So Blackmagic Design, maker of post production hardware and software, introduced their first camera at NAB 2012 – the Blackmagic Design Digital Cinema Camera. What to make of it?
UPDATE
On further thought, I’m being too hard on them. Here’s why.
First off – $3000. A phenomenal price point. And corners had to be cut, and the deadline was short, to get to this. YES, it uses their standard power source they use for other stuff. It isn’t ideal, it’ll fall out at inconvenient times, but they have’em, know’em, and can provide’em cheap. Same goes for some other decisions I’m sure.
To recap:
SUPERMODEL
-2.5K RAW
-$3000
-claimed 13 stops of latitude
-cheaper than buying a ProRes/DNxHD recorder, DaVinci Resolve, and Ultrascope by themselves! Consider it a “free camera with recorder/software purchase” – a killer deal!
-ProRes/DNxHD/uncompressed CinemaDNG recording
-reasonably sized sensor (not 1/3″ or 2/3″)
-12 bits!
-touchscreen on back
-can enter shot-by-shot metadata
-metadata flows into FCP-X via an XML – smooooooooooove!
-supposed to ship early summer
PURPORTEDLY NINJA CLAWS
-internal battery – non-replaceable
-when power runs out, need to hook up another battery, and attach it where?
-hope that power cord doesn’t fall out either
-who is going to make the battery to power socket cable?
-brick/fat paperback ergonomics
-twin handle ($200ish) thing doesn’t help you control anything, just gives you something to hold onto – hey guys, if there is time, keep rethinking this! A single handle underneath that allows operation with the other hand? Think old 8mm camera form factor….just sideways to show the screen and point the lens in the right direction!
-limited mount points (I think a 1/4-20 top and bottom or similar? But this is about the same/more as a DSLR. But this is a video camera ground up, not adapted stills camera)
-oddly sized smallish sensor – small enough to make a 11mm be a roughly 25mm lens – be hard to shoot wide on this. The famous Canon 8mm? About 18-19mm will be the widest you can get, AFAIK.
-no genlock
-no sync
-timecode sounds somewhat proprietary – record run only is what they described
-1/4″ audio inputs – not The Claws, but not what we really want
Ninja claws was overstating the case – the 5D that prompted that line was a line skipping, 30p only mess, only made worthwhile by the full frame, EF lens love. This is more like…she has sharp rings on. Mostly good, you can work with it and enjoy it, but rough edges, as it were. Grant and crew have made something that looks incredibly promising and could be total Awesome Sauce for the indie crowd – a helluva tool, just takes some working with to make it not lumpy and awkward.
In short – looks really exciting, could be super cool, form factor and other decisions annoying as hell. Of course, I gotta work with one before I can really say! These are just first opinions/reactions.
The workflow thought? All great- everything that happens in the camera with the image and bits I LOVE what I’m seeing them doing. In the physical, real world….the ergonomics and other decisions…Lack.
For the working professionals, it’ll be perceived as ninja claws. They like a low price, but they NEED certain features to do their job on non-simple shoots. For everybody else – smokin’ hot supermodel.
There’s been a lot of excited chatter on CML about this camera – a lot of enthusiasm, but also a lot of lamenting – no genlock, sync, proper timecode (including on the HD-SDI stream), non-locking power connector, no battery swapping, and other issues that make it so-close-but-dammit-not for “real” production. Then again, this camera (and price point!) weren’t designed for that crowd, so….get over it. For the money, if it does what they say it will, it’ll ROCK, and should do quite well.
end update, back to original article
======================================================
Just The Facts
My takeaway of the facts, based on that email and a discussion with Kristian Lam, the Product Manager for the camera on the show floor Monday:
$2995, ships June
The Sensor
-2.5K (2432×1366) sensor, 15.6×8.8mm
-roughly Micro Four Thirds sized, for comparison
-records up to 30p
-13 stops of claimed latitude
The Recorder
-records to SSDs that mount internally on the camera
-generic SSDs supported, you don’t have to buy BMD branded stuff
The Formats
-records to ProRes LT/422/422HQ, DNxHD up to 220 mbits, or to 12 bit CinemaDNG
-in CinemaDNG (an open standard created by Adobe), roughly 5MB/frame, about 130 MB/sec according to a BMD rep (presumably including audio)
The Body
-square box with lens on front, 5″ 800×480 touchscreen on back
OK, so where do you mount, um, anything?
-shown with a light hood and twin vertical side handle design – like a W with the V in the middle being the camera, or like this:
!__!
put the rectangular block camera on the __, about the same height as handles
-Canon mount for lenses
-fixed, non-removable internal battery, with 90-120 minute runtime
ports in: LANC-C (start/stop, iris control), headphone jack, non-standard power input 12-30V (or was it 12-18V?)
ports out: headphone jack for monitoring audio, 10 bit 4:2:2 HD-SDI out, Thunderbolt (for monitoring,recording, or downloading), LANC-C
-back touchscreen works like a smart phone – contextual screens, touch button keyboard, etc – allows for entering metadata
-plus a row of physical buttons at the bottom as well for record, rewind, play, fast forward, menu, and power
-a “step click” button at upper left of camera to control iris
-Autofocus? Not at this time
-XLRs for audio? Too big to fit
-Genlock? NO
-timecode embedded in HD-SDI? Not at this time, MIGHT be a future upgrade
-timecode works in a record run type modality, but there was talk about it being BMD’s own timecode -can’t take in a signal from elsewhere
-I heard 90-120 minutes, I heard 2 1/2 hours of run time on the battery
-if the battery dies dead as fried chicken, replaced under warranty (send it in) or replaced for a nominal fee (send it in still)
The Software/Features (that I’m aware of)
-peaking for focus assist
-includes Ultrascope for waveform/vectorscope/other monitoring
-includes a full seat of DaVinci Resolve for grading
-can enter shot metadata which gets included in an FCP-X XML file so your notes flow effortlessly into the app – NICE!!
Just The Analysis/Conjecture
I had a brief conversation with Kristian Lam, the product manager at Blackmagic for the Cinema Camera.
I asked a bunch of questions that he answered, and overall he was helpful, knowledgeable, and enthusiastic about the product. However, I was not as enthusiastic about all the answers he gave.
For example –
I asked about power – the battery is internal. I asked about what happens when it runs out – he said you can recharge it (lame if in the field), or you can plug in an external power source – AC line or another battery. Again, lame – now you have to figure out how to connect and attach a battery – where? There’s no mounting points on the camera. It is clean and sleek and…incomplete.
Along those lines, the camera itself really is a rounded brick – no handles of any sort. The proposed rig they showed had two side handles. OK. But who wants to hold it out in front of your face with both hands? And then how do you control, um, ANYTHING since both your hands are needed on those side handles to hold it up? One handing it? Not feasible with that rig. Where does the external extra battery go? How is that screen in daylight? What if I want to shoulder mount it to shoot for more than 10 minutes without my hands getting tired?
When I asked about audio, he said there was a headphone jack in and out for recording and monitoring. OK – what about levels on screen during recording? He paused, and my guess is that he was thinking about whether they had the computational horsepower to pull that off, or that they’d already discovered that they didn’t, and he had an unsure look on his face. Or worse yet, he hadn’t thought of that yet. Because his answer was something along the lines of that is a good idea, we’ll see (not actual quote).
Overall, this camera feels unbalanced – and I’m not just talking about the physical form factor. I think they deserve huge props for the price point, bundled software, Thunderbolt connectivity creative thinking, recording methodologies and post integration. AWESOME SAUCE.
However, BMD’s experience is in the above listed kudos. What they DON’T have experience in, and it shows, is making a physical device where the ergonomics, the physical human interaction is critical -and for a camera, it just IS. How to hold, monitor, power, connect, and add other things are critical if you care about your image. And while they’ve made a camera that can record a given image extremely well (RAW/ProRes/DNxHD/metadata), they haven’t shown similar care to how you GET that image.
In their press materials and talking points, they discuss the resolution…but never the sensor size. I actually think a smaller than Super35mm sensor size is a good idea for low cost cameras that are likely to be solo operated and not have a crew to pull full frame depth of field. But they seem to be most excited about what happens AFTER the image comes in the lens and hits the sensor, not what happens DURING.
Along those lines, if you’re going to call it a Digital CINEMA camera, should it not have digital cinema like properties, such as cinematic depth of field? I think the sensor size is OK/valid for this price range, but one of the characteristics that I use to define a digital cinema sensor, especially with so many Super35mm sized options is….a Super35mm sized option. Then again, this is their FIRST camera, so there’s always next year (or the year thereafter more likely).
Because ergonomics, and replaceable batteries, and a place to mount a microphone (oh yeah! That!) are critical if you care about your image crafting. Now, all the DSLRs have a lot of these issues as well, but they were graft-ons to existing tech, not a ground up approach. But when a new DSLR launches, there’s already a world of rigging around it from Zacuto and Red Rock Micro and the like. I didn’t see ANY of that in the BMD booth. Maybe they wanted to be stealth and quiet, I dunno – but there were rigs for other recently announced cameras, and rig vendors had a few weeks to throw something together.
Basically, I think what happened is this (TOTAL conjecture, alcohol may have been involved)
-last November, Red announced that Scarlet as a 3K for $3K camera was dead
-BMD noted and lamented that, then over beers somebody said “Why don’t we just buy a sensor and bolt it onto one of our recorders and call that a camera?”
-…and that was a good idea as a place to START
-they then launched into a rapid product development cycle, found a sensor, built a camera and by golly got it working in time to show at NAB – quite an accomplishment!
-But along the way, perhaps due to time constraints, that is ALL that they did – take the stack of boards that makes up a ProRes/DNxHD/RAW recorder, then slap one more board on the front of the stack that houses the sensor to punch through the hole for the lens mount, then put one more board on the back to drive the touchscreen and UI for the thing. Oh, crap, it needs power! OK, throw a battery in there and put it i a nice rugged sealed piece of aluminum like we do our other stuff.
But their other stuff lives in a computer. Or on a rack. Or is attached to a camera, but powered by the camera or other “main source” of power.
Some have said cameras are becoming just computers with lenses on the front nowadays, and this does nothing to disprove that theory – it is a box with a touchscreen on the back and a lens mount on the front.
Much like last year’s uncompressed recorders from BMD, these look like a minimal, perhaps rushed industrial design. Blackmagic makes a variety of inexpensive, powerful video capture/conversion/recording devices, and this looks like they took the boards to do that, put them in a stack, and put them in a box. Once they had that, it appears they decided to put one more board on the front with a sensor and a lens mount, and another on the back with a touchscreen.
I predict this will be awesome yet flawed, at least in the beginning – to steal a line from Stu Maschwitz when discussing the early 5D Mark II’s lack of 24p, it was “like getting a back massage from a super model…with ninja claws.” Oh so awesome in theory to think about behind the scenes, yet torturous in actual implementation. I’d love to be wrong, but the virtually unholdable brick form factor, inattention to details like pro audio inputs and levels while recording (remember, kids, sound is half the picture!), lack of mounting points, and fixed internal battery make me think this was a rushed, post-death-of-3K-for-$3K-Scarlet decision.
So where does this camera fit? For starters, it fits awkwardly – without the handles they were showing with the camera (two side vertical grips), it is an awkward box, like looking at the back of a small but thick book held sideways. On the good side, this camera appears to be going after the hole left behind when Red Scarlet’s “3K for $3K” went away and the Scarlet became a junior varsity Epic (“cut from the first squad” was overheard as a description). And that’s a good thing – a low cost, makes-a-good-1080p RAW shooting camera I think will be a boon to the indie filmmaker crowd.
The big sell on this camera in their booth is 2.5K, RAW/ProRes/DNxHD, $3000, 13 stops of latitude. Those are all excellent selling points, and if they work as advertised, I look forward to them. Interesting to note is that twice when I asked what the sensor size was, I was told the pixel resolution. When I asked about the sensor SIZE again, I did get the answer I was looking for – the physical dimensions of the sensor, 15.6×8.8mm.
This is not the Super35mm size of the Scarlet/Epic, Sony F3, C300, FS700, or any of the DSLRs. It won’t create the same kind of shallow depth of field of those cameras. Mild bummer. But I think as a practical matter, this will be GREAT for a lot of reality, documentary, one man band and other shooting forms that don’t have the luxury of someone to accurately pull that shallow focus. But for this price point…hey – it’s OK!
They say they’ll ship it in June or July, I forget which, but that’s an NAB July™®, which I have learned from years and years of experience at NAB to interpret as “September or thereafter.” June is right around the corner from mid-April – “Oh, we’re almost there!” From a sales perspective, it puts it tantalizingly in reach. That is the cynical view. Optimistically, as much as hopeful product developers might honestly mean that, it is so rarely the case. Some part gets held up, or a glitch gets found, and delays add up to…September. So no disrespect to BMD, this is just what I see ALL the vendors doing over the years.
I think the biggest challenge they are likely to face on this is going to be how is it as a CAMERA. BMD has a lot of history and experience making post gear, but this is their first image acquisition device. Much like Red spent a lot of time fleshing out the details of all the nitty gritty implementation stuff – sync and genlock and audio levels and and and and and….the bar is high these days for what is needed/required/expected of a camera. If they are going to bother trying to get 13 stops of uncompressed RAW image out of it, along with uncompressed audio, if you can’t properly monitor the audio live, you’re stepping on your own shoelaces.
For the money, I expect this camera to be a fantastic deal, tremendous bang for the buck. I also expect that more than the cost of the camera will be spent to make it shootable with rigging, microphones, SSDs, etc. – even mores than HD-DSLRs already on the market. BUT – maybe this will be like the SI-2K, another computer-with-lens-attached design – who eventually partnered with P+S Technik to get a much more shootable form factor – maybe next year we’ll see something a lot friendlier to work with – but this is an ambitious start.
So shooting – expect it to be not as easy as it should or could be. Post, on the other hand, ought to be pretty sweet – presuming the sensor has good enough color separation, detail, and dynamic range, easy flow ingest of metadata into FCP with ProRes, or at least ready-to-roll DNxHD for Avid, should be good. Adobe’s ACR support will hopefully support this (I presume, need to ask) – I use Adobe Camera Raw for a lot of my own personal work (mikecurtispictures.com), and I LOVE the control it gives.
Overall, by the specs, just on paper, it is like a baby Alexa – 2.5K, 13 claimed stops, ProRes/RAW/DNxHD recording. In practice, it will be incredibly far from the shooting experience of Alexa due to form factor, color science, integration with other shooting requirements (timecode, genlock, form factor, pro lens options, yadda yadda) and a zillion other details.
But I am concerned that either the price point will be too low for the effort required to really finish development appropriately, or that BMD won’t allocate enough resources to flesh out the needed feature set in a timely fashion (again, see Red’s early days in terms of more work to be done than they anticipated). BMD has a history or releasing hardware with features that aren’t quite properly supported by the firmware/software at first, and then trickling out the updates (sound like another camera company out there? Color?) There are just SO many details that need to be done right to make this both practical and take meaningful advantage of what the core purpose/strength/goal of this camera is – make pretty pictures that you have a lot of grading control over and that flow into post readily.
$3000 would have been a fantastic price point for a ProRes/DNxHD/uncompressed RAW recorder ALONE two NAB’s ago. To get a camera, recorder, high end color corrector and software scopes (hello MacBook Air HD scopes/waveform!) is an AMAZING deal.
Moves like this are good to push the industry forward – I used to be so frustrated by Canon, Sony, JVC and the like who put out cameras with seemingly intentionally throttled recording capabilities (Sony especially – hello F3!). The rush of inexpensive ProRes/DNxHD/uncompressed recorders in the last couple of years has sent a clear signal to camera makers – “You can no longer control the power/capabilities of a camera by throttling the recording – you HAVE to have a signal out for monitoring, and we can record over that with inexpensive devices.”
While they may still do this recorder throttling to try to segment markets, if buyers can get a ProRes recorder for $1000, that technique is effectively dead. DSLRs still limit resolution on recording output, but at least they have a horsepower excuse.
Anyway, I digress – what we have here is a fantastic recording device, a nice post workflow, a lumpy form factor, and a fantastic opportunity for growth.
Why am I being so hard/critical? Because I care, because I think this is close to awesome in the classic HD for Indies, power to the people way, and I want it to WORK. I’m not so analytical/critical if I don’t care.
-mike
PS – OH – and pre-RIP Digital Bolex – sorry guys, somebody bigger just did it instead. : (
I love their concept, but I think this kills it.
I mentioned this to someone in the BMD booth, and they commiserated, something along the lines of the saw the announcement of Digital Bolex and winced, with an honest “Ouch…sorry guys!” since BMD was already well underway when they saw the Digital Bolex announcement. Digital Bolex is to BMD what SI-2K was to Red – a bigger fish just swam into the same pool, you’re gonna lose.