DSLR Videography – Zeiss Primes – When will you make the jump?
Everyday I see new stuff on the web by people shooting with DSLR’s, and they almost always impress me in some way. Our company has thought about making the move, but we are still noncommittal for a variety of reasons. Not the least of which, is a less-than-ideal audio recording scenario and 12 minute record times. I didn’t grow up in the film world, so second source audio isn’t that appealing to me. Either is constant attention to record length. It feels like a step backwards. I know that strides have been made on the audio recording side to make it better and easier (Magic Lantern, CF recording devices), but it’s still no picnic. Once you’re done setting up your rig with all the stuff you need, it looks like a camera surrounded by scaffolding. Perhaps the day they make a DSLR that shoots great video and audio we’ll commit, but then isn’t that called a “video camera?” Today, Carl Zeiss lenses announced they’re making prime lenses for Canon EF mounts. The kit is about $20,000, and a PL mount adapter will be available too. This will only increase the momentum towards DSLR’s.
There’s no doubt about the fact that they produce absolutely stunning images for minimal cost though. I’d like to hear your opinions about the ground swelling move to DSLR videography.
Thumbnail Image courtesy of Philip Bloom
7 Responses
DSLRs have their place, but they should never be thought of as the only option. There are a number of issues aside from the sound recording problem. I’ll go with my main two:
1. H.264 recording codec. I find this to be the buggest problem with recording video on the 5DmkII or the 7D. H.264 is a web delivery format, not an acquisition format. It’s a highly compressed image that needs to be converted for efficient editing (in Final Cut; Avid makes an MXF file, so it’s not as much of an issue there). You also lose a ton of color information, making any chroma-keying likely to have remnants leftover.
2. Form Factor. The camera is a still camera, that’s its purpose. It is the size of a still camera, because it IS a still camera. When you decide to go into video mode, you can’t effectively do anything handheld as the size/weight doesn’t help steady the image. So in order to get a good handheld shot, you have to invest another could thousand dollars into good support hardware.
The point I continuously make with people is that this is a STILL camera first and a VIDEO camera second. I have many friends that have purchased this solely for its video/live view function. I find it depressing. There are far better video acquisition options out there with more functionality, cinematographic manipulation, and overall ease of use, without the major drawbacks of using a DSLR. Prboblem, they chose price over function. And that’s where it’ll end up biting them in the rear.
Great article Joe, I’ve seen some really amazing things done with DSLR’s as well. I do agree with Robert’s points; I think he is spot on in his analysis.
I have a DSLR that shoots 720p that can shoot some pretty nice stuff, but as Robert pointed out, unless it’s stationary on a tripod, or you have some sort of Steady Cam attachment it is going to be useless.
Robert, I tend to agree with all your main points. Which is why we still don’t have one. I do wish they would use a better recording codec, and I bet we see that someday. However, I just watched some raw stuff from a 5d at night on a 42″ monitor and although there was some noise, it was less than what I anticipated, considering the reputation of the codec.
My colleagues and I think that a major video manufacturer needs to jump into this market with at least a prosumer camera that allows attachment of EF lenses right out of the box. Red Scarlet seems to be perpetually “coming soon” and I’m looking for a more appropriate form factor. I realize there’s a ton of issues keeping this from market, but a boy can dream.
They do have their place. My station’s parent company is New Age Media and a station owned by them up in Maine charges extra for spots produced on their Canon Mark D5. My station has JVC video cams that shoot 720p, but so does my personal Nikon D90. Both look great but the Nikon DSLR gives you more exaggerated depth of field (or what paying clients think as higher production quality, thus the extra charges). Here’s a quick spot I produced where the interview shots and audio are with the regular JVC cam and the b-roll is all Nikon DSLR. DSLR’s can easily become your second camera, or if all audio is done separate from the shoot.
http://www.youtube.com/watch?v=_iQ9z8tb9Bk
Matt,
Thanks for your input. We love to get reader feedback. Especially examples. You can definitely tell that the Mark 5 added some real depth to the b-roll shots. A very nice addition. You could also tell how much sharper your JVC A camera was. I think you effectively used them both. I’m sure your client was happy with the piece.
The single biggest problem I have with shooting with my 5D Mark II is the inability to get reliable focus. It’s truly a double whammy, as focus is increasing in importance with these systems, yet more difficult to achieve! The video out ports are fairly useless, not true HDMI signals. I will be using it for an upcoming narrative short, mainly for it’s form factor. I can put the thing just about anywhere (with my Joby Gorilla Pod) and get some interesting angles for B footage and cutaways.
As a DP that is always trying to grow my ‘bag of tricks’… I’ve spent some time working with the 5d mkii and the 7d recently as crash cams and B-unit cams. In theory, its a great idea to be able to use a dslr for video, especially the full frame variety, such as the 5d. The problems you run into can be many though.
Problem number 1, as stated above is the record codec. H.264 is NOT a professional acquisition format, its a consumer delivery codec. The codec doesn’t necessarily look bad upon recording, but once you bring it into FCP, convert it to something editable, like ProRes, do a little bit of color work and output it to a delivery format, the codec really falls apart. What may be a great tool for a student film, doesn’t stand up to a professional delivery workflow with good results. A note about editing with the H.264 codec. We’ve cut 5d footage on FCP, Avid and Adobe. Adobe handled it poorly from what my editor friend said, FCP did a good job and we’ve done most of our DSLR work on it. Our most recent project we cut on AVID’s MC platform. It dealt with the codec the best in regards to speed of import, stability and quality. We transcoded to DNxHD 145 and 220 on import.
My second issue is the sensor and the method of extracting video from a chip designed to take 21MP stills, not video. The CMOS sensor, combined with the bayer mask and small lightweight form-factor of the camera lends itself to jello-cam images. Of course, if you’re shooting a nice static shot on a tripod, this isn’t really a problem. And unless you’re shooting a test chart, the lack of sharpness isn’t really noticeable. I’d compare it the the sharpness of a properly calibrated 35mm dof adapter.
Overall, I suppose the 5d mkii may be a good option for an low budget wedding video type shop, but when the lack of professional features on the camera (like recording outside audio) bite you in the rear, the money you are refunding your client would have been better spent buying real camera.
I will say that there are some situations where this type of camera could be exactly what you need though. Its small, light-weight and CAN create amazing images. Just remember, no matter what kind of camera you use, the pictures are only as good as the lens you put infront of it. The best camera in the world, shooting through a crap lens, is going to let you down every time.