Ehetyz

I'm very surprised of the versatility of this footage. A lot of Kinefinity footage I've previously seen online has been very brown and muted, this however is very flexible and easily on par with BMCC Prores. Perhaps even better - I'm seeing pleasing results much faster than usually with the BMCC. Getting very close to 5D RAW in terms of ease of grading. I was able to whip up a grade I'm happy with in a matter of minutes.

The camera never dropped from my hands. But shit happens on set and I need to have at least some assurance that when it does it won't instantly mean I become an owner of a 5000 euro brick. I've thought about the cage solution though. It might work. Other risks and caveats? The magenta issue. The screendoor effect (though this is nothing new, still irritating and seems to be more pronounced than on the BMCC), reported variance in ISO performance, getting a straight up bad one - Blackmagic's QA is all over the place. From my circle of friends there are multiple people who got faulty Blackmagic products. The BMCC has served me well, niggles and all. But I'm not putting blind faith on Blackmagic.

I don't think Ursa Mini will be the end of Blackmagic, but as a long time BMCC user, all the problems that have popped up after its release have made me wait and see if any of the competitors can put out a less risky proposition. The magenta issue was bad enough, but people bricking the camera by having the side handle take a hit is what kills it for me. I shoot often in difficult environments, do action shots and generally things that put the camera at risk. BMCC never failed me at that, it's taken a 1,5 meter fall to concrete like a champ, twice. I have no faith in URSA mini being able to withstand that kind of abuse. The Mini 4.6K is frustrating. The image seems to be pretty much untouchable at the price point - but with way too many caveats and risks.

With the raw recorder that's pretty much everything I'd want from a camera. Too bad it's about 3x the price I'd find feasible.

I have very similar memories of that era. I was at first very sceptical of HDV, because I'd seen only footage from the early Sony models that only did interlaced image with godawful washed out color. It seemed in some ways inferior to prosumer SD. I was actually thinking of picking up an XM2 at the very tail end of SD era, but then I looked into the HV30 and fell in love. That camera had the worst focus wheel, but other than that it felt revolutionary. I ended up shooting three indie features with it, and some footage still holds up to a degree.

Yeah I agree. That's a different post though and equating S-log2/3 from A7* with Log from Alexa is only technically true and a sure-fire way to set people up for disappointment.

Log coming out of an Alexa or Red (Or any camera with a proper 10+ bit codec for that matter) is a very, very different beast than Log in a 4:2:0 8bit colorspace and 100mbps bitrate.

I think the mythical "motion cadence", separated from rolling shutter, is down to how different codecs and cameras compress motion and how that affects motion blur. Pretty much without exception people pick RAW cameras as ones that have best motion cadence - because there's no compression or artifacting affecting the motion. Then after that come the cameras that do Prores or other flavours of high bitrate codecs, and dead last are the ones that compress the image to ludicrous degrees with interframe codecs such as Sony cameras or GH4. The best motion comes from a camera with a fast sensor readout and high bitrate compression or RAW. Imho that's all the magic there is to it.

I love 5D raw but this is kind of untrue. Having used both, BMCC definitely has the edge on 5D RAW when it comes to DR (unless you're using the dual gain hack, in which case they're neck to neck), but overall I think 5D RAW looks much more pleasing. It has better color handling than Blackmagic and you simply can't beat the full frame look. I've taken a habit of grabbing a few shots with both cameras on every shoot I do, and I always end up using the 5D Raw one. It simply looks better with a lot less grading work required. And I can easily see people grabbing the MK4. As far as video capable DSLR:s go, 5D:s are some of the most pleasing and hassle-free to use, with an image that's still mostly good enough for most low budget/internet delivery work straight out of the box. Canon's the Apple of cameras; they simply work, and they're easy and fun to use, which can't be said of the likes of Sony or Blackmagic.

Monitoring is SDI only - you can get scopes and preview image into a MAC through the thunderbolt connection, but it's not at all practical or useful. I've tried the SDI with one external monitor, it didn't work. Not sure if it's finicky or if my copy has bad SDI. Apparently the SDI gets fried easily if you connect it wrong or something. It's out of warranty so I decided to just live with it. AFAIK 1080p doesn't crop from 2,5k. Never had an issue overheating, it's rock solid in that regard. Have had a lot of days in +24-28 celsius summer heat and direct sunshine and had no problems. Had some minor things with very cold weather. In -10 celsius the startup gets slow and you get a lot of artifacts on screen during startup. They disappear quickly though and the shot footage hasn't exhibited any issues.

Going on a bit of a tangent here, but I'm constantly surprised by people claiming 5D2/3 RAW to be a hassle or difficult in post. I mean, sure, compared to 8-bit cameras with baked-in profiles it probably is, and you have one more step of transcoding, but compared to S-log or Blackmagic cameras, it's a breeze. Exposed correctly, 5D RAW is usually beautiful already in REC709 and only needs small adjustments, maybe an LUT, whereas BMCC and Sony require you to work over everything in the image to get usable results. In Resolve, I've found I'm satisfied with 5D Raw after a correction of 3-4 nodes. S-log2 and BMCC usually take 9-15 nodes. If I could get a 100% trustworthy Canon Raw camera, I'd never use anything else.

Am I the only one who finds the motion in several of those videos to look really strange? Especially in the Beautiful Riot and Akiyama-go - videos. Most of the motion is overtly fluid, like it's shot in 60i or very low shutter speeds. Or forced in post from PAL to NTSC. Very unpleasant look imho.

Well that's a known issue. You'll inevitably want to use some sort of ND solution and with that comes the need for an IR cut. The IR filter on the camera itself is so weak some people think it has none.

It definitely has a higher dynamic range. I think it might actually be the 13 stops they're rating it. I used to rely on gradual ND:s on pretty much every shoot when using 5D2, with BMCC they've been collecting dust most of the time, because even high contrast scenes tend to fit in the range of the camera. It has noise in the shadows at the 800 base ISO, and there are some common LUT:s that really bring the noise up - especially Visioncolor Impulz film emulation profiles. With those you have to overexpose intentionally to get a good result. But in most uses, the shadow noise is manageable and since it's not heavily compressed, you don't get the kind of macroblocky colour mess you see on HDSLR:s when starved for light. At 1600 ISO the image does get pretty rough with varying degrees of pattern noise.

I have very similar feelings and experiences when it comes to cinematic image, Canon vs Sony etc. Went from Canon to BMCC2,5K and I've been using it for about two years now. I'd say it definitely holds up, but there are a lot of buts. The good part is that the Prores HQ is still unmatched by any consumer codec when it comes to flexibility and detail, and in RAW it'll give you the most detailed and rich HD image at the price range. The dynamic range is huge and you can pretty much bend the image to your will in post. Having messed around with 5D RAW lately, the colour and the feel is very similar, but BMCC RAW is by far sharper of the two, at least when coupled with a good lens. On that note, you'll definitely want to use the Speedbooster. And here are the buts. With the EF mount, the center crop is poison - it's such a heavy crop any character a lens has will be instantly lost. This, lack of slow motion, and the ergonomics are my major gripes with the camera. The BMCC is a powerful tool and it gives you a beautiful image, but it's the most unfun camera I've ever shot with, save for maybe Red One. The ergonomics are awful and you'll end up building it into a rig inevitably. Having to go to menus to change shutter speeds etc is cumbersome. The in-camera monitoring is pretty bad as well, be prepared to stare at a Log image or a uselessly ugly video gamma mode all the time and just rely on zebras and the histogram for exposure. Also the center crop sucks the fun out of vintage lenses and makes flares look terrible. So definitely get the speedbooster, otherwise you'll be in a world of hurt. So yeah, the image holds up, it's very filmic, flexible and beautiful. But it's not a fun, intuitive camera in the way say 5D is.

Thanks! I'm sure one could easily get a cleaner image out of it by using a better capture deck (I had it set up directly from camera to a capture card via composite cable) and simply by using some fresher cassettes. The one I used was as old as the camera itself - from 1999! Probably taped over 100+ times too.

I've done something similar. Had an old Sony Handycam Hi8 camera from my childhood lying around at the back of my shelf, and one day decided to see if it still works. Much to my surprise, it did, and even the batteries were functional. Those things were built like tanks, I abused the hell out of it in my teen years shooting backyard movies, and over a decade later it works like a charm. So I fitted it with a cheap 35mm adapter, ran a few tests and then went out to shoot a small art short. Since the 35mm filter was mirrorless, I was practically shooting blind - the loop displayed a mirrored, upside down image, which really messes with your brain when you're trying to operate. Nevertheless, I'm happy with what we got. We dubbed it our punk movie, because it has that look of cheap, grainy 16mm film from old underground stuff.

This mirrors my experiences with the RX10 MK2. Slog in 8-bit colorspace is most of the time more trouble than it's worth - especially with the noisy sensor of RX10, which causes some insane, splotchy macroblocking in the shadows. I also tend to want to shoot saturated, contrasty images and most of the time the Slog just fell apart when pushed that way. I ended up using the Cine2 profile with Pro colorspace. Still not particularly fond of the look the camera produces, but it drastically reduced the headache I had in grading with it.

I have a tiny personal anecdote about the importance of 4K. I do a lot of food cinematography for a company that does video recipes, cookbooks etc. Their output is mainly online, on social media and the home pages of food companies. We've been shooting 90% on BMCC in prores 1080p. On one production we had to knock out a large amount of videos in a short time, so I rigged an RX10 MK2 to the ceiling of the studio to get us an overhead look of the process, giving me two angles simultaneously. I suggested the producer I'd shoot in 4k to give him room to crop the image in post. He laughed and told me even 1080p gives plenty leeway for what they do - they delivered in 720p, and in the case of instagram, 640xsomething. I've yet to have a client ask for 4K. I figure it'll be at least a few years before it'll be the norm or a requirement. And if you're talking about a fixed price point, the choice often comes between 4K or other attributes like raw/Dynamic range. I'll take the raw and DR over 4K anytime.

This doesn't come as a massive surprise. To me Digital Bolex always seemed like more of a specialty item than a workhorse camera (even though it may have well had the potential for the latter) and I didn't see any of my local colleagues using it, or very many productions in general. I think the camera's reputation was a double edged sword, it intrigued people, but also caused a lot of people (including myself) to not take it seriously, but more as an expensive novelty. Coupled with feature set that had more than a little similarity to the more established BMCC 2,5K and the lackluster footage from the early adopters (apparently caused by using wrong Resolve RAW profiles), the camera faced an uphill battle.