Right tool for the job...

[rpcrowe]

I don't use a screwdriver as a chisel and I don't usually hammer nails with a ball pein hammer...

I have never liked shooting video with a DSLR, although I do see a few advantages (specifically shallow DOF) but, these advantages are outweighed by many disadvantages (especially in the short throw of the focusing).

Although a person might use a DSLR for professional video production, I would expect that this series would be superior for professional video use...

http://www.bhphotovideo.com/c/promot...on-cinema.html

They are certainly more expensive but the price should not hamper pro video production. Especially since many professional videographers and cinematographers rent their camera/lenses rather than purchasing them outright.

[Manfred M]

What's that old saying; "When your only tool is a hammer, everything starts to look like a nail".

You reflect my sentiments exactly, Richard. Almost anyone who has shot with a high end video camera scratches his or her head wondering what the big deal of using a DSLR to shoot video is about. By the time you add a reasonable focus assist (Zacuto being the most common) and a decent sound recorder; you are starting to look at the same money as a dedicated video camera.

The whole DSLR video has been driven by two factors:

1. The large sensor and the narrow DoF opportunities it provides; and

2. The thought of getting something "free"; i.e. I bought a DSLR and I can shoot video too.

I always suggest that we are really talking the same issue as using a smartphone to take pictures. In the right hands; you can get wonderful images. Ditto for a DSLR and video; yes you can get great video too; but it sure is easier with a dedicated camera.

I essentially have asked, so if you won't use your smartphone to take pictures; why are you planning to shot video with a DSLR. That being said; for the most part, the DSLR shooters are getting good results, as long as they follow good shooting practices (like getting and using a proper tripod). I've seen too much amateur video the causes motion sickness....

[Colin Southern]

I agree Richard, although I was surprised at the quality of the House season finale that they shot entirely with 5D2 cameras - not something you'd normally associate with broadcast quality.

Obviously the camera men knew their stuff, but then again, we've been trumpeting photographer knowledge as the main ingredient for SLR photography for years. Perhaps when it comes to video on our SLR cameras, it's still a case of us just not knowing how to use the tool properly yet?

[Manfred M]

I read an interview of the folks involved in that episode somewhere. They interviewee said that they shot in a studio under ideal conditions and that was one of the key reasons that they could pull it off. The MkII's were fully fit up with matt boxes, follow focus gearing; i.e. what they used was in no way an out of the box camera; they threw a lot of expensive accessories onto the cameras. The used a focus puller just like they would with their regular shoots. They fully admitted that it took a lot of work by seasoned pros to pull it off.

The fact that this has not been repeated should tell us something. As a proof of concept, yes, it worked. Is it ready for prime time, no.

[NorthernFocus]

The whole DSLR video thing is a pet peeve of mine. How many extra hundreds of dollars does video add to a high end camera body? Surely it's no where close to being free. Statistically I wonder how many people use it? One thing I learned a long time ago about most any equipment, when it is designed for multiple purposes it does them all half.... way and isnt' optimal for anything.

I suppose there is an argument that video required higher processing/writing speeds then still photography and our DSLRs benefit from that. Maybe...

[Colin Southern]

The whole DSLR video thing is a pet peeve of mine. How many extra hundreds of dollars does video add to a high end camera body? Surely it's no where close to being free.

I don't use it either, but according to Chuck Westfall - yep - it pretty much is "free". It's only software -- no hardware changes.

[inkista]

Just me as a not-a-videographer-at-all person, being a boring geek listening to commentary tracks, but Joss Whedon used 5DMkIIs and 7Ds on The Avengers and Danny Boyle used them (and the 1DMkIV) on 127 Hours. Not as main cameras, obviously, but they seem to be hitting Hollywood in a big way--not just stills shooters who want video "for free." The House finale was far from an isolated incident. For mainstream Hollywood, these are the small, trashable cameras. If you need a dozen of them, dSLRs are the way to go. You can get 10 5DMkIIs for the price of a single RED. And no way does adding a stabilization rig and focus pulling gear add up to, say, the addition $28k of a RED over a single 6D. Not to mention the lenses are cheaper--even if they do breathe.

For some folks "HDSLRS" are the right tools.

I've watched an entire feature film shot on 5DMkIIs. Dave McKean shot his film, The Gospel of Us on 5DMkIIs, mostly because he was on a shoestring budget. It's a sorta-documentary (if you know Dave McKean's artwork, you know what I'm talking about) of the Port Talbot Passion. This was a two-day passion play by the National Theatre of Wales, starring and directed by Michael Sheen. McKean had to cover a live performance that was going non-stop for two days at multiple venues simultaneously. Hence the need for 10 cameras. It's kind of insane how good the film is, given the circumstances under which it was made.

I'm just saying this is hitting filmmaking and television at all levels from Hollywood on down. It's not necessarily jamming a square peg into a round hole any more. This is a viable filmmaking tool with limitations, but also with a very low pricetag and tiny size compared to the more capable gear. The main reason for its use on the House finale was that the story took place inside a very small enclosed set. The dSLR could get into tight spaces the more conventional cameras couldn't.

And of course, for the folks used to working in 35mm film--a full frame sensor has all the appeal of having an image plane twice the size of the one they're used to.

And whether the 70D's PDAF tech will change things even more, who knows? At least casual home movie makers won't need follow focus rigs and geared lenses any more.

[Manfred M]

You are right Kathy; I've seen a number of writeups and videos where they are used as the "B" cam because of their small size and low cost; stuffed up near the ceiling or on the floor. On the other hand, the Arri, Red and Panavision cameras still seem to be the "A" cameras.

They also seem to be catching on with some reporters on field assignments, where they can post stories with still and video images (quite important in the multi-media world we live in).

We are moving towards convergence and we will know we have gotten there when they start being used as the "A" camera. I find the Panasonic lenses that have a built in electric zoom to be the next step in development here. Once we start seeing some true parfocal interchangable lenses; that will get us closer.

The real issue is the amount of R&D money that these companies are going to be willing to invest in a developing a true crossover camera. The features that the amateurs are looking for (autofocus, auto exposure) are not going to be the same ones the serious videographers will be looking for (stepless aperture control, fast glass, phantom power mics, focus assist tools (like peaking) and parfocal lens design, etc.). Canon, Sony and Panasonic risk canabalizing their own lucrative pro video camera market if they go this way. Some of their recent video camera offerings seem to reflect this reality.

[inkista]

You are right Kathy; I've seen a number of writeups and videos where they are used as the "B" cam because of their small size and low cost; stuffed up near the ceiling or on the floor. On the other hand, the Arri, Red and Panavision cameras still seem to be the "A" cameras.

Yup. IIRC, the 5DMkII on the House finale was actually the "C" camera: the rover they weren't sure was going to yield any useful footage.

We are moving towards convergence and we will know we have gotten there when they start being used as the "A" camera.

I don't know. The price differential between the two sets of cameras is so huge that I'm not sure we could ever really expect convergence. $2k vs. $30K is still a big price leap to expect the same feature set, particularly since a lot of those features are lens-dependent, more than camera-body dependent. And, as you say, cannibalizing their pro video stuff is not good.

Probably the only place where dSLRs are and will remain the dominant main camera in filmmaking is going to be stop-motion animation or time-lapse where it's the stills quality, not the compressed video that's the output used. [I still crack up every time I think about the picture Peter Lord tweeted of "all the little cameras" they were using in Pirates!]

I think the farthest towards "convergence" we're likely to see in the near-term future would be 2K or 4K coming from dSLRs.

I'm also curious to see how the RAW video from MagicLantern is gonna affect things.

[Adrian]

It seems to me that, whilst interesting, pro use of DSLRs for film is a red herring, though a helpful little fish in that it shows what can be done.

We have bought two DSLRs recently, and the fact that they can do video was a large part of the decision. We produce web video every day for commercial posting (effectively disposable financial news) and the DSLR is a cheap, compact and flexible solution that delivers ample quality. If you compare a good DSLR with a consumer grade or semi pro video cam, the end result is indistinguishable. And as we still needed a stills camera the DSLR was a no brainer really.

We have invested in sound equipment - but frankly it's peanuts and some of that cost we would have incurred anyway with a video cam (especially the off camera microphone set up we are now moving to). A very capable digital sound recorder, such as the Tascam DR60D (http://nofilmschool.com/2013/04/tasc...ers-4-channel/) is under $500 and produces excellent results (as do broadly similar zoom products) and syncing is a piece of cake. We are experimenting with this kit at the moment before using it live, and so far I have been very impressed.

Clearly video has some specialist techniques and special lenses help, but in terms of accessibility to the wider market or even moderate intensity business use, the DSLR solution is very good value as a "multitool".

Adrian

[Manfred M]

The ones that I really love are some of these rigs; by the time they finish dressing up their DSLR, they could have gotten a proper video camera; although to be fair, you can dress up a pro video camera in much the same way, minus the finder:

http://www.zacuto.com/dslr-rigs

Adrian - you are shooting in quasi studio conditions (yes; I know you have set up in a corner of the office) and are streaming on the internet; so that makes perfect sense and is a good application for a hybrid setup.

Kathy - Other than wishful thinking, the primary problem that I see with the DSLR versus dedicated video camera that will keep convergence from happening is data processing and throughput.
What attracts people to large sensor cameras, we have discussed in previous postings. The reason that we can use DSLRs for video is that the data and processing channels align reasonably well in current cameras. If I look at the raw processing power side of things; pushing a 1920 x 1280p video image at 60fps through my D800; if takes virtually the same bandwidth as the maximum burst speed mode of the 7380 x 4912 at 4 fps (both run at around 145 MB/sec). As long as the processing requirements of video and burst mode shooting match reasonably well; a camera can handle both types of shooting. This is why the video functions are essentially “free”; because the basic camera circuitry can be shared and the only changes are going to be in software (although I do have a separate video trigger on my camera).

Increasing the burst frame rate on a DSLR also means that the mirror and shutter have to cycle more quickly; but this is obviously not a problem as we see in the Canon 1D or Nikon D4 frame rates; I suspect that these cameras are limited by the data and processor throughputs as well. Could frame rates and processing speeds be improved; of course; but that will mean more processing power, which in turn will reduce battery life and increased costs. While the 1D and D4 are large; they aren’t even close in size to my Panasonic AF100 video camera; so some of the basic engineering trade-offs. As an example; the battery that powers my D800 is rated at 1900 mAh while the one on the AF100 is 5400 mAh; which means it is roughly three times the size.

Further refinements in the chipmaking process are going to reduce the size and hence the power requirements of the chips, so the next generation of cameras are going to offer even more performance.

[RustBeltRaw]

[I still crack up every time I think about the picture Peter Lord tweeted of "all the little cameras" they were using in Pirates!]

48 Canon EOS 1D (MkIVs?) in that picture. About $150,000. Clearly film budgets are high enough to get whatever tools they want. Think they'd notice if one went missing?

There are some video features that I'd like to see from the manufacturer on DSLRs, even for photo-only uses. Focus peaking, live highlight clipping warnings, live RGB histograms, external white balance controls.... One of the neat side effects of DSLR video is that more and more still shooters are becoming aware of these remarkably helpful features.

[rpcrowe]

There are a few things that I don't like about using a DSLR for video work...

1. You must use live view which is difficult for me (and perhaps others) in bright lighting conditions. I like the choice of viewing using the LCD or eye level. Of course, a Hoodman viewer can make the LCD viewer approximately like eye level viewing...

2. The focus throw (turning radius between closest focus and infinity) is quite short (and rightfully so) in a lens designed for still photography. However, this short focus throw makes it difficult for me to "follow focus" smoothly. Smooth follow focus is another creative implement in the videographers tool box.

3. Lack of a motorized zoom. Although I mostly use my zoom in shooting video as a variable focal length lens and don't a lot of trombone-type zooming while actually shooting, there are times in which judicious use of a zoom adds to creativity. I can effect a smooth zoom transition much more easily when I am using a motorized zoom lens than when I am trying to manually zoom with a lens designed for still photography. However, I could do a better job of smoothly zooming with the manual zoom 12-120mm Angenieux cinema lens than with any still camera lens that I have tried.

BTW: as an after thought... When shooting video with two DSLR cameras with the intention of doing an A - B edit, it is handy to use a white balance card like the Whibal for aligning the white balance between the two cameras so the results can be cut together smoothly without a lot of color balance editing...

[inkista]

Kathy - Other than wishful thinking, the primary problem that I see with the DSLR versus dedicated video camera that will keep convergence from happening is data processing and throughput.

Completely agreed. Throughput (not resolution) is the only thing that keeps a dSLR from not being able to deliver 4k. But I have faith in Moore's Law holding out a bit longer. And while we may reach a limit on multicore processors at some point, image processing is one of those tasks that's kinda gloriously parallel and can take advantage of multicore/multithread processing easily without a lot of convoluted scheduling hassles (hence the dual-DIGICs in the 7D and 5DMkIII). Power consumption reduction, too, is something that feels inevitable these days, given the need/push for it from the cellphone side. And, iirc, the DIGIC is ARM-based, so its engineers are probably all on top of that.

Not saying it won't cost to up the cores and the buses. Just not thinking it'll cost enough to stop it happening, given the overall trends in SoC design. The throughput might actually end up eclipsing the physical burst rate capabilities of the shutterbox mechanism.

There are a few things that I don't like about using a DSLR for video work...

1. You must use live view which is difficult for me (and perhaps others) in bright lighting conditions. I like the choice of viewing using the LCD or eye level. Of course, a Hoodman viewer can make the LCD viewer approximately like eye level viewing...

Actually, I believe the solution du jour for this is a separate monitor. Hence the importance of HDMI liveview output from the camera. And, of course, MagicLantern adds focus peaking. But I think one of the appeals for the Panasonic GH cameras for video is that you can still use the eye-level viewfinder for video. Mirrorless does have some advantages.

2. The focus throw ...

Which is why the vintage manual lens market suddenly underwent a huge inflation in prices after the 5DMkII came out. The old Leica-R and Contax/Yashica Zeiss lenses in particular. HUGE, damped/smooth focus throws. When I purchased my C/Y Zeiss Distagon 28/2.8, it cost me a whopping $230. Today, it goes for closer to $500. Ditto my Olympus OM 50/1.2.

... BTW: as an after thought... When shooting video with two DSLR cameras with the intention of doing an A - B edit, it is handy to use a white balance card like the Whibal for aligning the white balance between the two cameras so the results can be cut together smoothly without a lot of color balance editing...

Wonder if that's part of the clapperboard setup these days. And, of course, the hot thing of late with Magic Lantern (well, previous hot thing. Current hot thing is the dual-ISO mode), is adding the capability of RAW video. Having to use the compressed h.264 MOV output was a real thorn in the side of a lot of folks. Like knowing you could shoot RAW, but being forced to do JPEG. For color-grading that must have seriously sucked.