It was shot on HD using Sony's CineAlta F900 HDcam camera with a non-prime zoom lens and output to anamorphic 35mm film. For the technical amongst you, HDcam records to tape 75% of the horizontal resolution of the luminance channel, 50% of the chrominance channels, and 8bits per pixel/colour vs. full spec. HD image information. The resultant anamorphic images also used less vertical resolution, obviously, so the results obtained were nothing less than astonishing and show what is possible with HD when it is used correctly and with care.

The DoP followed instructions well on how to set-up and use the camera resulting in high dynamic range images without crushed blacks or clipped highlights, ideal for the following DI process.

The approach adopted by the DoP was to not attempt to 'grade' in-camera (on set) but to capture as much dynamic range and detail as possible enabling decisions to be made in post, mimicking the standard 'film' procedure of OCN capturing a wide dynamic range with the grading and print processes producing the desired final image.

With this approach the DI process worked well with regard to the final image quality and colourimetry.

However, when shooting the sound engineer forced the production to shoot using time-of-day timecode rather than continuous timecode. This caused problems when loading the image data into the offline (FCP) system and the subsequent online autoconform. This is not something the client will allow to happen again.

Note: A related problem found on another feature shot HD (Avanim - shot in Israel and also technically supported by Digital Praxis) was the HD camera failing to pickup timecode correctly when recording continuous timecode, which caused the offline system to abort loading as it saw a timecode break!

The biggest problem facing Daddy's Boy was the online autoconforming of the edl. The problem was that the edl was not 'clean' causing a lot of issues with the online conform, resulting in something like 50% of the short having to be manually eye-matched to the offline video. Not too bad with a 10 min short but a nightmare with a full feature (see American Heart & Soul following).

As part of the DI process the ability to match online 2K image data with the offline video image is a real benefit. It's hard to imagine not being able to do this.

The second major problem was again audio. Although the production had been instructed to shoot 24fps HD they opted to shoot 25fps, which necessitated a 25 to 24 audio speed and pitch change. This proved to be next to impossible due to the difficulty in accurately matching the two stereo channels during the processing. This resulted in audio phasing errors with the left and right channels alternately becoming advanced or delayed with respect to each other. In particular it was the re-pitch that introduced the errors.

For dialogue it was decided to go with the non-pitch changed version, as the frequency change was not too noticeable. However, when music was involved this was not an option as it was immediately obvious something was wrong.

The only way to rectify this was to use the original non-speed changed music and re-edit the online as the offline had been cut to beats in the music. The saving grace here was that the DI system had full editorial capability enabling an easy re-edit without the need to go back to the offline.

The original HD image data, obviously shot linear, was mapped into Log space during the DI colour correction process and displayed via a Log to Print LUT, mimicking the print curve characteristics of a traditional film process. This removed the main negative trait of video captured images, and combined with the fact that the image had been captured with high dynamic range/low contrast provided a final image mimicking the look and feel of film captured imagery combined with superb resolution, stability and cleanness provided by the use of a digital capture medium.

The main negatives reported by the production team were associated with shooting, accepting the offline/online edl issues. They disliked the electronic fan noise produced by the camera when and found the camera back-focus slipped too easily.

The DI process itself was fairly painless, except for the edl issues, although getting the various film labs (two were used - one in Rome the other in the UK) to produce accurate film prints from the timed internegative proved to be something of a headache.