Color Management for Delivery to any Device
Color Management for Delivery to any Device
You’re a content creator, you’ve probably faced the issue of inconsistent image renderings at some point. Images look one way in your color suite and look a thousand different in the wild when viewed on your phone, computer, tablet, TV, or other screen in any viewing environment. It’s easy to feel like we have no control over how viewers experience our work. Isn’t it time to start pretending?
Well, the answer is yes. While we can’t control some of the most important factors in our work’s visual experience, there are a few things we can control or at least try to control. Let’s take a look at what causes image renderings to translate poorly first.
With that in mind, let’s look at some of the practical strategies that you can use to achieve more consistent results on different screens and different viewing environments.
Let's get started by defining our goal before we move forward. While it would be ideal to have every image look the same on every device and in every environment, unfortunately, that's not possible for various reasons that we'll discuss. So, what should our goal actually be?
As a colorist, my main objective is to provide the most accurate representation of the creator's visual intent to as many viewers as possible. Now, let's delve into the reasons why images don't always translate well across different platforms. The first factor is the varied display standards.
Every piece of content is mastered to a specific target gamut and gamma, which essentially define the range of colors and luminance/contrast. Different displays, capture devices, and intermediate containers like ACES have their own gamut/gamma pairs. For instance, common display gamut/gamma pairs include Rec709 Gamma 2.4, sRGB Gamma 2.2, and P3 Gamma 2.6, while common capture gamut/gamma pairs include Arri WCG LogC, Sony S-Gamut S-Log3, and RedWideGamutRGB Log3G10. The mismatch between the mastering gamut/gamma of a piece of content and the native gamut/gamma of the display it's being viewed on is a major cause of poor image translation across devices. For example, if I grade a film on a P3 Gamma 2.6 theatrical projector, it won't appear correctly on a Rec. 709 Gamma 2.4 TV.
Another factor is the varied viewing environments. It may seem obvious, but it's easy to overlook the fact that the same content will appear differently when viewed outside on a sunny day compared to a darkened room. The lighting conditions greatly affect the perception of brightness. So, considering these factors, our goal should be to minimize the discrepancies in image translation across devices and environments as much as possible.
This idea applies not just to the overall amount of light in a room, but also to the color of the light. If you view the same content on the same device in a room with natural daylight versus a room with tungsten lighting, it will appear warmer or cooler respectively. Our eyes constantly adjust to the environment, so they are always adapting to the white balance.
In an ideal world, we would have a consistent amount of artificial light that is not brighter than our display, and its color temperature would match the white point of the display. The further we deviate from this ideal, the more noticeable the difference will be compared to what was seen in the color suite.
A grading suite with controlled artificial light. The lighting is dimmer than the mastering display and matches its white point in color. Varying display accuracy We have discussed the importance of matching the mastering gamut/gamma and display gamut/gamma.
However, there is another aspect to consider: just because a display aims for a specific standard doesn't mean it actually achieves it. Display accuracy refers to how closely a display adheres to its intended standard. When a display performs poorly in this aspect, you may experience issues like color washes, shifts in hue, clipping of highlights/shadows, and oversaturated colors.
Displays that fail to meet their targeted standards are a significant factor in poor image translation.
Special instruments are used to measure the accuracy of a display's gamut and gamma, and adjustments are made to minimize any inaccuracies. In any color workflow, we need to consider the accuracy of two displays.
End viewer display
The average display that end viewers use is likely to be highly inaccurate. There are various reasons for this, but ultimately, accuracy is not a selling point for most displays. Consumers are often attracted to the brightest and most vibrant displays in stores, and manufacturers are aware of this.
While this situation is unlikely to change, we can still be hopeful about the increasing demand from consumers and image-makers for a mode or preset that prioritizes display accuracy.
This is exactly what the "filmmaker mode" TV setting aims to achieve, with the support of the UHD Alliance and renowned filmmakers like Martin Scorsese and Christopher Nolan. When it comes to mastering display, it becomes clear how crucial it is to have an accurate one, considering the unpredictable accuracy of the average end-viewer display.
Metadata can include various details about the file, ranging from the program title to the name of the camera assistant who loaded the magazine.
It can also include details about the color range and/or brightness curve of the content, which certain devices and software can detect and use to adjust the image.
Let's say we have an image that was originally created in Rec. 709 but mistakenly labeled as Rec. 2020. Some software and hardware will read this Rec. 2020 label and make adjustments to display it as accurately as possible within a Rec. 709 range.
Problems can arise not only from incorrect metadata tags, but also from the absence of correct ones.
For instance, if we create an image in Rec. 2020 but forget to include the appropriate metadata, many programs and devices will assume it's in Rec. 709, resulting in a less vibrant representation of our content.
Now that we understand what causes image translation issues, let's explore some strategies to fix them.
Specify your deliverables - Make a list of the platforms where you'll be sharing your content. If your content will be displayed on screens with different color ranges and brightness standards, you'll need to create separate versions that are encoded for each specific standard.
For example, if I need a Quicktime file for a DCP and another Quicktime for uploading to Vimeo, I'll need two separate assets: one encoded for P3 Gamma 2.6 (for DCP) and another encoded for Rec. 709 Gamma 2.2 (for Vimeo). If you try to use the same file for both platforms, you'll end up with at least one rendering that doesn't look right.
Make sure to use a metadata inspector/editor to check for any incorrect entries, like using Rec. 2020 flag when your gamut is actually Rec. 709.
Another helpful tool for reviewing and editing metadata is Digital Rebellion's Pro Media Tools (when you are on a Mac).
Lower your expectations and understand that not everyone will have the ideal setup. Focus on providing the most faithful reproduction of the creator's vision to the majority of viewers.
If you can't afford a professional colorist or calibrated display, there are still cost-effective solutions. Create a consistent lighting environment and use a color-managed workflow. Keep an eye on your scopes to ensure proper shadow and highlight details.
Optimize your workflow by implementing color management. This will reduce the strain on your inaccurate display.
Pay attention to your scopes to ensure that the darkest shadows and brightest highlights on your histogram are reaching 0 and 100 respectively, without going too far beyond. Also, make sure that the red, green, and blue peaks are balanced for a neutral white balance. Avoid colors that go beyond the edge of the vectorscope and get clipped out.
Take a gentle approach when making adjustments. It's better to make small, subtle changes rather than bold or aggressive moves, as they may not translate well in different environments.
Understand that there will always be some compromise when grading on an inaccurate display.
By accepting this, you acknowledge that the maximum possible inaccuracy of your content is significantly higher.
Happy grading Y'all.