Indie Film: Does a 4K finish pay off? (looking for examples)

[Elsie N]

And as a quick followup, I guess this is why RED did take a serious look at an alternative for cinematic distribution with REDray and .RED codec.
More or less:

Imagine what we could have done without the bandwith and compression restrictions of DCI.

Unfortunately they seem to have left that ecosystem, even though I think it would have been a good thing just to display what 4k projection actually could have been like.

That is RED's greatest failure IMO.

 

[David Battistella]

The revolution is over with cameras.

Distribution is the next frontier.

 

[Gunleik Groven]

The revolution is over with cameras.

Distribution is the next frontier.

While I agree that distribution is now a huge bottleneck for image quality, I also think we'll see some major evolution on the camera side the next years, until we reach some sort of common container, like we did in audio. That ended up as 32-bit float, and everyone and her grandmother can now record that on a home PC. THAT doesn't make all recording devices equal, just that we have ended up with a "big enough acquisition container". For digital to fully replace all varieties of filmstock combinations, I guess we'll end up in some similar domain:
Medium format 4:3 sensors and 8k+ acquisition for the high end and 32bit linear recordings. Not sure if the bayer pattern i sthe end to all image acquisition either. But it's the best for now.

People will be able to upgrade their computers for a while.

Now, does that imply that I think todays cameras aren't excellent? Nope. Not at all.
But it means that from a purely tech perspective, I think you should get the best acquisition you can get, even for an HD finish.

And if you go 4k for big screen now, make sure you shoot your movie in a way that the compression doesn't kill on the delivery-side.
Do tests!!!!!!

The fun thing with jpg 2k, is that it washes out high-frequency detail and color fidelity at higher compression rates, exactly what you'd hope to see more of in a highrez projection.

AND greens aren't good at all from DCP. Watched stuff from different cameras and 4k+ filmscans, and the grenes are noticeably less than vibrant in what we get to see. Even from a filmscan.

There are a lot more to the equation than 4k+ acquisition, but I agree, it all starts with a fat negative.

 

[Terry VerHaar]

Let's stay on the very narrow topic that Bruce has identified - examples only. No discussion allowed. We'd hate to have to have him scold the mods! lol

 

[Elsie N]

What's wrong with discussion on Reduser?

Kind of my thinking too. I've always viewed thread discussions as exponential rather than linear, although I try to wait until the threads original intent has been addressed sufficiently before going off on a tangent.

 

[Terry VerHaar]

What's wrong with discussion on Reduser?

My sentiments exactly!

 

[Gunleik Groven]

As far as I can see, discussions are only good. But I think it's not as easy as "4k or not" or whether "4k is harder or not". The matrix is a bit wider, also on the technical side.

And I think that my examples are relevant, even though they come with a bit of contextual and technical explanation.

 

[Milan Spasic]

Question should be rather: "does a 2k finish pay off?". What happened to all those indy features finished in SD?

If you are shooting 4k there's absolutely no reason not to finish in 4k unless you are making an epic alien interplanetary battle action film as your "indy" feature.

Instead of dubiously advising indy filmmakers not to finish in 4k, Alex should have advised them to come up with stories that don't require mega VFX.


Just like Bruce argued for putting story before camera resolution, he should be arguing for putting story before VFX.
Though neither really fit this discussion because good story and 4k finish are not mutually exclusive.

 

[Bruce Allen]

Anyone have any more concrete examples or details?

The Netflix feature film examples Peter Lyons Collister gave are very interesting. I'd be interested in how the process goes.

It sounds like it's for their in-house stuff? EG as part of them making the production happen, they mandate 4K?

Of course I spent the morning researching Beasts of No Nation and trying to Google about their 4K finish so that I could add one more to my list (we're only at 3 with concrete financial reward so far).

...then I found out it was Alexa XT with Panavision anamorphics. Oh.

So I guess they finished 2880 x 2160 anamorphic then blew up to 4K?

I thought everyone said Netflix doesn't do that, wants a true 4K finish, etc? But then I found out Beasts was bought by Netflix later, so I guess that fits that model?

EG you can sell something to Netflix that's not 4K and they will still bill it as "A Netflix Original Film". Just if you want them to help you with actual production, they want 4K?

Same as Universal or another studio picking up an indie movie and distributing it?

Bruce Allen
www.twitter.com/bruce_allen

 

[Michael Tiemann]

[...] So I guess they finished 2880 x 2160 anamorphic then blew up to 4K?

I thought everyone said Netflix doesn't do that, wants a true 4K finish, blah blah blah?! [...]

This logical inconsistency was discussed starting here. I have seen no definitive answer for how Netflix qualifies anamorphic-origin content. If I had a vote, I would vote for counting pixels after horizontal de-squeeze, but before any vertical up-scaling.

 

[Craig Parkes]

Another take might be the example of Tangerine - where the press from shooting the phone on an iPhone probably pushed the films sales further up than had the same project been shot on Epic.

Ultimately, for narrative features it's such a crowded market place that differentiating your project through UHD capture and mastering is really a long term investment on an item that has very little likelihood of either long term or short term returns.

In my case 4K mastering was hedging against the downside risk of low budget independent production - because it didn't cost the production significantly more it would have been stupid not to do it and pursue immediate sales because you need every deal possible to try and recoup your independent production. In the long run if it leads to even more sales (the film is a couple of years old at this point, so the window in which most indies achieve distribution is pretty much closing, but if there is a lot of demand for 4K content from a few new countries then being a title that has already sold to UHD VOD platforms offers some library potential).

If I was advising someone looking to invest in a slate of independent films at the moment - especially lower budget films that were potentially broadcast/VOD network friendly - I would advise them to all master to 4K unless VFX costs were prohibitive or a DOP with some name value was attached who could bring in greater sales but wanted to use a 2K/3K camera system.

But on a film by film basis - there are so many things that are levers to success ahead of mastering at 4K that if you are a producer or director and actually have to think at the start of production 'how can I finish this project in 4k?' then the energy spent on that question is likely wasted compared to thinking 'how can I attach that actor' or 'how can I attract the right sales agent' or or 'how can I improve the script' or 'how can I get more investment on board'.

(Unless you are a specialist/highly visual documentary - if you are a Planet Earth style documentary capturing unique imagery where the difference in detail is going to make the narrative more immersive, the sales difference for a UHD title AND the added library value of having that irreplaceable unique footage at UHD makes a UHD finish a no brainer.)

 

[Craig Parkes]

EG you can sell something to Netflix that's not 4K and they will still bill it as "A Netflix Original Film". Just if you want them to help you with actual production, they want 4K?

Same as Universal or another studio picking up an indie movie and distributing it?

Bruce Allen
www.twitter.com/bruce_allen

Beasts of No Nation was independently financed and acquired by Netflix. It wasn't commissioned by Netflix (The only Netflix co-financed and co produced features so far I think are the Adam Sandler Western Comedy Ridiculous 6 and the Crouching Tiger prequel/sequel - not sure if these have been done at 4K or not. IMDB reports Ridiculous 6 as digital, anamorphic with cameras supplied by Panavision - so that could be any number of actual camera bodies)

 

[Björn Benckert]

Here is a good example of 4k indy film... Won the Gold Lion on possibly the worlds most prestigius Film Festival, the Venice Film Festival 2014.

It´s all 4k and I can only say that it is quite a remarkable film and the visual language it uses, hardly any camera motion, clips long as 10 seconds, most on a closed down UP lens then that really put the picture quality to a test. And at such usage 4k is making a huge difference. I was working on some of the vfx scene and saw them both in 4k and 2k and simply put it´s better in 4k no doubt.

I think the fidelity of the imagery is very much one of the reason it was picked as a winner, the 4k / RED one MX was ofcourse not the soul reson, lightning, makeup and scenography as of course also playing a huge part, not to mention the acting and story etc. But still it´s the one film that I think pulls the 4k straw in the very best way possible.


http://www.royandersson.com/pigeon/

 

[Hrvoje Simic]

While I agree that distribution is now a huge bottleneck for image quality, I also think we'll see some major evolution on the camera side the next years, until we reach some sort of common container, like we did in audio. That ended up as 32-bit float, and everyone and her grandmother can now record that on a home PC. THAT doesn't make all recording devices equal, just that we have ended up with a "big enough acquisition container". For digital to fully replace all varieties of filmstock combinations, I guess we'll end up in some similar domain:
Medium format 4:3 sensors and 8k+ acquisition for the high end and 32bit linear recordings. Not sure if the bayer pattern i sthe end to all image acquisition either. But it's the best for now.

32 bit approach is useful only for reducing "last significant bit" type of computing errors. This is processing advantage and very specific post production work advantage. It is not an acquisition thing or media storage thing.

You cannot record 32 bit sound. Practical gain with theoretical 32 bit audio recording would be highly questionable. 24 bits already allow capturing more DR than you can hear and recording equipment can handle.

Regarding video technology, 16 bit container is already more than the current cameras can fill up without sampling noise and it already has enormous depth many are not aware of, with 280 trillion potential color tonalities. That's 280 thousand billion. Or 280 million millions. Or 280 thousand times of 10 bit recording's color tonalities. Seems enough.

If the camera signal was useful enough with full 16 bits you would theoretically be able to pull up the image buried in a small portion of the histogram without any perceivable loss in quality. With 65536 gray tones this is 64 times of 10 bit recording. 24 bit color sampling would be 256 times 280 trillion (65K grayscale) tonalities. 75 Quadrilion. Even if the sensor could capture that tonal range and A/D could supply the data, what would be the practical gain for required R&D and in the end processing power and storage ?

After reaching ultramegagodzillions, one might eventually consider diminishing returns and pull the line somewhere.
With 16 bits for imagery I'd say it's enough.


I don't think Bayer will stay.

 

[Gunleik Groven]

I agree (and know) about the float advantage. I also agree (and know) that current cameras don't fully saturate the 16 bits, and maybe 24 but linear vs 32bit half-float post is where we're going.
But I stand by that I don't think we're "there" yet. The whole chain isn't really mature to say "hey' this is it as to signal".

And I stand by that for cinematic release, the bandwith/compression ratio are the main hurdles to give it significant traction.
And also that even for 2k the current DCI specs are not really optimal.

I do think that a lot of things will develop in the delivery part of things the next few years though, and when that's done, it's probably regretable for some productions to not be able to have a 4k pre-compression master (or they'll need a system like the one I am working on to more easily re-conform all material to 4k, including VFX).

 

[Sergio Perez]

i actually directed a short film, exhibited at the Macau Arts Festival, shot on Dragon at 6K and Mastered and graded in 4K. A bit of a challenge to pull off, but I think it came out well for a low budget short:

https://www.youtube.com/watch?v=wil5KKeXmi8

You can view it in 4K. The grain and tone was intentional to fit the mood and shooting style of the film.

I think that if its not heavy on FX shots, it does make sense going 4K. This short film is about a unique culture and tradition of an endangered cultural minority. It makes sense future generations get it recorded and available in the best quality possible.

I am currently offline editing and browsing on a 5K screen. Possibly in the next 2 or 3 years this will be the norm. Doesn't it make sense the content being ready for now as well as tomorrow? I think it also depends on how you view the content you do, specially films. Is it a Macdonalds Happy Meal film, or a Nice Steak, or Chef's Special?

 

[Gavin Greenwalt]

The more urgent question to me is whether the bottom has fallen out of the film market entirely. I'm at the American Film Market right now, and it is a simply terrible market, just terrible. Attendance seems super low. Rooms are empty - a lot of the older, established players seem to have moved to cheaper suites. It ain't pretty. Quite frankly, I'd be questioning the wisdom of making any narrative indie film whatsoever, 4K, 2K, or any K in between, in the current climate. People want to binge watch TV series, and they watch web docs - viewing habits are changing altogether. The narrative film markets are in a huge deflationary spiral. Everyone is trying to figure it out. This is the real elephant in the room. Not resolution.

I'm going to take the exact opposite view, this is the best news in years. Netflix needs content. Netflix is committed to 4k. Perhaps long form theatrical content is in trouble outside of the studio system, but the studios aren't producing short form content. People want to sit down and watch a 10 minute narrative film. See Jon Stewart leaving television for HBO short-form content. I've for a long time felt that feature length projects are a bad fit for indies. Time is money and shooting a 90 minute feature is extremely expensive, it's difficult to get crew for a full feature length shoot, it's expensive to feed a crew for a full feature shoot, it's a logistical nightmare to have that long of a shoot schedule. 10-15 minute narrative shorts are perfect for web delivery and their production and post production demands are much easier to accommodate for an indie. If you're on Netflix great. If you're on YouTube, that means you are short enough to bookend ads without interrupting the film.

Another thing to keep in mind is that what matters is less screen size and more Field of View. The Oculus Rift is a 5.7" IMAX screen. When I hold up my tablet on the couch it's often being held at about the same field of view as my 55" tv from the couch. Most people hold their phones less than a foot from their face when watching videos. That's making for a big screen experience. And of every display technology available today Oculus most urgently demands 4k content.

 

[Craig Parkes]

Gavin - you are right that this is great for 4K shooters/content creators but the thrust of this thread is towards the independent narrative film market - so I'd echo the thought that a bad AFM is a bad thing for independent feature film makers.

I think short form content will always struggle to be as immersive as long form content - it's one of the reasons people Binge watch shows, they want to be immersed in a story long enough to get the full story. But at the same time unlike a movie you can generally be distracted from a show, not watching but just listening to the dialogue, and even if you miss an important part of that episodes arc you will still get enough of the series arc to keep watching. Binge television watching is distraction friendly long form content - films tend not to be. Youtube is distraction friendly short form content, it's great snack food and it's free, but it'll be interesting to see how their new subscription model goes to see if people are willing to pay for content developed by people coming from that ethos. (Ignoring pirated material on Youtube, which probably makes up the vast majority of it's long form narrative viewing).

However people's lives are often too busy to spend much time in immersive other environments. Even gamers totally immersed in a game also have chat going, and get texts on their phone and even though they might spend 2-3 hours doing a raid in World of Warcraft or something most of the time their attention will be split somewhat.

That's one of the hurdles I think VR will really face once it is consumer ready in a big way - finding the right duration of experience that people can be away from other distractions will be key, or integrating those distractions in an easy to digest way with the VR content.

From a purely technical perspective I agree with you - a move away to different screen types than TV's and Theatre screens will hasten the adoption of 4K workflows and increase value for practitioners delivering them.

But if you want to make narrative films - a collapsing market is never good news.

 

[Al Lougher]

I'm going to take the exact opposite view, this is the best news in years. Netflix needs content. Netflix is committed to 4k. Perhaps long form theatrical content is in trouble outside of the studio system, but the studios aren't producing short form content. People want to sit down and watch a 10 minute narrative film. See Jon Stewart leaving television for HBO short-form content. I've for a long time felt that feature length projects are a bad fit for indies. Time is money and shooting a 90 minute feature is extremely expensive, it's difficult to get crew for a full feature length shoot, it's expensive to feed a crew for a full feature shoot, it's a logistical nightmare to have that long of a shoot schedule. 10-15 minute narrative shorts are perfect for web delivery and their production and post production demands are much easier to accommodate for an indie. If you're on Netflix great. If you're on YouTube, that means you are short enough to bookend ads without interrupting the film.

All of my shorts were mastered in 4K. If the Apple TV was 4K enabled perhaps that may have opened up 4K delivery for short form content. What with the Apple TV development platform I'm sure we would see plenty of new apps popping up where you could play 4K short films at your heart's content. Heck, I'd even try writing one myself since I have a background in programming.

However, how I see it the average Joe isn't interested in watching shorts like us filmmakers are. The next best thing is a 4K web series, but as I have found myself first hand these are difficult to finance, at least to make something half decent anyway. There's no current return on short form content. No-one will pay to watch your short, web series maybe, feature yes.

 

[Jeremy Torrie]

Maryem makes a great point about AFM and where the long form narrative is presently at, and so does Craig. The attention spans of people these days is increasingly small -I worry about whether my kids will ever be able to complete a task in these days with Netflix, iPhones, iPads, and all the rest...thankfully we make sure they get access to these devices as privileges after completing chores, but other parents are not so strict.

Short form series are where it's at, especially when you can claim fees and/or rentals over several weeks compared to 4 weeks on a feature.

There are a few of us crazies left who will continue to aspire to produce features though...