Hello all,

Here is my submission for the Reduser Portrat Competition. It is a complete honor to be selected as a finalist. There were some really gorgeous shorts on display and seeing it all on 4k monitors all the way up to 12 feet was spectacular. Had to hold it back a bit, but this meant a whole lot to me.



Although still web compressed, I uploaded a 4k version to YouTube which can be found here:
http://www.youtube.com/watch?v=azNkijQMvoA

In the spirit of what I've been doing with my Scarlet-X "Skully" thread on Reduser.net, I thought I'd share the workflow and sort of the marathon-like week long journey that took place to make this short happen.

Postmortem, Project Details, and Post Workflow

Project Details:

Camera: Red Scarlet X #316, "Skully"
Format: 4k, 23.98, RedCode 6:1, 1/60th, ISO 320-3200
Runtime: 3:59:12 - 5,748 frames

Lenses Used:

Canon EF L Series Primes:
24mm f/1.4L, 35mm f/1.4L, 50mm f/1.2L, 85mm f/1.2L, 135mm f/2L
Canon EF Primes:
20mm f/2.8, 100mm Macro
Canon EF L Series Zooms:
24-70mm f/2.8L, 70-200mm f/2.8L IS
Leica R:
60mm f/2.8 Elmarit-R
Sigma EF:
14mm f/2.8, 300mm f/2.8

If you click on any image it will enlarge.

A Bumpy Start
I'll save the gruesome details, but my initial concept for the portrait was cancelled and I almost didn't shoot anything. While in the midst of post work on a commercial I suddenly found myself dealing with a catastrophic computer hardware failure. Luckily I back up regularly and didn't lose a thing. It took a bit of time rebuilding and reinstalling everything on my main color and edit system. I found myself about 10 days before the deadline wanting to participate yet due to that time crunch my original plan was out the window. After thinking it over and a few Redusers motivating me to work towards submitting an entry I decided to buckle down and get focused on what exactly I could pull off within the timeframe. The short above represents exactly 7 days of work. 24 hours of pre-production, 3 production days of shooting, and 3 post production days.


Pre-Production
I spent a day doodling, writing down ideas, buying some wardrobe, and building props. Started with a brisk brainstorming session which led to about 20 ideas. Then I went through a review process eliminating the impossible and not so good paths and found something interesting. After deciding which direction to go I started fleshing out the idea. Mathematically speaking, knowing the short couldn't be longer than 4 minutes; I made a delibrate decision about working with a 6, 3, and 1.5 second cut mindset based around a 160bpm music track (that would need to be scored later). I figured that I had about 50-70 cuts to get my idea across with approximate shot "bleed". That's not a lot of time to tell a deep story, but you can get a lot done. I kept this in mind as I started storyboarding and jotting down my notes.



I had to make some quick decisions about casting and locations. I have a list of actors I've worked with before and in this case it was important to have some familiarity with the talent as speed was going to be of the essence during the shooting days. I narrowed it down to three actresses. Sarah had the facial expressions and acting chops that I was looking for the role and I hired her for the 3 day shoot.

Similarly I wanted to shoot at locations that I was a bit more familiar with and I wanted to choose a few that were important to me in some way. I'm an explorer of sorts and I had this abandoned farm in mind that I found one day while roaming through the desert. I've done a fashion shoot there about a year ago, but I've always wanted to put it in motion. For the "woods flash-to" I had a great forest in mind. However, the entire fantasy world gag was dependant on shooting at El Matador beach which is about 150 miles away. Lastly I wanted to choose something interesting for the hanger. I have been instructing on Sundays at the Gnomon School of Visual Effects (which happens to sit on a movie studio lot) and thought the hanger there would be a kick for students to see. Oddly, that's literally up the street from Red Studios Hollywood.



After a brief run to my gear locker, an art supply store, and a thrift store for props and more wardrobe I decided to do a few camera tests for the magnifying lamp shots. I knew a few things, but was looking for some more information on reflections and how light would play on the glass. Because I was using fluorescent lighting I decided to shoot the entire project at 1/60th. As I occasionally have done before, I used my ugly mug for the reference.



After that test I grabbed some dirt from the backyard and used about 12 ounces of super glue (ouch my eyes!) to make sure that my hero prop of the "little world" wouldn't get destroyed in transit. I keep rock jars with rocks from every place I travel and I thought it would be appropriate to actually use my newer one for the piece. I went over the wardrobe and liked the black, color, black, color, black scheme I had with these garments. The original shirt I had in mind didn't look great however, and I had some wardrobe from the feature film "The Postman" laying around that I acquired ages ago. That top was aged "post-apocalyptic-y" and used in the film and had some nice thread detail I liked.


Fleshing Out the Concept
Knowing that this was to be shown at 4k to showcase Red cameras and post production workflow I decided on a few things early. The first thing was that I wanted to shoot at locations showcasing a lot of detail and hopefully incorporate that detail mindset into the props and wardrobe. Second, part of the challenge here was showing the viability of the 4k workflow. I've been working with Red material using Adobe Production Premium for a long time now and that personally wasn't the big challenge here. However, I wanted to up the stakes (go big or go home) and I made the decision to somehow incorporate what isn't so common in such a small deadline; 3D and 2D 4k VFX work. Third, I wanted to show the flexibility of the Scarlet X by shooting in both low light and high contrast/bright setups, hopefully to disprove some of the myths about what Red cameras can and can't do. Lastly, I wanted to see if I could pull this off by myself. Not for egotistical purposes, but rather to see if I could pull this off within the deadline and to see exactly how much work I could get done in a week. I am one of those nuts that likes the pressure of deadlines.


Lighting Design and Concepts
The exteriors themselves didn't have too much lighting design as they were shot in daylight. I did get lucky and unlucky with clouds that provided some nice diffused light here and there, but I had to balance that variation later in the color correction stage. A goal under those conditions was to show some rich shadow detail and the impressive dynamic range of the Scarlet X without the aid of HDRx. I made the decision to shoot the close ups and the very wide shots later in the day to get a bit of the magic hour effect.

The interior of the hanger is a different story. I did light the heck out of that. However, I lit it fairly dark and shot most of that at f/2-f/4. I wanted to incorporate the color red as this was a Red specific contest. You'll see it in the reflections, glints in the eyes, fill light, and lens flares. There's a lovely moment when you first strike on the florescent magnifying light that you briefly see the color red, which you can see above in the lamp tests. Hopefully that wasn't too "in your face", but it was something I thought would appropriate for this project. Here's a diagram of what the lighting was in the hanger:



That 55 watt tube on the ground was gelled and placed to show the last second of "daylight" as the hanger door closed behind her.


Production Shooting Days
One thing that bit me hard on the first day of shooting was the cold air and wind. While it did look nice to have the dead trees come alive with movement, it did effect my shots and shooting speed. The first shot of the day and the first shot in the short was low and at 300mm, and I knew then the wind would be an issue all day. I attempted to bag and flag the wind as much as possible, but because I was shooting this by myself with the amount of setups and lens changes ahead of me I just simply didn't have time to do this for every setup. That wiggle at the peak of the establishing shot with the jib looking at the farm in particular could have been stabilized, but I had bigger fish to deal with in post. Sadly the wind blew down the garage where I was going to have the pebble pick up insert shot, but I used it for the background of that sequence instead.



After a marathon first day everything else went rather smoothly. On the hanger day I came in two hours early and did my pre-light and built my little set. Talent arrived and we got to work. The last day was more about getting fairly precise shots and darting to the ocean to grab the giant boulder insert. Lot's of driving that day. For the slider moves I used my trusty Kessler Cineslider with the 500 motor. This allowed for the sliding move, pan/tilt move, and focus pull to be operated by one guy. However, on the first shots I did actually use a small dolly I built years ago for another production. I used PVC pipe and flattened the dirt under the track. Had to toss in a little DIY problem solving in there.



Under a crazy deadline and trying to get a handsome amount of setups done per day it was important to have fun where there was fun to have.


Post Production
I did my own data wrangling and kept things very organized. I imported my .R3Ds straight into Adobe Premiere Pro CS5.5 and made the selects extremely quickly. After that I created some dummy 12, 6, 3, and 1.5 second markers to assist with my edit and began cutting. I needed to keep that time base for the music scoring process later. The real challenge was editing and telling the entire story within 4 minutes at a pace I desired. Only 4 shots didn't make the final cut.



Color Workflow
I wanted to have fun with the color correction and grading process. I was a digital colorist on features for a few years and one of the reasons I chose to buy a Red was just how flexible RedCode Raw is. My first light color correction was performed in RedCine-X Pro. However, I decided on a slightly unorthodox grading workflow by using Adobe Camera Raw. I really wanted to flex the Adobe Production Premium muscle for some reason. This workflow required taking my clips and transcoding them to 16bit TIFs and then exporting the graded files into DPX sequences to re-link back in Premiere Pro.

I exported key frames from each shot and began my grading process. One of the techniques/tools that I learned back when I was a full-time colorist was to create a "color quilt". All that is a single image with frames of an entire sequence or sequences with thumbnails representing each shot. This gives you a visual index to see if everything is being pulled together and matching up nicely. Back in the day and this was something software needed to be written for as we were on Linux boxes using proprietary grading software at the time, but with Adobe Bridge color quilting is actually a very simple process. I have a customized an Adobe Bridge workspace called "quilt mode" so I can quickly review frames for such an occasion. Very quick and easy to make selects, group, and edit what you're looking at. I would have killed for this a few years back.



I had a great time grading. I pushed the post apocalyptic world the hardest via a tonal curve with the midtones and darks lifted a fair amount. I kept the lights to highlight roll off smooth. On top of that I smacked the red, green, and blue hues around. Then went in and pushed saturation around a bit. That was a good torture test. This is precisely the type of grade you can't do with other compressed codecs out there. The hanger interiors were more subtle, but I did enhance the sort of sterile light coming from the lamp as that was a point of interest. I did knock the red down a bit. The tonal curve a very slight S to it minus bringing the blacks down. I attempted to have similar hues hitting existing in both the miniature world and the fantasy world. Not too over the top, but hopefully enough to get the concept across. Lastly I went in and added some vignettes and graduated exposure adjustments to shots to help put the focus on specific areas of the frame and guide the viewers attention.



I don't own a Red Rocket, but my cut was under 4 minutes and I trimmed my clips to have handles of 12-24 frames depending on if they had transitions or not. This also gave me a small amount of wiggle room in case I wanted to fine tune the edit. During this transcoding and exporting stage it gave the opportunity for a couple hours of sleep.


The Alien
I wanted this alien being to fly around overhead. To just sort of exist and have it be something she was used to. She would glance up at it here and there, but I wanted it to feel like normal everyday life. We don't know if this thing caused the apocalyptic world or perhaps it is some sort of guardian. I wanted that open ended for the viewer. Some think it's scary, some think it's cool. I like that room for interpretation. I've been creating similar creatures for the artwork of my next music album, however I have never thought of bringing them into motion. Plus I needed to design new ones for this project that would actually work. I've been rendering those out at 5k to print HUGE, but they sometimes take up to 12 days for a single frame. The visual effects take up a total of 1072 frames, nearly 20% of the whole short. I needed many frames as these aren't exactly short shots and I had about 48 hours at this point. I needed to optimize a few things and change a lot of other things. I would call this experimental at best and started tackling some geometry and lighting concepts.

After wrapping on the first day of shooting I raced home and began on the concept art for the aliens. It's not my prettiest work, but I needed to get some ideas down on paper to start developing a look and visual feel.



I had a few ideas in there I liked. You can clearly see my excitement in my notes when certain ideas clicked and felt right. After wrapping on the second day of shooting I began to model and perform some look development. I wanted their skin to feel like a liquid trash bag.



Things were coming together and after wrapping shooting on the third day I began some quick animation and render tests. I wanted them to almost have a gliding/anti-gravity feeling to them. Mysterious and not too threatening. I moved onto a final lighting test using the same high dynamic range image based lighting techniques that I've used on major motion pictures. Which led to:



You'll notice I changed the red orb to more of an orange. This represents their life force and energy. If you pay attention during the wide shot with the windmill as the orb retracts into the body of the creature the shot itself seems become darker. Hopefully eluding to some potentially more weirdness that's left unexplained. I rendered out my animations at the end of the third day of shooting, which is also why I shot at the farm on the first day. By time I was ready to sit down and edit I had my 4k 3D animations ready to go.


Visual Effects Workflow
Awaking from my slumber after the color work had been transcoded I had to get cracking on the visual effects shots and integrate my creatures into them. That meant it's time to open up Adobe After Effects CS5.5 to get down and dirty with motion tracking, keying, compositing, matching focal blur, matching motion blur, adjusting levels, color correcting, and exporting. Not the quickest stuff to do at 4k on long shots. The keys themselves showcase yet again the advantage of having a a lot of dynamic range and shooting with a "fat negative" mindset. There was no green screen shoot here and these were wild extractions within the plates themselves. However, because there's so much information there, even when the alien needs to be behind her shirt (which gets very close to the tonal value of the sky) I was still able to pull off a seamless composite without rotoscoping individual frames. I did a few animated effects like the rack focus on the 14mm wide establishing farm shot to the alien to match the lens pull. I had to smooth out the tracking on that shot as parallax made some of the initial track not make any sense at all. That was a combination of a lack of consistent geometry within the frame to grab during the jib move and the distance of where exactly where I wanted place the alien in the sky.



Soundtrack Scoring
On the evening prior to the deadline I found myself with a completely silent short. However, I prepared for this by using a specific time base. I exported a low resolution proxy of the short and cracked open Image Line's FL Studio (long time user) and sat in front of my keyboard. First I synced up some quick clicks to get the pacing down. I knew I didn't have the time I normally would desire to knock out something like this. I started developing some sounds and created a couple samples to use for the longer wispy strings. My main goal with the track was to have a somewhat somber tone with a twinge of hope lingering in the background. One key point I wanted to hit was having the string section during the fantasy tree scene to also exist during the wide windmill shot hopefully showcasing some sort of duality about what serenity and beauty is. After I had my track written I launched Adobe Audition and did a bit of equalizing and mastering. Lastly I exported the stereo .wav stems and re-linked that within my Adobe Premiere Pro project.


Exporting
Everything was done at this point. I contacted Michael over at Light Iron to get their DPX specifications as I've never delivered content to them prior to this short. I verified my color management through After Effects and made sure things looked right. On my computer it took about 1 hour to export a full DPX which created about 200GB of data. I set the DPX, Uncompressed Quicktime, and Low Resolution Proxy export to kick off using both After Effects and Adobe Media Encoder then took a 3 hour nap. When I got up I verified the data, copied it over to a SATA drive and dropped it off at Light Iron prior to the deadline. Post work was about 66 hours of work within 3 days. I didn't log the hours I worked on the alien after each shooting day, but I can assure you they were long days. After the deadline I slept for about 15 hours on Saturday.


Last Thoughts
I want to thank Jim Jannard, Jarred Land, Michael Cioni, and all of the people at Red and Light Iron who chose the finalists and made the Reduser Party an unforgettable night. I viewed this project as a test and a challenge. This is the first time I've been able to use a great deal of my toolbox on one single project and it's amazing to see just how much you can get done in a week. A few years ago, this really wasn't possible. Now I really know. If this effort shows anything it's that 4k isn't something to be scared of and if you weren't there already, it certainly has arrived now.


I hope you enjoyed the short. It was a wild adventure getting it done, but I had a fantastic time. It's really something I can look at now with a big smile on my face. Seeing and listening to viewers reactions at the party at NAB 2012 as I "ninja'ed" through the crowd and spied on them is a memory that will last the rest of my life.

Cheers,

Phil


4 comments:

Shane said...

That was beautiful man.. I really enjoyed it!

Nicholas said...

Great stuff Phil! Thanks for sharing.

Tom Scholes said...

hell yeah dood ~!

Phil Holland said...

Thanks guys!

Categories

Blog Archive

Followers