peederj

For working Cinestyle, just load up FilmConvert. Cake.   http://filmconvert.com

None too shabby...   http://www.dxomark.com/index.php/Publications/DxOMark-Reviews/Sony-Cyber-shot-DSC-RX100-II-review-new-back-illuminated-sensor-promises-superior-low-light-performance/Sony-Cyber-shot-DSC-RX100-II-sensor-performance   12.4 EV DR, on the DXO scale (probably more severe than typical manufacturer claims)...actually in your pocket with lens included for $750.

Note that the 70D handles 7fps 20.2MP stills while the 6D handles 4.5fps 20.2MP stills. This makes it quite possible the SD slot on the 70D is upgraded to greater bandwidth.   Also, in the lens list above, note the absence of "II" on a lot of those models...those are the old ones. Canon has upgraded its AF accuracy within the lens motors considerably since around the time of the 70-200/2.8 II as documented by Roger Cicala. It's nice continuous AF isn't limited to the new STM lenses at least, though it may work best on those.

Two important things aren't clear yet before we write off the 70D:   1) What the maximum speed of the SD slot is (it could be, e.g. UHS-104 for a theoretical 104MB/s) 2) How the downsampling is achieved from 20MP to 2MP video (it could have more or less moire vs. resolution)   I understand clean uncompressed HDMI out will NOT be supported on the 70D. Though everyone's officially supposed to hate external recorders anyway (why?).   It's annoying that the stills shooters get all the info they want other than effective DR on announcement but video shooters have to wait for the bad news. Generally anything not revealed is bad news. But I think the SD speeds may be enough for an ML-improved video, even continuous 1080p RAW, because of the claimed 7fps stills throughput along with a buffer that can hold 16 20MP RAW frames...an awful lot of 2MP RAW frames in that case. I imagine the crippling then is once more at the downsampling step...the place ML can't get to.   Note how Canon rushed out a meagre (one lens only) autofocus implementation for the C100 just prior to this announcement. Avoiding confusion in the marketplace I suppose. I imagine the next-gen C200/400/600 will use this autofocus technology, and to be honest, I think having AF available is a damn good way to solve a bunch of problems. You can tell the pretenders by their insistence that "real filmmakers would never buy a camera offering autofocus."  :rolleyes:

What Panasonic (+hacks) gives the low-end shooter in resolution they take away in color. You will need some sort of LUT-based correction, but if all you've got is 8-bit 420 to work within it's going to be rough. The Panasonics don't offer Log gammas either iirc, and you might get in-camera sharpening/NR artifacts showing up too with too extreme a push.   The easiest and cheapest thing to do might be buying a 2nd hand Rebel T2i. You'll also get better DoF and low-light thrown in the package, and you can run Cinestyle on it for better DR. For all the complaining about Canon (some justified) the color is the best in the biz objectively and subjectively.

Streaming video (RAW especially) over wifi would be awesome...you'd have a HH wireless rig in a nifty compact form factor. Perfect for steadicam, crane, ceiling mount, etc. Especially if they have remote control focusing that works well. And for the RAW, you could have an SSD RAID on the receive end.   Video on this camera will live or die based on the downsampling they chose to give us. Unless ML gets full-res MJPEG working.

It would be awesome if the HDMI out was clean and uncompressed 4:2:2 (any word yet?). You could pack a Ninja into your satchel and record very nice video indeed handheld. The low-light performance looks outstanding in comparison to Mark I from sample stills. If the video is much better I may upgrade if I can sell the Mark I for a good price.

There are indications (wallpaper resolutions found in developer builds) that the next Thunderbolt display from Apple will be retina-density 5K 5120x3200 (double WQXGA). That would let you have a full-res 4K preview on on a screen with lots of edit controls alongside. Nice!

The articulating display is nice, and I imagine the crazy blue channel problem is gone thankfully. And 24p! A worthy upgrade and something to jump on for those who've been waiting for a truly pocketable instrument.

Well let's see if it's I or BMD and their fans that have misled the readership as the year proceeds.  ;)   I can assure you, if able I will apologize proportionately, and I do expect the same courtesy from my critics.

The de-bayer task is implicit in the 22MP JPEG stills process.   The key limiter I think is what ML has access to in the buffers, and at what frame rate. They only just discovered they had RGGB access in those buffers, leading to raw_rec. 30fps full-res 4K or 5K MJPEG seems well within the realm of possibility to me, as does 60fps 1080p MJPEG.

Lol I actually got a kick out of all that. Lots of pot-meet-kettle hypocrisy and "he doth protest too much." But amusing.   Anyway what it comes down to is that I'm giving my opinion, based extensively on evidence, on a camera manufacturer and its products and practices. An opinion that I perceive as undervoiced here and elsewhere. And you Mr. Essary are giving your opinion on me and my ability to do so. Ergo, I'm on topic, and you're at fault.   Back to topic.

The 5D3 has the onboard capacity to JPEG compress 22.3 Megapixel images over six times per second and write that data continuously to a CF card (even an SD card). This would suggest it also has the capacity to JPEG compress 2MP images over 60 times per second.    I know you disagree with that mathematical argument, but I've seen no credible basis for declaring it impossible. I should sign up to the ML forum and ask them directly but I figured they're already busy with RAW so I'll let that get done and then ask for 4K/HFR MJPEG.   I do think it's possible and even likely. They had been insisting that people not beg for RAW video and then they went and did it. They may not want to get Canon mad at them because it would essentially negate the 1DC. So maybe someone else has to take their open source code and get it done unaffiliated.

What I don't like is the combination of a lying, manipulative company with a online commentariat that steadfastly defends their every move.   I guess I'm an irrepressible optimist that thinks the Internet is the solution to fraud. And nervy enough to do my part. Here we have a company that's doing exactly the same thing two years in a row and people still eating it up.   "Fool me once, shame on you. Fool me twice, shame on me."   Meanwhile the Japanese giants the net.commentariat is determined to discredit at every opportunity just continue to quietly ship products on time that work and need no scheming or hype. Slower to roll out advanced formats and codecs perhaps, but turning a blind eye to the hackers enabling them on their hardware.   You have conceded the argument when you resort to ad hominems or simply tell or force your opponent to "shut up."

  If you say no one knows, then how can you accuse me of assuming wrong things? They've made clear it's the same sensor, it has the same specs, it quacks like the same sensor just a crop, maybe packaged a bit differently internally...   And they are "blatantly lying" about the 4K cam's shipping schedule. Unless you take the "especially" wink as getting them off the hook for mischaracterizing forward-looking statements. A public company trading in USA would face Sarbanes-Oxley action if stating that one of their most important products would ship in July when it's clearly going to be more like January 2014 or later.   As for 16mm lenses, this is one of those cases where "bad is good" as long as "bad" is at least 30 years old. It's hard to argue with someone desiring a negative, but I'd figure the Digital Bolex would give the whole retro experience better than the BMD cam, and without rolling shutter (very important problem for handheld, and decidedly not a retro flaw).

5D3 RAW crop mode clearly has false-color alias artifacting (a typical problem for any native res implementation, it was true on the 600D's 3x zoom feature for instance), you just repositioned the camera for the larger crop modes. A converging lines target would have shown that the artifacting remains for a given resolution, the concentric lines target just got enough space between the lines to not artifact when you dollied in.   The 1DC is oddly gray in this test (overexposure? uncorrected log gamma?) but shows no false-color aliasing which is a point in its favor.   After seeing the artifacting I never used digital zoom on the 600D and I doubt I'd use it on the 5D3 but there's always that rare opportunity it comes in handy. The clean HDMI video for the 5D3 showed you can also use digital zoom with the Ninja 2 on the 5D3 if anyone is interested in that option. It's going to look a bit different cross-cut with the unzoomed footage, more different than a decent telephoto lens but less different than another camera system entirely.   I still hold out hope ML is going to be able to do MJPEG on the 5D3. Which will open up much higher resolutions and possibly even framerates. Yes I am acutely aware they couldn't get H.264 to work at HFR...MJPEG is NOT H.264. If you do the math on how many pixels the stock 5D3 can throughput on JPEG compression they should have plenty enough power on tap.

Well I know you "don't care!" but fyi I did not pre-order the C100. Instead B&H and Adorama had stock ready to ship on the launch date in November and I got mine the next day. So there would have been no benefit at all to preordering...the C100 is so feature-complete the only thing I added to it was a Ninja 2 for ProRes that was also ready to ship that day. I'm sure you will sniff and say "that's because no one wants a C100" but it has only increased in respect since then while the BMCC has increased in reports of dropped frames etc. The only ways I can see to compete with the C100 + Ninja 2 are RAW (5D3), 4K and HFR (FS700 + R5/C500 + Odyssey).   I similarly think BMD will have plenty of stock of the Pocket cam when it ships as it is just using a crop of the same old tiny low-res sensor of their first-ever camera. I've addressed my concerns about the BMD Pocket here and on other threads, for instance, just how much active MFT support will there be? A lot of the benefit of MFT lenses comes from in-camera correction of distortion and CA...I assume those will not be active for RAW, but will BMD be able to implement them for ProRes? On launch or ever? Or will they have a database and presets for Resolve to do it? Will lens metadata even be recorded by the BMD? Check photozone.de to see just how awful MFT lenses are in their uncorrected form. Of course, the Pocket will be merely a center crop of their FOV anyway, giving us a miniature wide-angle problem just as the EF mount BMCC suffered from.   I think a sane person committed to the MFT mount (a rare combination I'll admit) would wait till the BMD Pocket showed up and then test it out carefully against the GH2 and/or GH3 to see if it really was worth it in practice before committing $1000 to it. $1000 is not an impulse price point for MFT users I wouldn't think. And for those not stuck with MFT lenses, as I mentioned, RX100...

A manufacturer has to determine how large a production run is or risk an inventory glut or opportunity loss, they or their distributors have to secure shipping containers to send the batches overseas (for many Chinese goods, the cost of a shipping container is much of the cost of the item, and small distributors have to share containers, etc.), they have to pay a lot of those costs up front, and as a result they need to not only gauge demand from dealers they also negotiate bulk preorders from them giving the largest block purchases with advanced payment a significant discount. Many of the larger dealers effectively become distributors for their region, paying for the shipping container and then selling to other dealers in turn while taking a significant markup on the risk. They are all co-opetitors...they don't want the public to know any of this, but it's what goes on down in the boiler room.   A campaign to hype a product, overpromising both feature set and ship date, can get the pump primed for a small manufacturer that needs to get those shipping containers filled up and the allocations prepaid for. It can be the difference between profitability and total loss of R&D and inventory for a soon-obsoleted product like a digital camera.   But if they get enough of you to bite just on faith, they can get enough of you to rationalize buying into not only the product but the ecosystem, and its halo effect for the dealerships and distributors. $ucce$$! Meanwhile your movies probably won't be very much better but you can feel good about yourself and your new toy for a month or two.   Sorry to dim spirits! But enough people attack my camera models here that some balance ought to be applied. If the 4K cam is a pleasant and near-term surprise I will praise accordingly.   But I will not be pre-ordering anything from BMDelayed. The only thing I've pre-ordered recently was the 5D3, and though I have (and don't care about) the LCD light leak, that worked out fine in the end.  ^_^

In turn you're assuming the dealers don't send money to BMD to get their preorder allocations. I would be surprised if that's not the case, if an item is "hot" whoever's got stock in the early days will cash in handily (including on acce$$ories). So a manufacturer would be pretty foolish to allow dealers to take preorder money on their merch and not take any themselves.   Anyway codecs do matter but not as much as getting the shot in the first place, and the RX100 actually being in a pocket and shooting in 2 seconds is going to get a lot of spontaneous shots that the BMD Pocket being tucked away in a case somewhere and possibly not having the right lens on etc. is going to miss. Run the RX100 footage through Neat Video and while you don't have "captivating cinema" quality you don't have chopped liver (cellphone video) either. If quality matters that much you're going to need to get something more substantial than a BMoireD anyway and the time and gear to set it up properly.

  I will be shocked if the 4K camera appears in quantity (>1000 units) this year. In fact I wonder if BMD is more desperate for preorder money than credibility at this point, steadfastly refusing to learn from last year's lessons.   Anyway the RX100 soundly defeats the BMD Pocket in: Stabilization (very useful for a camera that will nearly always be shot handheld by definition) True pocketability (again, whose pocket can fit any MFT cam with a lens on it? The RX100 slips into virtually anything worthy of being called a "pocket") and action mountability (between the MFT size/weight and the GoPro) Ready-to-shoot cost (you can add a cheap grip and a magfilter adapter to your $650 RX100, your total RTS price will be at least double when adding a comparably usable lens to a BMD) Moire and Aliasing (OK the RX100 does have a noisy blue channel, but that's more fixable than awful rainbow moire) Product Maturity (we have yet to see the ugly details on the Pocket cam in practice, but prior experience from BMD's first-ever camera wasn't promising) Stills quality (20MP quite DSLR-like RAW stills on the RX100, 10 times the resolution of the BMD, plus tiltable built-in flash) HFR Frame rate (50 or 60p depending on your region, very useful for action and stabilization too) Automation (real-time autofocus and autoexposure, very useful again for run-n-gun, and it also offers full manual mode) Would I trade my RX100 and Hero 3 black and cash for a BMD Pocket and an MFT zoom lens? Would I awkwardly pretend to be a "real filmmaker" on the cheap, throwing big cash at expensive manual MFT primes and absurdly rigging out some little toy streaked with moire? Do I buy yet another hypewave from the overpromise-and-underdeliver BMD and sit in preorder purgatory for a year? Not only no, hell no.

  I predict the Noise Reduction feature of Resolve full version would have been very welcome for the Pocket shooters who aren't likely to be shooting well-lit narratives and have only a tiny sensor to collect available light with.

This is funny:   https://vimeo.com/68895788 0:50   "Uhhhh...I will believe, you may, uhhh, probably you, you will see them in July for sure. Okay? Especially the Pocket Cinema Camera."   :D   40 days and 40 nights won't be enough to get the 4K cam into the stores. Try six months or more. But the "Pocket" (whose pocket can fit that?) camera probably does show up in limited quantity and with limited functionality for the first year. I would much rather have a GH2/GH3 if I wanted to use MFT...much better sensors, much more mature product. Or the RX100 for true pocketability. Note that you don't get a free full version of Resolve with the Pocket Cam, which is all the original BMCC owners really have for their money at this point just a few months after finally receiving the product.

It's such a satisfying impulse to blame the evils of society and corporate profit for one's personal failures and inhibitions.   It's also BS.

  To get it right, Sony really, truly needs to make the 4K cam E-mount. No one's going to buy A-mount lenses, and there's only so many Konica/Minolta enthusiasts in the world, and most of them have no money for a 4K-capable workflow. But a full frame E-mount (as seen first on the miserable VAG-900) would allow the entire project to succeed. 4K, one trusts, won't have awful moire issues from line-skipping etc.   Will Sony come to its senses and do the obvious next step?  :unsure:

  16mm is quite a bit wider than 18mm...when the numbers get small the differences get bigger. Also, do we know the actual angle of views yielded (the focal lengths on the barrel are often marketing-influenced rather than lab numbers)?   This lens certainly looks interesting, one minus is it isn't stabilized, but nice and fast for the money. I have the Sigma 17-50/2.8 OS and the Samyang 14/2.8 and since I shoot C100 I don't have much need for speed...the 5D3 covers me for DoF...but as a potential impulse purchase...