Posted on Jan 9, 2012 by Alex Fice
While post houses look for alternative revenue streams in distribution and archiving, Light Iron Digital in central Hollywood have veered the other way and are taking their expertise in processing RAW digital camera footage to the set. CEO and Founder Michael Cioni explains to JULIAN MITCHELL where their thinking comes from and how they are changing the role of the traditional post facility bit by bit
There is a movement in Los Angeles called ‘The Beautification of Hollywood’. A grand title for what seems to be an attempt by the City to make central Hollywood a nice place to live and work in again for its 350,000 inhabitants (celebrity Paris Hilton swept the streets recently as part of her community service sentence for the project). Light Iron Digital is right in the middle of Hollywood, not far from Hollywood and Vine the cross street usually designated as the centre of the district. The geography means a significant amount to founder and CEO Michael Cioni as he has to think of his clients and their journey and stay at his facility. It’s also about the rise in the number of restaurants, banks, and shopping in the area, again encouraged by the beautification project to reclaim the area from the gang graffiti and illogical placing of some undesirable businesses (adult bookstores etc…). Michael has got to know the person who is driving the project from the Major’s office and keeps a close eye on the project’s objectives.
“There are facilities that are on the West side, there are facilities that are in Burbank and there are facilities that are in Hollywood. Those are roughly where the concentrations of places are. The West Side I don’t particularly like, I’m a Mid west kid and like urban areas. The problem with the West Side is that it’s one-way, you go out there and come back, there’s nothing past it except the Ocean. My two biggest clients are Disney and Sony, Sony are in Culver City which is 20 minutes west and Disney is in Burbank which is 20 minutes north east. That’s a very reasonable drive from those places in LA. If you live out in Venice or Santa Monica it’s an hour.”
Smart thinking and attention to detail seems to be in everything Light Iron Digital do. Apart from this logical calculation of what his major clients would put up with as a journey time to get to him, the building itself was chosen to benefit the services he wanted to offer. “We have incredible theatres here that are as practical as they are beautiful. That’s really important to our clients, our biggest screen is 27 feet and for a facility that’s pretty significant.” The building is also very anonymous and is a ground floor to a 3,000 place parking garage, the signage doesn’t shout high technology but walk in as I did and come face to face with Director David Fincher sipping Latte and here to finish off The Girl With The Dragon Tattoo which is out for Christmas. (Light Iron have done the DI and Digital Cinema mastering).
It’s a modern LA building with construction that allows for columns to be separated by around 80 feet with 16 foot six inch ceiling, unusual for LA with its Earthquake threat but great for Michael as it gives him height for the theatres. Basic spec is 10,000 square feet with three theatres, two Barco projectors, two Christies, and four Quantel Pablo post systems.
Light Iron Digital has only been around for 30 months and was opened in the midst of the worst recession most people have known. Michael was previously running a post business called Plaster City but moved when shared ownership become a question for him (he was doing all the work and wanted some share, the owner didn’t agree). Michael set up Light Iron with most of the same people and managed to bring along with him 100% of his clients. He launched without a bank loan; they wouldn’t give him one even though he had guaranteed work; and has grown through the bad times to now specialise in dealing with digital camera footage. “I’m not concerned about surviving and making money today, at all! The business models of current technology, current tools and current philosophies don’t interest me. Our business model and our development strategy is exclusively based on future technology and executing pathways and workflow designs that will basically aid everybody in the future.
“To quantify that we basically do file based RAW image manipulation. That manipulation starts with capture, archiving, drive and data management, drive formatting and then it moves in to the working component of editorial, dailies processing, one light synching and then transmission of different components. Then there is finishing with DI and colour corrections, colour science, projection, compression and exhibition. Its sound like a lot but its really specific, file based RAW.
“We don’t have much experience with tape. When someone sees a tape deck I see a DVS Clipster. A Clipster can do anything and it can check my Email! Its our VTR equivalent.”
Decisions made for success
As a student Michael Cioni used his college loans for buying gear to make programming and other content for regional Emmys. Prizes included visits to facilities in LA which got Michael thinking about a career in the industry. His Dad had been an animator in Chicago and Michael had strung along to see how the process worked. His experience later at Plaster City definitely showed him that dealing with the future of digital cinematography was not just creatively satisfying but business wise made sense.
With his Plaster City experiences behind him, that had perhaps shown some naivety when he came to company ownership, he started Light Iron with his brother Peter who had Wall Street business experience. Today Michael has no problem in pointing out other post owners shortcomings when it comes down to business practices.
“When you struggle as a post house it’s not that you made one bad decision. To go out of business you have to make hundreds of bad decisions in a row. That’s why I think the blame for post houses that go out of business lie exclusively with the leadership. Its not the talent, not the location and usually not even the quality of the work as there are so many variations of work available.
“The leadership is failing to recognise the change in trends and philosophy that is required to survive. Yet the evidence is all around them. There isn’t one facility owner that doesn’t consider that tape stock is going to be a declining resource and therefore tape cameras will go away.
“For example if you read press releases from major post houses, you will find the most amazing, backwards…and forwards, up and down roller coaster of decisions. The amount of just blatant spin is remarkable. They have a building where they have multiple floors of telecine that they have just built! Investing in telecine in 2009 is probably some of the worst money you could possibly spend. This is a copy paste mentality,’we have telecines before and they made us money before, we have a new facility lets do it again.’ I often use a phrase ‘You have succeeded in becoming the city’s best typewriter repair man’ to some of my competitors who I like to challenge.
“I have worked with companies on the same project and we deliver a drive and they literally don’t have the connection for that drive, they say: ‘What is this connection?’. To me it’s like someone saying ‘What’s this 16mm version of film?’ they should know.”
No RED Without Panasonic
Michael couches the RED camera phenomenon in a previous IT based technology from Panasonic, he sees the introduction of P2 leading the way for RED’s tapeless revolution. “The RED story has to start with Panasonic. Panasonic released and cultivated an infrastructure for file based capture for high-end markets in 2005/6. I’ve been told that had the P2 system not been accepted, or failed as a format, that all the video departments of Panasonic would have then closed. They put everything in to file based capture. Sony was the complete opposite they put everything in to tape based capture, in fact after P2 was announced Sony manufactured probably five more new tape based cameras. Their motives were totally different. Sony’s first professional grade file based camera is the F65 which I’ve worked with and that’s nearly 2012 and you have Panasonic cameras coming out in 2005! So you have to give kudos to Panasonic for cultivating a file based beginning. What they failed to do is build a professional PL lens mount camera system. That’s their mistake. RED broke past all that because they made a PL mount camera and they move to wavelet compression technology instead of MPEG and the wavelet compression basically was the key for capturing high resolution images smaller. If you think back to the Dalsa camera, that was everyone’s favourite camera on paper but no one picked it in the real world. The RED that looked like it cutting every corner possible was actually what people needed, RED knew what people needed not perhaps what they wanted. They wanted uncompressed 4k that fits in a box on your shoulder. What they needed was to do that practically and cheaply.”
Michael hitched his wagon to the vision that RED was setting forth, but to do that he had to deal with the cynicism to start with but then deal with the footage which wasn’t so easy to do back in the day. “Most people wanted to discredit RED at the beginning saying that ‘they didn’t need 4K’, ‘you can’t see 4K’, ‘RED’s a bayer pattern so it’s not real 4K’. When RED launched in 2005 a very good friend of mine who I have been close with for years Ted Schilowitz called me right before the NAB of that year and said that he had something to tell me. We were saying ‘Yes the Kona card has gone 2K!’, that what we thought it was because he worked for AJA, but he said he was leaving AJA to do something completely different. He explained what it was and I thought that with Ted there it was going to work, he has a great track record.
“It was that year and moving in to 2006 that we said that we need to get a (Quantel) Pablo. We went to NAB to find the 4K solution for DI. This is 2006 so nobody but Sony had 4K projectors, (Sony actually had their first SXRD which wasn’t very good). Hard drives were only one terabyte back them which was really expensive about $400, you need a lot of those to do 4K. But we knew it would get better and we could predict how much it would cost in 24 months.
“We audited every system at the show, Lustre, Resolve, Baselight, Pablo, Scratch, Speedgrade. Remember RED still really didn’t exist, it was a booth with no cameras. RED was just a thing people talked about to make fun of. None of those systems were really 4K systems except Quantel. They were the only company that did two things, they recognised that 4K was going to be a future requirement and they were building a system to do it now. Nobody else talked to us about doing 4K. FilmLight wanted to push 4K for the scanning of film but we were interested in 4K for the options in digital cinema. But I wanted a company that was going gung-ho digital 4K and Quantel had custom hardware. A lot of the other systems used CUDA acceleration in their graphics cards for image processing. That’s not wrong but it’s not ideal. We didn’t want to be reliant on CUDA acceleration which means their power to a degree can max out when the graphics card is saturated. Quantel is different and they build all custom tools. They have custom boards for just geometry, boards for just colour or playback. We thought that architecture was going to work when it gets really hard in a few years time.
“I can’t compete with companies based on today’s standards, today’s technology and relationships. If you just flashback two years its all film and tape, in cinema there’s very few files. I can’t compete with people who have being doing this for in some cases 100 years. But when they go to files all of a sudden I’m the only one with the answers. In a way I’ve been working with files for my whole life. My entire professional career has been essentially tapeless. If you’re shooting film or tape I can’t do the job, in fact I don’t even want the job as I can’t compete with the people who normally do it. I want them to compete against themselves to ‘fix that typewriter’ and then we’ll manage RAW image capture. That’s what Quantel essentially allowed us to achieve because they like RED were building for the future.
“Coupling the Quantel with the RED in 2007 which was when the RED was released, we got our Pablo a little bit before that and it put us on the map. All of sudden we could do things that no one could do and we had pictures coming in that no one could capture. We were the only ones that said ‘Yes we can take it’. We were fishing with that bait.
“When we did RED for the first year probably third quarter 2008 we were criticised for concentrating on RED for our business too much because it was about 75% of our business at the time. People thought we were making a mistake concentrating on RED. I got worried but analysed the figures. I thought to myself ‘We are either getting all the RED work or the industry is becoming 75% RED’.
“If you look at a triangle and the top of it is the ‘tent pole’ movie, $100 million movies and above. The bottom is the no or low budget indies, the best part of the triangle is the middle because they have real cash, $10-30 million, they get real projects that get some form of distribution but they’re not so ‘tent pole’ that the politics get difficult to deal with. Plus they’re easier to do. A movie that’s $10-30 million has the biggest profit margin. The middle of the tier is the tier that RED started filling. The movies above $3 million. I believe 75% of those movies from 2009 to now are RED. RED was the right tool for the job.”
A New Term –‘Todailies’
Of course being a pioneer in a re-invention business like technology challenges you to keep pioneering especially in the Light Iron’s case where other Quantel competitors started offering RED SDKs and so in turn allowing other facilities to start biting in to Michael’s RED business. Michael’s conclusion and subsequent business direction is maybe a blueprint for other post companies, “The pattern of the trends that are happening are showing us that the facility is going to get smaller not bigger. The elimination of dailies is within 18 months. The end of the whole notion that you shoot something and somebody runs it to a facility to create dailies is 18 months away. We started realising that the computer acceleration in small portable machines is optimised better than a facility’s large infrastructure. A single MAC or HP Z800 computer can actually service file based manipulation better than a whole server room. You don’t need a render farm, it doesn’t work that way anymore. All that infrastructure was disappearing and getting crammed in to computers and we realised that it could be put on a cart with wheels.
“If you could put it on a cart, you could put it on set and if you could do that you could do the telecine right after it was shot. If you could do that it meant that the creative people and the business people on a show could go home with their dailies the same day that they shot them and it would be better, faster and cheaper. So we invented the system called Out post.”
Outpost is their on set mobile processing lab and their first job was on Avatar, talk about a baptism of fire. This was a RED shoot where Avatar was mostly to tape with Sony F950s but there was a component of some of the VFX that they wanted to shoot in 3D at 120fps and at the time the only cameras that did that were dedicated high speed cameras or the RED. “They wanted higher resolution than the Phantom camera could manage so they started shooting 3D frame rates with the REDs. Their requirement was to see the results right there, not a video tap and they don’t want to see it tomorrow. They needed to know whether it was working on not but if you’re doing it high speed it has to be decoded. They needed to ingest RED files very quickly, de-bayer them, colour them and play them back in 3D.
“We gave the creatives this solution without them realising it was special. They didn’t worship the cart but it felt so natural and that’s the best form of technology improvement. Outpost began to cultivate in to more customised solutions which were optimised not just for 3D but 2D, indies, big films and so on. I think what makes Outpost so successful is the thought process going in to it, our philosophy and our tool sets and our approach to this problem solving.
“We’ve probably done 40 feature films this way, all completely on set without a lab. Some of those movies like Spiderman a $270 million picture where all the back-up, 6,000 clip downloads, all 3D, convergence, colour, sound synching, triplication, transmission, AVID files, iPad files, web files were all made on the set while they shot the movie. We’re talking a reduction in budget of over $1 million, possibly $2 million doing it that way. This is evolution, the person who serves the creatives best always wins.”
Michael speaks eloquently and evangelises about his company’s achievements but there is a burning intellect at play here and the numbers add up. Outpost is now over 50% of his company’s business and next year his expansion takes in nine different cities worldwide and the really clever part is that he doesn’t crew anything, part of the design of Outpost is the fact that they don’t hire or own any operators. Light Iron has a free training course for IATSE members (union of motion picture technicians). He trains you for free and you take the cart.
Michael did mention that there was pressure from some of the studios to limit the number of DIT on set including FOX who apparently had announced a ban on DITs on all of their productions, (A quick tweet from us @definition2011 found at least one DIT happily working on the FOX lot in LA. The editor – @pietaricreative – was sat beside him and claimed that they couldn’t achieve efficient post workflows without DITs. A tweet from @RWDMTDIT said he had heard about this two weeks previously, “It does not include DMT types that are not doing onset dailies etc.”).
Being a product from Light Iron and from Michael’s own imagination, he is suitably effusive about Outpost even though it and others may mean DITs having to fight harder for their careers, he sees it slightly differently. “The only loser when Outpost succeeds with production people is the post lab. The production people aren’t going to protect the post lab, the post lab doesn’t have a union to lean on and may have to syphon money from other areas within the company to bridge the gap. I don’t have a post lab here, its DI only.”
Outpost is currently working on nine movies, none of that is happening at Light Iron in Hollywood. Michael has no employees involved other than the technicians that build the carts, on paper it’s good business, “So it costs very little. The overhead here is zero and the shipping costs are zero as its all done by the Teamsters.
“File based DI and learning more about computers, advance networking and how these files can be manipulated taught us to go back to the set. Some post houses think about moving forward towards handling more of the distribution. I don’t want to handle distribution, I use Deluxe, they can handle all that mass distribution. Also markets like long term archiving for libraries, I don’t think there’s money in that. The money’s on set!”
You may have thought that the recent huge surge in the use of DSLRs for video would be something Michael would herald as another example of the consistency of change, but he just doesn’t see it, “We never really invested in it. Companies like Technicolor did with a big investment in DSLR colour science, LUT management and pipelines. I didn’t think it made any sense as the DSLR will never penetrate that middle market I mentioned as part of the triangle. It sprouted up but never made it there, the only time I see it on a movie is when it’s used as a crash cam.
“The further ahead the good cameras get the more the DSLR is distanced. For student films, independent films, student projects there’s nothing wrong in using it and even cable TV. Network TV and narrative feature films will never be shot on a DSLR that has H.264 at 50Mb/s, that will never happen. Will they shoot on a camera that’s shaped like a DSLR, that works like it, sure. Today’s camera is not a professional narrative motion picture system without timecode, audio, monitoring, it just doesn’t work. It doesn’t matter how good the picture is, we’ve proven that with Dalsa. Dalsa had at the time the best pictures anyone has ever seen, but it didn’t matter because all the other things about it killed it.”