VFX TownHall Meeting

From Lee Stranahan: (twitter: Stranahan)

Online Town Hall registration is open - totally free, Monday March 29th, amazing panel
Link

Info from the site:

VFX Online Townhall
an open disccusion of a visual effects industry in transition

Effects driven films like Avatar are breaking box office records while visual effects facilities are forced to shut their doors and most VFX workers face long hours, no benefits and little credit.
It's time to start talking about the state of the visual effects industry -- where we are, how we got here and what can be done to move the industry forward in a way that's fair and thriving for artists, facilities and the studios.
We've assembled a world class panel with different perspectives and now we'd like to invite you to be part of the conversation. The discussion is open to everyone with an interest in the visual effects industry.

Panelists include
Chris deFaria -- Vice President, Warner Bros. Pictures
Jefferey A. Okun -- Visual Effects Society Chair and visual effects supoervisor
Scott Ross -- Co-Founder of Digital Domain & former CEO of Industrial, Light and Magic
The panel will be moderated by Lee Stranahan, a former visual effects artist and writer for The Huffington Post whose Open Letter To James Cameron: Fairness For VIsual Effects Artists started discussions all around the world.

This all started with the Open Letter to James Cameron

VFX companies continue to fall

[Update: 3/29/2013 See this article for a link of some of the companies that have gone out of business
Creative Cow VFX at a Crossroads 
and from Reddit
]

Last year we saw a few VFX companies close their doors (The Orphanage, Illusion Arts, Pacific Title, etc)

We're now seeing a few more, or in the case of Image Movers, a planned closing.

Image Movers
There are a number of ex-ILM people there and at 450 people it is not a small group.

Core Pictures I was in discussions of taking a project there a year or two ago.

It's always painful to see VFX companies close their doors, both from knowing artists and friends will be out of work and knowing how fragile the VFX eco system really is. Having owned and operated a VFX company myself (Dream Quest) it's tough to make it work financially. The pressure is always on from the studios and producers to work as cheap as possible yet VFX require specialized equipment and artists. It's also tough to have projects flow smoothly and keeping all artists busy. That means there will be times of no money so any budget you make will have to take that into account. It's not extra profit, it's covering expenses. With a small shop the owners can skip pay themselves for a certain period but for larger shops it can be impossible to cover the costs without some deep pockets.

Between the recession, out sourcing and short term fixes at higher levels, this is a bad time for VFX companies and artists.

Disney bought Image movers and now that new management is at Disney they want to get rid of Image Movers.
A large VFX company takes time and money to get going. It has a high operating cost. A studio or company has to be willing to accept that going in or there will be trouble. Disney did the same thing after they acquired Dream Quest and turned it into The Secret Lab. They closed it down after it was up and running.

Unfortunately outsiders rarely understand VFX and what it's about. Sure you need to make money to keep operating but some companies are looking for huge profits instantly. Studios and producers have this thinking that VFX are expensive because all these VFX companies are holding them hostage and making huge profits. Welcome to reality. VFX has not been a huge profit area.

Many studios had their own VFX departments in the golden age of movie making. Those were closed down when the studios sold off their backlots to make money in the short term. Star Wars and Close Encounters both required the productions to set up their own VFX groups from scratch. These were meant to only operate on these specific projects respectfully. With the success of Star Wars, George Lucas was able to keep ILM as such and could afford to set it up again in Northern California for his own projects. Apogee was the spinoff of some of the people from ILM. Many of the Close Encounters team went on to Buck Rogers and then on to Star Trek: The Motion Picture.

The studios always claim to have no money but it's not unusual for a studio to make million dollar changes to a production on a whim or creative impulse. In some case they're focused on the short term improvements and less on the long term. (See the Jay Leno issue the last few months) Sometimes the new group at a studio wants to make sure none of the projects of the previous group remain. They might not want a success that could have been created by the people they're replacing. This means in many cases they will cancel projects that are in development, even though the projects may be great and make a lot of money for the studio.

What many of us tend to think of as 'no brainer' decisions tend to be the hardest decisions to make for many businesses. I've seen this repeated time after time.

VFX have been making great strides the last few year and enabling filmmakers to create just about anything imaginable. Both the artist skill sets and the technology are making this possible. More people are trying to enter this field than ever before. The Visual Effects Society is getting larger and making improvements.
Just about every movie coming out has some VFX, no matter how small.

So the movie industry had the biggest year ever last year yet the studios are reluctant to produce anything beyond franchise tent-pole movies. The general public is going to the movies more in this recession but the banks are reluctant to loan any money to studios and filmmakers. The studios want not only more and bigger VFX but they also want them for 1/2 the price in 1/2 the time as the last project.

Welcome to VFX H*ll.

(I didn’t want to end this on a negative note. I still have high hopes for VFX and VFX artists but we’re certainly going through a rough patch here with a lot of conflicting forces at play. Lets hope for some good news soon.)

[Update: 9/12/2012 There have been a number of other vfx companies closed down since this was originally posted a couple of years ago. Just in the last 2 weeks alone we've seen the closure of Matte World, Fuel, Digital Domain's Florida facility and a restructure and sale of Digital Domain]

[Update: 2/16/2013 R&H, one the largest vsual effects companies in the US (who did Life of Pi and a number of other well known projects) has field for chapter 11 and laid off 200-300 people. (Who weren't paid) ]

[Update: 3/29/2013 See this article for a link of some of the companies that have gone out of business
Creative Cow VFX at a Crossroads
and from Reddit
]

Related:
Check the right side of the blog for lists of posts related to the industry under "VFX INDUSTRY - STATE OF THE INDUSTRY"

Pass Me A Nail - Problems of the vfx industry
Using the Nail - Possible solutions
VFX Business Models
Working directly for the studios
Visual Effects Service - The Big Picture

Bad Visual Effects Business practices
Oh, What a Mess We're In!

More VFX wage issues

There's been more posting and other things around the web regarding the state of VFX artists.

Here's a couple more related links:

Some Inconvenient VFX Industry Truths by Richard Kerrigan

A Visual Effects Digital Artists Guild? from the Animation Guild blog.

This just added:
Poll regarding VFX Union


I haven't had a chance to sift through all of this yet but hope to make more of an in-depth posting some point in the near future.

Motionographer interview with Lee Stranahan

More followup to the previous post regarding the VFX letter to James Cameron at Motionographer.

Open Letter To James Cameron: Fairness For Visual Effects Artists

For those who haven't seen this yet, Lee Stranahan has posted an open letter to James Cameron regarding the state of the VFX artists working on today's movies.

Click here to read.

There seems to be a consistent theme these days.

Commotion

Matthew Ingram has written up a great posting about Commotion here.  [Update: I think this site is gone so may re-post it here if I can]
Thanks Matthew!

Commotion was a software packages I wrote for myself while at ILM.
It started as 'Flipbook' and allowed realtime playback of images on a Mac II. (This would later be used as a design basis for other tools at ILM.) It also had a flowchart compositing system. You could actually setup multiple composites all types of other things.
(yes, before After Effects, Shake, Nuke, etc)

ILM had an SGI tool that did image processing but the command line interface made it less than useful for the artist. I created a Mac app that provided and interface and send the commands onto to the SGI. I had also written an app that would control Photoshop so it could be used on multiple frames. I was an VFX supervisor at ILM, not a software developer, so all this was to provide me tools so I could do my work without fighting the system and getting proper software written.

The originally released product was a bit different (roto, paint and playback) and certainly when the compositing was released it was aimed more at AE style than the original

[Update: fxGuide has an article entitled The Art of Roto: 2011that discusses the history of rotoscoping and some of the currently available tools. It also includes an interview with me talking about Commotion.]

More VFX wage issues

More bad news for Visual Effects Artists around the world. (and bad news for VFX vendors as well).
Keep in mind that most of the top money making movies of all time are heavy VFX movies. (Titanic, Avatar, Star Wars, Matrix, etc)

Consider joining the Visual Effects Society to help provide a more united front.


Via Twitter' from neonmarg

Visual Effects Industry gets weaker everyday

Working in China doing VFX

Woes of non-union VFX

The Meteor case has become a symbol of the shaky standing of the vfx industry and vfx artists in particular. Vfx artists have no union or guild, and no Hollywood union has consented to represent them. Meanwhile, some vfx studios are in such bad financial shape, with current income used to pay off past debts, that one analyst has called the entire business “a Ponzi scheme.”

Snippet from VARIETY article by By BRENDAN KELLY

Variety Article on Meteor Studios and the the Visual Effects Artists that worked there.

VES Annual Meeting

The Visual Effects Society had it's annual meeting last Saturday. There was a live feed to Vancouver and San Francisco.
It was also webcast throughout the world including the other sections of London and Australia.

The video of the meeting should be online in the next couple of weeks or so for all VES members.

As noted at the meeting the VES website is being totally re-written and is scheduled to be online in October.
Lots  of great stuff like forums, videos and user areas.  Headed up by Carl Rosendahl

Mike Fink gave a talk called 2000 years of visual effects in 30 minutes.
Discussed the history and art-form of imagery and how even early paintings are making an impact on cinema and visual effects.
Helped to illustrate the artistry and advances in visual effects.


Bill Taylor was awarded the VES Board of Director's Founder's award.
Bill talked about many things including:
1. Keeping production local
2. Having a visual effects trade organization for visual effects companies.
3. Having a visual effects union.

This was well recieved. These are some of the issues that the VES board has been discussing the last couple of years as we look to the future.
As Bill mentioned Jim Danforth had brought up some of these same issues years ago.

The VES is an honorary society for Visual Effects Artists.  It's been discussed that that these 3 things: Society, Union, Trade organization, make a fully rounded industry and each of those had different purposes.

The VES determined it was best to remain as it was designed to be - an honorary society, instead of changing into one of the other forms.

I think many members and VFX companies are interested in the future and want to avoid some of the issues that we currently have to deal with.  A union and trade organization are possible ways to deal with those issues.

Doug Trumbull was the main speaker at the event.

Doug is always interested in getting the greatest experience and felt that many of the limitations have been based on
decisions made years ago that aren't relevant now.

Doug felt that film will disappear (no time frame given) and that digital can equal the quality of film, even now.

FPS
Doug had said that in early motion picture development approx 50 fps was determined to be needed to show continuous movement and avoid flicker.  Soon there after multi-bladed shutters to minimize flicker issues. 16/18 fps were chosen for silent movies to use as little as film as possible.  Sound went to 24fps not for the picture but to make it possible to do sync sound.
He developed Showscan years ago which originally ran at 72fps and later at 60fps.
24fps results in a lot of blur.  It requires the blur for your eye to read it as continuous motion instead of flicker still images. He and Kubrick found this out when trying to shoot stars without blur on an animation camera. Faster frame rates don't need as much blur and tend to produce sharper images.
Even though 24fps is thought of as movies and 60fps is thought of now as video, the faster frame rates provide a more life like experience.

Projection
The current standard for footlamberts on a movie screen is 16.  This is much darker than your flat-screen at home.
With stereo 3D projection now a polarizer is needed over the projector lens which results in 1 stop loss of light.
Then the glasses the audience wears has polarizers which result in another stop loss.  End result is the audience is watching an image at 4 footlamberts which reduces the color and quality of the image, especially compared to real life.

The current digital projection systems use a chip size that limits the amount of light that can be projected through it so there still remains a limit to how bright and real the screen is.  Doug would like to avoid these limits and take the opportunity now that we're transitioning into digital projection to think in terms of the future and not just come up with standards that are barely good enough.

Worth checking out the video.

There were also reports from the various committees with a lot of progress being made in the last year.

Healthcare for all

Protect Insurance Companies PSA from Will Ferrell

Public Healthcare

Quite a few visual effects artists are freelance. They work from project to project. Some are contractors. Even when working at large VFX companies there are frequently large gaps when there is no work and the artist will be unemployed. Since there are no VFX unions there are none of the benefits of having united support. A few categories are covered under unions (modelmakers of physical models, matte artists and possibly a couple of others) but typically none of the CG areas are covered. I'm in the cinematographer's union (since I started before the days of digital I dealt with cameras) but some of the large studios refuse to recognize this since there's no official 'vfx supervisor' position listed by the union.

The problem here is trying to maintain health coverage in a field that has constantly changing amount of work and where most of the work is based project by project. That's why the Hollywood unions were set up to help artists working on films deal with things like health care, overtime and working environment. Studios are big businesses and as such they will try to save a dime where ever possible.

Fortunately the Visual Effects Society has signed up to the same health plan as the Producers Guild. So if you're a member of the VES it may be a little easier or cheaper to get health coverage than before.

However everyone in the US should be taking a look at the health care discussion. The US is the only industrial nation without some type of national health care support. The public supports schools, libraries, medicare and other things but the health of the people is not a priority. If you've had to pay for health insurance or medical care it's clear that the current system is very broken. The health insurance companies and the drug companies are making a mint off everyone.


Public Healthcare Video

Visual Effects Producer Book

This was written by Charles Finance and Susan Zwerman, both with the VIsual Effects Society.
Susan is also one of the editors of the VES Handbook that is in progress.

If you're interested in visual effects producing this is the book. Covers primarily the budgeting, scheduling and workflow issues but also covers basics of the technical issues along with pros and cons. 377 pages



Update:
Another book to check out is The VES Handbook of Visual Effects which was released end of July 2010.  More info.

Greenscreen plates

For those interested in practicing with greenscreen plates check out
Greenscreen Test Footage

Hollywood camera work

Tracking footage: 

Range of shots and difficulties. Tracker markers and other items in the shot.
I'd avoid using them in your demo reel (since there may be a number of other people with the same clips) but they should be an excellent learning materials.


New link added 4/7/2013 :
4 TB ORIGINAL 4K FOOTAGE AVAILABLE AS CC-BY

Reworking Blog

Reworking the blog and updating the template. Will have to restore many things so it may be awhile.

Previs thoughts

Just Misc Previs thoughts from a note I had sent someone months ago.

I think previs encompasses anything that is used to get a visual sense for the final product ahead of time. The majority today is 3D but video, stills, storyboards, film snippets, etc are all possible and useful.
For Star Wars and even some pictures today existing footage (from another film or stock) is seen as a placeholder and visual guide. For Phantom Menace and other projects temp actors have been shot, sometimes against greenscreen, to block in concepts and editing ideas. For some movies such as Amelia they used videos and stills from actual locations or sets to get a sense for how the scenes will work visually.

Misc thoughts

DON’T’T CHEAT!
When working in a 3D environment it is all too easy to make a change at the director’s wish. Scaling an object, moving the camera, placing objects, etc. Usually this is done to try to make the previs as exciting and interesting as possible but these add up to real production problems.

Examples:
Fantastic Four – A major sequence was to take place at the London Eye. The previs team decided to scale the London Eye to less than ½ of its actual size to make better compositions. For months we had been looking at these and all departments were basing their planning on these only to find out that it was all fictional. Since we were to shoot the real London Eye, create accurate 3D models for the final shots as well as potential miniatures that were the correct scale this was a huge problem. We now had to have the previs scramble and redo all of the shots. By now production was in a different country and another previs team had to do the work with a different software package.

Shots moving in the tunnel varied wildly in terms of camera speed, car speed and character speed. In a simple previs it looked right but impossible to make look right if it were to be shot like that.

For Van Helsing some of previs was done in a way that couldn’t be shot. In some cases the cameras were placed in areas they couldn’t possible be or doing something that defied physics. When the production crew is half way around the world 6 months looking at a previs on a laptop and the director discovers they can’t do the shot as planned, it’s very painful for everyone involved.

You don’t want to get to a location and find that you’ll need to jackhammer the road to place the camera where the previs was ‘shot’ from’

You also don’t want to find your lead actor is now somehow supposed to be 10 feet above the ground. On the location cheats are made (actor on apple box, shooting in a different direction, etc) but everyone there knows it’s a cheat and why it’s being done. When a cheat occurs in previs the previs artist may be the only one who knows it. It’s likely the director doesn’t even know that it’s taken place.

Purpose
Determining the purpose of the previs is critical. Make sure everyone, including the director is onboard.
We just need the previs to a certain point to understand how the sequence and shots work, the approximate action and timing, etc. Yet a director can easily spends weeks ‘directing’ the ‘actors’ in the previs. “No, he should smile here and then look toward the camera”.

Previs has a tendency to become like temp tracks to the sound track. Something that the director has been looking at so long that it is the only way the director sees the shots as being. On a non-prevised show the creative team may make full use of the here and now and compose the shots to their advantage. (some feature at the location, the light at that time of day, etc) On a prevised show the director may be unwilling to consider these alternates.


Speed
It’s difficult to get a full sense for speed in a previs. In early previs a car doing a drive by might have been just a colored rectangle moving against a simple background. The director of course wants it faster, faster. If the same scene were shot for real at the original speed it would have been fast enough. All the details of the car, backgrounds and motion blur would give the sense of speed. Even with today’s rendering there’s still some visual speed discrepancies.

Accuracy
When storyboards are done everyone understands these are the basic shot designs and placeholders. It is understood that the perspective and placement of the actual location will be different in real life. One problem with 3D previs, especially the more detailed they become, is everyone thinks this is the actual shot, even if the previs was done months before the actual location was chosen.

Design
A major potential problem is a lot of previs is started before any of the key creative team is hired. It may be just the director (or even just the producer or studio executive) and a team of previs artists. The amount of visual sense the director has can varied widely and the visual sense and experience of the particular previs artist can vary widely as well. On a non-previs show the director works closely with his DP, camera operator and others (production designer, VFX supv) to determine the best compositions. This is likely also based on blocking in the actors motions. Yet the previs may well lock the creative team into design decision that were made by the director and the previs artist in a vacuum. I’ve seen shows where the stunt coordinator was told that he’s to match the action in the previs. Imagine having someone like Jackie Chan being told that some previs person has already designed all the action and action shots and that all he needs to do is get his stunt team to do it that way.

A car stunt may require a special rig be placed and the stunt coordinator may know the best camera angles to capture it but the previs artist knows nothing of this and places the camera exactly where it’s going to be a problem. The director has now grown to love this and wants the stunt team to sort it out.

The same thing has started to happen with DPs but most of them have enough clout to stop it there and do it their way. So how worthwhile is the previs if it’s totally ignored? If the VFX has been budgeted based on the previs but the director, Dp, etc ignore it completely where does that put the budget and schedule?

If the previs is done ahead of time it’s difficult to talk the producers and studios into redoing the previs so the now hired DP, production designer, etc can be involved.

Science of Magic and Illusion

Wired magazine takes a look at the science of magic. (thanks to Eric Alba http://twitter.com/alba for mentioning it)

Click the title or use the link below:
Magic and Illusion

Many people think realistic vfx are much more difficult to pull off but in fact they can be easier because as pointed out, people don't tend to focus on anything except what's in the spotlight. The audience will easily accept something as being real if it's not fixated on and if it's a common thing. You can do a matte painting of a house along a street and you can do the same thing of a very abstract haunted house centered in frame. The quality of painting, lighting and compositing could be exactly the same but the audience will readily accept and even ignore the house matte painting. It's not calling attention to itself and since they've seen plenty of house they don't bother looking it over. The haunted house, especially if it's centered in frame and the full focus of the shot, will tend to be deemed as a VFX even if it were really built on the location. It's much harder to sell the audience on this since they have to suspend disbelief.

Even harder are things like dragons and other vfx that the audience knows are not real. Their subconscious knows it's not real but they readily accept all the other non-real things in a movie (dialog, actors, sets, etc) because those 'could' all be real.

Of course there are other things that can make realistic effects easier - you have a real reference(s) to the real item. By comparing photos or video there's no discussion needed about what it should look like and how each person imagines it. You have a direct reference to copy. It may take some work technically but you know once you've duplicated the elements that make up that image the audience will likely accept it.

VES Handbook

I wanted to let everyone know that there is a Visual Effects Society handbook being written by quite a few visual effects artists, myself included. There's already a publisher involved and the plan is to release it next year.

I have a few different articles halfway done for this blog that hopefully I'll be finishing up and posting once I have more writing completed on the handbook.


Update:
For full details on the handbook and it's release please follow link below.
Latest Info: Handbook has been released posting.

Cloud Tank effect

People have been asking me to provide some of the details of the Cloud Tank so that's been on a long list of subjects. Recently I had to write up some of the notes so thought i'd post it here as well.

When I start in visual effects I was hired by Doug Trumbull as his assistant. The first day I was asked to create clouds in a liquid for a film that he had just started working on called Watch the Skies. You may remember it by the final film title, Close Encounters. They needed to create clouds and they liked the look of pouring cream in tea or coffee.

I was given a 20 gallon aquarium and $20 of petty cash. I worked it out in a few days with additional ideas from Doug, Wayne Smith and others.

This same process was later used in a number of films and commercials. (Raiders of the Lost Ark, James and Giant Peach, ID4, etc) Actually most of these used the same exact tank as well.

I've attached a couple of primitive drawings to make it clearer. My photos are packed away deep in boxes otherwise I'd post some here. This was also covered to some extent in the Close Encounters book.

[Update: 12-22-11 To make things a little clearer I'm adding some notes here. The basics are a large all glass aquarium was filled halfway with salt water and then fresh water was carefully added to the top. Salt water is heavier so tends to stay on the bottom but you want to avoid mixing them if possible. That's what some of the details below are about. Also best to avoid getting the water too warm. It's very difficult to see any difference in the water once filled since it all looks the same (assuming you've also cleaned the tank and filtered the water).  under the right light and angle you may be able to just make out a slight difference.

Next white liquid tempra paint is injected in the fresh water portion (top), usually just a few inches from the dividing line of the fresh and salt water. Think of a large syringe with an aquarium tube going into the water. When the tempra paint is injected it billows outward like cumulus clouds and will tend to sink a bit. But the salt water prevents it from going lower so the 'cloud' tends to flatten it's base on the salt water line and and billow outward, similar to real clouds based on air pressure levels. Avoid going below into the saltwater since the clouds will just drop to the bottom of tank.

Once you're setup, light the water and create the clouds. You can film while injecting the clouds to get the large billow action or you can film once they're in place. Filming can be under-cranked (less than 24 fps) or over-cranked (over 24 fps) depending on the look desired. Over-cranked will tend to increase the sense of scale and by placing smaller clouds in background you can create force perspective. You'll need to completely drain the tank, clean it and repeat after each shot.]

Process in a nutshell as used on Close Encounters:
2000 gallon glass tank approx 7 x 7 x 4 feet deep

Water was purified and filtered in 2 large hot tubs/wine tubs(?) 1000-2000 gallons in size.
One had rock salt added (probably close to seawater specific gravity but I can't remember the details)
The salt water was pumped in first usually to halfway mark, dependent on the shot.
Sheet of thin plastic (visqueen) was laid on top. Think of the plastic used for heavy duty trash bags or painter's drop cloths, flexible but not easy to tear. This was longer than the tank front to back and would drape out the back of the tank.

A PVC pipe running the width of the tank was placed on top of this with a rig. Later to remove the plastic it would be pulled out and this PVC pipe was used to keep it down.
A PVC piping system that looked like a fork from above was placed just above the plastic and the fresh water was pumped in.
When done this was removed and the thin plastic sheet was removed.

Now it appeared as just a tank of water but there were two layers if you looked very carefully.
A large syringe was used with aquarium tubing to add small forced perspective clouds in the distance.
The mixed white tempera paint was injected right above the slat water layer. if it went below the tank was contaminated and would require a full redo.

Next an atomic arm was used to inject the main clouds during shooting. This allowed the operator to stand in front of the tank near the camera and as they move it around a brass tube was moving in the tank.
An electric trigger on the handle would cause the tempera paint to be injected as the arm was moved. The tube came from a pressure cooker filled with the paint.

Pressure was supplied from a compressor and an electronic valve was connected to the atomic arm trigger.
We also had a light source (with rotating colored lights) that went down this same tube as a fiber optic.

Frame rates could be under cranked or over cranked depending on the look.






Close Encounters Book

Close Encounters - BluRay

Close Encounters - DVD

Close Encounters VFX Video 1 - YouTube

Close Encounters VFX Video 2  - YouTube

Updated: Polaroid from setup 6-14-76

Cloud tank articles from others:

Single Minded Movie Blog

Fish tank for otherworldly environments

Update on Dallas Cook

When I started this blog I had posted news on a young man who was killed by a drunk driver.
Ryan Dallas Cook was in the band Suburban Legends, which my daughter saw quite a bit.
The drunk driver worked for Hyunda of Korea and the local company helped him to escape back to Korea.

He has finally been tracked down and extradited back the the U.S. for trial.

Here's the information from a few weeks ago:
News Story

(As a side note: If any one has any photos of the memorial they did for Dallas at Downtown Disney please let me know. His family would like to gather some of the pictures or videos. Thank you)

Fliparoni update

Just a note that Fliparoni (iPhone and Touch game) is available free for a very short time as a promotion on the App Store and iTunes.

Fliparoni

My newest iPhone application is now available.
Fliparoni, It's a video puzzle game for the iPhone. Improve your visual sense.
Click on the post title above to go to the website.

Now I'll be getting back to visual effects postings.

Squiggles

For those interested I updated my iPhone/Touch paint program Squiggles.

Click the Squiggles link above to see demo video and images.

This has even more advanced paint options (paint smoothing, opacity, paint modes),
stamps (color brushes using provided images), overlays (compositing with moving layers using the provided images), Text layers, eraser and Clone tools as well as distortion and interactive image adjustments.

Space Race

For those interested in a blast from the past (late 80's?, early 90's?) a ShowScan simulator ride I directed and Supervised is posted on the web. (Click Space Race above and scroll halfway down the page)

Showscan was a 70mm process developed by Doug Trumbull to run at 60fps. It loses a bit in the translation to 320 pixel web video. We shot this all in VistaVision at ILM and there is no CGI despite what the notes say. We built a 2-3 foot wide track that covered a 1/2 lap. This was shot motion control and redressed for each segment. All space vehicles are motion control models. This was probably the last all optical composited projects at ILM. Ned Gorman was the VFX producer (and Writer). Ned also spotted this video. Ty Ellingson was the Art Director.

The film played in both normal ShowScan theaters and in their simulator ride theaters that had motorized/hydraulic seats.
It played around the world including Toronto, Las Vegas and L.A. This was one of the most successful ride films at that time.

iPhone

I've been tied up lately with a lot of other things. (Hence the delay in getting new postings up).
One of those things is a paint program for the iPhone called Squiggles. The link above takes you to more info.
I frequently need to sketch little things to discuss with people on set or at a workstation. This also allows painting on photos so that useful for noting items. I have many more features in the work (along with others apps). Some these will probably be specific to visual effects so post a note for any feature requests to the existing app or to new apps you'd like to see.

Designing Visual Effect Shots

Designing Visual Effect Shots, Part 1

(This posting started getting very long and cover a lot so I’m breaking it into smaller postings. I’ll get into more specific details in future postings.)

The success of a Visual Effects shot is not only dependent on the technical aspects of the shot but also the creative aspects. It all starts with the initial shot design.

A well designed visual effects shot will have impact and help to tell the story clearly. A poorly designed shot may confuse the audience and at the very least will squander an opportunity. A poorly designed shot can actually cost more if the studio or director decides in the edit stage that it’s just not working. At that point the shot could be re-shot with a better design (unlikely) or many attempts will be made to fix the shot in post without a clear understanding of the problem.

Note that the design criteria for visual effects versus normal live action shots are primarily the same. The differences with visual effects shots are:

1. Usually the entire visual effects shot is not visible or apparent at the time of shooting. Images will be added later or the existing image will be modified. This requires pre-planning. A live action shot is usually working or not working on the set. If a camera angle makes a stunt look boring then they’ll know that when reviewing the video on set.

2. On live action the Director of Photography and Camera Operator are focused on the look of the shot in addition the director. The director respects their opinion. With visual effects the director and a storyboard artist may have designed the initial shots before the visual effects people are even hired. How much influence the visual effects supervisor and his team have on shot design depends greatly on the director and how much respect they have for the visual effects process. The better directors understand this and take advantage of the visual effects team.

3. Visual effects sometimes deal with design issues that don’t come up directly in live action. How to show the scale of smooth object floating in space? How to transform this paperweight into a creature?

4. Visual effects can be much more limitless. With live action you have set and equipment restrictions which may prevent you from doing certain types of things. A visual effects shot can have more freedom of action, movement of camera and lighting effects.

5. Visual effects can require a deft hand of design and editing just as a comedy sequence requires some finesse of timing, angles and specific phrasing.
Spider

Below are some of the many issues to keep in mind when designing a visual effects shot. These aren’t rules, just a set of suggestions.

Does the shot help to tell the story?
This should be a fundamental of any shot or scene in a film, whether live action or visual effects. Sometimes visual effects are only used as eye candy. The director wants to wow the audience with a car crash, explosion or a visual effects shot. If that can be done and still work to tell the story then that’s great. If it’s only purpose is eye candy to wow the audience then it may be a lost cause.

Audiences these days have seen a lot visuals between films, tv, video games and the internet. They’ve come to expect something new and different. Visual Effects are not as special and magical as they once were to the audience. There was a wow factor in the early days of computer graphics when things were new. It becoming more difficult to find techniques that provide the wow factor. Shot design is a major factor to making the wow factor even using standard techniques. As a case in point, THE MATRIX used ‘bullet’ time and most people thought this was the first use of it. There had already been at least one movie with the same effect (LOST IN SPACE) and a few commercials but the combination of art direction and design combined with the story made an impact.

Even in the early 80’s people thought much of what they saw was computer graphics. A number of visual effects commercials were designed to look like computer graphics even though many of these were done by traditional animation techniques. Logos would fly through the air with metallic glints. These were all done with a number of pieces of artwork and passes on an animation stand. For ESCAPE FROM NEW YORK at Dream Quest we used physical models of building painted black with white lines.

What is the point the shot has to make?
Maybe it’s to establish a castle or to show a creature emerging from a box. Whatever the reason it’s important to keep that in mind throughout the process. Given the length of time from the initial design to the completed final, the shot can veer off course quite a bit.

At the time of shooting someone may have a ‘better’ idea. Why don’t we frame it like this? Why are we wasting all of that frame area there? This is when the visual effects supervisor has to remind them that the empty area on the side will hold a creature in the final shot. Another reason why storyboards are essential is to provide a clear visual of the final shot. More likely it will be a subtle change that will have a big impact later. (Let’s put this prop here, lets add a real explosion in the foreground.)

In post production the editor may want to reframe the shot or use a different element entirely. The compositor may put in more smoke in the foreground. Everyone involved in the shot (director, supervisor, animator, technical director, etc) are likely to be focused on the details and lose sight of the purpose of the shot. In an attempt to make the shot even ‘cooler’ you lose the focus of the shot. It’s only when it’s cut in will the real problem become obvious again. The reason for the shot may now be so obscured that the audience will be confused and lost. If that’s the case it throws them out of the movie. Try to always review the shot in context and take a step back to check the intent of the shot.

Does the shot fit in the movie? Does it fit into the sequence?
Unless it’s a specific dream sequence, most visual effects shots are supposed to blend into the rest of the film. This is true whether it’s a period piece or a science fiction future thriller. The design of the shots, the camera motion and the lighting should match the live action. If you have a hand held action sequence and cut to a locked off visual effects shot, then it will stand out.

My suggestion to directors is to design the shots as if everything is really there. How would you frame and shoot this in live action? There’s a tendency to treat the design of even simple visual effects as different than the rest of the film. “We’re paying for the shot and by gosh we’re going to show it off” is sometimes the approach taken. If it’s a real building they might frame it from a ¾ angle and not make a big deal of it. If it’s a matte painted building then it’s likely to be designed to be shot straight on with clouds added to the sky. All of those are clues to the audience that something about the shot isn’t right.

It’s possible for a disconnect to happen since the director usually sits down with a storyboard artist months before shooting. These shot designs may be a different aesthetic than how the director of photography approaches the live action. The director is involved with both teams but there are thousands of choices to be made that may place them out of sync. There may be times a second unit director is approaching the shots differently than the main director. Sometimes in post the director realizes he can change a lot, especially on a virtual shot Focusing on a hand full of shots may cause them to shift away from the rest of the film.

Adjust the design of the shots based on their context and what they’re supposed to accomplish. If they’re supposed to be realistic backgrounds then all the more reason to fit them into the rest of the movie and avoid drawing attention to them. Once again, how would you treat this if it really existed? If it’s a dramatic effect then design the shot to take advantage of that and push it within context of the film.

(more design posts to come)

VES & Gnomon - Golden Compass Breakdown

Gnomon School Of Visual Effects And The VES
Welcomes The Visual Effects Team From The Academy Award Nominated, The Golden Compass

Feb 28, 2008 in Los Angeles at the Gnomon School

Academy Award nominated Michael Fink, Senior Visual Effects Supervisor, Susan Macleod, Visual Effects Producer, Bryan Grill, Visual Effects Supervisor for Digital Domain and Raymond Chen, Co-visual Effects Supervisor for Rhythm & Hues Studios will be will be discussing the planning and execution of many of the visual effects shots for The Golden Compass.

Click above on name of post for more info.

Depth of Field and VFX

Depth of Field and VFX

This is in response to a question from a reader about depth of field and how it relates to visual effects.

Models
With miniatures you need enough depth of field to hold focus from the front of the model to the back of the model. This means a lot of light and a stopped down lens. Lack of depth of field is one of the key things that give away the look of the model since in real life a large area would be photographed from a further distance and (possibly in sunlight) so depth of field wouldn't be a problem.

Greenscreen/Bluescreen
If you're shooting a greenscreen then you'll typically want to make sure all the foreground people and objects are in sharp focus. This is easily and frequently overlooked, especially by directors of photography. (And at times by the VFX supervisor) If you have a sunlit, exterior background image then it's likely to be shot at f8-f16. This provides a reasonable amount of depth of field. But when the DP lights the stage he probably won't be lighting to same intensity levels. If the lack of depth of field is apparent then it causes two problems. If the back end of the foreground is soft then you're forced to blur the background even if there's something important to see. It's impossible in real life to have only a mid-section that's blurred. If the director really needs to see what's back there then you're forced to try to sharpen the edges and back detail of the foreground. The end result will never appears natural. The other problem is the audience senses that this is unnatural (i.e. they only see this in vfx shots). An 'exterior' scene in bright sunlight with a normal or wide angle lens should be in focus throughout a normal shot.
You'll see animated films where they have cheated the depth of field. I find it best if you want your vfx or animation to appear natural is to use the guidelines and restrictions that a normal movie has. (depth of field, camera movement, etc)

Markers
The other depth of field issues for VFX people is soft tracking markers and very soft edge mattes. If you're on a stage shooting a greenscreen with a long lens then the markers may be so out of focus as to disappear. This is a big problem, especially if you're shooting a character from the waist up who's moving and jumping around. Since the markers are invisible, you have no easy way of distinguishing the camera motion from the character motion. Someone will have to manually work on that shot by eye until it looks reasonable. This can be very time consuming and require a number of takes.
Note that LED markers tend to hold up better for out of focus shots. These are markers using key ring lites (possibly modified) you see in the store. The point source of a red LED holds up better than an X piece of tape.

Mattes
The soft edges of the greenscreen matte (or a place you want the roto matte) will require delicate settings of the key. Any blur (from depth of field or motion) will cause some of the background to bleed through that area as if it were partially transparent. The blurred area then becomes more contaminated with the greenscreen. When rotoing a blurred edge it's sometimes a subjective question where the blur stops. If you include all of the blur then you'll be including some of the original background. If you clip off too much of the blur then it will look incorrect in the composite unless you blur the edges of the matte similar to the original. Fortunately software like Commotion could deal with natural motion blur so it was less of an issue. If you shot someone slightly soft in front of bright points of light (city, Las Vegas, etc) then in the edges of the blur you would have points of light that likely aren't in the new background. In these cases you may have to clamp down or paint out the offending lights.

VFX Schedules

More reader questions regarding schedules

What is a 'Shot Timeline'...is this some sort of
device used to tell the amount of time the shot should
last for or what. You also made mention of some types
of, magnetic boards, modified storyboards or computer
software which is most effective.

Scheduling is a big issue since you have a lot of resources and need to deliver in a timely manner.
There are at least 4 types of schedules. They can be corkboards, on the computer, magnet strips, etc. The specifics are
up the vfx production team (producer and supervisor)

1. Schedule of the shots. This is the true linear time estimate and is usually in a timeline. That timeline will be the length of the production.
It may list that shot KR030 starts June 14 and completes August 19. The shot itself is likely to be budgeted for less time but you have to take some delays into account. (changes from the director, waiting for feedback, shooting an extra element, etc)
Each shot laid out may also show the different stages (animation, TD, roto, etc)

2. Target shots - These are shots due in the next couple of weeks. These may just be names of the shots or their storyboards on a simple wall chart that's broken down by days.
The production team can review this and say next week they expect to complete KR040 on Wednesday. If shot RM125 won't be done this Thursday as planned then it will be moved to the next week at the likely day. This is so the team can focus on the immediate needs.

3. Storyboards - This is all the storyboards laid out in film order. Each has a breakdown of the different tasks and the initial scheduled dates. As work is done a colored dot may be applied that indicates which stage is done. (i.e. finished with matchmoving and layout, ready for animation) Red dot signifies a completed (finaled) shot or folding up the corner of the storyboarded breakdown. The storyboards may be replaced or augmented by stills from the actual footage. Most productions attach bulletin boards in the production office for this. This allows everyone to see the big picture. You get a true sense for what's been done, how much work remains and if there are some shots that are being overlooked.

4. Schedule of artists - Your key resources are the artists. There's a timeline with each artist (TD, animator, compositor, etc) that lists what shots they're scheduled to work on and when. An artist may be working on 1 to 5 shots at a time. After completing a shot the next shot for them is already scheduled. If a director adds a additional shots or something changes in the schedule then the production team will review this board/timeline to see who's available and what they should re-assign. If they need a shot sooner than expected it may have to be moved forward in the schedule and given to a different artist.


How are 'final shots' determined. If the feature
film has about 25 effect shots how can I conclude that
I am expected to get 5 final shots a day for 5 days.
Also, what happens if the finals isn't what the
director had in mind...will the shot be repeated and
isn't this waste of time.


A shot may be internally finaled by the supervisor but it's not truly finaled until approved by the director.
It's important to understand the director was involved in all stages of the shot (from the original design, the shoot and now post production) The director will have seen the shot tests from animation and at least preliminary renders and composites. This has to be done so they can cut it into the film and judge it context. Changes can occur at any point. If the director doesn't final the shot when it's expected to final it's usually because the final polish hasn't been done to their liking. Note that even after a director has finaled a shot it could still be unfinaled at some point later (studio hates it, new concept). In these cases that's a major change order (time and money).

Time - If you have 25 shots due in 5 days then you have to final an average of 5 shots a day. To calculate this you take the the number of shots left to do and divide by the remaining time you have (assuming 5 day week that the director can approve them). This will give you the average number of shots per day. You can just as easily calculate number of shots per week or average days per shot. The initial time is based on the date when you have the turnover of the shots (when they've been edited and production tells you these are the takes, shot numbers and details).
You're likely to start off woefully less than the average at the start of post production since you have to fill the pipeline and it can take some time to get the film scanned, cleaned, matchmoved and ready to work on. The number of shots actually finaled rises exponentially as you get closer to your finals date (contractual day you have to complete so they can get it in theaters). More things will have been worked out, the crew has hit their stride, the director and supervisor are now seeing through new eyes (200 shots in the next two week, akkkk!), you're waking up in a cold sweat at night and hopefully the studio has stopped fiddling with the shots.

Related:
Time's a wasting

VFX management

Here are a few reader questions. As I’ve said before each company and each production is a bit different. There are no standards of operation so there will exceptions to everything below.


How much management is needed for a small, medium and big sized operation?

In a small shop, management also works hands on. We had 6 people when we started Dream Quest in a garage, all of us co-owners. The amount of management of course varied per person. I was the president and would make sales calls as well and vfx supervise. We had over 60 employees when I left Dream Quest. ILM had about 120 at that time. (1985) Later ILM had more than 1200 people working at one time.

As a company gets larger you start getting more support people and employees. It also becomes more difficult to balance managing and working hands on shots. Legal paperwork, payroll, computer support, coordinators, etc. start requiring a management structure in place.

Small companies are more likely to have people who are multi-purpose. The individuals may do everything on the shot from start to finish. As a company gets larger you’re more likely to go to specialists for each craft. Typically you’ll have multiple departments or groups of people as you get larger ( Technical directors, compositors, roto artists, animators, etc.)
If the size of the company is only working on one project at a time then you may have a lead per department that balances working hands on as well as managing the people within his group. As the company gets larger and works on multiple projects at a time (a large company may be working on as many as 6 to 8 projects at once) then a department head will be assigned to each department. This person may also be a lead on a specific project or may strictly be a manager with no hands on activity.

Note that most companies consider supervisors and producers managers in addition to department heads. An added benefit for the company is they don’t have to pay overtime to any ‘managers’.

The greater the size of management the more overhead the company has to add to the budget. It’s easy for companies to become too top heavy with management (in some cases several levels deep from the company headto the artists). Since management controls employment they’re much more likely to layoff the actual artists than management. Most of vfx management is made up of people who were once vfx artists or vfx producers themselves. Not everyone who’s a good vfx artist makes the transition to manager since it’s a different skill set. Unfortunately some people are promoted to a job they’re unable to do well. This can be a real problem.


How many leads?

Number of leads is dependent on size of the show. If it’s a large show with big sequences you may have a sequence lead for each sequence. You may also have a lead for each discipline. An animation lead for each sequence or for each main character, a compositing lead, a TD lead, etc. A lead may be assigned a different sequence after they complete their first one assuming they’re not concurrent.


How do you balance between creativity and the budget?

A few key things to note here:
Creativity isn’t directly proportional to budget. We’ve all seen very expensive movies with little creativity and visa versa.
Budget does provide: More R&D for new vfx, more concept work, more shots and/or more complex shots, more time and effort to finesse the shots.

The director controls the creativity and the visual effects team serves the director. Some critics and internet users think the vfx team does it’s own thing and just delivers it at the end as if the director has no involvement. The director is very much involved in all designs, all the shooting and all the post. The only time this doesn’t happen is if the project is over-schedule and/or over-budget (or if it’s with a specific, nameless studio where the studio executives control all the vfx) The other case is when the director turns over all the action design and execution to his 2nd Unit Director. Since the 2nd unit director usually isn’t involved in post production this can be a problem.

In commercials and television work the director usually isn’t involved in post production. It’s in the hands of the creatives at the advertising agency for commercials and with the producers/writers for television.

From a VFX standpoint we work with director in pre-production to create concept art for what the final shots will look like and what the creatures/objects look like. We also try to be heavily involved in the storyboards and previs work. Many directors are very eager to get the most out their vfx and vfx team and this works very well. They’re open to new ideas and the vfx team is more than hapy to help. In other cases you can provide a totally new concept or idea that would be a perfect fit with the movie but it’s ignored.

My suggestion is to initially design as if the budget didn’t matter. Brainstorm working with the director and come up with the most powerful shots for the movie. If the budget doesn’t support that then the director will have to reduce the number of shots, ask the studio for more money (which the concept art may allow them to do) or be willing to simplify the shots.


Who is responsible for what?

Each vfx artist is responsible for the specific work he’s been assigned on a shot or model.


What are their roles specifically?

There’s an endless list of jobs and job descriptions. Here are some of the common ones: Technical director (lighting and rendering of 3D), Compositing (combining multiple images), animator (animation of a character or object), roto (someone who traces to creates mattes), painter (painting out unwanted items in frame, fixing frames), 3D modeler (builds the model), texture painter (someone who paints the 3D models), model maker (builds physical models), rigger (builds the 3D skeleton for the characters), skinner/enveloper (works on the skin of the characters (flexibility)), dirt removal (paints out scanned dirt on images), layout/matchmove (creates 3D representation of the live action), particle animator (works specifically with partical systems), previs artist (creates simplified animation before production), motion capture actor (creates moves the animators can use for a character), motion cature artists (work with the data from motion capture), set surveyor (record information when shooting), coordinator (gathers and disperse information, help with schedules), Production assistants (anything)


Does the vfx supervisor worry about the creative only?

No. We worry about everything. Supervisors are always very involved technically and at the end of the day they have to be worried about the budget and schedule even if these are directly overseen by the VFX producer. If you run out of time or money because of previous choices then you won’t be able to complete the project.


Is the supervisor responsible for managing his crew directly?

The Supervisor reviews the dailies of all the TD’s and Compositors and provides both creative and technical feedback. An Animation Supervisor reviews the animation dailies. These artists consult with their leads to discuss details or solutions. The supervisor provides the creative guidelines for the artists (based on the directors vision) and deal with the large issues. The supervisor may only be able to interact with a specific artist once or twice a day (such as dailies). This is because there are a lot of artists and there may be many meetings. The leads have less people under them so are more likely to check in on all their artists more frequently.


Does the supervisor have a say about workflow and how things should be done technically - or he/she responsible solely for the creative side of effect?

It’s all a question of details. Normally the supervisor oversees the basic technical aspects of the shots but the specific settings and details are guided by the CG supervisor or leads. The supervisor is usually the one to define the basic approach to a sequence or the shots. (i.e. matte painting or model, greenscreen or CG, etc) Whether to use a specific plugin or version of software is up to the department, lead or the artist.


In solving problems and making decisions, how does the crew structure help?

When bidding the supervisor meets with his leads or department heads and discusses his proposals. If there’s a better solution or alternatives those are discussed. During post production the artist works out the details of a solution themselves. They decide to use another mist element to blend on top to give some depth to a shot. If there’s a problem with a roto then the compositor or TD talks to the roto artist. If they’re having difficulty getting the look correct then they’ll check with their lead or a fellow artist. If that’s unable to resolve the issue then it’s brought to the supervisor.


What are the limits for the crew members themselves, their responsibilities?

The crew member is responsible for taking the elements provided and completing their aspect of the shot. They will make adjustments themselves based on what looks correct as well as feedback from dailies. If there’s a serious subjective or creative decision they’ll call on the lead or the vfx supervisor to make a decision. They can also opt to do it the way they think is right and review in dailies. If there is technical problem they may check in with the lead first.


Can you talk more about production vs. creative process?

Part of the issue is when is a shot is done? From a creative standpoint you could tweak a shot for months to make slight improvements. From a production side you want it to be complete and approved by the director as quickly as possible. What if there’s a better idea halfway through completing a sequence? Is there time to do it? Is there money to do? Filmmaking at some level is always a compromise.


What about chain of command?

Top level: VFX Supervisor (creative and technical), Animation Supervisor (animation), VFX Producer (schedule and budget)

CG Supervisor (big picture view of the computer resources required and how to achieve the different looks on the computer)
Leads (specific to a sequence or task, oversees the people working in that area and helps to mentor the artists)
Artists (These are the people doing the actual hands on work)

Related links:
VFX Producer
VFX Supervisor
Visual Effects Positions
Bad Visual Effects Business Practices

Close Encounters Book

There's a new book on Close Encounters of the Third Kind. This was the first film I worked on and the author covers the behind the scenes, including the visual effects. I still have to get around to covering some of the vfx on this blog, including the cloud tank but he's got most of it covered.



[My cloud tank effect posting describes how we created the clouds in Close Encounters]

Close Encounters VFX Video 1
Close Encounters VFX Video 2

VFX Supervisor article in Variety

Click on the post title to go to an article in Variety about visual effects supervisors. Some productions are catching on.

Here's the link if you have trouble:
http://www.variety.com/awardcentral_article/VR1117977459.html?nav=news&categoryid=1985&cs=1
link

Searching for VFX supervisor may bring you to this page but a more in-depth article is here.

Visual Effects in India

NPR did a story on vfx animation in India last week. Click on the Post Title above to go to the article or use the link below.
http://www.npr.org/templates/story/story.php?storyId=17103636

There's text and a 7 1/2 minute audio interview primarily focused on Rhythm & Hues's group in Mumbai.

Audio Reminder

This is just a reminder for those who subscribe to the audio feed to check out the latest articles posted here.
http://www.effectscorner.com is the base address that links to this blog.

I have a few other articles in progress as well as outlines for some video blogs. We'll see how much free time I have available.
I also sould mention I've made a few additions and changes to the links on the lower right and added a few items to the stores for crew gear.

Comparisons

When you go to an eye doctor he asks you to compare two different lens choices. “Is this better or this better?” Step by step he refines the specific lenses that you need. The visual effects artist goes through similar comparisons to arrive at the final shot. They compare their shot to real references, they compare the shot to surrounding shots and they compare of changes while they work on a shot.

If you don’t have any reference to compare you may drift off course and end up with something that doesn’t work

Real References
During each step in the process the VFX artist should be comparing their work with any real references from the original shoot or research references (stills, video). What needs to be done to match this reference? Does the creature move like the real creature reference? Does the lighting match reference photos?

Sequences
You also want to be able to reference your shot in comparison to others in the sequence. Cut the shot into the sequence and view in context. Does it match the other VFX shots? Does it match the live action shots?

Color
For color balancing, film clips are sometimes filmed out as a wedge. These show a range of color and brightness values and will be used for digital color balance reference. The DP or supervisor may make a selection to use as a guide for a sequence. This is similar to doing color variations in Photoshop. What looks good by itself may not look like the best choice when you can compare it to other variations.

Within the Shot
In the pre-digital days the visual effects artist would create wedges and shoot film tests. These might be checking settings, changes, exposure, and focus or animation tests. It’s still done for miniatures and other photographic effects.

With digital effects you have the advantage of saving multiple versions, undoing/redoing and seeing the results instantly in many cases. The VFX artist takes advantage of this by experimenting and refining. If you add a filter or element you can toggle it on and off to see the result even on a single frame. This would be like the Preview button in the filter dialog within Photoshop or the layer visibility. You also have the option to Undo/Redo to compare any change you just made.

Depending on the software you can load in a previous version or take a snapshot and do the comparison. Some software allows you to do a split-screen to compare 2 versions of an image within one image.

All of this allows the VFX artist to refine their work and make choices.

Changes
When a director or supervisor asks for changes it’s important to make large enough changes so it’s evident looking at the shot. Many artists will make minimal changes and slowly build up to the desired look, take after take. Unfortunately this wastes quite a bit of time. Comparisons are good for you but if it’s not possible for the director to tell the difference without seeing them side by side then it’s not a large enough change.

It’s best to make large steps, which ideally includes going too far. If you had to blur something instead of going by single pixel increments for ten images, it would be better to go by 10 pixel increments. By coming up with an image that goes too far (this could be color, speed of animation. filter, etc), it will allow you to know the range to work in and get a better idea of what the director wants.

A typical phrase in VFX is to “split the difference”. (ILM even had a comic poster of this). This is likely when you’ve gone too far but the previous version or another test didn’t go far enough. In this case split the difference is a way of balancing those two. From that result you might need to split the difference yet again. This is actually a fast way to hone in on the desired look and uses the same algorithm as some computer sorting routines. As you proceed with these adjustments you’ll be comparing the previous versions. If you get to the point where you can’t see the difference without doing a split then you’ve hit the point of diminishing returns.

LA Shows - Matte Painting, Illustrations

The Academy is having a couple of events in the next 2 or 3 weeks of interest.

Motion Picture Illustrators exhibit in their gallery until Dec 16.

Matte Painting show by Craig Barron Dec 10 at the Linwood Dunn Theater. $5 There will also be exhibits in that lobby.



Click on links for more details.

Mac Widget for Effects Corner

Click the link of the title and download a Widget for the Mac to display the highlights of the last 10 postings here. Click on the article title in the widget and it takes you to the page.

Photo Real and Realism in Visual Effects

At the start of most projects every director requests their shots be photo real

What does Photo Real mean? Are realistic vfx shots a lot more difficult than fantasy shots?

Photo Real means to create a shot that looks as real or true to life as it possibly can. It can mean that the visual effect is so totally convincing that the audience doesn’t see it. Another term for this is Invisible effects.

The lighting, textures, detail and compositing try to mimic a scene that the audience will think is real. To do this the visual effects supervisor shoots as much reference as possible when shooting the plates (live action footage that will be augmented or modified with visual effects). This may mean filmining a physical prop in the same lighting and setup. The clock in The Mask and a piece of fur for Van Helsing were photographed for the TDs (Technical Directors) to use as a look and lighting reference.

As much information is recorded at the time of photography as possible. This includes lighting diagrams, measurements of the set, camera and lens information as well as filming of the chrome and gray spheres.

All the vfx artists that work on the shot will have access to that material and be able to use it directly as well as to use it as a comparison. From this they try to create and integrate the images as much as possible.
The visual effects animator may film or obtain reference footage of people or animals moving to use as a guide, even if it’s a fictional creature.

The trouble is you may not even be creating a shot that will ever be believed.
No matter how well the animator, TD and compositor do their job if the script calls for a pink feathered whale in the sky it still won’t be considered Photo Real by the audience. There’s nothing wrong with creating shots like this since that’s the requirements for the film. The vfx artists try to add as much of real life as they can into shots like this to give it a more solid foundation. This may mean moving the whale slowly and adding in a haze layer to set the scale.

This leads to our second question regarding the difficulty of creating realistic effects.

There’s the impression that realistic effects are very difficult and much harder than shots dealing with fantasy or science fiction. Yes, realistic effects are difficult when you’re trying to create something very complex such as a held close-up of a human, moving cloth or flowing hair. But the difficulties of a shot are usually more dependent on the specific shot and less on the context of the content.

Let’s consider an effects shot: a man on crutches coming toward camera and is missing one leg. The audience will easily accept the man missing part of his leg, especially if it’s an unknown actor. The audience knows this could be real and doesn’t require a suspension of disbelief. Now consider the same shot but instead of missing the leg this man is missing his head. The headless man however doesn’t exist in real life so the audience instantly knows it not real. It will be in the back of their minds no matter how well the vfx are done. A large part of the reality of a shot is based on the perception from the viewer. From a difficulty level these are similar and use the same techniques. The headless man is probably more difficult because you have to create and track the inside of the collar.

Suspension of disbelief plays a large role in film. To some degree everything is a bit unreal in film. The basic story is a fabrication and the dialogue is hand crafted. The director of photography doesn’t necessarily match real life. He lights it to go for a specific style and to make it cinematic. Note that this can be a real problem when you’re trying to match greenscreen people with real outdoor backgrounds). Stunt people rig ramps to make cars spiral in mid-air. Not necessarily real, but certainly visually exciting (cinematic). This is the same thing with the winged spaceships and hearing explosions in the Star Wars films. Not real but cinematic.

Hopefully the story will keep the audience engaged and there will be nothing to force the audience to fixate on the effect. Anytime you give the audience a reason to suspect something, they will find it. You could have a real shot and if the audience thinks something has been added they’ll happily point out several things that are wrong with the shot. A real shot can seem fake under the certain circumstances.

People think since something exists and they know what it looks like they could certainly judge the quality. The reality is most of these effects when correctly deigned pass by audience members unless they’re very poorly done or there’s something to arouse their suspicion.

If you have a matte painting of a stylized or haunted house and center it in frame as the only thing in the shot then it’s going to be suspect. If you add a matte painting of a normal house to fill in a vacant lot on a street and then have the actors in the foreground the audience is unlikely to think about the matte painted house, especially if it doesn’t play a promenant role. Most people think a matte painting has to be super detailed but the primary issue with matte paintings is to get the lighting and perspective right.

The advantage of creating something real is you have reference of the real object or creature to constantly compare to while working on the shot. It may take a lot work to get your CG model or other items to match the real thing but you always know how close your are and where it falls short.

With imaginary shots there are frequently doubts and changes to the design since you don’t have anything to compare to. Some people think the creature should move it’s arms in one way and another group thinks it should move them in a different way. The director may switch his/her thoughts as well. A real reference gives everyone something to lock into.

If you recreate one of the NASA rocket shots now you have reference to the original material and people will accept it. They may know that you created it but it won’t remove them from the movie. If you showed the same footage 50 years ago people wouldn’t accept it as real since they had nothing to relate to.

Old movies had shots done on stage sets that were supposed to be outside. People in cars were placed in front of a rear projection screens. By today’s standards those shots don’t hold up as well because we have a different level of realism in films. It’s not that people at that time thought they were real, it’s that they accepted it much as a theater audience accepts a stage play in front of single sheet sets.

This also applies to camera moves. If you move around a model like a helicopter then the audience will accept it more than if it’s totally static from 1000 feet up or if it moves a mile in 2 seconds. The director may have wanted a photo real effect but in the effort to spice it up by moving the camera faster and further than it could in real life, you destroy the illusion of reality you had created. If you zoom out to space and then back down thousands of miles away like on Google earth then that a style decision but not one that will help the feeling of realism.

The ‘though the engine’ shots in Fast and Furious or the slit-scan shots from 2001 are pure stylized shots. They may have had a high tolerance as to what they looked like but they would still require a fair bit of effort to make work.

Some directors want to try to convince the audience that something is real by focusing on the effect and doing what they can to show it off. In the headless man shot they want the camera to fly around the man and then down through the collar to prove that there’s nothing there. This is like the magician that moves the hoops over the levitated assistant floating over the floor. But the difference is most of the time the visual effects are used to tell the story. By trying to convince the audience the shot may actually come across as more fake.

In summary, the vfx artist tries to make every shot as real as possible. In some cases that may not be possible due to the subject matter itself and in other cases may just be a style choice. Creating invisible effects is usually more dependent on the subject matter and the design of the shot than the execution difficulty.

Visual Effects Society

Just a reminder to check out the Visual Effects Society (VES) sometime. If you're a professional with 5 years or more experience then you can qualify for membership. They have member inductions twice a year. There are now additional sections in London, Vancouver and San Francisco area.

Students and others interested in visual effects might want to check for events in your area. There are presentations and other events during the year.

If you're a student in the L.A. area then check into the mentoring program. I and a few others spoke to a group of students last weekend.

There's also the VES award submission coming up in 2 or 3 weeks. More details here: http://vesawards.com/

Budgeting VFX

Budgeting VFX

I covered some of the basics of this in the post/podcast Bidding and Preproduction
I’ll be going into more detail in this posting.

As previously discussed budgeting visual effects can be very difficult. You have the possible issues of developing new looks and techniques. How long will this take and how many people? Add to this mix the director’s creative changes that happen on any film project. Should you calculate 2 takes? 5 takes?

I’ll be discussing feature films here. Budgeting is somewhat the same for television or commercials with a few key differences. Television shows may have a small permanent staff (or at least an assigned vfx company) if they’re vfx heavy. They also tend to have very little time or budget. There’s usually no time for storyboards in television. In many cases they’re told what has been budgeted for the effects and have to work within that. On a commercial much of the final work may be done on a Flame™ or similar system. These are expensive high-end systems where the client sits in the room and guides the artists to achieve the effect they want. Note that feature films use Flames and other equipment as needed where the vfx supervisor is the client. For high-end 3D work it may go through a similar pipeline as a feature film but with a smaller team. The budget and post time for visual effects on a commercial can be much higher on a per second basis than a feature film since a commercial is only 30-60 seconds in length. Commercials have detailed storyboards and may have previs because they use it as a selling tool to the final client. Both commercials and television used to be primarily done at standard television resolution (which made it a bit faster and easier than feature films) but with the advent of High Definition they’re now done at resolution at or close to film resolution.

The average Hollywood feature film is now approximately $100 million. A VFX film can run even higher. For a ‘visual effects’ feature film the vfx budget can be 1/5 to ½ of the total film budget so it’s critical to get the budget done correctly.

Why are visual effects so expensive? VFX are very time and labor intensive. A large project may take 200 people to work on over the course of several months or a year.

The first step in the process is someone at the studio plugs in a guestimate for the VFX. Years ago a producer with minimal vfx experience might pencil in a budget that was an order of magnitude off. These days most of the large studios have a VFX head in charge of budgeting and assigning the work.

Initially the script is broken down by the VFX producer on the film (likely with a supervisor involved) or it may be done by a VFX company’s producer and supervisor. A rough count of the number of estimated vfx shots is created based on the script text only. Note that even if production has done scene numbering these don’t break down into the specific shots, which is what’s required to budget the vfx. Shots may be designated as hard, medium and easy to allow setting an approximate cost for each shot or as a quicker process an average cost may be assigned to all shots. A studio film usually has an average shot cost in the $10,000-60,000 range. Low budget projects may have $1000 or less per shot.

These ballpark estimates are submitted to the studio to evaluate. The VFX company may meet with the director and the team to refine the bid further by dropping potential vfx shots that they expect to do with stunts or clarifying some of the planned shots.

During the pre-production phase the director works with artists to create storyboards and/or previs animation. These resources will be used to create more accurate bids. The production or studio VFX head will submit a package to any VFX companies bidding on the work along with parameters. Parameters might include average length of shot (typically 5-8 seconds), handles (4-16 frames on each end of the editor’s cut shot to allow some adjustment), delivery dates, delivery formats (film, digital, both), temp screening dates, etc. Each shot may also include specifics or assumptions. Due to times constraints sequence bids may be worked on as soon as storyboards for that sequence are done.

In some cases the studio employs a company or multiple companies to do some concept work and/or some R&D. The studio may require this before proceeding to the next level and to make sure the look and approach is what they want.

With storyboards and information in hand the visual effects supervisor determines the best methodology. He/she works with the vfx producer and the team (lead compositor, lead TD, CG supervisor, etc) to review the approach. If it’s an animation heavy production there’s likely an animation supervisor as well. Each shot is then broken down into detail regarding the amount of time required for each task. Ideally this involves all key personnel but could also be done by the producer and supervisor if they feel they have an accurate idea.

The animation supervisor may review the shot and determine it will take 8 days for an animator. The CG supervisor may think it will take 5 days for a technical director (TD) to light the scene and work out the look. Estimates are made for composting, rotoscoping, paint work, specialized modeling, matte painting, simulations, dirt removal, etc. Each of these goes into an Excel spreadsheet the vfx producer has. Modeling (CG or practical) is budgeted (and required texture maps or paint work) along with R&D. When you’re doing hundreds of shots this process can take days.

All of these estimates are based on actual experience and gut feelings. The complexity and difficulty of the shot is taken into account. A company that’s been operating for some time should have averages from other shows to refer to (although they seldom do refer to them). Of course all of these estimates have to take into account some changes and doing multiple takes. Some companies bid the lowest possible number (as if everything were perfect and there would be no changes) to get the show and then force production to have change orders for every change, big or small. This just creates a very awkward and painful process for all involved. It’s always best to consider some time for reasonable adjustments or changes. If the VFX team has worked with a director or given studio before they make some adjustments to the estimates based on number of changes on the last production.

The biggest danger in creating the estimates is being too optimistic. Your first impulse is to say that it can’t possibly take that long to do that shot. It’s just a simple blue screen, etc. This is especially true since the person making the estimates is likely an expert and can do the shot quickly. The reality is there are no simple shots, the original plates won’t be as good as you’d hope, there will be changes, etc. You also have to take into account the average person working in that position at that company. The range of quality and speed between employees can be vast. One person might be able to do something in a day and another will take a week.

In some cases the vfx producer and supervisor may modify or pad the budgeted shot. This can be tricky and is best to be discussed at the time of the group budgeting. If the producer feels a certain task is under bid then that should be brought up. The danger is if the estimates are modified after the fact then the person who provided the initial bid may be held responsible for it, even though it was changed for more or less afterwards. If you’re bidding a show I suggest always keeping your own notes if there a discrepancy months down the road. The other issue is if the producer pads the bid and the people originally providing the estimate are padding their estimates you end up with a double padded bid. This may mean you lose the project.

When planning a show the producer and supervisor try to cast the show much as a director casts actors. Who are the best people available for the different tasks and different leads? Is there an animator that would be perfect for a specific character or a modeler that is great at the organic modeling required? Unfortunately schedules and any other shows may prevent the flexibility of having a choice in all of this. It’s not unusually for modelers to be tied up on another company project. This delay in getting the models started ripples through the entire production causing more overtime. All of this needs to be taken into account internally at the company.

The vfx producer calculates the cost of the shot based on the time estimates and the average cost for that type of person at that company. That rate will likely include not only the employee’s salary but also their related expenses such as health insurance and pension. Note that most vfx artist work more than 8 hour days so the overtime has to be taken into account for the estimated. Some companies include in these time estimates the entire overhead and profit margin. In some cases they may fold in the R&D costs or the model costs. If a 10 shot sequence has a $100,000 model budget then they will add $10,000 to each shot in the sequence.

The cleanest and best process though is to keep these costs separate. If you include a lot of extra costs in the time budget then things get very wacky as changes are made. In the example above if the 10 shot sequence is cut down to 1 shot then you only have 1/10 the model costs. If the sequence is dropped before it’s started but after the model is built then you have to find out how you can cover the price of the model somewhere else. If a 100 shots are added that doesn’t necessarily mean all your overhead goes up dramatically. Likewise if production requires a lot of overtime work at the end of the production then those calculations will be out of line.

Overhead includes all the labor and materials not directly related to a shot or that can specially broken down into shots. The basic staff of production assistants, coordinators, vfx producer, vfx supervisor, etc. are all part of the overhead. The cost of the computers, software, sales people, human resources etc. need to be included as well. The time the vfx artist spend in meetings or general prepping (getting the pipeline worked out) also fall under this category. The amount of down time when the company keeps people on payroll without a project has to be incorporated as well.

An estimate is also calculated for plate photography which will have a vfx supervisor and possibly a small team of matchmovers or coordinators.

A percentage will be added to the project to cover profits. Contrary to studio belief the markup for visual effects is not astronomical. (i.e. not as much as most retail)

If you’re a freelancer or a new, small shop, try to do the work on a time and materials basis if possible. You’ll have to provide a basic estimate but if the director or the vfx company that’s contracting you changes anything you won’t be forced to cover the change costs yourself. First time freelancers make the mistake of charging what they were being paid working for another company. The problem is none of their overhead is included such as health insurance, computer and software expenses. You’re also assuming you’ll be working full time which won’t be the case for most freelancers.

The vfx producer has to lay out the linear time required to do the work. If you have 15 TD’s and it will take 20 weeks of work but production only has 12 weeks of post then there has to be some adjustment. Either more people need to be brought on, more overtime is added in or the number of shots the company can do must be limited. As mentioned if there are other productions at the company the resource allocation (for both people and computers) can be a real problem if a production adds or drops a 100 or 200 shots on a project (not that unusual unfortunately).

Visual Effects are usually feast or famine. Either there are too many projects and a company will have to turn down work or there’s not enough work and the company has to do what they can to get the project. In some cases the company may choose to bid the project at their cost (no profit) or even as a loss. The smarter companies know that if they have to lay people off it will cost them a lot to rehire people and bring them up to speed with their internal systems.

In the end the vfx company provides a cost for each shot (not broken into each task) along with an overhead budget, model budget and R&D budget. Any assumptions the bidding team made should be spelled out clearly for each sequence and each shot. (The car will be a stunt car, the fire will be provided by the onset special effects crew, etc) In commercials there are standardized bidding forms but there’s no standard in visual effects for features. The studio will likely add up all the costs and divide by the number of shots to get an average shot cost. They can use this as a rough comparison with other bids as well as to help ballpark additions or deletions of shots. Of course the studio is likely to ask the vfx company to sharpen their pencils and provide ‘better’ numbers and they may ask the director to cut shots or elimatinate a sequence based on budget considerations.

Do not be surprised if you’re asked to do this process a dozen times by the studio. Storyboards and concepts change which require a re-bid. Bidding occupies a large amount of the pre-production time.

Multiple vfx companies bid on the work. The vfx producer for the show may choose specific companies to bid on specific sequences. If a company has a specialty (such as matte painting) they may only bid on those shots. The studio may have a list of companies they’re willing to work with that the vfx producer has to use.

After the bidding has settled down the studio will award the show to a company or companies. In some cases the studio makes the decision based purely on cost and in other cases the cost only plays a small role in the selection. The quality of work, past experiences of the studio and the ability to accommodate many more shots are all taken into account.

At this point the studio and the companies work out their contracts. In some cases the contracts may not be official until the end of production but of course the best approach is to get it locked into before production. Note that the contract may lock in a specific bid even though additional changes and storyboards arrive after that time. It’s critical to have an experienced entertainment lawyer review the contract since the studios have teams of lawyers who specialize in this. You need to make sure everyone is on the same page with regard to the assumptions and delivery schedules. You also need to be clear about the payment schedule. Some are done by weekly allotment and others are done when hitting milestones (per shot or by major sequences)

During production the supervisor and producer flag the studio when changes or additions are made that might affect the budget. Note that the movie release date will not change so the only way to handle additions is to add more people, more overtime or additional companies.

Once the film is shot and there is a rough edit that can be turned over, the vfx company will likely review the bid and compare it to what they actually have in hand. Quite a few things can change from the plan to the final result and this is a final reality check before the work begins.

Hand in hand with the budget is the schedule. A preliminary schedule is made when the budgets are done. There is usually a schedule for turnovers. This is where production (director and editor) gives an edited sequence to the vfx company. If the turnovers happen later than planned (a frequent occurrence) then the entire schedule and budget may be affected. Time is money as they say. Also the finals date (when all shots are supposed to be done) may vary a bit but it’s very unusual for a release date for a movie to change. If it does change, it may be for the worse (i.e. earlier)

In scheduling the work each step of the process is taken into account. The vfx editor will have to provide negative numbers, the film needs to be scanned, then the matchmoving (or layout) will need to be done. Next the animator starts, then the TD and then the compositor. These last three overlap a lot or a little depending on the shot and pipeline. Most people are working on 2 or 3 shots at a time. Even though a shot may be budgeted as 2 weeks it may take 3 or 4 weeks to complete since there will down time while waiting for feedback from the director or while other changes are made. In some cases a shot may have to be put on hold for a month or more while a change is made or until additional elements are shot.

Ideally each task for each shot is tracked either by timecards or by a database. If they’re not tracked it becomes difficult to determine how much progress is being made relative to the schedule and budget since linear time and budgeted time are different. If the work isn’t monitored you may not realize how over budget or over schedule you are until the last phase of the show when it’s too late.

The poorest process is for the vfx producer to say we’re $5000 over budget on a specific shot or to dump an inch thick document of raw figures on the desk of the supervisor. At that point the shot is already in the red and you’ll have to spend time figuring out what that amount means (which task is causing the problem, is the dollar amount with markup, etc). The best process is to monitor the shots and flag them as they reach critical stages (50% done on a task, shot to be completed that week, etc). As an example: If the composite time is already 75% used up and there’s a lot more work to be done then the supervisor may want to review the shot with the compositor and determine if there’s anything that could be done to simplify or complete the work in the budgeted time. There may be another 200 similar shots yet to do so if there’s a major flaw in the approach then it should be modified or discussed with production.

Dailies are held everyday to review the work in progress. The team creates weekly targets for which shots should be completed. These are reviewed and frequently adjustments have to be made. A shot that was scheduled to finish that week has to be pushed to the following week because the director kicked it back or because there were larger problems than planned. As production nears the finals date the weekly reviews become daily reviews. Don’t wait until the last minute to alert the film production company if there are scheduling issues.

If there is a crunch time at the end of production then the costs will start to skyrocket with overtime and other rush charges.

Related posts
Getting the most out of your VFX Budget
VFX Wages
Why do Visual Effects cost so much?

Related books (In the VES Handbook I cover budgeting similar to this site)