By now, the dust has settled after Blackmagic Design dropped their latest bomb, the 6K version of their Blackmagic Pocket Cinema Camera that uses Canon EF-mount lenses. The product was ready to ship upon its announcement (compared to waiting months after that 4K launch), and just as quickly, this space got flooded with anxious, speedy reviews that tended, as usual, to be binary: buy it, or not?
Well, let’s slow down and most importantly watch a formal test shoot, closer to real-world usage. The featured video here builds upon practices learned (often painfully) using the BMPCC4K, because they’re almost exactly the same camera. Also importantly, it’s shot in the way I’d shoot something I care about (as I’ll be using much of this in a new addition to “Whitman on Film”): following formal cinematography principles in framing, exposure and color. As I note in the optional commentary (or, you can just mute it!), the footage is essentially raw from raw: I ingested the Blackmagic RAW (BRAW) 6K footage using Autokroma’s BRAW Studio plugin (about which I created a long-form tutorial), to stay within my preferred NLE Adobe Premiere Pro, but I did not apply any plug-in, or even Lumetri Color. Given the nature of raw video in a log profile, it’s a mandatory step to coddle the footage with source-level adjustments to luminance and chrominance clip-by-clip, so I essentially color-graded — and applied the official Blackmagic LUT — at the BRAW Studio level by only adjusting the raw metadata. It’s really a faithful representation of the camera’s base capability without interference. Normally I’m tempted to stack something like FilmConvert for stock emulation (even though that causes most of my crashes), but this was a camera test, not art.
So, the images speak for themselves (and I speak over them, about some key issues). But the following written sections dive into a few trailing subjects that you may be interested in.
CANON EF MOUNT & SUPER 35MM-ISH SENSOR
The permanent adage holds, that you will/should spend way more effort + money on building your lens arsenal, than any given camera body. So the choice of camera is radically affected by your current lens kit, right? Of course. That’s the number one hugest factor why you’d go for the BMPCC6K, apart from the bump in resolution and sensor size, if you’re already a Canon user. In general, investing in EF-mount glass (especially for cine lenses) is the better proposition, because it adapts down to just about anything (E-mount, Micro Four Thirds, etc.) due to its outdated gigantic flange distance that was originally designed to accommodate deep inset single-lens-reflex prismatic shutters (i.e., something irrelevant to video). Also, lots of EF-mount glass is full-frame (compared to EF-S crop), which gets you ready for that inevitable time when we’re all finally shooting full frame, like I do on my Sony a7 III whenever I want better low-light sensitivity and depth-of-field latitude, if I can live with Sony’s inferior 8-bit 4:2:0 H.264-based codec.
So yeah, longer flange distance, and bigger sensor coverage, are in the plus column for EF-mount and the BMPCC6K. But in practical application, I see little difference between the Pocket 4K and 6K mounts, and sensors. I already have a toxic stew of EF, MFT, and E-mount glass, and since these Blackmagic cameras lack in-body image stabilization at the sensor, in-lens optical image stabilization (O.I.S.) is especially critical given the Pockets’ ornery form factors (e.g., no third point of contact without an eyepiece). MFT lenses have a distinguished history of nailing O.I.S., generally better than EF lenses, and if we really do care about the Pocket being (nearly) pocketable, MFT lenses are also more portable. (Many blogger types also emphasize here, that focal reducer mount adapters, including but not limited to Metabones, further close the gap between these models. Yet I’m personally, permanently averse to sandwiching more layers of glass between sensor and lens, and risking incompatibility where modern lenses rely upon electronic control of aperture and focus. Basically, native-lenses-or-bust is a smart practical decision; life is short.)
As to sensor size, it was never so simple. I was getting bad vignetting at the 12mm end of a MFT zoom lens on the BMPCC4K, because that camera’s sensor is simply bigger than a typical MFT sensor, especially if you’re shooting in DCI aspect. Meantime, Super35mm was never an exact measurement, including variances like APS-C, and so the difference in sensor size further narrows — both still far away from full frame. And after increasing resolution on the BMPCC6K sensor, the photosites are more dense and less sensitive. All told, low-light capability is basically a wash. Same for control over depth of field. Whichever aperture value you choose relative to lens quality, is practically everything (especially given the bonus feature of dual native ISO).
TOO DARN HOT VS. HVAC
One of the biggest surprises, arriving into the BMPCC6K, was its substantially bigger fan/ventilation, compared to the BMPCC4K which actually didn’t ever produce reports of overheating. Maybe the marginal resolution bump demands enough extra number crunching to push the chassis into furnace territory, but an obvious concern now is weatherproofing. I don’t shoot for NatGeo or anything, so maybe it doesn’t matter much, but what I did run into is the fairly universal need to mount the camera onto a tripod, using a common quick-release plate such as the Manfrotto seen at right. All that’s left is to poke it out the back, to avoid blocking circulation, which really throws things off balance especially if you’re mounting a heavy lens. But there’s a little good news, at least: the fan is whisper-quiet, even when it’s hot outside and when your shot runs long.
COMPATIBLE STORAGE MEDIA
Let’s start with the complaint that CFast 2.0 cards are just too expensive and too scarce, if we have other options. And we do. The obvious choice is simply to plug a portable SSD into the BMPCC6K’s USB-C port. I’ve done that on my BMPCC4K, but it required some awkward rigging, with both a cage, and a proprietary SSD mounting sled — going even farther away from a “pocket.” But for this shoot, I really needed to run-and-gun, powering off those short-lived LP-E6 batteries by simply stashing one-half dozen best-in-class knock-offs into my pockets. I actually burned through only half of them; when shooting short bursts (e.g., gathering coverage, etc.), freed from the liability of missing one second out of unrepeatable minutes, those Canon batteries work out just fine (and the new battery sled designed for the 4K, will work on the 6K). But, what about portable storage media?
Same strategy: buy a bucketful. Strangely, there’s just one sensible option on the market at this time, pictured here. Reason is, if you avoid CFast 2.0, and go with UHS-II as the internal alternate, it’s critical to buy the speed class labeled “V90” which guarantees a minimum sequential write speed of 90 megabytes per second; theoretically, this Transcend reaches 180 megabytes per second for headroom, but you’ll need as much speed as you can get for both ProRes and BRAW at 6K resolution. (And it doesn’t hurt that the read speeds, when properly connected via USB 3.2 Gen.2 UHS-II card readers, go through the roof as pictured at right.) Weirdly, at this spec Transcend maxes out at the 64 GB capacity, but it’s stunningly cheap at around fifty bucks each, while competing higher-capacity cards offer bad cost-per-gigabyte value — and if you can manage it, there’s wisdom to breaking apart your shooting day into multiple smaller cards, reducing risk of loss anyway. Given the extraordinary efficiency of BRAW at reasonable compromise ratios like 12:1, you’d be surprised how much can fit onto a 64 GB card (I got about a one-third-hour) — everything you see in the video here, fit onto about three of them. But…
WHICH CODEC? THIS TIME, IT DEPENDS
I’m on a mission to kill ProRes, which is an unspeakable offense among the Church of Jobs, but let’s get real: it’s a monstrously inefficient codec, originally designed as a technology compromise for slow Apple computers to keep up in Final Cut as an intermediate codec. It was never meant to be an acquisition codec (i.e., shooting video), but for a while, it was the best we had, starting with the Atomos Shogun. Yet now we have BRAW, which fundamentally moots ProRes — especially now that Autokroma’s BRAW Studio gets the stuff seamlessly into Adobe Premiere. Case closed!
Well, no. When you really think about it, raw means raw: if you want to store what you shoot into a format that doesn’t match the native sensor specs, given any manageably simple conversion formula . . . then it’s raw no more. So with this BMPCC6K, the odd result is that you can shoot full 6K footage in BRAW, but if you want to shoot standard 4K-DCI or 4K-UHD footage, you’re stuck with ProRes. You can see this in the available menu options above, between resolution and codec combinations. It’s really frustrating! When you consider how 4K-UHD is now the standard benchmark resolution, having this camera that can’t shoot natively into its benchmark BRAW codec, is a real limitation. And it offsets the argument that many are making (and have acted upon), to upgrade from a BMPCC4K to a BMPCC6K.
Yet ultimately, are we any worse off shooting 6K?
6K: WHY NOT (AND WHY)?
The first, gut reaction to any 6K camera arriving in 2019 — Blackmagic is soon joined by the Z CAM E2-S6 and the Panasonic S1H — is that no one watches content in 6K. Of course, but once you’re past that, it comes down to over-capturing resolution into the actual video footage, to reap the benefits when you finally render out to UHD or even HD (besides the other benefit of being able to crop in post). Storing more resolution than you’ll eventually watch, is better than: in-camera reading out more resolution than you’ll watch, and down-converting it to a lower resolution for storage. That’s literally the history of bad video cameras, dating back to the high-megapixel sensor of the Canon 5D Mark II designed for stills, bayering badly down to two-megapixel HD video footage with harsh aliasing, etc.
One of my favorite reveals to film school students, is that the classic ARRI Alexa that still shoots the majority of all movies we watch today, doesn’t even record in 4K. Does that mean my Samsung Note10+ shoots better video at UHD+? The weird conclusion is, resolution matters and yet it doesn’t: the reason the Alexa looks so amazing, is that its sensor and processor deliver such immaculate definition and color science for every precious photosite — fewer, but better. But! When you don’t have/can’t afford an Alexa, there’s no getting around the fact that a wide shot with lots of tiny little tree leaves fluttering in the background, look better from the BMPCC6K, than from the BMPCC4K, when their respective maximum resolution footage gets scaled down to HD. This compensates for its lack of elite ARRI juice.
True, 6K footage takes a lot more grinding to churn through. Last month I built a new workstation around the epic Intel-killing AMD Ryzen Zen2 3900x (ignore Puget Systems), paired with Nvidia’s new RTX 1070, and this 6K footage flies fast off my PCIe SSDs. But that’s not a typical workstation, and you might end up needing to use proxies. Even with my Death Star machine, Adobe Premiere crashed — of course — reporting that it was a GPU error, which is just Adobe saying as usual, “not our fault!” But I had applied no GPU-accelerated effects, besides Adobe’s own Warp for which they hold sole responsibility. Point being, computer editing technology keeps crawling neck-and-neck, barely keeping up with this Moore’s Law-like rise in video resolution where 6K is becoming the norm — 8K is next. We keep needing to carefully balance our demands and values, against the rage of tearing our hair out.
GETTING EXPENSIVE: PARTY OVER?
Speaking of which, in life, when things seem too good to be true, they freaking are. The BMPCC4K is $1,295, and the BMPCC6K, double that. What’s different, performance-wise? A little more resolution — but not even doubling down across the whole feature set. Another fact: notice how the 6K is being shot out like T-shirts at a baseball game, but the 4K still has a line going outside the stadium since the past year? Business is business: there’s more profit in a camera that rakes in twice as much, but costs hardly any more to make (with most of the R&D already sunk into the first one).
It’s totally unfair and unreliable to prophesy that Blackmagic is the next Apple, becoming a beloved boutique brand that simply cannot afford not to become unaffordable (!). But I’ll never forget seeing that fancy new keyboard at NAB, thinking to myself “Blackmagic being Blackmagic again!” expecting a $100 price tag . . . then, seeing it’s the most expensive QWERTY keyboard known to humankind, at $1k. If RED folds under the weight of their extraordinary arrogance soon, Blackmagic will further dominate the market (and most of us cannot afford ARRI anyway). Is China next, after these years down under? I’ve got Z CAM’s new E2C and will be posting a review here soon.