Demystifying cinema DCP
What stands behind the standard for film projection in theaters?
I decided to write this article after encountering false beliefs among filmmakers, film producers and film distributors, towards the functionality, features and use of DCP.
As a person who is fond of evidence and facts, I thought it could be nice to explain DCP in plain language.
So there we go.
First, what is DCP
DCP stands for Digital Cinema Package. Basically, it’s a set of files, grouped together in a closed package and organized in a specific structure.
Practically, it is the standard format used in professional digital cinema theaters around the world.
I will not elaborate on the package structure. Aaron Owen from Cinematiq does that in an excellent article which you can read here.
Owen’s article explains how can a DCP be used for different scenarios (e.g — one theater playing audio 5.1 with subtitles and another one playing audio 7.1 without subtitles). He also reviews the process of creating a DCP.
My advice: if you need to create a DCP with more than one possible format (called “version”), such as 2K and 4K, or different subtitles to choose from — let a DCP service do it for you. It’s complicated and you’ll also find it hard to quality-check it. But if your DCP will be just the encoding of your single master file with subtitles burned-in or no subtitles at all, you can do it by yourself.
Therefore, my post will not relate to the creation of DCP.
I will, however, write about the things that are a great source of misunderstanding and confusion (hence the demystification):
- Quality.
- Playback and security.
- Floating / soft subtitles.
Quality
That is the #1 false belief among filmmakers and festival programmers.
The common belief is that DCP always provides a better quality. It comes from the more common perception in our society that bigger is better. We see this in many other fields. For example, people think that large companies are more stable than small ones, that bigger vehicles are safer and large organizations are more reliable. Following this, a DCP, which weighs on average a few hundred gigabytes even for a small film, ensures high quality.
Why it is false (or rather — not always true):
- Because converting a low quality movie file to DCP won’t improve its quality. Want to improve the quality of an old film? — work on the colours and the sound.
- Because the size of DCP comes many times from the fact that it’s inefficient. Efficiency in picture and sound quality is the ratio between file size and its quality. If you can reach to a high quality image with size z, then creating a file of 10xZ in size in inefficient. In that sense, DCP files are totally inefficient. But… no one cares about efficiency when it comes to DCP. God knows why.
For example, a DCP can be composed of video made by TIFF files. One TIFF file for each frame. Being a format which was created in the 1980’s, TIFF is a highly inefficient format in terms of size. It is uncompressed, but compression algorithms today are very very efficient anyway and difference between an uncompressed TIFF and a high quality JPEG will not be visible to the human eye. It is true that most DCPs are made using ProRes, which is lossy (weighs less) without that being perceived by the human eye as such (“visually lossless”).
My argument, and I know that many people disagree with me, is that this whole discussion is for the purists: many movies look great on MP4 (yes, MP4) if you just allocate enough data rate to them. MP4 is just a container, it’s not a “quality”.
Why can it look good? — First, because as I wrote before, compression algorithms are very efficient, even the relatively old H.264. Secondly, because many movies were not shot with the best cameras and lighting conditions, so converting them from their anyway-lossy and basic quality origin to a lossless DCP will not increase quality.
To sum up the issue: if you have a chance to screen your documentary on MP4, it’s not necessarily a compromise.
Playback and Security
A DCP can be played from a dedicated projector or a software (note on the software below). Speaking of projectors, it’s definitely not a home projector. A DCP projector starts at a few ten of thousands of dollars and more.
A reasonable solution for low budget film festivals who think DCP is better (or have no other choice, because that’s what right-holders provide them) is to play the DCP with a software. There are a few solutions, the most known of them is DCP-O-MATIC, an open source project, which also has a software for authoring (creating) DCPs. A list of software to playback a DCP on a computer is available here.
As for security, there’s a very strong point in favor of using DCP. For studios and distributors, it is critical to handle security in an efficient manner and DCP provides that option, if it’s encrypted.
Encrypted DCP, which are a common practice among large studios, require two elements:
- The projector must be professional. A professional projector is always DCI compliant. DCI (Digital Cinema Initiatives) is a standard created by a consortium of Hollywood studios and projectors manufacturers. Just like with DRM, the anti-piracy for streaming, which has agreements between content providers and hardware manufacturers to prevent screen recording, with DCI, the ability to play an encrypted DCP, requires the use of a DCI compliant projector. That compliance in part of the projector hardware. That serves as a kind of a threshold against illegal screenings, because those illegal screenings will usually use a software to screen the DCP file. And since none of the software mentioned is DCI compliant, it just wouldn’t work and the pirated film will not play.
- A KDM. KDM — which stands as Key Delivery Message, serves more or less like password to play the DCP. In encryption and decryption, a “key” in not exactly a password, but let’s leave it like that now for the sake of simplicity. When creating an encrypted DCP, the creator also generates a KDM. A KDM is generated on a per-screening basis, it’s not a generic password. It’s interesting to note that the key concept is also used by us on Movies Everywhere encrypted streaming solution: we break each video to very small segments and allow to playback each segment with a different key. The keys for those segments are provided from a separate server only in our player and only in user-authorised sessions. The logic is similar to that of a KDM allowing to playback a DCP. I described that process in details in another post.
Those two components - encryption (which prevents non-DCI compliant projectors from screening the film) and KDM (without a KDM, the movie won’t play), help right-holders fight against piracy. Those elements are part of the DCP authoring process.
Floating / Soft subtitles
I mention this topic here, because it’s relevant to most of my readers, which are film festival organizers, arthouse distributors and indie filmmakers.
This group of professionals frequently encounters a special situation: they need to arrange subtitles for only 1–3 screenings.
Now, if the subtitles for that screening were not included in the creation process of the DCP, it’s a lost case. You can’t edit a DCP — it’s part of its protection (and protocol). You can only go back to the source files and create it again - a time consuming and costly process. It is not worth re-authoring the DCP only for those few screenings.
So what can they do?
The solution is “floating” or “soft” subtitles: those are subtitles that are projected in a physical area in the theater, under the screen. The subtitles are synchronized to the playback of the film (“timed”), just like any other subtitles file (e.g — SRT), not being part of the movie file (burned-in or as a metadata channel), but screened independently with a dedicated projector.
That’s about it! The article only touched briefly the process of creating a DCP, as it was not the main topic. If you wish to know more about that process, consult one of the many links I integrated in the post. I hope I helped clarifying some misconceptions about DCP quality and the ways in which it ensures security.