Jump to content

jcs

Members
  • Posts

    1,839
  • Joined

  • Last visited

Everything posted by jcs

  1. ​Hi Giulio- what profiles settings where you using? (Slog2+SGamut, Slog2+(Pro|Cinema|Rec709, etc.)). What was white balance set to? How was exposure set? Blue lights? If so what kind? (would like to test them).
  2. The FS700 buttons+UI are also over complicated and inefficient to use. However, the UI itself is not slow. While I'm sure the FS7 is improved over the FS700 (IMO the A7S looks better than the FS700 in some ways), it would be helpful to see the FS7 compared to the FS700 shooting the same scene to understand exactly what has improved. I haven't used my FS700 in a while- have been waiting to see how the FS7 looks and if it's worth $4+k to sell the FS700 and upgrade. In one test comparing skintones, the FS7 did really well against the C300 and A7S (though IMO in a few shots the A7S had the nicest looking skintones). Here's a humorous test I found which also shows decent skintones with the FS7. Regarding your motion cadence comment: guessing you are referring to timing regularity between frames? For example, always 1/23.976 seconds between frames, with no variance or appearance of dropped frames / stutter? Toward the and of this test video (1:17), there's significant stutter, guessing they did this in post (24fps played back at around 12fps). Could you explain more by what you mean by 'video cadence'?
  3. ​Did Sony release math/equations necessary to create a LUT/matrix which transforms A7S specific SGamut into Rec709 for the A7S?
  4. jcs

    A7s and ENG project

    For paid ENG work definitely get a workhorse ENG camcorder, especially if quickly and reliably getting the shot is most important requirement. The A7S with something like the SEL18200 lens and an external mic (perhaps with the $700 XLR attachment) could work, however it will be more fragile with more points of failure. ENG camcorder IS, AF, smooth power zoom, and pro audio is much more suited for this kind of work.
  5. I think what hmcindie was saying was to mask a window that was very bright and not clipped. You can do this with the sky- after shooting it looks white. In post, the sky can be brought down, red & green pulled, etc., and now we have blue sky and clouds too (can't mess with red & green too much else clouds will start to blue too). If it's clipped, this will still work, however clouds will be gone and a gradient will also help make it look real. In the case of a window, you could shoot plates of just the windows exposed properly, then in post mask and create alpha channels for windows, then compose the window plates back in. If the scene outside doesn't fit the story, this opens up the option to shoot plates elsewhere, etc. In some cases the simple solution suggested by richg101- ND film placed on the windows will make the most sense and be the most cost effective.
  6. SGamut is Sony's 'top-of-the-line' ultra-wide color space. Without a proper color transformation, SGamut is time consuming to color correct. After a bit of testing, including using a simple 3x3 linear color matrix (thank's dhessel), I found the best results with 3D LUTs. These will work in Resolve, Speedgrade, and PPro (load with a Lumetri Effect): 3D LUTs from Sony: http://community.sony.com/t5/F65/NEW-F65-3D-LUT-s/td-p/135903. Since the A7S gamma in Slog2 isn't full range, using the included Slog2+SGamut to Slog2+Rec709 3D LUT is a good starting point. Once converted to Rec709 color with the Slog2 curve, a custom curve can be used in Resolve, then the final look saved as a 3D LUT which can then be loaded into Premiere Pro (real-time effect). Here's Art Adam's article comparing Slog2+SGamut2 to Slog3+SGamut3: http://www.dvinfo.net/article/acquisition/sonyxdcam/sony_sgamut_vs_sgamut3.html . Alister Chapman's A7S 3D LUTs: http://www.xdcam-user.com/2014/08/set-of-20-cube-luts-for-the-sony-a7s/ac-slog2-to-709-luts-v2/ . These are pretty nice too, and have versions for various levels of exposure. These might be helpful as Slog2+SGamut changes color behavior with exposure (fixed in Slog3+SGamut3). In my tests so far, the F65 LUT from Sony for Slog2+SGamut to Slog2+Rec709 appears to be working OK for skin tones. Using Resolve 11 with the above LUTs it's possible to get some pretty impressive looking color and skin tones with SLog2 and SGamut. Also a good opportunity to learn more about Resolve (more impressive the more I learn about it). I'm currently comparing custom Rec709 picture profiles to SLog2+SGamut to see if it's possible to get the same looks using Rec709 in-camera (much easier to expose, focus, etc.). In other words, are there any practical advantages to using SGamut vs. Rec709-based color modes for the A7S?
  7. Isaac Asimov's take on this topic is one of my favorites, "The Last Question": http://www.physics.princeton.edu/ph115/LQ.pdf
  8. Those papers unfortunately don't reveal any of Canon's color processing algorithms past de-Bayer to RGB and YUV. There's only one simple linear equation (luma calculation). The papers only apply to the C100, C300, and C500. The 1DC uses a different sensor.
  9. ​Right on- another good one! Looks like he turned that into a series too.
  10. ​I liked the 5th Element more, however I agree Léon: The Professional rocked. http://www.imdb.com/title/tt0110413/ (IMDB agrees with you: 8.6 rating).
  11. For serious Luc, I didn't like The Big Blue, however IMDB gives it a 7.7 vs. 7.6 for Fifth Element. Perhaps time to watch it again (great cast; I did like the music: Eric Serra): http://www.imdb.com/title/tt0095250/. Once I started paying attention to color/filmstock, I began to notice how many of my favorite films were shot on Kodak 5248 (including the Fifth Element). Color helps with story/emotion: http://shotonwhat.com/?s=5248 The Transporter series is less campy- has done well (Luc as writer).
  12. An H.264 option would make a lot of sense (with a bitrate high enough to match H.265 (about 2x)).
  13. ​Indeed, used a bunch of images and video from Hubble/NASA in the philosophical doc; helps put things in perspective and our place in the universe.
  14. The VO/story and Hans Zimmer-like music are much stronger than the imagery, which is also lovely, just not as strong as the other two. The gear is really irrelevant and is a good sign of a good story. Would not be surprised if Zenpmd frequently made meaning of life posts . Facing death is the most life changing event one can go through (next up is experiencing death, the ultimate life-changing experience!). After such a moment (or moments), the meaning of life becomes much more clear, and what makes us happy and peaceful changes forever. Film can help those who have not had one of these experiences to see reality in a different way, as can studying Zen/Buddhism (at its core, not a religion; a study of the mind and the nature of reality, guided by the same principals as math and the scientific method). Life has more meaning the more we let go of materialism and ego/selfishness and increase kindness and connections to other living beings. A couple couple of books: http://www.amazon.com/Peace-Every-Step-Mindfulness-Everyday/dp/0553351397/ http://www.amazon.com/Heart-Buddhas-Teaching-Transforming-Liberation/dp/0767903692/ My first documentary was related to the meaning of life topic: https://www.youtube.com/watch?v=BLbZ_nwm8-I .
  15. ​Thinking as well as creating tools, technology, and processes. As a developer I would sometimes go to extreme measures to ensure the code I was writing would be ultra-reliable. For example, for the first XBox live game, I used custom tools to test the reliability of Microsoft's TCP implementation (which was custom running on top of UDP). Microsoft's implementation broke fairly easily, so I wrote a custom TCP-like implementation on top of UDP. My implementation was much more efficient and reliable, but occasionally our game would fail and it pointed to a low-level network problem. I wrote custom tools to try to break it in a controlled way with tons of logging (running for days at a time). Finally it failed with full logging to help find the issue and the bug fix was just one character of code. Much to the surprise of everyone, our little company was first to ship a game on XBox Live. As a manager I would only allow certain developers to build systems this way after they earned the risk to take the extra time. Many times software is disposable, and there's no point over-engineering it. However, more times than we'd like to admit, systems are built fast with disposability in mind and some systems live long, painful lives (for the developers tasked with keeping the bug-ridden spaghetti code barely running!). Perhaps some camera systems have been built in this way, which could explain lack of quality/innovation. ​16-bit RAW CMOS, very low noise and aliasing, wider color gamut than film(!) https://pro.sony.com/bbsc/ssr/show-highend/resource.solutions.bbsccms-assets-show-highend-F65.shtml . They used to call it "8K", but perhaps don't(?) after being busted when the RED Dragon came out: http://camerarentalz.com/sony-f65-6k-sensor/ (more like 6K). I will say for Oblivion and Lucy, the F65 has a magical quality to color that may just exceed the Alexa (it did for those directors who can use any camera). By magical I mean evokes a very positive emotional reaction, perhaps in the same way Technicolor did back in 60's. Check out 'Robinson Crusoe on Mars'- amazing color: http://www.imdb.com/title/tt0058530/. Wow, only 12 features shot on F65 according to: http://shotonwhat.com/?s=F65 . After Earth also looked pretty good (average movie, but still worth watching). The F65 is a bit of beast for shooting and post- that may explain why it isn't used very often. If Sony put F65 technology and color science (or better) in an F55 sized package (or smaller), that might shake up the high end! Red Dragon keeps improving color science too. Speaking of magic and emotion for images, the A7S (and Sony in general) can sometimes have colored noise present in just the right way that makes an image look magical, like some film stock (Kodak 5248 is my favorite), typically in outdoor, direct sunlight shots. This colored noise is even present in highlights- might be possible to do this in post to make blown highlights look more pleasing. ​Have you seen 'The Fifth Element'? One of the silliest and most fun SciFi movies ever! That's what Luc loves to create, with deep, epic storylines. Lucy and The Fifth Element are very similar in that regard (though I found Lucy to be far more serious in comparison). The idea that either through mind expansion of humans or through AI that we'll rapidly figure out the nature of the universe is intriguing. Kurzweil calls this rapid expansion of knowledge The Singularity (http://www.singularity.com/). IBM and Google are racing to be first with a functional AI system with beyond human intelligence. Why not just 'human level' intelligence? Because the theory is human level intelligence in a computer will rapidly evolve far beyond human intelligence. Douglas Adams had a more humorous idea of what this meant with The Hitchhikers Guide to the Galaxy: http://www.imdb.com/title/tt0371724/. If you have seen the BBC version of it, it's wonderful (amazing given the low production quality of the time- great story and ideas!). I added AI+Cameras to the storyline for our short SciFi Delta after shooting began (in VO). The idea being that whoever controls the AI+global-camera systems controls the world (in Delta, a planet in parallel universe).
  16. ​If Canon = red bias (R), and Sony = yellow bias (R+G), then taking the difference D = Sony - Canon, D = (R+G) - R, D = G: the difference simplifies to a green bias. The green bias with Sony is mostly with SGamut color mode. Cinema color mode has less green, Pro has even less. I can now fairly well match 5D3 RAW with the A7S skin tones with custom tweaked profiles, never using SGamut. While SGamut make work OK if corrected in post, I suspect without the proper color science workflow in post, when skin tones are corrected to look less green, the background may end up magenta, etc. Some lights have color spikes/bias, especially low CRI LED, FL and CFL, adding to the challenge. More info here: http://www.eoshd.com/comments/topic/7762-lighting-white-balance-and-skin-tones/
  17. Part of the challenge to getting good skin tones is lighting. Many lights have a green or magenta bias (cheap LEDs, fluorescents, CFLs, etc. (some better than others)). This can be especially troublesome when the lights are far away from full spectrum. You can think of light as a lookup table multiplier and real-world objects having coefficients which determine what color they reflect. For example a fire truck has mostly red with less green and even less blue. If you shine a blue light at a fire truck then try to correct it in post, it won't look very good. If you were to shine a blue laser at a mostly red object, you'd see mostly black. Here are some example spectra: Now it should be clear why shooting in daylight or incandescent we get the best results. The color multipliers are more uniform across the spectrum, and all the subtle off-shades are captured. The worst example is fluorescent (some fluorescents are much better than this example, especially those designed for use in photography/interior design). White balance (WB) is most simply applied to an image as an RGB multiplier, for example (1,1,1) (white, won't change the image, (1,.8,.6), a WB to warm up an image, etc.). It's clear that if we process the WB this way white may look white, but other colors will be sorely misrepresented due to spikes in the spectra (the next more complex step would be to use an offset in addition to the multiplier). Some colors will be way too saturated, some the opposite. Do modern camera systems use (3D?) lookup tables or complex algorithms to try to compensate? I would expect that presets could do this, and by looking at the color temp they could guess at the type of light source (low K = the incandescent curve, high K the daylight curve, then interpolate for in-between values). Halogen has a yellow peak, cool white LED has a blue spike, and warm white has a green peak, etc. Ideally we'd want the camera system to compensate for these cases too. It's likely the cameras have algorithms for figuring out the spectra of the light from histograms and other means, where some cameras to a better job than others. Canon has paid particular attention to skin tones. This attention could be all the way down to the sensor+hardware behavior, which would put less of a demand on the software side to correct (lower power as well). 5D3 RAW appears to show this. One of the reasons computer generated imagery can look fake is the simple light model (just R,G,B multipliers for lights and object materials). Realism is improved when taking into account spectral properties: http://en.wikipedia.org/wiki/Spectral_rendering In the early days of computer graphics, we had to write our own renderers from scratch, the most basic models using simple scalar coefficients for both lights and materials. Even with fancy texture maps, specular modeling, filtering, and post processing, the CGI had a fake quality to it (even all they way up to Pixar's Renderman used in many movies). Now we have renderers which take the spectral characteristics of light into account, where the final results are looking much more realistic (some would argue photorealistic): http://www.maxwellrender.com/, http://en.wikipedia.org/wiki/Metropolis_light_transport So when our real-world images start looking fake or plastic, the simplest first step would be to take a look at the lighting. If we set white balance so white is white, skin tones might be off. So we adjust skin tones to look more correct, now the background and even whites might be off. That's a good example of low quality lighting and/or weak color science in the camera. One reason RAW cameras can look so amazing is that much more powerful color science can be used in post (too complex to run on the camera). Currently the easiest way to deal with lighting issues is to shoot a color chart, then using Resolve's Color Match feature. This can work remarkably well in getting two cameras to reasonably well match. It's more work to shoot the charts, and it doesn't always result in a pleasing look, especially for skin tones. Kudos to the camera makers with good skin tones under most lighting conditions!
  18. Higher ISO = more noise. Is it possible the reason they are forcing higher ISO for Slog is to create noise for dithering to prevent banding? I've been noticing a lot of banding in my indoor low light tests with the A7S when not using Slog2- basically noise free and clear banding in mid-dark regions. Next tests will boost ISO to see if the noise will help reduce banding. Agree re: tradeoffs for Slog- skin tone is much more important to me than highlight and shadow behavior (though I'll take it all when available ).
  19. Yeah, Oblivion is my favorite F65 film to date. Beautiful color and sound+music- great acting and cool story too. Even on Apple TV at 1080p and highly compressed, the detail and colors come through. Sony can do great color- is it the result of pro colorists taking extra time for features, or is Sony holding back? The F55 doesn't look as good. For example, Marco Polo on Netflix was shot on the F55. The first few episodes didn't look very good, though it did seem to improve with time (cool series- well done). The older F35 also has the Sony 'magic' colors- some folks picking those up used for a fraction of what they went for new ($250K). I will say with a bit of time and tweaking and good lighting, the A7S can look pretty good. Perhaps it's time to get the FS700 out and try applying what I learned on the A7S to the FS700. A lot of other folks don't like the FS700 color- I'm curious with more experience can folks get better color. Sure, it's more work than Canon cameras, however they can't shoot 240fps. Initial examples of the FS7 are looking pretty good against the C300. Did Sony further improve the color science, or are people just getting better at matching Sony to Canon. That's how I'm learning- using 5D3 ML RAW with mlrawviewer straight to ProRes 10-bit 444 as the reference and tweaking the A7S to look as close as possible. Folks might wonder why anyone would spend so much time figuring out how to make the A7S look like 5D3 RAW. Well, once figured out, it will be easy to apply during production. The A7S can do 60p at full 1080p (crop mode), ultra low light, less noise, tiny files, can use autofocus lenses, etc. It's also helpful in learning more about how to work with color. In the software industry there's a saying, "Lazy Programmer", where a programmer goes to extraordinary lengths to work out a solution to problem that will save much more time later. The concept applies to filmmaking too.
  20. Looks like most of the shots were on the F65: http://www.creativeplanetnetwork.com/news/shoot/superhuman-power-and-sony-f65-processing-creating-spectacular-visuals-lucy/606871 Car chase was on Red Epics and other shots Arri Alexa. Really impressive look and fantastic color. Movie was pretty good too- Luc Besson writes, directs, and shoots good stuff! Curious how the F65 color science differs from F55 on down the line. For this film Luc put the F65 above the Alexa for color (not just 4K).
  21. ​ML RAW Mark III provides 1080p and more, but not 4K. The 1DC provide 4K with a 1.3 crop. Currently, only the A7S provides full frame and 1080p internal, 4K external. At the highest end is the ARRI 65: 65mm! (priceless: rental only).
  22. Filmic: shallow DOF is helpful to blur the background- easier to do with larger sensor. For example, I have to use an F.95 lens the GH4 (2-2.3x crop) to do about the same as an F2 lens on full frame. Filmic also relates to how highlights look when they approach clipping (and clipping itself). Once clipped, there's no way to fix in post. Shadows are typically of concern for noise, and tend to get tweaked in post for the filmic touch (making shadows slightly teal, green, red, etc., for the desired emotional effect). ARRI desaturates color as it approaches clipping, here's an example showing how to do this with the F55 in Resolve (using the lum vs sat curve): http://www.dvinfo.net/article/post/making-the-sony-f55-look-filmic-with-resolve-9.html. Any camera that does this in-camera means a more filmic look (and less work in post). Filmic also means absolutely no digital artifacts with a pleasing organic looking noise pattern. Too sharp is also less filmic (fixed in post or with lens filters). No noise with 8-bits is a problem due to banding (something I see from time to time with the A7S). This can meaning using a higher ISO to increase noise and reduce banding before compression (banding can be reduced in post using added noise/dither, but never as good when done before compression). Film has a rather unusual color response- doesn't really look 'real' (more so older stocks); it's off a bit and that's a good thing- helps create the 'unreality'. A very big part of the filmic look (perhaps most important) is filmic/cinematic lighting. In simplest terms, it means light only where it's needed, and using darker scenes with controlled light on face/eyes goes a long way. Even outdoor sunlit shots never have super-high specular reflections- nothing is super bright. Cameras which spread out highlights smoothly tend to look more filmic (ARRI, Canon to an extent). This can be helped with lighting and makeup (and in post).
  23. jcs

    5D Mk3 Loupe

    I purchased one of these for $129: http://www.amazon.com/gp/product/B00EAXU6UM/ (price has gone up). Optically excellent if the mount works for you. The viewfinder is held securely by a magnet, with no tape/glue on the camera. The magnet came lose (super glued from factory?), however easily fixed with clear tape (could also use super glue, epoxy, or perhaps the new UV cured liquid plastic). The tape is only need to keep the magnet from sliding out (other axis held by metal). I mounted a handle on one of the rails- makes for a compact, stable platform.
  24. dhessel created a matrix that might work: http://www.eoshd.com/comments/topic/7145-grading-s-log2s-gamut-in-adobe/. I didn't try it as I don't use sgamut (haven't seen a clear reason why it should be used over other color modes). If I were to use it, I'd find a way to use it in Speedgrade or Resolve so a LUT could be used otherwise in PPRo it won't be hardware accelerated. PP7/Slog2 gamma with Pro or Cinema color modes and added saturation can work well. In other words, get the color to look as good as possible in-camera using the Slog2 curve: better results (so far) and less surprises in post. Here's Kholi's example (the biggest change is using Pro color mode with saturation +8): http://www.dvxuser.com/V6/showthread.php?326049-A7S-PicPro-Settings
  25. Hitfabryk- in my tests PP4 works best in lower light conditions (as does PP3), PP5-7 in brighter light conditions. In their examples, IMO the A7S looks best in a few cases: color & noise. toxotis70- I found this test via a search. Here's the full info: http://www.keyframe.tv/wp/2014/sony-fs7-alpha7s-and-canon-c300-color-science-comparison-part-1/, http://www.keyframe.tv/wp/2014/sony-fs7-alpha7s-canon-c300-color-science-comparison-part-2/
×
×
  • Create New...