2.6 Using Infrared

22 Jun, 2017 | PFCLEAN, TUTORIALS

In our previous extended study material for the Level 2 PFClean training (2.5 Fixing Film Dirt) we were looking at tackling dirt in the Chanti clips. Both the chanti shots were suffering from large amounts of dirt and striations. We used the Workbench’s Auto Dirt Fix effect to fix or reduce these problems.

In this tutorial we are going to be looking at an alternative way of tackling dirt using the Auto Dirt Fix effect and an Infrared scanning pass in the Workbench.

If you haven’t completed the previous 2.1, 2.2, 2.3, 2.4 and 2.5 tutorials, I would suggest going back and starting from there as we will be using some of the material and profiles we generated.

NOTE: To follow along in this tutorial, you will need access to the clips provided for the class. If you don’t have them to hand, below you will find the link to download the footage again.

Tutorial Footage

The footage used in the training.

Footage: PFClean_footage_level_02.zip

Download

01. What is Infrared?

So what is Infrared? The principle of infrared (IR) is relatively simple. We scan using RGB to capture our images, with some specialised scanners an additional pass is run using infrared light to generate an alpha so that our images become RGBA. Infrared light passes through colour film but not through dust and scratches. This information is processed and turned into either a black and white matte or a greyscale map showing the exact location of the dust and scratches in the frame. Black and white film stocks containing silver halides will not work with this process, halides will reflect the infrared light and provide false results.

02. Loading Clips

In this tutorial we will be covering many of the things we covered in the previous study 2.5 Fixing Film Dirt tutorial. Let’s start where we left off in CHANTI_SHOT_001.  In the detection panel you will notice a greyed out button called Use Defect Map. If we had an IR pass available to us in this clip, this button would become active.  In tutorial 2.1 Degrain and Regrain,  we did import some other clips which do have an Infrared pass.  Let’s go to the workflow manager and double click on the Workbench under Media bin 2 to load the Adam and Eve clips.

03. Viewing the Alpha Channel Data

These clips have been scanned with an IR pass and the clip includes a defect map in the alpha channel. Click the A key to toggle between viewing the alpha channel and the RGB data in the Cinema. Scrub to frame 15 and compare the frame’s alpha channel with the RGB. You will notice that the visible piece of dirt corresponds to a black area in the alpha channel.

04. Setting up the Defect Map

Click the thumbnail at the top of the Effects Stack to open the Clip Manager to access the clip’s defect map settings. Since defects are marked as dark spots in the defect map, click the black alpha alpha_pfclean button. Click the View button to get see the areas identified as dirt in the defect map highlighted in the Cinema. For greyscale defect maps, like the one in the tutorial clip, you can define which values should be interpreted as dirt with the threshold slider. The default of 0.5 works well for this clip. Decreasing this value will increase the amount of dirt and potentially grain and image area that is pulled through the matte. Once we are done let’s switch of the View button.

We can also assign whether the clip has a black or white defect map in the media admin panel, but calibration of the map can only happen in the Workbench and Digital Wet Gate toolsets.

05. Infrared Pass and Auto Dirt Fix

Click the + button on the left to create an Auto Dirt Effect which is under the Repair Operations. As soon as the effect is created, it reads the clip’s defect map and fixes the dirt in the current frame.  It is worth noting as no detection is required processing will be quicker.

The parameters in the Fix section define how the Auto Dirt Fix effect fixes the areas that have been identified as dirt. The Type parameter defines whether dirt areas should be fixed spatially, that is by using neighbouring pixels to fill the area, or temporally, using motion compensated information from neighbouring frames. By default, small areas are fixed spatially and larger areas are fixed temporally. The Threshold parameter defines up to which size spatial fixes will be applied. If you de-select the Spa. button, all areas will be fixed temporally. Accordingly, all areas will be fixed spatially, if the Temp. button is de-selected. De-select the Temp. button to see what a spatial fix of this dirt area would look like.

We can also assign the grain profile we generated adameve_light in extended study 2.1 to further refine the results.

You can find more information about these and all the other parameters in the Auto Dirt Fix manual, which will open in your web browser when you click the help-effect button.

Using Infrared dirt maps in either the Workbench or Digital Wet Gate are a powerful way of working, removing large quantities of artefacts with ease.

Related Posts

0 Comments

Submit a Comment

Your email address will not be published. Required fields are marked *

Share This

Share This Post On: