2D to 3D conversion is the process of converting films from 2D (normal film) to stereoscopic 3D (sometimes called S3D) film for viewing with a stereographic viewing systems such as anaglyph (red/blue), polarized and shuttered glasses or using a special lenticular (or other ‘glasses free’ autostereoscopic) screen.
A good conversion is labour intensive and it requires an experienced eye because the process is one part science, one part art. The best analogy is making a suit of clothes – anyone can cut and sew material but you need a good, bespoke tailor to produce something that looks really great.
Film that is converted well is indistinguishable from film that is shot in stereo or “native” 3D and it is an open secret that most – if not all – 3D films contain converted material whether or not it is acknowledged. By way of example, James Cameron has always been coy about whether Avatar contained converted material but his collaborators have been more forthcoming. In fact, it is increasingly not a discussion about WHETHER to convert or shoot, but HOW MUCH to convert; when locking down a 3D budget. Avatar, Pirates of the Caribbean, Judge Dredd, John Tucker, Transformers Dark of The Moon… they have all included conversion material – and for the latter three it has been at least 50% of the content.
One of the reasons most 3D films contain converted material is that shooting in stereo usually means that there is loads of footage to correct in post-production and often the easiest and best solution is to simply convert. Also, converting film well allows for much greater creative and production flexibility; whilst being a lot less expensive.
CONVERSIONversusSHOOTING STEREO OR “NATIVE” 3D
Some people are prejudice in favour of “native” 3D and believe that is superior to conversion in every way which – as with most prejudices – is not true. Clash of the Titans is always cited as a film that justifies the prejudice but it’s a misleading example because no one claims it is a good example of conversion. The latest Transformers is a better example because it contains native and converted footage in the same film, and often it is intercut seamlessly.
Before going through a list of pros and cons, here’s an anecdote that neatly illustrates the fact that most people have no idea what they are watching – the viewing public just like good 3D.
In 2011, we went to BAFTA to watch a preview of a 3D movie with plenty of sword play. The director was well-known and declared his love for 3D movies telling the assembled industry professional that he would only make movies shot in native 3D because conversions were a terrible abomination that could never produce anything of quality.
Having made his impassioned speech, the eminent director sat down and the preview began. The 3D looked excellent and received a rousing ovation as well as many favourable comments from the assembled directors, producers, heads of vfx and post-productions during the Q&A that followed.
Our team was smiling from ear to ear with quiet satisfaction because we knew that the preview had been almost 100% converted. There had been so many problems in post-production that in desperation one of our team had been recalled from his holiday to supervise a conversion team.
Everyone was happy and no one had been any the wiser, especially not the director who had left his film in the tender care of the post-production team.
Whether shooting in 3D or converting later, a different approach to 2D is required and the key to getting the best results is to plan how 3D can be used for best effect. Martin Scorsese’s Hugo 3D is an obvious example of a film that has been shot to maximize the 3D experience.
Usually, the best approach is to use both processes and currently this is how most productions are planned. Here’s a guide to help you form those decisions:
Shooting in stereo – as a rough guide allow 30-50% increase on a 2D budget (which is why most films mix and match with conversion which costs 5-10%)
1. Easier to capture nuances of complex 3D scenes such as crowds, rain, snow, leaves, smoke, reflections, rain, leaves, etc.
2. Preview stereo images on set and on location
3. In edit, footage can be reviewed in stereo making editing decisions easier
A. Cost: it’s more expensive to shoot in 3D because it requires specialised equipment and personnel (and involves more time)… plus twice the amount of data to manage, wrangle and archive.
B. Specialised camera rigs are required and they are bulky
C. Specialised rig technicians required
D. 2 digital cameras are required which further increases the bulk
E. More set-up time (cost and convenience) and less set-ups accomplishable, per day
F. Lens choices are restricted. For example, long zooms don’t work on a rig because the image is flattened and the 3D effect is lost, plus zooms can be tricky to align correctly
G. Depth is essentially locked-in. It cannot be changed greatly in post. The only way to change depth (as it is really a factor of the interaxial distance) in any meaningful way, is to convert one eye.
H. No guarantee that what is viewed on set will translate to the scene (see BAFTA anecdote above)
I. No option of using 35mm film (unless we go back in time to the Friday the 13th dual camera 35mm solution, or similar – and anyway, 35mm grain, flicker, natural degradation when presented multiple times and it’s inherent expense, is not really ideal!)
J. Non-parallel rigs require a convergence puller on set who determines the convergence points during filming. In edit, it is vital to have the correct convergence point and yet the scope for changing convergence in post is limited.
K. Each camera can and will see colours, lens flares differently which must be corrected in post and this can be complicated and expensive. Often the solution is to convert one eye.
L. Frequently the alignment will be different and this must be corrected in post and this can be complicated and expensive. Often the solution is to convert one eye.
M. Polarisation will mean that there are image differences when filming reflective surfaces
N. VFX requires more work and longer render times. Simple things like rig removals can become very complicated and time consuming (as they discovered on Resident Evil, where what was expected to take a day, was taking a week, or two!).
Stereo Conversion – as a rough guide allow 5-10% increase on a 2D budget
i. Cost: it’s cheaper
ii. Shoot as normal but plan for 3D to maximize the effect
iii. Complete range of options. Can shoot on film (see latest Star Trek) with any film camera or use any digital camera
iv. Complete range of camera lenses
v. Complete flexibility to add depth even when using telephoto lenses
vi. In edit, complete flexibility to set the convergence
vii. In post, complete flexibility to set the depth and volume for each and every element in each and every scene.
viii. In post, complete flexibility to add or delete elements
ix. VFX is as per usual
Stereo Conversion – CONS
a. Crowds, rain, snow and smoke, sparks are more difficult to convert, but certainly not impossible.
There is a huge range in quality and price so it is important to choose a good company offering sensible prices, who have a great reel.
Originally posted at The 3D Company: “Native” 3D versus 2D to 3D conversion: Pros and Cons | THE 3D COMPANY