Jump to content

William Loekken

Basic Member
  • Posts

    5
  • Joined

  • Last visited

Profile Information

  • Occupation
    Colorist
  • Location
    Denmark

Recent Profile Visitors

The recent visitors block is disabled and is not being shown to other users.

  1. This is some good feedback David, I am by no means making a claim that no one is trying hard enough - I don't know much about what post houses and studios are doing to be frank. I think their work is incredible, and agree that many films intercut film and digital seamlessly. But no, I haven't seen a film shot entirely digital that I felt could have been film - I also don't know if it's even possible to do, nor if it's desired by enough filmmakers for it to be relevant. But I'm hoping to learn more about that here. Wouldn't any attempt at color science would be 'another film look' due to the fundamental differences between the capture formats? Personally I would find the look of films finished photochemically (e.g. as an IP), digitised and matched to the print a great look. I do believe it could be paired with a modern workflow - keeping the color and tonal reproduction for the overall look as a startingpoint while enabling the tools available in the DI. Maybe it's already being done - but wouldn't it be cool if it was available to those who don't have major post houses as an option as well?
  2. Hi guys,This post is about recreating the look of photochemically finished film and the options for achieving this look for filmmakers, as well as future ideas. I'm interested in hearing your opinions on the topic. My background in color scienceFor the last almost 10 years, while working as a DP and colorist, I've been working on creating color science for emulating different film looks for different cameras. I've conducted countless test shoots from small setups with a Blackmagic and 135 film to fully controlled studio setups comparing Alexa and 35mm motion picture fim.I attempted building a system to emulate several types of photochemical film looks and hired several image programmers to write algorithms to perform novel image processing tasks not available with current software. (I stopped for financial reasons)A while back I posted an article on color science which led me to teach color science at the National Danish Film School and a mailbox full of inquiries.And yet after all this, I am very far from developing a system that is complete. Developing color science is very complex and very expensive.The ChallengeIn my eyes Steve Yedlin is an outstanding artist, and his statement is inspiring - that we should be authors instead of shoppers and create our own look with or without the help of a color scientist / post lab - the idea sounds beautiful, but the more I work with the subject the more I cannot help but come to the conclusion that the vast majority of individuals do not have the money, time or programming knowledge to create color science from scratch that can live up to the quality standards achieved by Kodak. Which leaves us back where we started, as shoppers - with the services of a major post production house as a new addition to the shopping chart.Besides this, I see a few other challenges for those looking to get the photochemical look in their work Those who cannot afford to shoot film on a project probably cannot afford to hire a qualified post house either The post house may not have the color science to do an accurate transform in all of its aspects It may not be possible to work with the desired post house on a particular project, due to availability, deadlines, physical location etc. There are no available options for indie filmmakers providing accurate results Available OptionsTo my knowledge there are currently no available options for fully recreating the look of a photochemical pipeline.I have not yet seen a film shot digitally that I felt had the look of a film shot and printed to film.As far as I know the best options are some of the leading post houses who work with inhouse color scientists.My own color science may be decent, but I do not have the capacity to handle more than one client at a time, and I only have color science for one type of film look - 5219 scanned on a Scanity to log. It's an approximation of a film DI and not a photochemical finish. There may be other colorists like myself who have more or perhaps better options.Then there are the plugins like FilmConvert, VisionColor, Dehancer. There are LUT packs and film overlays.These are great tools for some people I am sure, but they do not provide accurate film transforms or a controlled pipeline.The options are few, and yet as far as I'm aware the demand is high.DemandI don't have any numbers, but from my limited experience and research it seems to me that the majority of filmmakers want to be able to recreate the look of photochemical film. Most people I work with bring me references of films finished photochemically, I also get a bunch of references from films finished in a DI (of which many, to my knowledge, have had print-film-emulation from major post houses applied as part of the pipeline), and less often references of films acquired digitally but processed by a skilled posthouse with a certain type of film look.I've tried doing some field research and contacted some of the leading DPs working today, and even academy award contenders have told me that given the option, they would use color science that accurately emulated a photochemical process.And almost everyone I know personally in the business are interested in either working with film or processing their digitally acquired footage to look as if it was shot on film.IdeasI believe film has a richness and a magic that brings stories to life and while I don't think film can be replaced by digital - I do think it's possible to create software that recreates the film look with enough accuracy to bring out a similar cinematic feeling and satisfy the creative goal, and that it's possible to make this film look available to all filmmakers, not just the elite.There is also much potential for digital post processing to improve images, and we are not stuck with only 3D LUTs and film overlays to do the work for us. Algorithms and inventive programming can help us a lot.Here's a simple example many readers will already be familiar with: X digital camera records 10 stops into the shadows and 3 stops into the highlights before clipping, but has a low noise profile. We can move middle gray so that we have 6 stops of shadow DR, and 7 stops of highlight DR at the cost of some video noise. This noise can be removed with a digital noise reduction algorithm. This reduces accuracy in the captured details in the shadow area but gives us more highlight latitude. The spacial algorithm makes it possible for us to get a more filmic result with the camera. There are many opportunities for novel algorithms to perform tasks by analysing the image. But we don't necessarily need to go that far. A 3D LUT created from enough color samples to cover the color locus at all relevant exposures (e.g. -9 through +9) with interpolation, implemented correctly could create results far better than many of the available options today if paired with a quality texture and halation algorithm. My questions to you Are you interested in recreating a film look when acquiring a project digitally? Are you currently doing anything to emulate film when shooting digital? If so what's your process? Would you be interested in using a software that could accurate emulate a range of film looks for all relevant cameras? Would anyone be interested in participating in the development of such as software?
  3. Thanks guys That's interesting, David ! You wouldn't know of anywhere I can find examples of what such exposure ramps or charts look like would you?
  4. I'm about to shoot a project vertically for social media, we're shooting s16mm - on an Aaton XTR Prod Does anyone have any experience with this? I'm worried the camera is not build for this and it may cause the film to run incorrectly through the mag because of the 90° angle Seems nothing's mentioned in the manual about it which makes sense, I guess, as the camera was created before vertical screens
  5. Do you know of any diagrams or text on the tonal reproduction of a traditional film pipeline? I'm trying to figure out how camera stops translate onto the screen in the final projection of a 'by the book process'. E.g. 5207 shot at normal exposure, processed normally, contact printed to 2383 and projected on a standard cinema screen. Kodak publish how camera stops translate onto the negative - I'm looking for something similar just for the final image If there is no diagram out there already. Is there a way to calculate this? Or another way to previsualize how camera stops will fall in exposure once projected? - If the deepest blacks occur at -5 stops or at -2.5 - does middle gray fall at the equivalent to the-digital-512-in-10bit-viewing in the projected image? and how bright is +4 stops compared to +8 I found the Jones Diagram through the ASC article 'The Color Space Conundrum' - which has a diagram of the print-through curve. - How does this print-through curve compare to what's seen on screen? I tried mapping the -8 through +8 onto this print-through curve and the result looks quite flat with very soft highlights and seems to look different from the references I have and doesn't add up with the ASC manuals comment on blacks falling somewhere between -2/-3 and whites go white and lose definition around +4/+5 I hope someone can help with some answers! or even better examples
×
×
  • Create New...