So this has been covered a bit before but not enough to be sure.
I'm working with Alexa footage and have previously applied LUTs to the input.
It gives different results to the output and is easier to grade as ones grades aren't being then modified by another 'grade' downstream.
However, previous posts have suggested that applying a Lut on input doesn't give the maximum benefit of Log C footage - is this true?
What are peoples experiences of this?
Furthermore a DCP author has also suggested that he prefers to take a LogC delivery to code but obviously the placing of the Lut would be critical,
I usually use color transformations like LogC-Rec709 etc. in between the grading node tree. For Alexa you can generate different, LUTs from their website and play with them What creatively and in point of digital image math suites your workflow. For DCP delivery I grade all in Rec709 and when rendering out I use custom gamma correction LUT (2.2 -> 2.6) on a "track" nodes and render in XYZ option in render. For my Rec709 calibrated CRT and plasma it works perfect for DCP projections. At least in two different near cinemas on some tests.
But generally LogC LUT I use one or two operations before the and of node tree. Like having 5 nodes, the 4th node is LUT. But it is not exactly the same for all the projects. You can be a bit creative with this color transformations in what order of grading you use them. Generally using it as input LUT it just gets your options of manipulating image a bit more narrow.
You mean for cinema delivery, but what about television / web?
What do you mean with the need of applying LUTs in two operations?
Shouldnt you just apply an (alexa) output lut on, lets say the second node, that gives you the opportunity to grade before and after the Lut?