Jump to content

Sean Cunningham

  • Posts

  • Joined

  • Last visited

Everything posted by Sean Cunningham

  1. Not totally insignificant too is that the BMCC is also writing 12bit log from a 16bit linear capture.  
  2. That's what you took away from it.   You're all set then.  Be merry.
  3.   It absolutely is the principle of the BBL.  The DI implementation is a means to achieve an approximation of a look that used to be done totally in-camera.  And while I don't know that they invented it, the pre-DI body of work from both Ridley and Tony Scott is filled with the the orange and blue palette that is practically synonymous with the BBL.    The basis of the DI BBL is warm light on talent, cold light in the background, which affects shadow color.           ...besides the whole complementary aspect, even going back to the Old Masters paintings, the eye moves from dark to light and, given a color scheme mixing cooler and warmer colors, from cool to warm.   The latest DI trend just gives it a hue-shift and does it in perhaps a more self-conscious way.
  4.   Hmmm, not really.  This is the palette of Blade Runner, though not always at the same time, in the same frame.  Scenes with characters were predominantly warm.  Effects and process shots of the rainy, urban vistas had mostly cool light.  Characters were almost always shot with warm light, sometimes backlit with cool light...creating the blockbuster palette.           This color palette can and is achieved even pre-DI by the thoughtful mixture of color temperature on your lights with respect to what stock you're using.  Warmed tungsten sources on your talent with tungsten film, using daylight balanced sources in the background for moon and fill and, voila, blockbuster color scheme without DI.   Bay and loads of commercial directors achieve it in daylight scenes by lighting talent with tungsten fixtures and shooting daylight balanced film...expose for a deep, rich blue sky while pumping up the warm light on the talent to compensate.
  5. Here's maybe the best YT-specific article I've seen:   http://www.streamingmedia.com/Articles/Editorial/Featured-Articles/Encoding-for-YouTube-How-to-Get-the-Best-Results-83876.aspx     ...it confirms (what I think common sense should already dictate) that uploading close to YT's target bitrate is a bad idea.  It goes further, from YouTube's perspective, identifying average and high quality upload settings.  My own, @ 25Mbit, are half what they consider high quality 1080P uploads.  Interesting.         It's important to think of the files you upload to YouTube as golden masters, as they will be used as source material to generate video streams for years to come. Simply put, the better the quality of the file you upload to YouTube today, the better quality the viewer's experience will be throughout your video's life on YouTube.   This contradicts a lot of what's been said elsewhere.  In YouTube's own guide for "TV/Film Partners" they're suggesting bitrates now that are twice what's stated above.  100,000kbps for 1080 and 81,000kbps for 720P with 30,000kbps for 480P.  Yowza.
  6.   Perhaps, but I've never found a good, complete document of all the settings.  Not on their site, not the wiki, they all seem to reference settings that aren't present in all implementations or fail to mention others entirely.  I think my current setup was based on taking bits and pieces of four or five "how to" tutorials out there, none of them complete, and some experimentation until I got the dreaded brightness changes to go away.   Dunno if you're Mac or not but I looked at where presets were saved and it's in here:   https://docs.google.com/file/d/0B1fj5b6y6BODM3FyV0EtcWU5YkE/edit?usp=sharing   ...copy that to your user Library/Preferences folder and it'll be accessible by switching the encoder to "x264Encoder", hitting "Options" and then "Load 1" down at the bottom of the dialog.  Then in the main interface I switch on both "Multipass" and "B-Frames", Quality slider to 100% and set my maximum target kbps to where I want it.     For 1080P content I'm encoding at 25000 kbps.  For 720P I'm going between 15000-20000 kbps.  Giving them something at or near their own maximum bitrate results in less quality than even they are ultimately capable.  If they would just give people the right information, so that they could encode to a real spec, then they could just push out new content without the burden of re-compression and the damage that it does.  Until then,  I won't upload anything that's not at least twice the bandwidth they're ultimately going to make it.
  7.   Interesting you should think so.   David Slade told me, via Twitter, that (at least on the episodes he's directed of Hannibal so far) he felt the need to sharpen Alexa footage because 1920x1080 looked soft to him.  For budget and time they're shooting in its ProRes 444 mode to avoid external recording and the costs (time+money) associated.   But I see what you did there ;)
  8.   What would be interesting would be a post-decode grain applied by your receiver.  For a brief period, I was convinced this was a feature on my early HD cable receiver.   Early HD broadcasters, starved of HD content, realized the power of grain, used well, which aided some to broadcast previously non-HD content without drawing attention to itself.  Low-rent channels from Turner were doing stupid stuff like taking SD episodes of the X-Files and doing a "fit to comp" so to speak, creating a fuzzy, stretched image.  The results had Scully looking pregnant through the entire run and morbidly obese during her pregnancy episodes.   Meanwhile, over at Universal-HD, they were running episodes of The Equalizer, from the 1980s, as well as other film-originated shows, uprezzed but sharpened and a subtle grain added as a final step.  I could tell the grain was added as a post process because it's wholly monochrome, appearing as light and dark dots in mostly the midtone range.  It gave the perception of increased resolution, a technique already common by this point, in the early-mid 2000s, mixing DV blow-up footage with either 35mm or HD.
  9.   No, they decreed that 16mm film grain and their encoding don't go that well together.
  10. When was the last time you saw a movie trailer, one shot on film, where the grain was a contributing factor to a bad looking stream viewed online?  Not just a bad stream, poorly encoded, but where you could tell it was because of the grain?   I've seen plenty that were just badly encoded, often uploaded by random people using, whatever, likely some preset.  But also consider digital TV.  Something like U-Verse is streaming HD content at 6-8Mbits which isn't much.  Still, The Walking Dead, shot on Super-16mm, doesn't suddenly look bad, or worse, than any other content seen before or after it.    Given they don't switch to some other technology besides H264 for the best looking content for each venue or source the real issue here is mastery of the encoding process.  As codecs get cleverer and cleverer I highly doubt they're moving towards a fire-and-forget existence.  
  11. My problem with that apology for him is he's making purely visual, short-form documentaries for the most part. If his work is all about the image then it should be all about the image. He's not doing investigative journalism or following activism or doing war correspondence. Granted, most of it isn't going to be viewed anywhere but online, so there is that. edit: imagine Baraka or Samsara if the filmmakers had just set out around the world with a compressed DSLR. Sure, there'd still be lovely moments captured but...he's free to do what he wants, it's just that now he's maneuvered himself into a position where he doesn't have to work with such engineered compromises.
  12. LOL, I didn't notice. I just couldn't believe the stuff was going for more than retail. I feel evil for saying so but somehow it made me picture "Dr" Phil selling old psychiatric books and paraphernalia.
  13. They broadcast material acquired in Super-16mm, guaranteed.  It's broadcast all over, in H.264 HD transmission.   I'm a fan of adding grain.  It helps with the failings of compressed origination, for one, but it also adds texture, which I like.  It doesn't have to be full-on 100% dialed up.  But only the most antiseptic of industrial would I ever deliver without grain anymore.   You're looking at sources with grain when you watch a BD or movie over digital cable/satellite.  You see sources with grain in H264 pretty much guaranteed, every day, if you're watching TV or streaming content on the internet.   Part of the problem is h264 isn't a codec with a single dial that the higher you go the more quality, the lower you go less quality.  Of its seemingly dozens of flags and settings it has a few directly related to handling grain, whether you want to squash it or preserve it.  Tuning H264 encoding seems to be a mixture of science and art.     The compression guidelines posted on YouTube and VIMEO by interns (I'm guessing) aren't going to get you there.  I've had to go digging far and wide for settings that I'm more or less happy with for now.   And I've come to the conclusion that digital origination, no matter the camera, needs grain for DCP.  
  14.   It was a story that hinged on time travel.  There is no need to say anymore.  If he really wanted to reboot the very concept of Star Trek with any hint of being different or better or any reason at all to exist in light of the previous offerings, most of them bad, the first order of business should have been no stories, ever, involving time travel.   The ST franchise has abused time travel so many times I'm instantly turned off by any new film exploring this concept.  I don't care.  I've come to associate it with lazy, bad writing.
  15. I was trying to stay done with this thread and this topic but...I just had to say...I love you man. I haven't heard or seen anyone use that phrase in years. I missed it. edit: PS, you wanna get really annoyed? I saw in a thread on a GH2 facebook group that PB is ebaying a bunch of old gear at the moment and the stuff is going above MSRP, because he touched it I guess.
  16.   Of course, there'd be no need for BMD camera users (or concerned citizens) to rain on these guy's 5D's parade if 5D users hadn't been repeatedly showing everyone their "O" face every five minutes, making absurd declarations and moving through the five phases of grief in repetition.  This place and everywhere else DSLR related has become a splash zone of premature emissions.
  17. No worries.  Thanks for even looking into it!  After seeing your demo it's going to be hard to look at the built in Exposure effect in AE and not be thinking to myself, "you suck, Exposure."
  18. Given that it works in Premiere, have you tried making it work in AfterEffects?  I liked your demo video a lot but I don't do any rendering or finishing in Premiere.  I would use the hell out of it if it was available for AfterEffects, I expect.   :)
  19. It is confusing, even looking through the ML thread. Near as I can guess it was the earliest versions, perhaps, that were successfully writing using the YUV422 and raw wasn't working but at some point perhaps they switched to writing true raw. Somewhere in that 30+page thread there might be the "eureka!" moment. I didn't find it. Confirmed though, what they're grabbing from, at least, is 14bit RGGB. Stu didn't like the early mix of terminology either and wasn't going to stand for it so he got a direct response: prolost (http://prolost.blogspot.com/2005/05/log-is-new-lin.html) Also something to keep in mind, that's been distorted some through some overzealous comparisons. This is 14bit linear color. The BMCC records 12bit logarithmic color. It takes 14 linear bits to equal 12 logarithmic bits so they're essentially recording the same precision from a math standpoint (with theories over logarithmic sample steps being a better representation of some aspects of photography while linear steps offer advantages in others).
  20. If you have a crew and you're shooting in 40sec or 49sec or even 1min bursts I feel really sorry for your crew and your talent and everyone else associated with being in that goat rodeo. It's a waste and you should be ashamed of yourself for wasting all of those people's time and effort. And that's the last thing I'll say about such foolishness.
  21. Really? You think under these lighting conditions DOF is a contributing factor, assuming no heavy ND was employed. They shot at ISO160 and I'm betting at or near f/22 but, sure, the two cameras have different depth of field. I'll even go so far as to agree they might have completely screwed up shooting a broad daylight landscape focused to infinity. The files are there for you to inspect as well. LOL..."butchered".
  • Create New...