QC Series – Tackling Pixel Errors During Quality Control

In the world of film and television production, precision and attention to detail are of utmost importance. Every frame, scene, and shot must meet the highest standards of quality. However, despite the use of advanced equipment and meticulous planning, technical issues may arise. One such issue is the presence of Pixel Errors in your footage. In this article, the fourth and final part of our QC series, we will guide you on how to use Silverstack to identify such defective pixels. So let’s have a look at why Pixel Errors affect the viewer experience.

Pixel errors: A customer’s eye view

Pixel errors, visible in dark scenes or on a character’s face, can distract customers. In the final scene of Red Bull Media’s documentary¹ on kayaker Nouria Newman, a persistent white pixel from the main camera appears (1:20:00 to 1:22:00), impacting the viewer’s experience despite Nourias outstanding performance. An example of what this may look like is shown below.

Understanding pixel errors

To address pixel-related issues, it’s important to first understand their physical background.

How light becomes a pixel

In an image sensor, a photocell absorbs incoming light which generates an electric charge proportional to the light’s intensity. This charge accumulates over the exposure period, reflecting the overall brightness of the scene. 

The accumulated charge is then converted into a voltage signal, which is digitized and collectively forms pixels in a grid on the image sensor. The digital information is used by the image processor to reconstruct a visually accurate representation of the captured scene resulting in the final image.

How pixels become faulty

Defective pixels on a sensor can arise from manufacturing defects, environmental factors, physical damage, or the natural aging of camera components. While manufacturing processes aim for minimal imperfections, a modern 4k camera sensor with around 12 million photocells will still have some defective pixels, but these are corrected by the camera’s internal image processing and are not a problem.

Exposure to extreme conditions such as lasers, environmental conditions such as heat, and physical impact can contribute to the development of malfunctioning pixels. Over time, wear and tear on the sensor can also lead to defects. 

Camera image processing techniques such as camera image sensor calibration, and black reference images are utilized to identify and compensate for these defects in the camera. And of course, it is possible to fix pixel errors in post.

Defining types of pixel errors

As is so often the case when it comes to naming image artifacts, different terms are used to describe the same phenomenon. Some people refer to the problem as “stuck pixel,” which is the same as what we call “hot pixel.” However, others use the term “hot pixel” only to describe pixels that appear when the ambient temperature is high. In this article, we will use the Netflix QC glossary’s definition to distinguish between the two terms.

We can distinguish pixel errors by their brightness (dead/hot). And secondly, whether they are constantly visible or appear irregularly or even periodically.  

Dead pixels and hot pixels:

Dead pixels are non-functional pixels that do not absorb any incoming light or generate any electric charge or voltage. They therefore appear as tiny black dots in your image or video.

Hot pixels are pixels on a camera sensor that generate an abnormally high voltage signal, resulting in bright white, red, green or blue colored dots in footage that do not match the surrounding scene’s brightness. The color or appearance relates to the defective pixel color on the sensor.

Dead and Hot Pixels can either manifest as permanent or intermittent pixel errors.

Permanent and intermittent dead/hot pixel error

Permanent Pixel Error either appearing as a dead or hot pixel error, is when pixels persistently appear in an entire shot, maintaining a consistent state without variation. Once identified as defective, these pixels remain unchanged throughout the shot. 

Intermittent Pixel Error is a term used to describe the occasional appearance of dead or hot pixels that appear and disappear intermittently during image capture. These errors may be related to the image content and may only appear in areas of contrasting elements or edges. Alternatively, they may be related to environmental conditions, such as temperature or operating hours. These types of errors can be more disruptive, as they may appear to flicker or blink.

How can we use Silverstack’s software features to detect pixel errors given this background knowledge?

Pixel error QC with Silverstack

Quality control is crucial for identifying and managing defective pixels. This is how Silverstack helps you track defective pixels in your Footage. Let’s talk about the best decoding configuration and playback settings to find pixel errors and then look into visual inspection and technical grades to improve the visibility of pixel errors.

Decoding resolution

At first make sure to set the decoding to full resolution to avoid pixel interpolation and keep each single pixel visible. This can be done in the playback HUD or in the playback mode menu in the top bar. 

Special playback settings can further assist you in pinpointing pixel problems.

Playback view: Fullscreen playback

Silverstack’s special playback view, or the keyboard shortcut Control-Command-4, can help pinpoint pixel problems. Especially when activating the fullscreen player.

In order to make the video playback more immersive, you can click on the player icon located on the top bar next to the video controls or use the keyboard shortcut Command-Shift-F. This will expand the video to fit the entire screen. Additionally, you may choose to hide the sidebars to view more of the video content at once.

Playback mode: Every frame

In the same menu as before, set the playback mode to every frame. This ensures that in case of intermittent pixel errors, no frames are missed.

Zooming

When dealing with pixel-related issues, zooming can be a powerful tool. By zooming in on your footage, you can inspect it at the pixel level, making it easier to identify individual pixel errors. Silverstack’s zooming capabilities further enhance this process. 

However, it’s important to keep in mind that you should view the image at least at 100% zoom level, as interpolation may cause defective pixels to no longer be visible at smaller zoom levels. If you want to zoom in on an image, simply scroll to adjust the zoom level.

External video output

If you’re using an external HD-SDI monitor for quality control, it’s important to ensure that it displays the signal in full resolution. You can configure the output by accessing the Silverstack settings and selecting external video². 

Visual controls and technical grading for better visibility

Silverstack offers visual control settings that enable you to meticulously examine your footage, facilitating the identification and resolution of pixel issues. By carefully observing each color channel, you can more effectively spot pixel errors.

The channel view

To access the visual controls, you can either find it on the top bar next to the timecode or use the shortcut Command-F11. Once you activate the visual controls, you can check the quality of each color channel individually for any defective pixels. Remember that bright white dots indicate hot pixels, while dark dots indicate dead pixels. 

It’s important to note that the kind of pixel error you can see depends on the image content. Hot pixels are better spotted in dark areas, while dead pixels have better visibility in bright areas. It is also easier to spot a hot red pixel in an image with a blue or green background instead of red. For more information, please refer to the final chapter on checking for camera pixel errors.

Focus assist

Silverstack has another useful feature known as focus assist. It helps to determine if an image is in focus by highlighting edges with a red outline. Furthermore, it can be utilized to identify faulty pixels, as they usually exhibit sharp edges that are accentuated by the focus assist. By toggling this feature, it becomes easy to identify distortions in the image and locate any defective pixels.

Using a technical color grade for better visibility

Technical grading manipulations like saturation and contrast adjustments amplify the visual characteristics of pixel errors. These adjustments can intensify the visibility of pixel errors or irregularities in color, brightness, and edge details, making pixel errors more conspicuous in the processed image.

When using Silverstack Grading Mode, activate CDL and LUT grading to adjust saturation and contrast settings. Remember to adjust the grading settings for each color channel individually for optimal visibility of the pixel errors.

Saturation Increase:

When saturation is increased, colors become more vibrant and pronounced. Pixel errors, especially hot pixels, may stand out more prominently as they fail to blend seamlessly with the surrounding colors.

Increased contrast:

Enhancing contrast boosts the differentiation between light and dark areas in an image. Pixel errors, especially those with irregular brightness, become more apparent against the heightened contrast, making them easier to spot.

Playback controls via shortcuts are invaluable for pixel error detection*, allowing users to meticulously review and analyze footage frame by frame, making it easier to identify and address any visual anomalies.

Now we know how to find defective pixels and make them more visible with the SST features. What can we do to test the camera?

Pixel perfection: Checking for camera pixel errors and post-fix solutions

Camera check for pixel errors

In case you notice defective pixels in your footage or you want to check your camera during prep, you can setup a similar test during the preparation phase, as described in a previous blog post³.

When shooting the reference images, make sure to shoot not only a bright white surface to detect dirt, but also a black image with lens coverage, and also a bright, luminous red, green, and blue surface to test the individual color channels. 

You might want to point the camera at a bright display or use an RGB LED light, defocus the lens, pan, and tilt the camera slightly while recording. Record a video sequence like this to have all the colors included⁴. 

Certain cameras provide a black shading calibration that can potentially resolve any pixel errors. If this feature is available, it is recommended to perform the pixel error check twice – once before and once after the calibration is executed. This will enable you to compare the recordings and determine whether the image quality has improved and the number of pixel errors has decreased.

If you find several pixel errors during this test, it may be time to contact your rental or camera manufacturer regarding a calibration service for the camera image sensor.

Fix it in post

To remove dead pixels in your recorded footage, a VFX paint fix must be applied. It is suggested to paint out the pixels in the non-graded archival master (NAM) to ensure that any fixes made will be reflected in all subsequent deliverables, including the graded archival master (GAM), video display master (VDM), and all final master deliveries.

Conclusion

Dealing with defective pixels can pose a significant challenge in film and TV production. However, with a solid understanding of underlaying technical background and established quality control processes you can effectively manage these issues. 

Silverstack, with its advanced playback settings, zooming capabilities, special playback views, and general playback controls provides filmmakers with the tools they need to detect and address pixel problems. This ensures that every frame meets the highest standards of quality. By diligently managing these pixel anomalies, you can deliver a final product that captivates your audience with its visual excellence.

*General Playback Controls (toggle, rewind)

These SST playback controls will help you to thoroughly review your footage and efficiently identify and address any pixel anomalies. 

You can control playback using the J, K, and L keys, which is similar to many editing applications. This allows you to speed up playback, slow it down, play it in reverse, and step through individual frames. This functionality, also known as playback, shuttle, and jog, can be especially useful in spotting intermittent pixel errors.

Frame Stepping is available via arrow keys. The shift and shift+control keys allow larger steps. Combining frame stepping with zooming into the Pixel Error is very helpful to check for if a pixel error is intermittent in a clip or between clips in a timeline.

Read more about playback features you might want to visit this article in our Knowledge Base⁵.

1. Red Bull Media’s documentary on kayaker Nouria Newman ↩︎
2.  HD-SDI Output in Silverstack Knowledge Base ↩︎
3. Sensor Check Blog Article ↩︎
4. Dead Pixel Tester Video ↩︎
5. Playback features 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