Photos by Amit Ashar

& kind permission of Rose Movies

Rose Movies, a major Indian film production company with more than 40 years experience, together with highly successful main stream producer Eros united forces to make Goldie Behl’s Drona, a film based on a modern tale of one man's journey through the labyrinths of mystic myths and legendary legacies.

With such a visual film, I was presented with numerous stimulating challenges during production, and in post production.

Principal photography was split between Mumbai, Namibia, Prague and Rajasthan, while the visual effects were created in six of Mumbai’s most qualified facilities: Eyeqube Studios,  FutureWorks, Prana Studios, Prime Focus Plus, Redchillies and Tata Elxi Visual Computing Labs.

In my role as visual effects producer and supervisor I had looked at many facilities, their show reels, got to know the professionals and facility set up’s, while asking for estimates and creative proposals for numerous vfx shots from them, before settling down our choices after competitive, creative, and critical time-to-completion bidding on these.

It was a stimulating shoot as we invented practical solutions for complex shots - how to do a “motion-control camera track” underwater, shooting a stunning action sequence on a running train - and even getting our hero (played by the famous Indian star Abhishek Bachchan) to jump onto the roof of the moving train - you guessed it, on horseback!

I have never worked on a film with so many dynamic and breathtaking visuals, and I was well occupied over the twelve months stretch in which I followed the film, helping to come up with solutions for each and every one of these.

Following on to the well shot film material, we wanted Drona to have a pristine and sharp overall look, so we decided to investigate the possibilities of doing the digital intermediate in 4k in several facilities that trumpeted “4k “ workflows in Mumbai.

We had seen some previous Indian blockbusters go the 2k route in 1:2.35 cinemascope format, and we weren’t happy with the overall soft image that is derived at that resolution - in fact, centre extraction of a 2k scanned image gives us less than a thousand lines of resolution, and I personally think that this is not quite adequate for wide format films on the big screen.

So, I explained to all the potential facilities what we wanted, and warned that we would have to accommodate editorial changes and last minute vfx versions into the two and a half hour film in the final stages of grading, although this has become practically the norm with most films today.

We were shown various opportunities, systems, colourists and theoretical workflows by quite a number of facilities.

I smiled quietly to myself as some of the various proponents sported a series of amusing verbal acrobatics to describe the way that they could achieve the best results by scanning in 4k, then down-rezzing in 2k in order to do the grading and editing with 2k equipment, and then up-rez again to 4k, and other variables of a similar nature.

To be truthful these techniques are indeed eminently useful for getting more out of less, and I have used them personally many times in the past, but together with Goldie Behl, the director, and Sameer Arya, the cinematographer, we had already decided that we wanted nothing but the best for Drona, and that we weren’t going to accept any compromises on the film’s image quality, and that actually meant that we wanted to do the entire process in full-blown 4k; this can accurately match the original camera negative when calibrated correctly, and it ends up making the DI process more exiting as there is practically no quality compromise for using all of those wonderful and creative colour grading tools that DI has become justly renowned for.

At the same time, I had the responsibility of finishing the film on time for the release date (October the 2nd 2008) which I did indeed manage in the end. that, in itself meant making a choice that would  not fall over for the huge amount of data involved and require more time for the DI process, a particularly important point since this was one of India’s largest visual effects films of it’s time with more than 1500 visual effects shots and some very spicy fast cuts and speed-ramped edits to boot - a tribute to the craft and skill of the film’s editor Shyam Salgaonkar (affectionately known locally as “Sally”). We also needed to keep the total time of the DI down to a bare minimum in order to gain every last drop of time for the visual effects, and that meant a rocket-fast system for on-lining and grading.

I had personally experienced many digital intermediate sessions with most of today’s available systems in the recent past, so I had a reasonable degree of first-hand knowledge on the pro’s and cons of each of them; and as some of the less experienced proponents learned to their consternation, it was not quite so simple to pull the “digital wool” over my eyes.

Funnily enough, many of the proponents I had discussed the project with focussed primarily on the grading aspects of the DI, as if this was the be all and end all of digital intermediate work.  Personally I think that there’s a whole lot more to DI than just colour grading. I was still surprised to see such little general understanding of all the other no less important issues; while I certainly didn’t and don’t underestimate the need for the best possible grading capabilities, and, indeed we did not want to compromise on these at all - quite the contrary - the fantastic world of Drona demanded it, and in fact, powerful grading tools indeed helped to achieve the sophisticated and vibrant visuals that Goldie and Sameer desired for their film.

Once the various proponents had understood our full 4k requirements, it seemed that many were underestimating the sheer speed and volume needs for an effective 4k pipeline, and based their proposals on unproven and theoretical workflows using proxies, down rezzing, still to be tried and tested storage area networks, and moving material around from editing to colour grading systems backwards and forwards more or less continuously. Many sported flawless theoretical assumptions but actually little or no real world practice in this 4k “heavyworks” arena.

Perhaps with time and experience I’ve become less inclined to embark on games of musical chairs to satisfy some of the workflow proposals that combine different systems. I prefer seeing all in one suite, and I think this also makes absolute sense; we just cannot afford slow and cumbersome processes, transfers back and forth of huge volumes of data with the dynamics and logistics of film finishing these days.

I guess that, as usual, it really boils down to good old experience in the end, and while I certainly can’t claim to know all in this field, my experiences are certainly useful to me in avoiding as many pitfalls as possible. Since ‘95 I have supervised the DI processes of a fair number of feature films, and I’ve learnt quite a lot about the process with each and every one of these, using different professionals, technologies and workflows around the world.

Anyway, to cut a long story short, the choice of system was actually pretty much a no-brainer for me, since I had watched the evolving Quantel Pablo and it’s ever more sophisticated grading panels ever since Quantel’s top management had shown me the first working prototype in Newbury on the first working day of this millennium.

In this last decade I have seen a constant dedication to the needs and requirements of the film industry and an ever more effective system come to light. Indeed I had quite often specified feature sets myself, and have appreciated the adoption of many of these - deliverables have become one of the most important aspects, and these are better catered for on the Pablo than on any other system. The sheer quality and lightning speed with which data-heavy 4k work can be performed on the system, coupled with the dynamics of constant updates and new feature sets must prove a moving target that Quantel’s competition must find fairly hard to catch up to, judging on the evidence of what the competing systems are capable of doing realistically today.

As an independent visual effects producer and supervisor on Drona working for production, my concern was to get all of the effects work underway as early as possible, and I got to know FutureWorks first for some of these - we had originally planned to do the DI elsewhere with other proposed technologies, so I had not initially looked specifically at the opportunities that FutureWorks could offer in this regard. 

My simple but tried and trusted method of explaining our technical and creative requirements for each shot and then comparing the cost estimates and creative proposals from the various facilities for these, coupled with some viewing of show reels and interviewing of operators, led me to realize that FutureWorks was an interesting opportunity for vfx as well as the DI, so we decided to do a fair number of effects shots in the facility (by the film’s end, we had done over 400 vfx shots there).

FutureWorks is also a facility that was planned and built from the ground up for 4k work, and the presence of Quantel’s 4k Pablo and Gene Pool had evidently contributed to making the vfx department very effective and better equipped for all important quality control than some of their competition were at the time. I specified the fact that Goldie, Sameer and myself wanted to see and judge all the film’s vfx in a calibrated high quality projection room, so as to negate the need for constant outputting to film, and this was provided for with the excellent film-calibrated Barco DP90 in the facility.

The actual workflow set up in FutureWorks was based on scanning with the  Spirit 4k & Bones, Pablo and iQ with Gene Pool for editing and grading, and Arrilaser for film recording. The vfx were sent to the various afore-mentioned facilities and were also directly available on-line in FutureWorks itself, after which the finished effects went back into the facility’s 4k Pablo. After the grade and edit in 4k, film recording to internegative was done on the facility’s 4k Arrilaser, and for Drona Steve Shaw consulted and supervised the DI process; Steve is no stranger to me, having set up Cinecittà Digital with him in 2001, so I was happy that he could contribute his creative and technical flair to this project as well. He co-ordinated FutureWorks own in-house DI grading team on the film, and worked directly himself on setting up all the looks. There’s nothing like experience for getting the job done well, and follow this link to see some of the before’s and afters in the grade on Steve’s website.

Drona pushed the bench mark ever higher in the fast growing Indian Film Industry, and I’m proud to have been part of the team actively involved in having pushed the quality envelope.

Back to Home Page