Jump to content

tupp

Members
  • Posts

    1,148
  • Joined

  • Last visited

Everything posted by tupp

  1. Other than the camera/sensor mentioned by @androidlad, scientific and machine vision cameras likely exist that have a greater dynamic range than an Alexa (and an A7S). The Panavision Dynamax sensor supposedly took three different exposures, which yielded a 120dB DR. I usually don't make qualitative judgements on the results of such early tests, especially since whoever blazed this trail into HDR was a pioneer, lacking the benefit of years of HDR tweaking by others who subsequently jumped onto the HDR bandwagon. However, I wouldn't call it "bad" nor does it seem to be "muck," lacking in contrast. It seems exceedingly "Dragan-esque," with an unnatural, tone-mapped feel. It's actually interesting, but not the look into which HDR eventually developed. Obviously, they weren't striving to match the look of "Citizen Kane" (which is a phenomenal film, by the way), nor were they trying to apply "ideal" lighting. They were merely testing a new idea. Please give them a break.
  2. There is a bit of overlap in terminology, so It's confusing. By "triple exposure video," I think that OP intended to mean "HDR video with three different exposure levels." Historically, the terms "double exposure" and "multiple exposure" meant exposing a single piece of emulsion in-camera, two or more times, to combine different images into something like this: With digital imaging, the multiple exposure process is a little different, because each exposure is a separate image and, thus, a separate file. Some digital cameras offer various ways to combine files and to create multiple exposure images, such as the Canon 5D mkIII and the Olympus OMD cameras. Undoubtedly, most who create "multiple exposure" images today are combining the images in post/editing. Magic Lantern offers two HDR video methods: Dual ISO (which you mentioned) and HDR video. I don't remember which method appeared first. ML's Dual ISO is a technique in which every other row of pixels is given a different exposure/gain. So, for example, all "even" rows are given a darker exposure while all "odd" rows are given a brighter exposure. The separate exposures are "blended" to make a single HDR image. This method can induce aliasing/moire. Our own @ZEEK made a video tutorial on setting up a Dual ISO raw video on a Canon EOSM: Incidentally, "dual ISO" has morphed into a term for more recent cameras featuring a sensor that can be "set" to one of two native ISOs. On the other hand, ML's HDR video simply gives a different exposure/gain to every other frame. For instance, every "odd" frame is given a darker exposure while every "even" frame is given a brighter exposure. This technique doesn't suffer the aliasing/moire of ML's Dual ISO, but I seem to recall reports of motion artifacts. By the way, prior to ML's Dual ISO, Panavision touted their Dyanmax HDR sensor which was claimed to feature three different, nested pixel arrays, with each array having a different exposure (somewhat similar to ML Dual ISO). Panavision has remained mysterious in regards to why they abandoned the idea (and subsequently sold their sensor foundry). It was done in 2010 with two Canon 5D mkII's and a beam splitter (probably the mirror type): Here is another early method from 2011 with a single Canon 7D and a post process: For three cameras, one could possibly use a prism beam splitter. If one could put the beam splitter and three shallow-mount cameras in a light-tight enclosure, it might be possible to use a single "taking" lens.
  3. There might be ways that a cinematographer and editor could "interpret" a performance. And, perhaps, lighting could influence a performance, as could lenses, DOF (if the actor/performer could see those results on set). However, it is not the cinematographer's job to influence a performance -- that is the directors job. In regards to an editor or cinematographer "interpreting" a performance, both crew members exist to serve the director's vision. So, they can suggest ideas and execute their craft, but the director necessarily has the final word.
  4. tupp

    Scanning film

    Yes. That requires a telecine or film chain. If it's serious work, best to bring the footage to someone who has a decent set-up.
  5. tupp

    Scanning film

    A cheap slide copier attachment works for small batches: Some slide copiers have a negative holder. Here is an inexpensive 22 megapixel scanner that outputs jpegs. In addition, I have used a hi-res flatbed scanner for negatives, and it worked just fine.
  6. It would not be easy to do this with auto focus.
  7. Not sure if Canon has enough snap to do something that makes sense.
  8. Are you going to review the Metabones GFX Expander?
  9. A complete camera rehousing has already been done with a BM camera. Due to BM's "design aesthetic," there also have been several other mods to their cameras. Remember the Wooden camera BMPC lens mount mod? How about their current BMPCC6K lens mount mod kit? That would be great! By the way, a Chinese company has already added the hinged screen.
  10. What is the full name of the company again?... Blackmagic Design is a textbook example of a small, newer company focusing on form over function.
  11. Blackmagic isn't exactly a large company -- film at eleven... Not so sure about this. Casual shooters usually want stabilization, good autofocus, and a flippy screen (from the factory).
  12. Yep. It's ugly, but we got used to it. However, there is definitely an ID person's heavy hand involved in its form. It's doubtful that is the reason for its form. Most likely, a designer who doesn't work with cinema cameras was faced with the task of creating a housing that would accommodate a bigger "non-flippy" screen, a handle with "ergonmic" buttons and the main circuit board. It was easiest for the designer to just make it DSLR style with a fan. The CEO who is guilty of signing-off on all previous BM design blunders probably fawns over this designer's work, and, so, the BMPCC4K came to be it's current form. The question is: How many users hand hold the BMPCC4k without any stabilization? Incidentally, because of its shape, the camera doesn't work very well on many gimbals. By the way, @leslieyou quoted my question, but you never answered it: which camera would you prefer -- the original BMPCC6K or a streamlined blocky version that works on gimbals and that can accept a lot more lenses, adapters and speedboosters?
  13. Beat me to it! Here are some Series 9 holders.
  14. I think that I found the post, but I am not sure exactly what is going on. I went back a couple of pages to see what preceded the post, and I still can't figure it out.
  15. I don't recall that. If there is a link, please post it. Thanks!
  16. None of this is surprising. This is what you get when you have narrow minds at the top. Hope the employees are not suffering too much. It's obvious that BM's CEO has one of those typical personalities that adores "ID" -- which means that he has a pedestrian view of design and that his company puts stylishness/form over function. That attitude shows in their cameras, as almost every model has at least one major functional design blunder. We've seen this malady in other organizations, such as Apple. Stylishness can certainly sell products regardless of functionality (or regardless of the lack of functionality), but sometimes stylishness isn't enough (AJA Cion). Of course, the fact that BM's cameras are priced lower than their competition doesn't hurt their sales. There's no question that, performance-wise, BM cameras are one of the best values in the market. The BM CEO (and other manufacturers) need to realize that function should rule completely over stylishness in the professional camera market. Is the Alexa Mini stylish? ... how about the Sony F35? ... the Panavision Millennium? None of these cameras would be highlights at fashion shows, yet they all function well and can produce exquisite images. Let's say that BM offered a second version of the BMPCC6K with a non-stylish, blocky, but more streamlined form -- like the Kini-Mini. Both versions have the same capabilities, except that the second, blocky version also featured a shallow, interchangeable lens mount (with EF lenses performing just as well as on the original BMPCC6K). Both cameras have the same price. Now, which would you prefer: the original stylish BMPCC6K; or the second, blocky but more streamlined and versatile version? By the way, the BM CEO boasted that his company produced the BMCC because none of the camera companies would listen to their requests in camera features. Ironically, BM exhibits more hubris and "NIH" syndrome than any other camera company that I have encountered. They are dismissive and condescending regarding outside suggestions, and they don't take criticism very well. In regards to Canon, who would have thought they were steeped in corporate BS? /s Go to the Canon booth at a trade show and see how one of their sales people react to the mention of MagicLantern promoting their brand by unlocking amazing features in their cameras. I have heard nothing but hostility from Canon regarding ML.
  17. Isn't the IR cut filter on the GH5 stronger than the one on the P4K?
  18. No. It's not. He is just doing stitch/shift (he cutely calls it "twinspin"). Shots are captured as the sensor is shifted around the focal plane to cover the larger image circle of the lens -- the lens/camera is not panned. On the other hand, shooting a panorama with a normal camera involves panning the entire camera between shots -- there is no shifting of the sensor along the focal plane of the lens to capture more of the lens' image circle. I don't mean to be a downer, but there is nothing special about what this guy is doing in this test, and he makes most of the same fatal mistakes as other "format comparers" who came before him. What he calls "twinspin" has been around for awhile known by another name. He could have bought a more versatile shift/stitch bracket, instead of using that rig. The most popular line of shift/stitch brackets probably are the Vizelex/Rhino-Cam offerngs: By the way, our own @maxotics made his own shift/stitch bracket!:
  19. You might be able to adapt the Rawlite Ursa Mini OLPF for other APS-C cameras. There have been a few lens filter moire/aliasing solutions over the years. Here are the Caprock filters. Here is a filter set solution from Kristian Ponttopidan. Here is "variable" anti-aliasing filter from Kristian Ponttopidan.
  20. There is no theoretical limit on bit depth relative to dynamic range, as they are two independent properties. On the other hand, there certainly are practical limits (and ideals) as to how much bit depth is mapped to the amplitude range (bit depth is mapped to the range in amplitude -- not to the dynamic range). Applying bit depth numbers that approximate the dynamic range has become standard practice for feasibility reasons. However, there is no connection between bit depth and dynamic range -- they are two independent properties. Bit depth is simply the number of digital intervals mapped to a amplitude range. Dynamic range is essentially the portion of the amplitude range without noise. Amplitude range is the total range of signal level (including the noise). One can have a camera with a bit depth of 3, but with a capture dynamic range of 16 stops. Likewise, one can have a camera with a bit depth of 16, but with a capture dynamic range of 3 stops. Indeed, there are many cameras with multiple, selectable bit depths, yet the capture dynamic range remains unchanged in each of those bit depths. By the way, the lack of beer functions was very disappointing.
  21. The mythical "Sharpene" is improving! I wonder how this "AI" product compares to typical wavelet sharpening. The Topaz Labs JPEG to RAW AI seems to get rid of macro-blocking. Topaz Labs offers six products, some with apparently overlapping functions. Yes, resolution can be traded for bit depth while maintaining the same color depth, but those figures are not mathematically correct. Keep in mind that bit depth is not actually color depth. Digital color depth is a product of resolution and bit depth: digital color depth = resolution x bit depth. The color depth of an image cannot be increased, unless something artificial is introduced (which is evidently what some of the Topaz Lab products do). So, scaling down the resolution of an image to increase bit depth won't add any color depth for grading. On the other hand, if you merely scaled down the resolution without summing the binned pixels and then just increased the bit depth, you likely threw away a lot of the color depth. The pixels have to be binned/summed (or averaged) to retain the image's color depth, when trading resolution for bit depth.
  22. I would buy this camera just for the "beer functions!"
  23. Glad to hear that Olympus is alive and well and advancing M4/3. 3 stops. Yes. 25 > 50 > 100 > 200.
  24. I am not so sure that's the situation with my similar difficulty. Apple's HFS+ file system does some crazy things and is notoriously delicate. Don't throw away the drive. A Parted Magic torrent is free, and it takes about 15 minutes to download it, install it on a thumb drive and boot it on your Mac. If it sees your T5 drive, you're set.
×
×
  • Create New...