Jump to content


  • Posts

  • Joined

  • Last visited

Everything posted by maxotics

  1. Goes to show you how much I follow who-does-what anymore. "What We Do" blew me away. It's difficult to do a 2-minute video gag. 90 minutes? One of the best films I've seen in the past ten years. I watched it twice with my kids. One of them has probably watched it three times.
  2. That's the thrill of movie-making, talent saves no one from flop-dom! Indeed, when one looks at all that can go wrong, it's a miracle good films ever get made. I didn't mean to malign them in the least! They failed for me, that takes nothing away from anyone else, I hope! Well, I know! Hollywood does better with films I don't like.
  3. A friend once remarked to me that he enjoyed reading Edgar Allan Poe because he could tell when Poe wrote one paragraph in a good mood, then the next day, he's in a crap mood, and would then write a depressing paragraph. In other words, our emotions change hourly, daily, and they will change our emotional choice in colors. So not surprised you're driving yourself crazy There is NO perfect color. LUTs are just another person's color opinion, emotion. As @hijodeibn said, calibrating your monitor is your first step. I have a ColorRite but it creates more problems than it solves for me, so I just use a color chart I find on the internet. First use your software's calibration of brightness/contrast though (gamma). If you don't know a lot about how cameras create color, I did a video on YouTube, search "Camera RAW is Color Blind". After that, try to decide on a look, create a few reference shots, and grade to them.
  4. I'm a total sucker for treasure movies too, but not for fakery. I only watched ten minutes. That was all I could take, sorry. A similar movie that did work for me is "What We Do In The Shadows". It too, is shot like a documentary, but it is written for laughs and I cared immediately about the characters. In a real documentary, the camera doesn't move around to give a sense of movement, as it does in TFTW, it moves around because it must keep up with the subject in the frame. Right away, I knew it was a fake documentary (camera movements too studied) and I found that obnoxious; again, fakery for the sake of looking truthful instead of lying to make a genuine point. So from the start, I felt like I was being manipulated. It was just too obvious. The people interviewed in the film (real, characters?) sounded idiotic, whoever they were. Every frame shouted "look at a great filmmaker I am". What is art? Well, we could talk about that all day. What is fakery? That film. It's tough for me, being alone, not being part of the crowd. The older I get, the less tolerant I am to lying, artistically, though I am more open to different kinds of art! On the bright side, that film strengthens everything I hold important. It makes me remember to talk up the art that don't get enough recognition. It makes me appreciate other documentaries like "The Artist Is Present", where the artist recognizes that they straddle the line between political nonsense and having something to say. Sorry for the negative review @Jonesy Jones But thanks for the suggestion.
  5. Very nice work! My opinion is that you've already bought the perfect camera for what you're doing. If you want your documentaries to have a more "film" look than you need a camera that shoots RAW. Only RAW will give you that individual pixel grain/color look that is washed away by 8-bit video compression. You could either get a Canon 5D3 and use Magic Lantern, or get a Blackmagic Pocket Cinema Camera BMPCC. With the Panny 12-35mm/2.8 lens you should be all set. It's small, perfect for what you're doing. However, this technology is NOT easy or fast. It requires a fast computer and patience. If you want to see if you can see a difference, then look at videos that were shot using RAW cameras. Here's something I found searching "Brazil and BMPCC", not best shooting, but gives you an idea
  6. The drawback to that system of color checking is it uses 8-16 color samples at a one level of exposure/dynamic range. I believe it works well enough. Indeed, I don't think it that hard to match footage is you shoot standard profiles (with minor tweaking) with white balance set custom on each camera to the same reference gray card. Do many people care about this problem? Does anyone? I'm not suggesting you made an unimportant post @webrunner5 . I'm really curious. I've been developing a system that can use 100s of colors to perform what I believe would be a very precise match. But I've lost interest. Maybe I'd put more time into it if I felt people would use it, have an interest. So, does anyone have a real-world problem matching their cameras? Or, as you know, I've explained how LOG profiles are destructive of color information, yet everyone seems to use them EVERYWHERE. Maybe that makes matching that more important? Again, curious what people think.
  7. Let's not shoot the messenger? He's already short one eye What the heck is happening with EOSHD? I get the sense Andrew is burnt out. Not that I don't sympathize. It seems consumer video has reached a plateau.
  8. None of the cameras you've used, I would call a cinema camera. My solution to your problem, which is similar, is a Canon C100 for video. I can also shoot RAW on a 7D through Magic Lantern. I've made this decision because I believe a sensor can be optimized for STILLS or VIDEO but not both. Mostly a physics thing (pixel size). Also, good audio often requires XLR, so you get big in a hurry. If you are going to use one camera for both, then the D850 or A7RIII are both quite capable. All depends on what you want to shoot of course.
  9. Welcome to my world When I read people talking about getting 14 stops of DR with a camera, in video mode no less, I can only roll my eyes. Nikon/Sony cameras can get sort of close to 14 stops at ISO 100, but nowhere near at ISO 400/800 say (which is where most video is shot) To extend what you're saying. Even if you ignore the camera's suggested exposure, which I agree, is tweaked according to what mode you're in (unlike what a light meter would say) , you have to consider the sensitivity as pixel exposure decreases as you go from brightest to lowest light. So if your middle gray, or wherever the sensor can most accurately read light, is moved up or down, it effects the NOISE 2 stops below it. That is, if you you expose to avoid clipping you may add noise below the image. If you clip, you may get better saturation (less noise) below--AT ANY ISO, because again, ISO just considers your center exposure. Cameras today are truly marvels. That said, it's still very difficult, for me at least, to really nail exposure. It's an art, for sure. Getting a really good exposure, both in setting up lights, and setting the camera up, is no trivial matter. I believe the best photographers really SWEAT those details--getting the center 6 stops in the optimum range. Like Andrew's latest video. The way the lights were blowing out drove me a bit nuts. I wonder, why can't the camera just figure that out? It has a computer in it But it can't. To get the image Andrew really wants, I assume, he'd have to replace all those practicals with low watt bulbs and then bring up the room light to the right ambient level for the camera. The camera, any camera, is really the least of one's problems So when I read people saying they want 10bit, or LOG gammas, I think, I have enough trouble getting 8bit normal gammas working well! I mean think about it, why should we have to set a LOG profile and all those parameters. Why can't the camera analyze the scene and give us some options? Maybe one day. For now, I agree with your last post. You either A) sweat the exposure if you're looking for a certain level of color fidelity or B) put it on Auto and focus on composition. Doing both at the same time. I can't do it. Not even close.
  10. I sympathize with your comments. I don't shoot enough with high end cameras to speak from experience. From my analysis, however, I believe that the "don't-shoot-well in low light" reputation of BM cameras is a distortion. All 8-bit video requires a fair bit of noise-reduction. Add to that the chroma softening inherent in 422 or 420 video compression, and you're going to end up with an image that is going to look fairly similar from good light and down. So on an FS5, for example, you won't see much of a color difference between low light and good light, only a difference in noise. With the BM cameras, (or any RAW camera) the god light image is going to look significantly more detailed than the FS5 image. On this I have tested over and over again. I can see the difference between RAW source and compressed. Not a difference most people would recognize, but if you work with video you can see it. So when you look at low-light RAW on a BM and low-light on the FS5, the FS5 will look more quickly usable, but I'm not sure if it would look that different from the BM if the BM was built to process 8-bit NOISE REDUCTION video. Or, if the FS5 did shoot RAW natively, not sure it wouldn't get a bad reputation too. Put another way, the kind of detail a BM camera records in RAW is something one wants to maintain, but is impossible in low light. It feels a negative of the camera, but it's really a matter of EXPECTATIONS LEAD TO RESENTMENT
  11. Yep, LEDs get hotter and hotter and they still have color spikes, etc. I have Fiilex lights and I notice in some clips that the color temperature shifts here and there. That never happens with Tungsten. And of course, again, tungsten, though reddish, delivers a nice smooth color spectrum. Of course you know all that. But for any newbie reading, if you see some cheap tungstens jump on them! I see some lowell kits going for near nothing. Except for the heat and their hunger for watts, the quality of the light cannot be beat! Young filmmakers have no idea how lucky they are to have all these quality lights available for pennies on the dollar.
  12. Sorry to interrupt this fine threat, but did you guys notice @TheRenaissanceMan's signature "Tungtsen fo' life" HA HA HA! (and agree, at least in the winter)
  13. I'm pretty sure DPAF is a combination Canon sensor design and software, so don't see Panasonic being able to do that UNLESS/HOWEVER, now that they have two data channels coming from the sensor, for their dual ISO stuff, it's possible they could use one of those channels for better autofocus. I'm shooting from the hip hear, but what the heck. If something like that is possible they could perhaps get super fast auto-focus in 1080 or 30P 4K. If something like that was possible, I believe it would take some time so can see why it wouldn't be available in the first model with that sensor design. So maybe GH5S2 could have a new focus mechanism? Another indication that it is something they're working on is the lack of IBIS. Pretty sure that mechanism would confuse the focus issue. Who knows? Anyway @Trek of Joy I think it smart to wait for the next model to plunk down yer sea shells. It's a new sensor, my guess is that it's not completely built out yet.
  14. First, LOG gammas reduce color fidelity. They stretch out brightness DR at the expense of mid-tone color. It should be easier to match LOG on both cameras, but only if you go for that low-saturation "Handmaid's Tale" look. Your problem is exactly what I warn about on this forum. You can't match colors that are no longer there. First, put both cameras on a neutral, standard or portrait profile, make sure you have both white balances to the same number, might be 5000K, whatever. then shoot some test clips and compare them. YOU MUST FIRST get an idea of their natural differences. Then, in your editor you can try something like this (someone's else's approach) on the Sony: Orange -17 (make orange hues more red) Yellow -30 (de-emphasize yellow hues) Green +10. My guess is that in their native profiles you will get the cameras fairly close. After that, in shooting, you can try on the Sony: Creative style setting "vivid", sharpness +1, saturation -1, contrast 0, during picture taking will also produce more realistic 'Canon like' colors.
  15. I put the Excel file where you, or anyone, can get it, if interested in the code. It's a freakin' mess. Lost of old stuff in there, but if you're like me, you just need to poke around to get what bits you want. The hardest part for me was figuring out how to get the RAW data out as integer values. Turned out it could be done in a few lines of Python code. Excel with VBA Download: https://drive.google.com/open?id=1d6-tBeVcbyiVOCXgSRiV2eNe80ipfruR PYTHON import rawpy import numpy raw = rawpy.imread('D:\\Files2018_Maxotics_Analysis\\RAWExplainDR\\7D_RAW\\DNGs\\M08-2120_DNG_000001\\M08-2120_000520.DNG') rawdata = raw.raw_image numpy.savetxt("D:\\Files2018_Maxotics_Analysis\\RAWExplainDR\\7D_RAW\\DataToCSVs\\dng520.csv", rawdata, delimiter=",")
  16. Very interesting! Thanks! Now I understand why the published contrast ratios don't work in "real life".
  17. In my ongoing self-appointed role as "LOG gamma" police, I've created a video that shows how RAW is processed into images we use, both for photography and video. It drives me nuts that people talk about 14+ stops of DR, as if noise, or sensor pixel sensitivity, stays constant from lower to higher values. After 5 stops, IMHO, noise makes any extra DR recorded by the sensor quite undesirable. Further, even in 5 stops of DR there is a huge difference between how well cameras record DR even in fairly well exposed image. That is to say, the GH5S isn't better only because it can record in low light. It can be better because even at ISO 800 it will get a richer image in the shadows. However, it is still 8-bit. It still throws out gobs of sensor data. Well, enough of that. Another video to put you to sleep
  18. I say this until I'm blue in the face. I'd add how a good Gaffer shapes "good" light. Day-for-night is an example of how cinematographers used good light to get the look they want (as if night) instead of a super grain/noisey image they'd get if they shot in real night. Another way to say what you're saying, is the only difference between an A7S/GH5S in good light shaped to look, say "romantic" and an A7D/GH5S in low light which is naturally "romantic" is the second will be full of noise. Both can look equally "romantic" if the gaffer knows his stuff. Shooting in low light should ALWAYS be avoided at ALL COSTS unless NOISE is what you're after. I understand most people will use these cameras outdoors, or indoor where they can't control the light. For those purposes, those cameras fit the bill! But for studio/set work? Put your money in lights and modifiers--control light/noise so your camera DOES NOT STRUGGLE, don't become a slave to the camera's signal/noise ratio
  19. I don't know if I mentioned this before, Bill Claff did a DR test of some Sigma cameras and found a DR of 6. He was flamed off some forums for reporting those findings, which I believe true. But they do NOT make a difference, at least to me. We don't look at images beyond that range. The flip side of the question is, aren't the Nikon images too low in contrast (too washed out?). Nikon has the widest DR tested. But again, the question isn't how much DR of the physical world a camera captures, but how well it can deliver sensor values into our 6 DR viewing space. In my tests, when I've tried to match bayer images with Sigma images I have never been able to get the same saturation look. The bayers always look slightly washed out in comparison. That's to say, if you do get the Sigma camera and have the time to work each RAW image to match contrast, the Sigma will still maintain that 3D look--definitely against an APS-C sized sensor. As for why no one uses it in a studio. Let me be blunt. A studio is for fast, efficient professional work where the end-user is almost never looking for the kind of quality that separates cameras. Therefore, the Sigma is too slow. Second, most photographers are technically weak; that is, they don't really understand what the camera is doing under the hood. Third, they need to justify in their mind why their camera is already the best. To be fair, as I've said, I would not use a Sigma cameras for professional work. I'd only use it as a fine art photographer. I'd DEFINITELY use it for portraits, but again, not for typical professional work--it's just too slow and time is money! If you're just starting out and don't have the money for a D810/D850, the Sigma will give you a Medium Format look that I believe will impress people. I did some tests I did 5 years ago trying to match bayer to Sigma cameras. I couldn't do it. But again, if I don't know any better (like most people ), bayer images are fine! http://maxotics.com/2012/11/26/sigma-dp2s-vs-sony-nex5-with-18-55mm/
  20. I believe there's a zoom assist, probably buried. As for radio control, you can get something like this https://www.adorama.com/us 934638.html?gclid=Cj0KCQiA7dHSBRDEARIsAJhAHwjCws--uxALtZ9Oln3j_78EvcqcNnWm4ZzbuMSq5-hBycBDGr18caEaAgYMEALw_wcB
  21. Really? You know I gotta say it, again. Slow? Sigma? The Sigma colors above are significantly richer, IMHO. Though, the Nikon will always be fine. Still, I can never get that rich look with bayer cameras. I take it you didn't buy the camera, just tried it out?
  22. You might look at this too. He did one for a high end PC, and one for $800
  23. Apple has a target data rate of 884 Mbps for 10bit 3840x2160 UHD ProRes 4:2:2 HQ, The Blackmagics can do that, and even 4:4:4:, the GH5 cannot. The 400mbps for the GH5 sounds like a huge data rate, but it's to reduce temporal inter-frame compression,. It isn't true 10-bit color depth, as I would judge it. Simply put, Panasonic and Sony can't match the data processing power of a Blackmagic. Of course, the BM cameras are bigger than they look in pictures, a lot bigger, like "I'm not carrying that thing around" bigger! For the size of the GH5/S, one can have no complaints.
  24. Exactly. That's the story here! The A7RII and A7SII are the end of the tech line for light sensitivity in full-frame sensors, now the GH5S is the maximum a MFT sensor can do. One can read between the lines in how important Panasonic believes light-sensitivity when they take out IBIS to accommodate more data channels for split pixels gains--though that assumption might be wrong. Or it might be as IronFilm just said, the size of the sensor. Either way, the manufacturers seem to be up against many limits to the technology. Sony made the right bet a few years ago (in going large pixels). Panasonic may be too little, too late here. Sony A6300s can be had in good shape for near $600.
  • Create New...