At the risk of sounding like the cynical old man that I actually am, I wanted to take some time to talk about LUTs because of the way the idea is being presented to creatives. Serious videographers have been aware of LUTs for a long time because they have been around for a long time, while many still photographers have the impression that this idea is brand spanking new. As the topic can get rather deep technically, I referred to the superb article here. The people over at http://lightillusion.com are experts in this field. Any factual screwups are my own. I also recommend the excellent folks at http://mixinglight.com for their site and tutorials.
What is a LUT
A LUT, or Lookup Table, is a mathematical algorithm used to perform a transformation on a source to produce an output in some consistent manner. Filmmakers have used LUTs for a long time to give a look and a feel to footage.
Let’s look at two examples from TV from about ten years ago. If you were to go find an episode of CSI Miami and watch it solely for colour treatment, what would you discover? You would find an orange tint in all the outdoor shots. The idea is that the association with orange is warmth, so a subtle application of orange makes the scene subconsciously warmer. On the first few episodes, the folks went a bit overboard and the actors started to look jaundiced, so they backed off. In a relatively short time, viewers reported that they felt warmer when they watched the show. This is not a new idea. In the distant past we used filters on camera lenses and gelled lights to produce similar results. They work best when they are subtle in general, but the density of the effect is up to the director and the colourist.
If you were then to go find an episode of CSI : New York, especially one of the first episodes in Season One, and watched it closely for colour, you would discover that it has an overall bluish cast, and is a bit dark and a bit contrasty. The creative goal was to create the sense of a large city that was edgy and not termperature cold, but cold in emotional context. It was overdone at first and then corrected, because initial viewers said they hated the show because it was cold and dark.
These are two examples of what we will find in today’s digital LUTs. Another approach for a digital LUT is to create the look and feel of an older film stock used when films were actually shot on film. Certain film stocks were known for a colour bias, a level of contrast and a level of softness.
In order to recreate these looks, LUT designers would create these algorithms and then package them up for easy reuse and consistency, saving time and money during colourization.
Let’s Dig In
LUTs can be called primary LUTs, meaning that they impact all of the colour space, or secondary LUTs where they only affect certain Hues and Hue Ranges. A LUT that creates the look of film is often a primary LUT whereas the now overused Orange/Teal LUT is more specific to highlights and shadows and two Hue Ranges.
We also encounter 1D and 3D LUTs.
1D LUTs are very simplistic and don’t have enough information to be used in full colour grading exercise, so tend to be centred on a single attribute, such as the white point of a display, or an overall colour balance adjustment. We will typically find that 1D LUTs are referred to as uncoupled, meaning a change to one channel has no impact on another channel, so if we raise the value of Red on an inspected area, it makes no change to Green or Blue. This means that for any input, the output can only have one value. Consequently we might see 1D LUTs used for global adjustments to brightness or contrast. If you want to alter the different channels differently, you use more than one 1D LUT, for example three 1D LUTs for each of Red, Green and Blue.
Prior to 3D LUTs, a 3x3 matrix was often used to describe the values from 1D LUTs in a single package. The challenge here is both in application and that even in the matrix, you do not have all the control you might want. They exist, and this type of LUT is often used in what industry professionals call “Look LUTs”. To put this term in context, think of a simple preset that you might download and apply. Effective, but not creative because the output is pretty boring.
This is why 3D LUTs are so popular. They considerably more complicated mathematically but get to a more refined output more quickly. A 3D LUT, often called a CUBE or CUBE LUT represents the alterations from source to output in a three dimensional space. 3D LUTs have enormous levels of control at a very fine level, but using them will typically require more computing horsepower. 3D LUTs are coupled, so a change to one channel will also engage some changed in other channels. This can produce much more powerful and less stark results. When we use software such as DaVinci Resolve to generate a LUT, we always get a 3D LUT.
We can see in the graphic that a 3D LUT describes transformations in three channels at once, based on the intersection of the three planes for each origin, and then the transformation for the output. This means a level of fine control in a non-linear fashion in what is called a volumetric colour space. For really precise transformations, a professional colourist will use a 3D LUT. But wait, there’s more…
As we talk about the differences in a file whether 8bit or 10bit video, or 8 bit vs 14bit for stills, the size of the LUT also has impact on its effectiveness. To get really tight colour matching across different systems, we would need very large LUTs. They get big very quickly, because however many levels are assigned to each individual colour, because this occurs in three dimensional space, we have to deal with cubes. Thus a very simple 5 step LUT will have 5^3 entries or 125 entries in the LUT file. If we went up to a LUT with 64 steps per colour, we would be dealing with a LUT file with 262,144 entries. The larger the LUT, the finer the control but the more processing is required.
You will often find that when you download a LUT you will not know about the number of steps in each colour in the LUT. This is why you will get different output with different LUTs that do ostensibly the same thing. You will get what you pay for when you buy professional LUTs, but it is important to measure your return on investment.
And That’s The Problem
Now that LUT is the buzzword of the day for still processing, we are awash with free or cheap LUTs, but just like buying presets, they tend to be limited in capability because a really comprehensive LUT is going to be harder to build, a larger file and require more processing capability at the CPU or GPU level. This flies in the face of the Instagram look model, which is a look that gets applied very fast and is rarely ever adjusted after being slapped on.
There’s NOTHING creative about slamming a preset on anything, and there’s nothing creative about slamming a LUT on something either. You are applying a mathematical transform and your ability to finetune a LUT is going to be rather limited unless you really dig in and learn how to build LUTs. And by build, I do not mean click a menu item that says “Generate LUT”
LUTs in Photoshop
A while ago, Adobe modified the Color Lookup Adjustment Layer, to allow the user to insert a 3D LUT and have it applied to an image. Because Adjustment Layers come with masks, have opacity controls and a selection of blend modes, using really nice 3D LUTs in Photoshop does not have to look like you ran around with an ink pad and a rubber stamp.
LUTs in Lightroom Classic and Lightroom CC
You can use LUTs in both versions of Adobe’s Lightroom but it’s a pain. Neither Lightroom can handle a .cube file as Lightroom Profiles need to be in the XMP format. You can go online and buy Lightroom-ready LUTs from places such as Lutify.me but seriously, go back a paragraph and do it with regular CUBE files in Photoshop using a Color Lookup Adjustment Layer.
LUTs are brilliant tools for consistency in output, they can very quickly apply a look to footage or a still. Think of the sickly green overlay in all scenes in The Matrix when the characters are “in” the matrix and the very different look when the characters are outside the matrix in the submarine for example. The initial colourists had to work very hard to create a consistent look that could be used on all kinds of different footage. Once done, only then could they build a LUT for it. And that LUT is brilliant and will work a charm if you want every scene to have the same look and feel from the colourist perspective. The creativity was in the creation, not the application.
It’s one of the reasons I get so ticked off when I hear someone call themselves a colourist because they figured out how to stick a LUT on something. These people are no more a colourist than an ankylosaur is an aircraft carrier.
Do you have an idea for an article, tutorial, video or podcast? Do you have an imaging question unrelated to this article? Send me an email directly at email@example.com or post in the comments. When you email your questions on any imaging topic, I will try to respond within a day.
If you shop with B&H Photo Video, please consider doing so through the link on thephotovideoguy.caas this helps support my efforts and has no negative impact whatsoever on your shopping experience.
If you find the podcast, videos or articles of value, consider clicking the Donation tab in the sidebar of the website and buy me a coffee. Your donation goes to help me keep things going.
I'm Ross Chevalier, thanks for reading, watching and listening and until next time, peace.