Originally Posted by John Schuermann
Sorry I've been away for a while - in pre-production on a film so have been distracted. Lots to get caught up on here and some stuff to clarify, so this will probably turn into a long post. FYI, some of my comments are cut and pasted from another post I did on a similar subject.
Much of what is discussed here has to do with resolution in terms of pixel counts and potential resolutions, but there are numerous hard facts of the movie-making and post-production processes we keep coming up against. As I've posted here before, one of the things we are discovering is that 35mm film really does not support much in the way of visible picture detail at scans much above 2K, especially if the original was shot Super 35mm (4K scans are likely overkill). Scope films shot anamorphically, on the other hand, do have a bit more visible (and therefore usable) resolution. I've sat in on comparisons between film sourced 4K scans and 2K down-rezzed versions, and the visual difference is almost zero. Unless the film in question was shot with a native 4K / 5K camera or sourced from 70mm, there seems to be little visible benefit to going to resolutions much higher than 1080P.
I also quite often hear from people who (quite understandably) would like to see their favorite blockbusters released in 4K. However, we now come up against another limitation in terms of how movies are typically assembled in post-production. In the case of almost every FX heavy movie over the last 10 years right up until today, the FX have been rendered at 2K. So even if you went back and re-scanned the film at 4K, you would have to totally recreate every CGI effect and composite from scratch at 4K resolution - something that would be incredibly expensive and time consuming. As others have pointed out in other threads, even those films that possess very little in obvious FX work are edited and graded at 2K resolution levels. As R Harkness pointed out in another thread, "...even the average romantic comedy can have numerous CGI fix-ups, composites for lighting alteration, adding or subtracting buildings, changing parts of locations, etc. It will perhaps be the rare movie which *doesn't* have any 2K rendered FX elements incorporated." The downside here is that any movie that was finished in 2K will need to be totally re-assembled in 4K with all new FX renders. All this means time and money.
What I'm really hoping for is to just get the Movies re-released on "4K" Blu-ray at 2K but with the higher bit depths and wider gamuts that will be available. Those are really my two biggest issues with Blu-ray today, banding and that I don't get the same gamut as I see at the theater. Heck I don't even really care if they upconvert them to 4K or sell 4K scans (with "2K" VFX), though I know the media/forums will cry out at that being a scam.
Here's where it gets even more complicated. At 4K / 5K resolutions we are literally getting to the limits of what not only the human eye can perceive on a display at reasonable seating distances, you are also getting to the limits of what can be resolved on the source medium. For example, if you move the camera (especially at 24 frames per second, the motion picture standard), motion blur becomes such an issue that fine detail gets destroyed. If your shot is even slightly out of focus, fine detail gets destroyed. If you aren't using camera lenses that resolve 4K / 5K resolution, those fine details won't even be captured. If you are shooting a landscape on a hazy day, if the camera shakes, if your shutter speed is too long, if the photosensors in the camera don't actually resolve 4K, the list of things that can destroy fine detail levels goes on and on.
This is why I'm a lot more excited about 4K displays than I am 4K content (outside of the bit depth/gamut/HDR/etc that are supposedly coming with "4K).
So, how does all of this impact next generation video and Blu-ray formats? So far, what has been ok'd and approved by the BDA still forces us into a situation where Scope content is letterboxed. AFAIK
, there have been no provisions set aside for native 21:9 / 5K video. We have demonstrated to the powers that be that there is a clear visual benefit to Folded Space technology in going from the 810P of letterboxed material to 1080P anamorphic. As things progress toward true 4K, that benefit in terms of content sourced from film elements
becomes a bit less impressive.
I kind of wondered if this would not be the case. With 4K you're already about at the limits of the source material without going "anamorphic", I kind of figured Folded Space would be a hard sell for 4K.
Our position is that image capture technology is certainly going to improve to the point where acquisition and display of extreme detail levels is going to become more and more practical. Technologies that support this are better quality camera lenses, improved digital image sensors, and higher frame rates. With all of this technology currently being developed and actually in the pipeline, why constrain Scope images between two black letterbox bars? Why not utilize this space to store extra resolution for the displays that can actually take advantage of it? The beauty of Folded Space is that it is simple to implement, does not require dual inventory, and creates a letterbox version as part of the process.
And here is where we start getting resistance, because it requires a post-production pipeline that supports 5K and anamorphic processing. Movies would need to be captured and post-produced at 5K (5120 x 2160) resolutions in order to take advantage of a process like Folded Space. With all the post suites currently sitting at 2K (2048 x 1080) and only gradually upgrading to 4K (4096 x 2160), an additional change to workflow is required - to 5K (5120 x 2160). Now, let's think about how that impacts other aspects (no pun intended) of film post-production work. This means:
- Digital composites need to be rendered at 5K
- Post-production color correction and editing processes need to be rendered at 5K resolution
- All digital FX have to be rendered at 5K resolution (even a recent CGI heavy film like GUARDIANS OF THE GALAXY had its FX rendered at 2K, as the cost and time involved to render FX at 5K massively increases the cost).
So, the long and short of it is that 5K post-production is not really feasible or cost-effective at this time. However, eventually it will be.
This is why we still feel Folded Space is a forward thinking proposition that has its place with next generation video.
I offer all of this here to give a complete picture (again, no pun intended) of the challenges of getting something like Folded Space accepted within both the movie studio and consumer electronics worlds. These are the kinds of objections that get throw out there, and ones we have to address.
At this point we are still working with various players to bring Folded Space to fruition. The challenge remains getting all the different players on the same page in terms of seeing the benefits. Some do - and are actively encouraging us - while others do not. The challenge as we move forward is to persuade the majority to see things our way
I do wish Folded Space all the best and I'd love to see something like that supported on next-gen Blu-ray (again, hopefully the industry will be open minded and not limit it's use to just 4K sources, but give us higher quality 2K sources), but I'm not sure I really see the benefit. Or, more accurately, I understand the uphill battle to sell it. Discussions like these always bring me back to this article with some good info from Dr. Soniera:
We can see about 100 pixels per degree. At SMPTE reference 3 picture heights, that's just a bit over 18 degrees, meaning we can see about 1800 lines of resolution. Now this is significantly more than 1080p and especially 810p, which jives with the findings you're reporting where Folded Space (1080p) shows a visible improvement over letterbox (810p) and where 4K (2160p) shows an improvement over HD (1080p). But letterbox 4K will still be over 1600 lines which is very close to the limit of what we can see. You have to be at more like 2 picture heights, where we can see about 2600 lines that there should be much visible difference between 1600 and 2160 lines. Granted technically 3 picture heights is already across that threshold but that's probably roughly the line.
Point is, as you say "anamorphic"/native scope mastering will require new/additional mastering steps/costs, and with the minscule market that will actually be able to appreciate the difference (those sitting less than three picture heights), I just see it being a hard sell.
Originally Posted by GetGray
They tried, it flopped.
That will tend to happen when you charge more for an inferior product. Those 2.35:1 machines cost more than better 16:9 machines with ISCO class anamorphic lenses.