Howdy! There’s a newer version of this script here: 2016-05-23-nuke-script-read-node-from-write-node

Written in Python, with Nuke 6.9 in mind, generate a Read node from the selected Write node.

The script will attempt to load an image sequence or a single file (such as a movie file), based off the selected Write node. If the first and last frame can not be determined, it will fall back to the project settings’ frame range and throw a warning.

The script comes with one limitation; it assumes you render out any file sequences with a frame padding of four digits surrounded by period signs:

  • .####.
  • .%04d.

Supported real world examples:

  • lotr_seq1_shot1_v001_fa.####.exr
  • blah.%04d.jpg
  • hello.mov

Download: readFromWrite.py

  • v1.4: Better sequence detection. If you had problems with 1.3, try this.
  • v1.3: Write nodes with expressions now gets evaluated correctly.
  • v1.2: Takes the ‘premultiplied’ setting from the Write node, auto-detects movie files frame ranges, better logic when detecting a frame range
  • v1.1: Support for filenamefilter callback
  • v1.0: Initial release

Place the Python script in the /scripts dir inside your NUKE_PATH (see my previous post 2011-10-28-nuke-63-small-studio-setup-for-windows-osx on setting this up). Add the following to your menu.py:

import readFromWrite
nuke.menu( 'Nuke' ).addCommand( 'My file menu/Read from Write', 'readFromWrite.readFromWrite()', 'shift+r' )

You should now be able to select any Write node and hit Shift + R to generate a Read node!