Smoothing a Shaky Camera Move in Fusion
Inspired by the Shake “SmoothCam” tool or F_Steadiness in Nuke I’ve written a plugin for Fusion that allows you to automatically smooth or stabilize a shaky camera move. Fortunately I had found a public domain program by a Finn called Jarno Elonen that determines an image’s transformation (scale, translation, rotation) based on a variable number of points. Without knowing anything about “reduced echelon matrices“, “least square fitting” or the “Gauss-Jordan Elimination” (those Wikipedia pages give me the creeps!) I managed to translate the code to LUA and it worked perfectly.
The secret is to interpolate the motion vector image down to as little as 2×2 values. These can then be fed as points to the algorithm. Even my naive approach of using a garbage matte to simply zero vectors that have distracting motion seems to work.
There’s also a video on YouTube about it as well. It’s a demo of my beta version that has an outdated interface but the way of using the Fuse is mostly still the same.
I don’t know how robust it is to various kinds of shaky, jittery, wobbly footage and some GUI decisions might seem odd. But on more than one occasion I was limited by what Fuses can currently do. Still, I think it works well enough to publish it to the Fusion community.
Download the plugin here: SmoothCam_v1_0.Fuse or read the manual on Vfxpedia. Photo credits for icon: CC-BY Nayu Kim
The Hobbit – Yet Another Disappointment in 2012
I really would have wanted to end 2012 with a nice movie-going experience. I tried to ignore people lamenting about 48fps or stuff that wasn’t part of the book. I had never read the book.
Before the movie started there was a trailer for the new World of Warcraft update called “Mists of Pandaria”. And it had Kung-Fu-Pandas in it, fighting orcs. It was the most ridiculous thing I had seen in recent months. Little did I know that this was foreshadowing the movie I was about to watch.
‘The Hobbit – An Unexpected Journey’ is basically the same thing but roughly 3 hours longer. It’s an attempt to blow up a tiny story to not only 9 hours but to Lord of the Rings epic-ness while making it look like a video game cinematic. The term cinematic is actually quite ironic. While video games have tried to look more and more like movies (by their themes, camera angles, animated or life action cut scenes and the use of machinima) it seems like the future of blockbuster movies is to look more and more like video games:
Level 1 is the Shire. Go on a journey, battle some foes, meet some allies… until you reach Level 6 – The Goblin Cave! Press A to swing your sword and B for a special move to decapitate your enemies. The Level Boss is the Goblin King himself! Attack his vulnerable spot and when your energy level drops low, press Up-Down-Up-Down to make Gandalf appear and save your ass.
Sorry, Bilbo, the princess is in another castle. Stay tuned for parts 2 and 3 – available next year.
For a movie essentially geared at a young audience who might not even have seen LoTR the movie is an astonishing mix of childish themes, brutal (yet blood-less) hacking and slashing and dialog scenes that drag on for way too long.
The movie’s first 15 minutes are filled with shots of dwarfs eating cheese, juggling plates and two musical numbers.
After two thirds of the movie I had to accept that nobody’s going to get injured or die even after falling downhill for hundreds of meters. I accepted that one of the dwarfs and the goblin king looked like “Fat Bastard” from Austin Powers. I was no longer surprised when Gandalf just showed up and saved everybody at the last moment using his magic powers again and again – this happened at least three times during the movie.
And the HFR thing?
For a movie that is so intend on selling an experience and showcasing new technology (instead of, you know, making you feel sympathy for fictional characters on screen) “The Hobbit” actually tries hard to make you loathe it. The high frame rate irritated me every other minute with its “sped up” effect that you might have heard about. It’s an optical illusion and my fellow movie-goers didn’t notice it but to me it felt like watching a TV documentary about the movie, not the movie itself.
The 3D felt forced as well. I might be from a dying generation of movie-goers but it still irritates me when there are elements in front of the screen while being cropped at the edges. Fast-moving sparks, butterflies or gold coins still are a flickery mess to me even at 48fps. And landscape shots still have that miniature look to them because directors and DOPs insist on using an exaggerated interocular distance.
In a way it’s comforting to know that even huge productions like this suffer from that shit that James Cameron successfully avoided in Avatar. But that’s probably because one disappointed moron in the target audience of 16 year-olds (“omg the 3D was non-existing I could have left my glasses off”) weighs heavier to any producer than somebody who is pulled out of the movie by miniature landscapes.
All of this overshadows the fact that the VFX are of course top notch. Except for one or two scenes you never think about the fact that Gandalf and the dwarfs are composited together for their difference in size. Closeups of wargs and eagles are great and the level of detail in the dwarf city or goblin lair is breathtaking.
In hindsight I should have watched the 24fps 2D version to apprechiate all of this.
6/10
“Damn Damn Good:” Movie Bob’s positive review of “The Hobbit”
“I hope the worst is behind us”: Red Letter Media’s more negative review of “The Hobbit”
Hänsel & Gretel: Witch Hunters
HO – LY – SHIT. That one looks like it’s gonna be fun 🙂
Or… it could suck like all the other movies of that genre that try to re-invent fairy tale characters in a world tainted by the Twilight saga (“Red Riding Hood” comes to mind but also “Van Helsing” although it predated this decade’s vampire craze). But at least “Hansel & Gretel” looks like it’s going to be honest: no hollow claims of answering questions about the origin of life (“Prometheus”) or re-inventing time travel (“Looper”).
Just “this is what Hansel & Gretel would look like in a steam-punk movie for teenage boys. Hell yeah!” *
Who knows if it’s gonna be good. Maybe it’s witty like “Brothers Grimm”, maybe it’s trashy like “Underworld” or “Resident Evil”. But with stills like these, it’s definitely on my watchlist even though this mentality has led to disappointment most of the time 🙂
The VFX look much much better than in Resident Evil though. That hero shot at the bottom looks really well done. Light is engulfing the actors quite nicely and although a real explosion would be even brighter the actors are underexposed enough to sell the shot.
Compare this to a shot from the Resident Evil trailer I’ve criticized in the past where the integration of explosion and environment is abysmal.
For people who are not so much aware of what constitutes effects shots nowadays (and why movies get more expensive all the time…): I’m pretty sure the fog in the topmost screenshot was added later on in postproduction. Even if there was such a forest with the right kind of humidity at the right time of day to produce a suitable fog, no sane producer would allow a shot of the main actors to be restricted by such unreliable weather conditions. Doing the fog in post also allows the director creative freedom to add details like those godrays. All it takes is a matchmoving artist, a bunch of roto slaves and a 4-digit amount of dollars. If the director requests half a dozen versions until he’s satisfied with the color and density of the fog, the VFX company probably won’t break even. That’s the state of the industry.
*) yeah, I’m using the term “steam-punk” loosely. But any movie that mixes gatling guns with 18th century German timber-frame construction qualifies as steam-punk to me.
Keying Window Reflections, Smoke and Explosions
This post is about a keying technique that saved my ass on a couple of difficult shots recently.
Most keyers are all about pulling a matte from a green or blue screen plate which is then used to merge a despilled image onto the desired background. This technique has its limitations though and who doesn’t know the tedious task of combining and layering several keyers in different parts of an image until the result looks right.
There are, however, keying issues that are hard to get right even by combining multiple keyers: things like glass or glow. Both phenomena are additive in nature – they add light on top of what’s reflected from the green screen into the camera lens – which makes them different from semi-transparent things like smoke or motion blur. In compositing terms this means that there is no proper alpha value for them but since they change the color of the green screen behind them (reflections on a glass window are a good example) traditional keying algorithms will calculate an alpha value > 0 anyways.
click to enlarge
The technique I’d like to demonstrate is simple. I didn’t invent it, by the way. I first came across it in a post from the Shake mailing list years ago but didn’t fully understand it. A compositing artist on “Vanilla Sky” (if I remember correctly) said that he pulled a soft matte and used it to push the color-corrected background through the foreground, preserving grain and fine hair.
These are the steps you need to take:
- Use a color difference formula to pull a soft matte, which is similar to despilling: use the difference between green and the average of blue and red, or the difference between green and the maximum of blue and red.
- Instead of manipulating that matte until it’s fully black and white, use it as it is to add the background image onto the (despilled) foreground image. This will lead to a strange low-contrast mix of foreground an background. But here’s the trick:
- Crank up the contrast of the background image until the result looks right. This of course only works in floating point since it requires pixel values outside the 0..1 range.
This technique will preserve edges perfectly as well as glow, reflections or smoke up to a degree where the greenscreen is but a tiny green tint in the background. On the other hand, it will also “imprint” any imperfections of your green screen into the result: folds, shadows, tracking markers, uneven lighting. This means that you either need to do some serious cleanup work and degraining beforehand or combine this technique with traditional keyers (e.g. just for fine hair, or inside a window pane).
I’ve turned all of this into a handy Fusion macro. For Nuke, you can use the “Despill” gizmo from Luma Pictures, which served as an inspiration for the UI of my macro.
I’ve called the tool “SpillMerge” since it performs a despill first, then merges a background image into the despilled area (it basically treats the screen as a giant spill-affected area instead of trying to pull a solid black/white matte).
Download SpillMerge here or read the manual at vfxpedia.com.
