post and the D800 needed a 3rd party filter and assuming you corrected for both, they both are great cameras, noise issues aside for the d800.
--- End quote ---
I think you're remembering the CameraStore video (the 3rd in their series of 3). They did specifically say that the D800 can't be sharpened without really bringing out the artifacts. And their example shot certainly backed that up.
That said, they didn't have a pre-production Mosaic filter like Bloom did. They also weren't recording to an external, higher bitrate recorder - which was what was done in the example psolberg was referring to. They just tried sharpening the 24mb/s internally recorded D800 footage
I think it's funny how adamantly people are STILL arguing over who's dad can beat up the other guy's dad. As Bloom said, these are both GREAT cameras. Both require a fix - the D800 requires the filter if you don't want moire/aliasing - the MK3 requires sharpening in post to bring out the details. Noise in low light/high ISO on the D800 isn't fixable, but if you usually shoot in controlled conditions, and don't ever need to shoot above 3200, no worries, the D800 will be great.
The one thing Bloom didn't talk about, which I found very surprising in the CameraStore video, was the weird jumpy laggy picture when you magnify the D800 live view to check focus. In practical use, this would probably drive me more crazy than all the other gripes combined, because I use that constantly in a shoot. Then again... I'm sure you could get used to it.
I think Philip is right - they're both great cameras. If a close friend of mine buys a D800, and another buys a 5D3, I'll tell them both the exact same thing... "CONGRATS!!! Now shut up and go shoot something"
bp:
--- Quote from: psolberg on May 24, 2012, 07:37:42 AM ---Andrew over at EOS HD mentioned he'll be doing a shootout of the D800 soon so stay tuned for his comparison since he posts still images of frame grabs which is the proper way to compare video short of the original mov file.
Meanwhile I would refer anybody curious about the uncompressed HDMI out (which is far more interesting than the resolution) to this video (download the original and not the vimeo embedded video)
https://vimeo.com/40788982
--- End quote ---
Man... Great demonstration of that extra tiny little bit of quality you can get when you're not relying on the internal codec. Canon definitely screwed up by not giving us clean HDMI out. I'm SO happy actually, that Nikon did in the D800, because just that could be the impetus for Canon to do the same via a firmware update.
Now I just wish Nikon had added zebra and focus peaking so Canon would also have to follow suit with those as well. heh
Navigation
[0] Message Index
[#] Next page
[*] Previous page
Go to full version