Strange Encoding Times

[Herbert]

I have two versions from a documentary feature, 139 Minutes and 120 Minutes.
Technically both are identical:
1920x1080 ProRes4444, 6 discrete Surround tracks in 24bit/48kHz PCM.
I still use the defunct Final Cut Pro 7 for editing several versions
from one master because
- as long as you stick to editing and do not apply any effect-
there is no generation loss:

http://www.niwa.nu/2013/05/generations- ... em-or-not/

So both versions are the same, besides one version missing 20 Minutes.
But funnily the longer version took shorter to encode to a Flat DCP.
Settings were the same, 200mBit/s, audio untouched.
I even checked twice, once both versions with resizing to 1:78 (the source format anyway),
once with no resizing... the longer version took almost 4 hours less to encode
BEW-THEA is the short version (120Minutes), BEW-DIR is the long version (139Minutes)

[Carsten]

There may be other reasons for it - but, in part, encoding times are source complexity dependent. Low complexity images will compress faster than high complexity (high detail) images. If you don't see these differences between your two masters, another possible reason could be that, if FCP7P doesn't recompress (what you describe as generation-loss), it may be that the final master contains source material from e.g. different cameras. Depending on the compression characteristics of different codecs/codec settings/cameras, these may also lead to different uncompression/encoding times in DCP-o-matic.

Again, just guessing, one would have to look closer at the source material to find this out. Sometimes, changing small environmental conditions (e.g. external hard disc, USB connection) etc, can have a major impact on conversion speed.

- Carsten