QC Series – Aliasing and Moiré: Navigating Quality Challenges in Film and TV

Previously in our QC Series, we highlighted the significance of quality control (QC) in the film and television industry and looked into Exposure QC.

In this article, we look at an image artifact, called aliasing, also known as moiré. Aliasing might occur from scaling or compressing an image, but can also be present in the original camera negative (OCN) from the point of capture. This makes it a somewhat complex or bothersome QC issue, which might require some troubleshooting to determine the source of the error.

We will first look at the technical background of the image errors and how they happen. Then, in a brief digression, we will look at the viewing conditions that are important for QC. And finally, we will focus on how you can use Silverstack to determine the origin of the image artifact, and how to prevent these errors from occurring. 

What is Aliasing and Moiré?

The alias effect is a phenomenon that occurs when digital cameras fail to accurately reproduce detailed patterns. This often results in unusual visual artifacts in photos or videos. For instance, when photographing narrow striped or patterned clothing, new wavy or swirly patterns may appear in the image. In addition, color noise, tone jumps (i.e. random color changes), irregular edges, and pixel artifacts on lines and borders may also occur as a result of this effect.

When you take a photo of a monitor, you may notice an effect where the pixel grid of the image sensor doesn’t align with the pixel grid of the monitor. This results in image disturbances appearing in the form of moiré patterns. This effect is quite common and can be seen in various situations where two grids with similar spacing are overlaid.

When a digital signal is inaccurately displayed or distorted because of a low sample rate, this typically leads to visible artifacts, or audible noise in audio signals, known as aliasing. This is related to the Nyquist frequency.¹

The Nyquist frequency refers to the highest frequency that can be accurately captured and reproduced through sampling. It is equivalent to half the sampling frequency of a digital signal. For instance, if an audio recording has a sample rate of 48kHz, then the maximum audio frequency that can be recorded is 24kHz. Any sound higher than 24kHz cannot be recorded effectively.

For the purpose of this article, we will discuss the limitations of image recordings. A sensor with 4K resolution can capture a maximum detail level of around 8 megapixels, due to its horizontal resolution of 3,840 pixels and a vertical resolution of 2,160 pixels. This means that finer structures with a higher resolution in front of the camera cannot be accurately recorded. The Nyquist frequency sets the limit that determines how fine structures can be captured in the digital signal. Anything beyond this limit is either not recorded or is recorded inaccurately, resulting in aliasing artifacts.

It is important to understand the difference between aliasing and moiré patterns. Aliasing occurs when a signal is not sampled correctly, resulting in incorrect frequencies appearing in the output signal. On the other hand, moiré patterns are caused by the interaction between the pattern of the image and the raster pattern of the sensor or display. This creates a visual phenomenon where two regular patterns overlap and create interference.

To simplify things, we will use the terms “aliasing” and “moiré” interchangeably in this article, as they are also used interchangeably in the movie industry.

Now that we have a better understanding of these technical terms, let’s take a closer look at what these image errors can look like in practice, and where they tend to occur.

What does Aliasing Look Like?

Image defects such as aliasing can have negative effects on the color and clarity of an image. Aliasing can cause “stair-stepping” or “jagged” edges on objects or lines that should be smooth. It can also result in iridescent “color streaks” or “rainbow artifacts” on edges and detailed areas due to variations in how the image’s color channels are resolved. When grid structures are recorded, interference can cause unwanted, wavy patterns to appear over the image, distorting its details and reducing its overall quality. Image defects are particularly noticeable on fine structures, especially those with higher contrast, such as actors’ costumes or set surfaces. Netflix’s QC Glossary defines aliasing with the Error Code I-303:

“Aliasing” is a visual artifact caused by limited spatial sampling during image processing and/or poor compression or a bad conversion. Object or text/graphic edges appear “jagged” or pixelated. Can also refer to a moiré (or “screen door”) pattern across certain parts of the image. This can be achromatic (it affects every color channel equally) or chromatic (it affects color channels independently).”²

Netflix QC Glossary

If aliasing effects become visible, it is important to identify their source and determine whether or not they were already present in the OCN. The next steps would be to initiate a reshoot as soon as possible or discuss possible retouching with post-production, including any associated costs. 

To recognize such image errors, it is necessary to control the viewing conditions to achieve a meaningful QC.

Viewing Conditions for QC

In our series so far, we have omitted one topic that plays a major role in the assessment of image quality: the viewing conditions for QC.

This is important to ensure accurate footage assessment and consistency across the production pipeline, which is particularly important during QC because image artifacts can be easily missed under poor viewing conditions.

To ensure proper viewing conditions for QC, you must consider at least the following five aspects.

1. Controlled Environment: QC should be conducted in a dedicated area to minimize distractions. For example, a Digital Imaging Technician (DIT) van or a specialized tent can be used for this purpose.
2. Ambient Lighting: In a controlled environment, you can set ambient lighting to mitigate glare and reflections on the monitor. Soft, diffused lighting is preferred to eliminate undesirable influences on color and gamma perception.
3. Lighting Consistency: Consistent lighting conditions should be rigorously maintained throughout the QC process in a controlled environment. Any fluctuations in lighting could potentially impact the on-screen appearance of the footage, influencing judgments related to exposure, color, and overall image quality.
4. Viewing Angle: Proper positioning of the monitor is crucial to ensure a clear, undistorted view for all roles involved in the QC, including the director, cinematographer, and Digital Imaging Technician (DIT).
5. Calibrated Monitors: Monitors used for QC should undergo regular calibration to uphold precise color and gamma representation. Calibration entails adjusting monitor settings to adhere to industry standards and establishing a dependable reference for color and exposure accuracy.

If the viewing conditions are appropriate for QC and we see aliasing artifacts in our footage, it is necessary to determine the source of the issue. As a first step, we will investigate our Silverstack setup*.

Relevant Silverstack Settings for Detecting Aliasing

As described above, it is important to rule out whether the aliasing artifacts occur in the OCN or are caused by the respective software and monitor setup used for the QC.

To prevent aliasing, Netflix states in its Help Center: “Render at the highest quality and resolution available whenever possible. Do not apply any unnecessary compression to source footage⁴.” It is therefore essential to ensure that the settings regarding resolution and decoding are set correctly.

Otherwise, the error can be easily missed, despite it being present in the OCN, which causes confusion for the production team.

This chapter provides an in-depth explanation of the key Silverstack configurations that are necessary for effectively identifying and managing aliasing artifacts. The following settings play a crucial role in maintaining exceptional visual quality and resolving any problems you may encounter while assessing whether aliasing is present in the OCN or not.

Decoding Resolution

To detect aliasing artifacts, it is essential to understand how decoding settings influence playback. With Silverstack, you can adjust the decoding resolution to optimize playback and transcoding performance. For instance, playing back 8K resolution footage at half resolution reduces it to 4K, while a quarter reduces it to 2K, making it easier to handle on slower systems. 

However, scaling down the image resolution may cause aliasing artifacts. Therefore, when checking for aliasing on the OCN, it is crucial to set the decoding resolution to full. 

We strongly advise you to check this setting first.

Zooming

When you’re using the playback HUD, you can easily select a zoom setting for the current clip. However, you need to be careful when zooming because it can cause aliasing. If you notice any aliasing issues during the QC check, double-check the zooming settings and set it to 100%. This will help you determine whether there is any aliasing in the OCN. 

It is also possible that aliasing image errors are not visible due to a reduced image level on the “fit” setting.

Checking the zoom settings should be the second thing you do if you notice any aliasing.

SDI Monitoring with Correct Resolution: 

Serial Digital Interface (SDI) is a commonly used method for transmitting high-quality video signals. When using external SDI monitoring, it is important to set the appropriate video format based on the resolution to accurately detect aliasing problems. This ensures that the image is evaluated at its intended resolution and therefore quality, providing a reliable representation during QC.

When displaying clips through an SDI device, it is essential to verify the external video out settings of Silverstack. You find them in the top bar menu under Playback > “Show External Video Device Settings.”

There are two options to choose from:

1. Scale to fit: This option fills the output display with the image to compensate for any difference in resolution. However, it may introduce aliasing.
2. 1:1 Pixels: This option displays the image in full resolution, which can crop the image if the output display has less resolution than the source. If you want to check the OCN on your external SDI monitor, 1:1 Pixels should be active to avoid any scalings.

Ensure that the 1:1 pixel mapping is also enabled on the external SDI monitor to prevent further scaling of the footage in case the monitor has a lower resolution than the source footage. 

In case you see aliasing on the SDI Monitor compare it with the image on your computer display to verify the source of aliasing.

For further reading see the HD-SDI in Silverstack Knowledge Base article⁵.

Anamorphic De-squeeze Settings

Anamorphic lenses are used in filmmaking to create a cinematic look for the footage, but, they can also introduce distortion. To achieve the correct aspect ratio and a distortion-free image, it is necessary to de-squeeze the footage and scale it. Proper configuration of de-squeeze settings is important to avoid issues such as aliasing. You must ensure that the configuration is accurate to maintain the visual integrity of the footage and prevent the artifacts from getting worse.

Further details can be found in the Anamorphic De-Squeezing Knowledge Base Article⁶.

Render Preset Settings for Transcoding

In case you notice any distortion (aliasing) in your rendered files, it could be due to certain settings that affect its presence. Therefore, it is essential to carefully check and adjust the following parameters within the transcoding settings:

• video codec
• resolution size
• bitrate
• source decoding resolution
• compositing resizing

Adjusting these settings properly in Silverstack is crucial to achieve optimal rendering and reduce aliasing artifacts. You can use the transcoding preview mode to see how changes in these settings impact the preview image. 

However, it is necessary to check the transcoded files in the end to ensure that they are alias-free, as the preview cannot guarantee 100% accuracy.

Summary and Outlook

Detecting the source of aliasing requires a clear understanding of the technical background of digital processes and consistent viewing conditions. Check the software configuration to avoid false positives within the QC report, and be aware that there is a significant difference between aliasing in the OCN and visible aliasing due to certain software and monitoring configurations.

If a quality problem arises, involve all relevant departments, including post-production, and check if they are also experiencing aliasing issues. The QC report generated by Silverstack will help to facilitate communication between the DIT, cinematographer, director, and post-production team in these cases, allowing for quick troubleshooting.

In the final part of this series, we will explore how defective pixels can be tracked down more easily using the QC features in Silverstack. We hope you’ve enjoyed the series thus far, and have learned some useful things about quality check in Silverstack!

*Pro tip: Use a conversion LUT to transform the LOG image into Rec.709 or Rec.2020. This process can increase contrast and saturation, making aliasing artifacts more visible during QC.

1. Nyquist frequency ↩︎
2. Netflix QC Glossary aliasing definition ↩︎
3. Read more about Allan Nielsen’s DIT Cart ↩︎
4. Netflix Help Center ↩︎
5. HD-SDI in Silverstack Knowledge Base ↩︎
6. Anamorphic De-Squeezing Knowledge Base Article ↩︎

All posts in this series:

• Silverstack QC Series
• QC Series – Checking Exposure: Measuring Light in Silverstack
• QC Series – Aliasing and Moiré: Navigating Quality Challenges in Film and TV
• QC Series – Tackling Pixel Errors During Quality Control