Dekam

Is this behaviorally equivalent to converting the 4K footage to 10-bit, then bilinear downsampling to 2K 10-bit? Or has some additional special sauce been introduced?

As expected, the users over at dpreview came unhinged over all this talk of video. Did they not even ask about Magic Lantern? It'd be useful to publish a list of "questions we asked that they outright refused to answer..." since those can be a bigger scoop than the softballs they did answer. And citing the EF-M 11-22 as an example of lens innovation? A lens they won't even sell in the United States for some reason? How about the obvious follow up "so why aren't you selling that lens worldwide if it's so great"? Baffling.

[spoiler] Shooting on film is like collecting vinyl records, or buying the stereotypical "artisinal fair-trade organic locally-sourced small-batch hand-made" whatever.   It's an appeal toward something more authentic, and it's always a silly affectation when you look at it too closely.  People do these things as a form of conspicuous consumption.  You shoot on film not because it produces a better image, but because it gives you a story you can tell your buddies about how you like to shoot on film.  Same for buying vinyl records.  Same for buying hand-made beeswax zen candles at the farmers market, or whatever it may be.   Calling any of these things a movement or revolution is odd.  A revolution in targeted marketing demographics I guess.  Now McDonalds wants to sell you Angus, not a plain old hamburger. [/spoiler]   Unhelpful meta-trolling removed by me. :)   It feels like the animosity in this topic stems from the marriage of video functionality onto the modern photo-DSLR.  Since this site is a DSLR Filmmaking blog, it's important to think about the problem from that perspective.  If you're angry about video features affecting your photographic camera of choice, you may have chosen the wrong place to complain about it.   At this time it seems like Panasonic may be the only maker prepared to experiment with reversing the traditional balance of video and photo features toward video.  What percentage of customers are buying the GH3 just to take photos?  We can call the Japanese makers incompetent until we're blue in the face, but surely Panasonic must know the answer, at least outside of Japan, is "nobody".   In a recent dpreview sewer thread, someone said "maybe in the future all cameras will just shoot video and we'll pull our stills from the video stream", which of course lead to a bunch of name-calling and "that will never work" backlash.   The new Lumix DMC-GX7 shoots full-res raw stills at 40 fps.   Let that sink in for a minute.   (The fine print is that it can only shoot 9 raws continuously, unlimited jpegs.  Almost certainly a buffer or storage limitation that can be addressed in various ways on a future video-oriented body.)

  Totally agree on this point.  RAW is great, but the BMPCC has shown that simply moving to 10-bit makes all 8-bit cameras look like a joke.  For the big players to have not gotten on-board by now is utterly baffling.  Especially since ML has proven the existing hardware is completely capable.  Some codec licensing deals and a few months of firmware dev would pay for itself 50x over.         When we talk about GoPro we're not talking about their existing product line, we're talking about a hypothetical future "cinema hobbiest" line that would, in theory, come with a set of 4 adorable little swappable lenses, offer manual exposure control modes, shoot 10-bit CineForm ProTune Log-Curve at 2.7Kp30 and 720p120, and cost something reasonable (say under $650?)   Maybe even bonus features like mic input and monitor output, depending on cost.   Why would GoPro do this?   1. Zero legacy baggage.  Canon and Nikon have pressure from multiple directions to not innovate in video.  They have existing professional video products which will be affected, and they have bread-and-butter wedding photographer customers that get angry that their photo cameras are being complicated by video features.   2. Unlike those same players they have CineForm.  No messy codec licensing issues to worry about.     3. Most Importantly:  They are a marketing powerhouse, and they seem to have an effective supply chain and retail presence.  They've shown they can make a compelling argument why people need to buy their product, and they can manufacture and distribute it.   -EDIT- 4. They've already saturated their own market.  They can keep improving the existing GoPro by making it smaller, but is that going to bring them substantial earnings growth?  Their days of quadrupling revenue in a year are long over unless they come up with something fresh and new.     Advantages #1 and #2 also apply to BlackMagic, but BlackMagic is straight up terrible at #3.  BlackMagic has Da Vinci Resolve, which is fantastic, but they also need to offer a hobbyist-level workflow like GoPro Studio if they ever intend to sell into that market.       People buy cheapo DSLRs for many reasons, but there's a giant market segment that just want to take better pictures than their cell phones are capable, and don't want to blow $3000 to do it.   How many people walking around with the T3i Kit from Costco are actually getting better pictures than their cell phones?  Not many.  Why?  The kit lens is junk and won't give them the low-light capability and holyshit BOKEH they expect from a "better than a cellphone" camera.     So maybe the future of the low-end DSLR is Sigma showing up on the scene with a light mirrorless body offered with a wide-aperture kit lens, and low-noise sensor that's actually worth a shit in low-light and gives people their insane bokeh, and has on-camera shooting modes that make amateurs able to use it effectively (imagine a Canon 'creative mode' that actually works properly).  Then actually backs this product up with marketing savvy to tell people why these features will make them an instagram hero.

1. Does GoPro fit into this?  Remember around NAB time, it seemed like everyone just assumed they'd usher in a new hobbyist cinema camera revolution, but nothing came of it.  Their GoPro Studio software gives them potential "easier to use" conversion workflow for CineForm.  Look at the trouble Dave Dugdale had getting up to speed with raw video on the BMPCC and you'll see why Canon can't "sell" raw video to its hobbyist userbase with the current shitty software landscape.   2. Has Canon even acknowledged Magic Lantern exists?  What do they, at a corporate level, think of it?  Getting one of their reps to talk about it might be a great scoop.

Everybody saying "I don't care about 4K" needs to realize that we still don't have 1080p on any of the Canon bodies, unless you count the soft garbage we're getting today as 1080 lines of actual resolution, which you should not!   If 4K, as a feature, finally gives us a way to achieve sharp 1080 (either by postprocess downsampling or superior on-camera processing), then I think we can all agree it's a good thing.   So please, get on board, and quit complaining about 4K like it's a problem.