Eugene Lehnert Posted April 20, 2012 Share Posted April 20, 2012 If I have an HDCAMSR tape that is 1.85 or 2.35 how is it converted to a DCP? Is the HD image uprezzed to the 2K size or does the HD image sit inside the large 2K frame? And the color conversion is a simple 2D LUT to go from Rec709 to XYZ? Link to comment Share on other sites More sharing options...
Benjamin Kantor Posted April 22, 2012 Share Posted April 22, 2012 Below is a sample workflow (I work at a post house where we occasionally make DCPs): 1. Ingest the HDCAM-SR as an uncompressed Quicktime (probably a Blackmagic 4:4:4 codec, in our case). 2. Scale the image to meet the DCP 2K standard (for 1.85:1, it's 1998×1080, for 2.39:1, it's 2048×858). To be clear, the image fills the entire frame, there is no letterboxing or pillarboxing. 3. Export as a TIF sequence. 4. We use OpenDCP, which semi-automatically takes the TIF sequence and converts it to a JPEG2000 sequence in XYZ color, and then wraps it up in an MXF wrapper for delivery. Link to comment Share on other sites More sharing options...
Eugene Lehnert Posted April 23, 2012 Author Share Posted April 23, 2012 Thanks. I would imagine the uprez to 2k would not be that noticeable? I guess it depends on what algorithm you use. What are your thoughts? Link to comment Share on other sites More sharing options...
Benjamin Kantor Posted April 23, 2012 Share Posted April 23, 2012 Yes, from a scale perspective, we are talking about 9.4% to get it from 1080p to 2K, which I personally think is negligible. Also, if your film were playing in a theater off an HDCAMSR deck feeding to a 2K projector (like a Barco), it would be getting upsampled to 2K by the projector hardware anyway, so I don't know that there's a difference in the end result regardless. If you really wanted to fine-tune the scaling algorithm, then something like B-spline is going to technically look better than Bicubic. However, it complicates the workflow. The four step workflow I mentioned before could probably be covered in 4 hours for a 2 hour tape, using just FCP or Premiere and a DCP application. Adding a B-spline step could more than double that time, as it would add a completely new, time consuming process in a different application (we would probably use Nuke or a Photoshop action). Link to comment Share on other sites More sharing options...
Eugene Lehnert Posted April 23, 2012 Author Share Posted April 23, 2012 Ok great. Thanks for the help and advice. Link to comment Share on other sites More sharing options...
Recommended Posts
Create an account or sign in to comment
You need to be a member in order to leave a comment
Create an account
Sign up for a new account in our community. It's easy!
Register a new accountSign in
Already have an account? Sign in here.
Sign In Now