How to DCP or: the “video to film to cinema” manual

Since it was the first time this year but not the only time remained that our moving pictures came to the cinema, I would like to write here a small, compact guide and share the painstakingly acquired knowledge of how to prepare a video so that it runs wonderfully “on the very big screen”. Of course there is software for everything and with money you can buy everything, but there is always another way and thanks to open-source a way to create a “cinema-ready” DCP file for free.

Captured on celluloid and unknown color spaces

Traditionally, video and cinema have been somewhat at odds: we remember legendary buzzwords such as 35mm, film and “captured on celluloid”. The history of cinema also evolved parallel to photography from the analogue to the digital age, albeit somewhat later. I don’t want to go into technical details, even if they are admittedly exciting, such as the XYZ color space, and will limit myself to the pure process of converting a video that was created with the usual tools(Lightworks, Premiere, AfterEffects, etc.) to (digital) “film” for the cinema.

Video Killed the Cinema,
OpenSource triumphs over propriety: the tools


At the heart of the workflow is the small but powerful tool OpenDCP, available for Windows, Linux and MacOS. That’ll do whatever it takes. Of course there are also commercial solutions like easyDCP Creator from Fraunhofer Institute – where the mp3-format was invented and the online revolution of music was proclaimed – but with more than € 2.300,- it is not exactly a bargain and was therefore out of the question for our purpose and need.

First of all, it should be noted that in the DCP cinema format – 35 millimeters say hello – we traditionally have only 24 frames per second, while video usually has 25 frames per second or “frames per second” (fps) in full screen (i.e. “progressive”, no (“interlaced”) fields). This is important because otherwise at some point the picture and sound will no longer run synchronously. This adds up: after 25 seconds of video or “film” the sound is already a second behind, so watch out! If in doubt, simply pitch the note minimally (make it faster) at the same pitch. Therefore in the following always pay attention to the number of frames, OpenDCP always shows them. If in doubt, just export the video again at 24fps and extract the audio.

The picture or video is mostly in FullHD with 1920 x 1080 pixels, which can be – with minimal margin – directly brought into the“Cinema 2K” format (1998 x 1080 pixels “Flat” or 2048 x 1080 “Full”). If the production was in full “UltraHD” or also called UHD with 3840 × 2160 pixels, the “Cinema 4K” format is of course the best choice. However, there the edge becomes somewhat larger, because in the 4k-Cinema Full Flat format the number of pixels is 4096 3669 x 2160, i.e. larger edge than with HD-to-2k, unsightly “stretching” or scaling and cropping … Alternatively, simply work in the correct resolution. On the safe side is who uses a resolution of 1998 x 1080 (cinema standard 2K; 1 : 1.85; flat), not all cinemas use 4k projectors.

The way to the cinema in practice:
From video to film in eight steps

First download and install the said OpenDCP and for editing the audio we need the fine audio tool Audacity then the rest goes quite easily in the following steps: OpenDCP-tool-to-make-cinema-movies UI screenshot

  1. Export the desired video as an image sequence. I have always preferred the TIF format, but BMP or DPC also work. This can be done, for example, with Premiere via the normal export.
  2. Then we export the audio track in the same way as uncompressed WAV in stereo with 48kHz at 24bit resolution.
  3. To optimize or prepare the audio track first
    1. open the WAV-file in Audacity, then
    2. click on the small arrow at the top left of the envelope curve to split the “stereo track”, now there are two mono tracks
    3. Double click on the first track above to select it (left channel) and under “File” go to “Export selected audio” to save it with a meaningful name (e.g. “MySoundtrack….1-Links.wav”). Important: we need a 24bit Wav! This can be done via “other uncompressed files”, then set “WAV (Microsoft)” as header and “Signed 24-bit PCM” as encoding. It says .aiff but it is a .wav
    4. After that – the attentive reader already suspects it – we repeat the same with the second, right audio track, to be found in Audacity under the first one and call it “….2-Right.wav”.
    5. This concludes the editing of our soundtrack. It is interesting to note that we can use the same or similar methods to bring not only stereo but also 5.1 surround sound or even 7.1 sound, multiple audio tracks (e.g. for different languages) and the like to the cinema. Straight Surround is of course only sexy but standard in the cinema hall.
  4. Then we start our OpenDCP and convert the TIF-Sequent into the JPEG2000 format, has the extension .j2c and can be found in the first tab of OpenDCP. Important: in case of doubt, always keep the default setting for all tools and do not change anything, unless you know what you are doing…
  5. Then we pack our J2C sequence into the MXF format. This is probably a container format similar to Matroska, AVI or Quicktime. The function can be found in the second tab under “MXF”, the category “JPEG2000” and an output name must be selected.
  6. We repeat the same with our two mono audio files: same tab (MXF) this time category “WAV”, sound input “Mono”, sound output should be clicked “Stereo” and then select the two input files, left for left (…1-left.wav) and right for the right audio track. As output name “MyFilename…-Audio.mxf” is a good suffix. Click “Create MDX” and the cinema sound is ready!
  7. Now we have everything we need to create the actual DCP file. To do this, select the fourth, last OpenDCP tab called “DCP” and pre-select
    1. try the “title generator”: this should be self-explanatory and better not skip it because otherwise the cinema technician can’t assign your film correctly and it can be difficult to find it on the cinema server, so take some time for it…
    2. Then select the image track (select the video .mxf) and the audio track (“…-Audio.mxf”). Now you can see if the picture and sound are synchronous: if the number of frames differs (because 24 vs. 25 fps…), you will see that now and then it’s over at this point and you have to optimize your soundtrack again and get it right as described above…
    3. If everything is correct, select “Copy MXF” and then click on “Create DCP”.
  8. Your movie is ready!

Before the human player in the technical room of the cinema scolds, we recommend the free (test version) of the Fraunhofer Player called easyDCP Player which is available for download here, the registration can be kindly skipped. The player only plays the first 15 seconds, but that’s enough to see if the movie is playing at all. It also gives out informative messages such as:

Code 35189 from Wed Nov 01 20:01:14 2017:

Audio and video length does not match.

Audio track duration is 00:01:59:21

Video track duration is 00:02:00:00

Which is now for me and now means fix the soundtrack because three frames are missing and end this post…. ;)

I am looking forward to further tips & tricks, experiences and comments and wish you a lot of fun!

If you want to delve deeper into the subject, you can find a more comprehensive description in English and many more details on the subject here on The Film Bakery. The “cinema”photo is by Pedro and published under CC-BY-2.0 license.