
Kino
Members-
Posts
248 -
Joined
-
Last visited
Content Type
Profiles
Forums
Articles
Everything posted by Kino
-
I do not see the C500 support as particularly high on the Atomos agenda. I could be wrong, but the Inferno is simply the wrong device for capturing C500 footage in all its different output modes, even with the two SDI connectors. They may get there someday, but it's a huge gamble to assume that everything is coming in future upgrades. DVXuser is one place to look for how the C500 works or doesn't work with Atomos recorders. The Odyssey may be a bitter pill for some, but it is one you have to take to work with the C500. The extra fee for the RAW license is unfortunate, I agree. Of course, without the license you can still shoot the camera in 4K ProRes (up to 30p) and 2K 12 bit RGB 4:4:4, which upscales incredibly well to 4K as both are derived from the same 4 x 2K lattices I mentioned above. Compare 4K and 2K (here only at 10 bit) starting at 1:57 in this test by Hurlbut: Aside from the Human Voice trailer I linked above, here are some other examples of 2K 12 bit RGB 4:4:4 (shown in HD without upscaling): There are many more of those lens tests from Cooke Optics using the C500's 2K RGB mode on Vimeo. Cooke also has some slow motion examples shot in 60fps (10 bit RGB) and 120fps (10 bit YCC) and recorded on the Codex. The Odyssey for its part records onto Samsung 850 EVO drives that cost $320 for 1TB or you can go for the Samsung 850 PRO 1TB at $420 (or any of the 850 Pro line that are smaller and cheaper). That will give you one hour of 4K RAW at 24p, 2.5 hours of 4K ProRes 4:2:2 HQ, or 5.5 hours of 4K ProRes 4:2:2 LT. In terms of cost per GB, I would say that is a very competitive price when compared with RED mini-mags or CF 2.0 cards, for example. You can also take your SSD, stick it into a drive bay, and start editing right away. I do agree that the battery solution is not ideal, but the optional battery adapter on the back works with Canon C300/C500 batteries and gives you around 1.5 to 2 hours of continuous recording from everything I have read. It's a much lighter solution than attaching a V-lock on rails. You would be carrying the same battery that the camera uses, so I think it is an ideal setup in that sense. You can also mount the recorder to the camera in numerous ways depending on if you want to go with rails or not.
-
CinemaDNG is Adobe's RAW format (for example, the BMPC-4K records RAW using CDNG), but it is only promised in a future Atomos firmware update for the Inferno (as mentioned on their website). When it is released, you will be able to record up to 30fps from the C500 in CDNG on the Inferno. Nothing else has been specified regarding CDNG recording in the Inferno. For example, will it retain the 10 bit logarithmic RAW signal, which is crucial to unpacking Canon Cinema RAW in post? With RMF, you can process your files to ensure proper unpacking of the 10 bit log. I'm not sure how it will work with CDNG. What will happen with the four lattices of the Bayer sensor data I referenced above? These data streams must be recorded properly to ensure effective demosaicing in post to retain the C500's outstanding resolution and anti-aliasing performance in 4K. As for 2K 12 bit RGB 4:4:4, it is not even supported in the Inferno, let alone ProRes 4:4:4:4. Some of the links I posted above go into the specifics and unique attributes of Canon cinema RAW, which is quite different from almost any other RAW implementation since the Canon RAW has much larger data rates (moreover, ISO is baked-in and white balance is set). For this reason, I would prefer to remain within Canon's RMF workflow (or DPX, as the Odyssey records the 2K 12 bit RGB 4:4:4 into this compatible format) as it is crucial to extracting the most faithful recording from the 3G-SDI data streams. Here is a section from one of Canon's white papers on Canon Cinema RAW in the C500: "Management of C500 Output Data Rate – Logarithmic Encoding Considerations of RAW data rate outputs from the EOS C500 differ from that of most other single sensor cameras – and they are two-fold: 1. In close consultation with no less than five separate digital recording manufacturers we established a consensus on a maximum camera data rate output that could be managed by each (even though each had different recording strategies) 2. Very important – Canon specifically wanted to use the SMPTE ST 425-1:2011 serial 3G SDI interface as the transport mechanism for our RAW data – in order to facilitate: o A standardized (and universal) serial digital delivery to each of the disparate recorders o Passage of the camera output RAW data into broadcast 4K/2K/HD infrastructures that utilize standardized 3G SDI system elements (camera CCUs, routers, encoders etc) Accordingly, the RAW data rate output is constrained to less than 3Gbps for both 4K and 2K having frame rates up to 30P – so only one 3G SDI interface would be required. This data rate was achieved without resorting to compression by constraining the camera output to a bit depth of 10-bit for the 4K RAW and to 12-bit for the 2K RAW – as outlined in the following: Canon 4K: 4444 Bayer RAW Data rate @ 30 fps 2048 x 1080 x 4 x 10 x 30 = 2,654,208,000 bits/ second Canon 2K: RGB 4:4:4@12 -bit Data Rate at 30 fps 2048 x 1080 x 3 x 12 x 30 = 2,388,787,200 bits / second Both of these data streams fit comfortably within the 3G SDI interface standard. For higher frame rates – up to 60P – the elevated data rates (still less than 6 Gbps) will require two 3G SDI interfaces. To meet those two bit-depth constraints without compromising restoration of the higher bit-depth of the image sensor output in postproduction resort was made to an alternative bit rate reduction strategy – namely, logarithmic encoding of the image sensor linear representation (according to the mathematically prescribed Canon Log). Specifically, the high bit depth of the image sensor digital outputs are logarithmically transformed to a 10-bit depth for the 4K mode, and to a 12-bit (with 10-bit as a selectable option) for the 2K mode. This is a completely reversible process. While Canon does not disclose the bit depth of our A/D converter – we do affirm that the linearization process (de-Canon Log) can reproduce the linear representation at 12-bit, 14-bit, or 16-bit (DPX or Open EXR)." http://learn.usa.canon.com/app/pdfs/white_papers/White_Paper_originatinghighqualityc500.pdf (bold emphasis added)
-
While most people associate the C500 with famous American DPs like Hurlbut or Jeff Cronenweth, we should not forget to mention the C500 work shot by CML founder Geoff Boyle, BSC: It's impressive to read that this was all shot with EF glass. Here are some of Boyle's lens tests using the C500: http://www.cinematography.net/CML-CMIR-Lens-Tests.html His over and under exposure results as part of the 2015 CML camera tests are also available: http://www.cinematography.net/edited-pages/C500-uwe-2015.html There is a wealth of useful info there on the camera, especially with regard to the differences in overexposure when dealing with tungsten vs. daylight. Unfortunately, the 5,000 ISO test of the two low-light champs, the C500 and the Varicam, is no longer available through the online link.
-
No problem. I'm glad to help in any way I can. The C500's HD 120fps is output at 10 bit 4:2:2 YCC in DPX (uncompressed) and MOV formats that should be supported without the RAW upgrade. However, I would check with Mitch Gross or someone else at CG to make sure, as it seems important for you. From what I have read and seen, the RAW 4096 X 1080 is the best 120fps footage from the camera. This is different from 120fps "Half RAW" 4096 x 2160, which leads to a loss of vertical resolution.
-
4K RAW @ 60fps is recorded in a frame-interval system, where even frames go to one SSD while odd frames end up on the other. That's the only way for the C500 to record HFR in Canon's RMF (Raw Media Format), as it utilizes two 3G-SDI connectors to transfer the massive data rates I noted above. Recall that each 3G-SDI connector in the C500 is outputting 4 2K streams, which have to be recombined to create the 4K or 2K RMF file recorded by the Odyssey: At 60fps using frame-interval recording on dual SSDs, you have two sets of these 4x data streams. That would explain the insane 700 MB/s data rate. [Btw, the two green data streams or lattices, which form an over-sampled "super green," also have implications for suppressing aliasing and enhancing resolution and MTF, properties that are evident in C500 footage at both 2K and 4K: http://www.provideocoalition.com/nab-2012-canon-c300-image-processing/] The Odyssey allows for this dual SSD recording as does Codex, for example. Atomos states that the Inferno will not ship with any type of RAW recording. Instead, they are promising 4K RAW in future firmware updates, but only up to 30fps. The Inferno does list 4K ProRes up to 60 fps: https://www.atomos.com/shogun-inferno Aside from the lack of 4K RAW 60fps and HRAW 120 fps on the Inferno, you would also be missing a whole bunch of other functions, including 2K 120 fps (10 bit YCC) and 2K 12 Bit RGB 4:4:4, which is one of the C500's headline features. Both are included on the Odyssey (120fps 1080/2K is not listed here but definitely supported): https://www.convergent-design.com/canon-eos-c500 I understand that the Inferno will have a lovely HDR display, better sound options, and a friendlier interface, but it is simply not suited to unlocking the C500's numerous recording modes. More info on the C500's workflow can be found all over the web. However, not all of it is reliable or up-to-date so it's best to read as broadly as you can and to check for Canon and Odyssey firmware updates to see what features are currently supported. A primer on Canon's unique approach to cinema RAW can be found at these links, including info on the diagram posted above: http://blog.michaeldanielho.com/2014/02/understanding-how-canon-cinema-raw-works.html http://blog.abelcine.com/2012/11/05/what-exactly-is-canon-raw/
-
The C500 does 60fps in RAW DCI 4K (4096 x 2160), but it requires two SSDs and a data rate of over 700 MB/s! By comparison, that is 7x the 1DX II's 60fps data rate of 100 MB/s (800 Mbps). But at 120fps HRAW, indeed you are only getting half the vertical resolution at 4096 x 1080, with two options as to how you conform the material in post, as Andrew mentioned in his review: http://www.eoshd.com/2016/06/4k-raw-120fps-3k-say-hello-second-hand-canon-c500/ The problem with the Atomos recorders is that they only have one SSD. This means that they can't record the C500's 60fps RAW or 120fps HRAW. I believe Atomos lists 60fps ProRes recording from the C500, but I have not seen any user reports confirming this. In any case, if you own a C500 (or one of the Sony cameras that does HFR RAW), you need the Odyssey. As for Scorsese film, I only know that they used the C500 on the drone shots. It's entirely possible they shot that slow motion sequence in HRAW 120 fps. I'll have to go back and look at it, but I don't remember any distracting aliasing in the image (which I would expect, depending on how they dealt with it in post). Suffice it to say, you would not want to use too much footage in HRAW 120fps. A little here and there is fine. The motorcycle documentary I linked to above used the C500 HRAW 120fps extensively, if you want to see what it looks like in a production.
-
I only mention it because I am very close to cancelling my Scarlet-W order and going with the C500 and Odyssey 7Q+ instead. I absolutely love the C500 image and with the recent price drop, I think it's a much more attractive setup for me (and a little cheaper). I just have to look more into rigging options for hand-held use. For some reason, the Canon engineers removed the hand grip when implementing the 4K output into the C300 body. I've been researching the C500 recorder options for a while and have read about problems and quirks with other recorders. For example, Jonathan mentioned using the C500 with the Atomos. According to what I've read, unlike the Odyssey, the Atomos cannot record the 120fps RAW signal from the C500. That's why I mentioned to Jonathan that he should know the exact combination of the recorder he wants to use and confirm that all the functions are enabled with that recorder. Of course, it's great to hear about your problem-free experience with the Odyssey. With the RAW upgrade package, it is a $2700 recorder so it better work without a hitch!
-
https://***URL removed***/news/5855300360/sony-announces-alpha-99-mark-ii "The a99 II can capture 4K at 100Mbps (using XAVC S) with full sensor read-out and no pixel binning. A Super 35 option is also available, with 1.8x oversampling. A 'Slow and Quick' mode lets users jump between 1 and 120 fps at the push of a button. All of the capture tools you'd expect are available, including zebra patterns, time code, S-Log2 and S-Log3 profiles and 4:2:2 output over HDMI." The XAVCS internal is a shame, but necessary because Intraframe would probably be impossible in a body that is much smaller than a Canon 1D series. At least there is HDMI clean out. The 4K downres from 8K makes it sound like a Sony F65 for your pocket!
-
I would cross out the Scarlet-W right away. You are looking at $15K minimum to get up and running and the waiting time is very long. I know because I ordered many months ago and I'm still near the back of the queue with no allocation in sight. Moreover, the RED cameras are not ideal for documentary, have very poor on-board sound options, and would introduce a steep learning curve for someone who shoots primarily with a Sony A7RII. I would also stay away from the C500 + RAW recorder option unless you have used this exact combination before and know all the quirks involved. There are simply too many things that can go wrong here. In addition, adding the recorder means that you lose the hand-held advantages of the Canon bodies, so it really defeats their purpose for me. If you can afford the C300 II, and you don't need 4K HFR, I would definitely choose that option. Otherwise, the FS7 is also an excellent choice and has much better slow motion than the C300 II. You can't go wrong with either one of those. Best of luck!
-
I think we should also mention that, in the hands of a real pro like Canon's own Brent Ramsey, the C300 II is capable of stunning internal 4K: It's amazing to think that this was shot in the camera's internal XF-AVC 10 bit 4:2:2 (YCC). Canon always gets the best possible colors, skin tones, and DR out of any 8-bit or 10-bit codec. There are many RAW cameras that don't deliver this type of 4K image quality. Here is a useful article about image processing in the C300 II: http://www.thedithouse.com/on-set-data-management-transcoding-dailies-systems/canon-digital-slr-workflows/ Unfortunately, a lot of the C300 II 4K footage available online does not do justice to the camera's full potential.
-
Internal. It's right there in the C300 II menu. And, yes, it is really gorgeous!
-
The winning bid was $6,000, which sounds unreal! The owner must have been really desperate to sell. The Sony F5 is one of the reasons why the C300 II could not really compete at $16K (the other reasons are the FS7 at $8K and Canon's own C100 II and C300, of course). 16 bit RAW vs. 10 bit 4:2:2 (with YCC color space) is not much of a competition. Unfortunately, the best mode for the C300 II is the 2K 12 bit RGB 4:4:4 (similar to the C500): It's a shame they can't make that available for 4K internal, since it would make for an incredible camera. Dogtown, if you are going to the $10-15K market, and you want a cinema camera as opposed to a doc/ENG/event cam, you should definitely check out a used Epic (MX sensor). They are still producing some of the best footage in this segment:
-
The C500 produces beautiful, filmic imagery and is outstanding in low-light compared with most cinema cameras. Musicbed used to shoot their "Artist Spotlight" series on the C500 (before they switched to RED) and there are some really nice examples of 4K and HD footage there. You should definitely check out their Vimeo channel: I'm sure you've seen the Hurlbut C500 tests, so I'm not going to post them here. As demonstrated in those tests, the DR in 4K mode is lacking compared with the best of the modern cinema cameras. But I don't think the more limited DR takes away from the overall image quality, considering that the C500 has a smooth and pleasing highlight roll-off (as visible in the above clips). Shooting in the 12-bit 2K RGB 4:4:4 mode also yields DR benefits in addition to numerous advantages in color grading, as seen in Human Voice: Here is yet another kind of 4K look or grade from the camera that is punchier than the Musicbed shorts (set to a fitting Schubert piece we've heard before on the Barry Lyndon sound track): Meanwhile, an example of HFR (up to 120fps) can be found on the documentary Why We Ride, which was shot on the C500: For $7K plus the $2K Odyssey, it makes for a great package and is competitive with other cameras in its price range, so long as you don't mind carrying the camera and the recorder everywhere you go. If you're only shooting events/docs, I would just pay a few thousand more for the C300 II as it has the DPAF and records 4K internally. The C300 II also has a newer sensor and better C-Log options for an expanded DR in 4K.
-
Well, I thought we were past the personal attacks, but I guess you just can't help yourself. Indeed, some people never learn or "wise up." I'm not calling the UM 4.6K "off-the-shelf," as you already assured me that it is not. It is custom designed and exclusive to BMD. Fair enough. The Fairchild sensor, on the other hand, is listed as available to the public and is by definition "off-the-shelf." I never said ASICs are related to sensor design, but they are part of the image processing chain, which I wrote about back on page 6 of this thread: "It appears that BMD has to involve third parties like Fairchild in sensor design, suggesting that they don't have the same mastery that RED currently has. This lack of expertise may lead to problems such as we have seen with magenta-gate, considering how the sensor and its integration into the surrounding circuitry and the larger image processing chain would have to be designed with absolute precision." http://www.eoshd.com/comments/topic/20526-ursa-miniis-this-the-end-of-blackmagic/?page=6 You apparently forgot to read this part. My point is simply that when you design all the components in-house and have demonstrated expertise in sensors and image processors, it is easier to avoid such problems in the first place and to diagnose and repair problems as they come up. RED had numerous mishaps in their 10-year history, as you well know. BMD is just starting out and has to go through a similar learning curve as with any company that is new to designing cameras. That's about it. ARRI, RED, Canon and Sony all have numerous patents relating to their cameras. BMD needs to complete its patent applications, which I cited above, in order to put them is the same category. They are surely aiming for just that kind of recognition after designing an advanced camera like the UM 4.6K. At least, I would hope so.
-
I own and have used the BMPC-4K. Nowhere do I make any claims about using the 4.6K, which I don't own. I have discussed on this forum my experience with the BMPC-4K but never anything else. As for the UM reviewers and the magenta problems, I've seen a lot of evidence of flawed units posted online by individuals well-known to the community on BM Forum and bmcuser. I've also looked at camera files that were made available. What I don't accept are the examples of "magenta" with suspicious parameters such as f16 or beyond, where all digital cameras look very poor and display noticeable color shifts. For a scholarly analysis of the magenta issue (if there is such a thing), I would have to take samples myself from dozens of cameras, if that is what you mean. I would also have to have access to BMD internal files on the matter.
-
Well, I suggested that very possibility on page 6: “And, yes, it is entirely possible that BMD contracted Fairchild for the 4.6K with an arrangement to create two versions of the sensor: one for BMD and one off-the-shelf version.” http://www.eoshd.com/comments/topic/20526-ursa-miniis-this-the-end-of-blackmagic/?page=6 To which John responded with the following on page 7: "Of course you are trying to perhaps have me confirm a technology partner of BM, when you know full well I'd never be able to disclose that without breaking an NDA, but I can assure you, the sensor used in the UM4.6K is not an ‘off the shelf sensor’, nor is it one that you can just go order from whomever you think the vendor is. I know this because of actual personal involvement with its development. Please explain how I can have that so wrong and your version be more correct?” http://www.eoshd.com/comments/topic/20526-ursa-miniis-this-the-end-of-blackmagic/?page=7 I think that is best explanation anyone could give on the status of the 4.6K sensor and it's a lot more than I expected. Let’s just leave it at that because we can’t ask John to say anything more. He has already been generous enough on these matters with all that he has told us here. Anything more and legal teams will be descending on this forum and watching everything that he says. I personally don't want to see that. Moreover, this is why I prefaced that section with "Let's say for the sake of argument . . . " In other words, in a hypothetical scenario where BMD designed the sensor and owns the patents/licenses . . . Ultimately, the only thing that is going to solve this is the patents, which I am sure will be forthcoming from those involved (if they have not already been filed)
-
That combination is actually the essence of the term "Blind Peer Review," which is what gives scholarly journals all over the world their credibility. Otherwise, reviewers would be tempted only to accept articles by their close associates. It's the substance of the argument that matters, not the identity of the writer. But, certainly, there are different protocols for a forum and I can understand why you say that. A casual discussion is different from a journal.
-
The Fairchild 4.6K is "off-the-shelf" in that it is available to the general public. John has stated that the Ursa Mini 4.6K sensor is not "off-the-shelf." It is of course possible there are two versions, one for BMD and one for Fairchild, as I noted a few pages ago. John also maintains that BMD has designed the sensor, which is fine. I believe him and I even cited the lapsed BMD patent application that may be relevant. We will have to wait and see when they file those applications again. That's the only way to know for sure.
-
John, You have a very selective memory when forming your arguments about what I’ve written. And speaking of diversion, your latest tactic is to cut-and-paste some posts out of context and out of a timeline that went on for months as we waited for the Ursa Mini 4.6K’s release. These bmcuser posts from 9-12 months ago also have nothing to do with the current topic of BMD’s design patents or its quality control. Moreover, anyone can be quoted out of context: “You're looking pretty silly. You're inferring a conspiracy theory that those that have shot with an Ursa Mini are hiding its flaws until...it get released...and....we all get found out ?...” http://www.bmcuser.com/showthread.php?15520-Brief-thoughts-and-a-bit-of-footage-from-the-URSA-Mini-4-6K/page140 That was what you wrote on January 14, 2016. This is after some had already noted the magenta problem in the beta footage and had been ridiculed and attacked on bmcuser. An NDA that prevents you from reporting the magenta publicly is one thing, but deriding and discrediting those who saw the magenta problem early on is another. And guess who leads those online attack parties on the BMD dissenters, John. Who are the ringleaders when the wagons are circled? I guess RED has had them, so it’s okay for BMD to do the same now, right? I thought you guys were better than them. So yeah, you’re right John, the 4.6K was released and everything was “fine,” which is why they had so many returned units and great reviews. I’m not even sure how long the magenta thread is on bmcuser: it probably requires its own server it’s so huge (much of it is silly, of course, like people shooting at f16 and wondering what went wrong with the image quality, but there are some real issues there otherwise). In my case, there were important reasons why I started to see better results from the 4.6K footage. Let me remind you since you obviously forgot. Here is what I posted back in January as we saw the first RAW footage: “Speaking of appreciation, I will be the first to say that the Mötley Crüe footage was absolutely gorgeous and some of the best that we have seen so far, along with the dancer video. Both were reportedly shot in RAW, so I am not surprised that we are finally seeing the full potential of the 4.6k sensor unleashed in terms of DR, shadow detail, highlight roll-off, and color science.” http://www.bmcuser.com/showthread.php?15520-Brief-thoughts-and-a-bit-of-footage-from-the-URSA-Mini-4-6K&p=192053#post192053 I changed my views on the 4.6k as the first RAW was finally released. To accuse me of “back-flipping” only makes you sound like a politician trying to discredit his rival for office. Our host Andrew has done a complete 180 degree on the Canon 1DX II, from downplaying the camera on this forum to stating that he wants to buy one after releasing his new C-Log picture profile. I’m very surprised, but there is nothing wrong with that. I’m very happy for Andrew and I hope he enjoys the 1DX II, because it looks like an awesome machine. People change their views all the time as new footage is revealed and new aspects of a camera are made functional. I think such evolution is something to celebrate, not ridicule. As a matter of fact, the 4.6K ProRes beta footage had not impressed me so much. With the RAW, I could see a lot more potential, better colors, and closer to the advertised DR. A lot of forum members also noted an improvement in the footage. I was not the only one who thought the camera was finally living up to its potential for “prime time.” There was then (near) universal acclaim. As for the Xyla discussion, I thought we all decided that it was pointless, since it just went in circles with no end in sight: “The. Lights. Are. On. There's no way you shoot that chart with the lights on. It totally invalidates the test. For real. JB.” http://www.bmcuser.com/showthread.php?15520-Brief-thoughts-and-a-bit-of-footage-from-the-URSA-Mini-4-6K/page140 Here was my response: “These men generously went to the trouble of setting up a Xyla test, capturing an image, and displaying it on Resolve. Then they started to move things around and turned on a light. By the time of the photo, the Xyla image is already a captured file on Resolve, while the chart has been moved out of the way. I don't know why this is so hard to understand. There is no reason for them to capture an image and view it in Resolve if it is taken when the light is on since you would not see a damn thing with that spotlight on it! You would not get any kind of Xyla image at all, and certainly not the one displayed on the monitor. When I said that we may be looking at a legitimate Xyla test and waveform in the Resolve file, everything went crazy here because it doesn't match what people want to see.” http://www.bmcuser.com/showthread.php?15520-Brief-thoughts-and-a-bit-of-footage-from-the-URSA-Mini-4-6K/page145 You then asked about the identity of the person in the photograph. I was the only one who responded with the answer: “The gentleman in the Xyla photo is a world-renowned cinematographer, Affonso Beato, from Brazil (he is at the computer). He shot Almodóvar's All About My Mother (1999) among other films in his incredibly long and distinguished career. I'm sure he had nothing to do with the Facebook post being pulled as he is not affiliated with BMD. It is more likely that BMD has a problem with a Xyla image from a pre-production camera out in the wild before the camera is finished.” http://www.bmcuser.com/showthread.php?15520-Brief-thoughts-and-a-bit-of-footage-from-the-URSA-Mini-4-6K/page149 You forgot to mention that part didn’t you when you said the following: “He didn't know the image WAS an independent source, namely the individual ASC technical committee members facebook page, but whatever...” Now let’s brush aside your ad hominem tactics and selective quoting from a year ago (because anyone can play that game), and get to the meat of the matter. I asked if you could point to any of BMD’s design patents on the 4.6K and here was your response: “Anyway, last I heard patents weren't the only way to protect IP. In fact, the most enduring way is to keep it secret...like Coke and Colonel Sanders do. Having or not having a patent is indicative of nothing at all. Once again, a very old fashioned out of date view about how IP is protected. Of course you are trying to perhaps have me confirm a technology partner of BM, when you know full well I'd never be able to disclose that without breaking an NDA, but I can assure you, the sensor used in the UM4.6K is not an "off the shelf sensor", nor is it one that you can just go order from whomever you think the vendor is. I know this because of actual personal involvement with its development. Please explain how I can have that so wrong and your version be more correct?” Fine John, let’s say for the sake of argument that it is not the Fairchild 4.6K sensor and was designed by BMD with no significant help from anyone else. Do you really believe that sensors and camera components are not necessarily patented nowadays and that it is simply sufficient to keep everything a secret like the flavor in Coca Cola? I would respectfully disagree. A $10 million sensor is not like a soda flavor or chicken recipe. Electronic devices and cameras involve numerous patents to make certain functions possible, as you well know. You posted no link to patents or evidence that BMD had any. But it is easy to find out that they do make patent applications, at least for the entirety of the camera. Here is an example of Blackmagic Design patent files I found with a search at the Australian Patent Office: http://pericles.ipaustralia.gov.au/ols/auspat/quickSearch.do?queryString=blackmagic&resultsPerPage= Interestingly enough, they are listed as “lapsed” for “renaming” suggesting that the company is going to renew the patent application for a camera originally filed on April 10, 2015. The camera is not named or described in any detail (we may speculate that it could be the Ursa Mini). There is also no mention of the sensor, but the application is very brief and simply for reference purposes. So BMD does make patent applications like I would expect them to, but has not secured the 4.6K patents as of yet, at least not with IP Australia. I would also note that camera patents are not always filed right away. BMD may be in the process of filing its 4.6K sensor patents, if they are in no way owned by Fairchild or any of their possible sensor partners. To write that history, I would need access to BMD’s internal company files, which is the only way for me to publish an article on that topic. That’s the way these things work. And, yes, there are different standards for what goes into a peer-reviewed academic paper and what I write on here as a Blackmagic camera user and former post-production expert with “opinions” about the quality of BMD’s cameras. As both a peer reviewer and writer of scholarly papers on film history and film technology, I’m well aware of the differences. It’s very tedious archival work to write on studios or camera companies and I assure you it has almost nothing to do with putting this stuff into Google. I’ve spent more than a decade in various archives going through internal company records to write such original studio histories, including on the history of cinematography and camera technology. “Does he even know what an ASIC is I wonder? Does he know that not all cameras use ASICs? Does he think that using an ASIC is the only way to make a camera, as Sony and RED apparently do, which of course means that Arriflex have no idea how to make a camera or design a their own sensor? Or Panasonic or Canon for that matter.” I never said that all cameras use ASICs. But image processors often use ASICs, as they are “application specific.” There are a few other options to the use of ASICs in image processing, but they are all very similar. More to the point, you are now just splitting semantic hairs, as what Land is saying in that quote is that RED designs its own image processors and sensors, which is largely correct. Here is an FCC teardown of a RED Epic-X that demonstrates exactly what Land is referring to when he claims that RED designs its own ASICs, which in this case are the Image Signal Processors or ISPs: http://www.extremetech.com/electronics/113331-red-epic-x-5k-camera-tear-down/5 Where you can fault Land is that he should have included Panasonic as part of “this market” since they also design their own image processors and sensors for cinema cameras. Canon’s image processors, the DIGIC series, are well known, but they use third-party components. Land should have also used the term “image processor” instead of ASIC, as not everyone who is reading the article will make the connection with this esoteric term. Indeed, I addressed RED’s mastery of the entire sensor and image processor chain on page 6 above, in case anyone was confused as to the function of ASICs. Did you miss that part as well? As I also stated above, of the cinema camera manufacturers only Canon, Panasonic, and Sony can do everything from sensor design to fabrication in-house, since these companies have their own plants. Samsung also has sensor fabrication plants, but it doesn’t make cinema cameras. Toshiba used to make sensors as well, but recently sold its plants to Sony. Is there currently a difference between BMD and the other cinema camera manufacturers when it comes to expertise in camera design? Based on the available evidence on patents, it would appear so. ARRI, Sony, Canon, and RED all have numerous patents relating to their cameras. We can list them here and their significance if you think that would help. BMD’s camera patents are perhaps forthcoming, as I demonstrated above. BMD’s camera division is also very new and it will take them a few years to file patents and have those approved. It can be a long process. I’m not an advocate for any company (well, except for BMD—see below) and I am mostly brand agnostic. You took an off-the-cuff remark that was really an aside and tried to turn it into the basis of a whole tirade, just because someone dared to criticize BMD’s quality control or its experience in designing cameras. You also came on this thread and forum out of the blue, like Superman coming out of the sky to rescue a stranded cat. Once here, you launched into a personal attack on me (“idiot,” “doesn’t know sh**”). Such a hostile approach doesn’t do you or me any favors (and, I admit, I also went over the line in response to your attack by using the “snake oil merchant” moniker: that was uncalled for on my part and I apologize). You are obviously very passionate about BMD as someone who has contributed input to their products and you have done your best to ensure their success. Moreover, Blackmagic’s continued innovation can actually be important for independents everywhere. It has already forced Sony and RED to respond by offering more and more features at lower prices. They have my respect in this regard, as was already noted above. But next time when you are quoting me, you might want to look at all those instances where I defended Blackmagic for all the good things they have done, including what I wrote here before you even posted on this thread. Never mind all the times I’ve introduced students to BMD cameras and told them about the amazing price/value of BMD RAW cameras for their first investment during or after film school. Yes, I’m also guilty of “peddling” BMD’s products but I always lay out the advantages and the disadvantages whenever possible. “It's very usual for camera manufacturers to partner with companies making sensors to also spin off the development costs to amortise them for other uses. I bet you can't name me another sensor company that has done this after developing a sensor that never came to market.” There are likely a few candidates. Hmm . . . is there multiple-choice on this exam, Professor Brawley? Based on that description alone, I would have to go with the Kodak CCD sensor that was adapted for the Ikonoskop and later the Digital Bolex. Okay, sorry for the length, but there was a lot of material. Let’s just try to keep things civil and respectful from now on.
-
Speaking of "sideways arguments," John, instead of addressing the issue of whether BMD contracted Fairchild to design the 4.6K sensor, you have resorted to another sideways debate on whether I know what ASICs are and how that relates to sensor and camera design. You then suggest that I consulted Google when forming my arguments and that I’m otherwise an “idiot” for accusing BMD for not designing their own sensor. If BMD designed the sensor instead of hiring Fairchild, where are BMD’s patents and why does Fairchild hold the rights to a sensor with the exact same specs and release date? It is easy to disprove your claims by demanding the patent documents or looking at the facts. But you don’t like facts. You like to hurl insults and use distraction and straw man arguments (e.g., ASICs), which simply display a lack of reason, logic and tact. It is also too easy to question your neutrality and credibility on the issue when you are so closely associated with BMD. But let me also give you some advice as you have no idea who you’re talking with here. You’re simply out of your depth on this one. I choose to remain anonymous here because people in my line of work don’t generally come on camera forums. In fact, they almost never do. That’s not because we have no expertise on film technology, cameras, film history, or the film industry. Rather, the opposite extreme is true. You might say that I’m a little bit of a “Professor” or “Dr.” on these things. Do you know what happens when you put my real name (hint: it's not "Kino”) into a Google search? As a matter of fact, aside from my industry credits from the 1990s and 2000s, you get a list of peer-reviewed academic publications on the film industry and the history of film technology and cinematography. Even your local university library contains academic journals with articles I've written on several of these topics. You might want to read some of those academic film journals before throwing the material back into my face on film cameras or technology. In short, I don't cite Google, but Google cites me and sells my book on the Hollywood studio system on Google Play. It's also available on Amazon.com and Amazon.com.au for your convenience as well as on iTunes, Barnes and Noble and the rest, not to mention from numerous university libraries and hardcover distributors around the world. Now I suggest to you that you conduct yourself with respect towards others online, no matter who they are or how you may agree or disagree with them. And this is not about insulting people you don’t know or who unbeknownst to you may have doctorates in film or media or years of experience and accumulated knowledge, but simply a matter of decency towards all the forum participants. And, please, cut out the snake oil merchant show. If you want to reduce yourself to peddling camera products on these forums, at least provide an ethics statement so people know what your exact relationship is to BMD. It's just a sad spectacle, in any case.
-
Land claims that RED is the only cinema camera company other than Sony that designs its own sensors and ASICs (integrated circuits). He makes no claims about where the sensors are fabricated. Of the cinema camera manufacturers, only Sony, Canon and Panasonic have their own sensor fabrication plants and have the potential to do everything in-house. Like BMD, RED would have to "outsource" the fabrication itself, which involves many steps, including the wafers from a semiconductor plant and the various other components that are then attached. The finished sensor just doesn't come from one place or even one factory. It's an elaborate process that is almost always outsourced for companies like RED and BMD. The difference is that RED designs their own sensors and boards and those are exclusive to them. You can't buy off-the-shelf versions of the Dragon sensor made by Fairchild or any other semiconductor company. It appears that BMD has to involve third parties like Fairchild in sensor design, suggesting that they don't have the same mastery that RED currently has. This lack of expertise may lead to problems such as we have seen with magenta-gate, considering how the sensor and its integration into the surrounding circuitry and the larger image processing chain would have to be designed with absolute precision. Perhaps they will get there eventually, such that they will have custom designed sensors that are exclusive to their products and not sold simultaneously as off-the-shelf sensors as with the Fairchild/BMD 4.6K. At that point, I would expect better quality control, which was the entire basis of our discussion. And, yes, it is entirely possible that BMD contracted Fairchild for the 4.6K with an arrangement to create two versions of the sensor: one for BMD and one off-the-shelf version. But the fact that they have to go to Fairchild would be proof that they did not design the 4.6K sensor on their own and do not hold the exclusive license or the patents involved.
-
From the Kinefinity owners I've read about or heard from, it seems that they have excellent customer service. They are also an honest company and tell their customers the truth, which is a great relief after what has happened with that other budget camera manufacturer. Distribution issues aside, I have much more faith in Kinefinity than any other camera manufacturer in that price range and I'm sure the 5K version will be along shortly. The Apple ProRes certification is also really difficult to obtain, so maybe they will push that to a future firmware update. Of course, everyone has to make a decision that meets his or her needs, budget, and production time constraints. If you require a camera right away, well then the Terra 5K is obviously not for you. In my case, I'm sitting on the Scarlet-W order/deposit whose allocation date is unknown but likely in early 2017 at the slow rate that Dragon production is going these days. This is despite the fact that I ordered months ago. In which case, the Terra 5K is not a bad proposition, considering that the only other (new) 5K camera in this market is going to require the same kind of waiting period or more if you ordered one today.
-
Yes, the 5K has improved rolling shutter over the 6K, which is exactly the same as the Kinemax 6K. I corrected the sentence above in case it was vague or unclear.
-
The Kinefinity rep at NAB claimed that the Terra 5K version will have a "dual gain" sensor (for an expanded DR) as well as an improved rolling shutter mode in contrast with the 6K, which has the same specs as the Kinemax 6K. He made it very clear that the 6K version will perform exactly the same as the existing Kinemax 6K: A lot of the Kinemax shooters seem to love the golden 3K that is for sure. Here is a very useful review of the Kinemax 6K by someone who owns and operates the camera for a living: To me, 5K is absolutely the perfect resolution as 6K simply results in little visual improvement at the price of much larger file sizes and storage requirements. Just ask some of the RED Dragon shooters. Meanwhile, the improvements with DR and rolling shutter performance make the 5K a much better option in those areas. If I can only see some footage soon, I will consider cancelling my Scarlet-W order and moving to the Terra 5K. It has an amazing feature set and is almost too good to be true.
-
Nice, but where have I seen this type of modular design before? Recall that the Kinemax 6K has the worst rolling shutter of any cinema camera unless you shoot in the 4K "sport mode," which then defeats the purpose of buying a 6K camera. So I'm not sure I would go for the 6K version. The Terra 5K, however, has the potential to be the best value in this price range, with the switchable global shutter and a 5K ProRes option. 5K also makes a lot more sense than the oddball 4.6K resolution in its competitors the Raven and Ursa Mini when considering the advantages gained in debayering for 4K delivery.