Multi-Camera Matching and Creating a Streaming Picture Profile Using the Sony PMW-300K1
By using a vectorscope and a waveform and the process described in this article, you can limit or eliminate distracting color-matching issues between cameras and ensure your audience has an immersive experience--at least as it pertains to the quality of the video.
Our primary focus in this article is to lay down a process to help you match your cameras for multi-camera live streaming production. If you’ve ever produced a multicamera shoot, you may have noticed variations to blacks, whites, and colors between cameras. These variations can subtly take the viewer out of the “moment,” stopping them from fully immersing in the event and/or story you are producing.
There have been times in shows when we have “matched” cameras to our eyes, and everything seemed to be the same and then someone with a bright magenta shirt would walk on stage and that color magenta would be different on all the camera angles. This obviously comes across to the viewer as a mistake. By using a vectorscope and a waveform and the process below, you can limit these variables and ensure your audience has an immersive experience, at least as it pertains to the quality of the video.
You can apply almost all of this process to any video camera, but for the sake of specificity, we’ll demonstrate this process on the Sony PMW-300K1, a camera that we use frequently and one that’s well-suited to streaming production.
The Sony PMW-300K1 is the upgrade to the venerable PMW-EX3, a game-changing video camera for low to medium-sized budget productions. Released in 2008, the EX3’s price point, form factor, and lens interchangeability, among other features, quickly vaulted it to industry workhorse status.
With the release of the PMW-300 in October 2013, Sony added better signal processing, MPEG HD422 50Mbps recording (plus XAVC-Intra 100Mbps and XAVC Long 50/35/25Mbps recording in 10-bit), and a new Multi-Matrix mode to picture profiles among other enhancements. The new Multi-Matrix mode allows for even finer tuning of color than the EX3.
Understanding Waveform and Vectorscope
When matching cameras, the two most important diagnostic tools available to the operator are waveform and vectorscope. A waveform monitor displays the luminance (brightness) values within an image, while a vectorscope displays chrominance (color) values.
Waveform
Figure 1 (below) shows the Leader LV5330 test monitor’s waveform monitor, displaying luminance values for the 3 color channels within the image (RGB). This method of displaying color channels separately is referred to as Parade mode, and is necessary for accurate color matching. As denoted by the “RGB” in the top-left corner, the monitor shows Red, Green, and Blue from left to right. It’s important to pay attention to the letters’ order of appearance, as some monitors will display Red, Blue, Green from left to right, in which case the notation will read "RBG." Luminance level is displayed on the Y-axis, as an IRE value from 0 to 110, with 0 representing absolute black, and 110 representing pure white overexposure.
Figure 1. The Leader LV5330 test monitor’s waveform monitor displays luminance values for the 3 color channels in traditional red-green-blue (RGB) order.
In Figure 1, the lines on the waveform are all relatively straight because the image being analyzed is of a white card that is evenly lit. The waveform represents the exposure values across the image from left to right. That is because the waveform is being displayed horizontally (referred to as a horizontal sweep). Waveform monitors generally allow for waveform viewing vertically (i.e., with a vertical sweep), from top to bottom. Both vertical and horizontal viewing are necessary for proper lens shading.
Vectorscope
Figure 2 (below) shows the LV5330’s vectorscope displaying chrominance values for the incoming feeds. A vectorscope has six points of color reference around the inside edge of the circle, each denoted by a box with a set of crosshairs inside.
Figure 2. The LV5330’s vectorscope displays chrominance values for incoming feeds.
Starting as close to center as possible at the top of the circle, and moving clockwise from there, the colors are Red (R), Magenta (M), Blue (B), Cyan (C), Green (G), and Yellow (Y). These points represent color standards that have been set across the film and broadcast industries.
A vectorscope should be used only with an accurate chip chart (Figure 3, below), which displays colors to represent each of the above-mentioned color points on the scope. Depending on the type of chart being used, the number of color chips displayed may vary, but all accurate charts will at the very least display the 6 vector colors. Most charts will also display an 11-step crossed grayscale, showing 9 values of gray between white and black in both directions.
Figure 3. The DSC Labs chip chart.
More advanced charts display a large range of color values, which allows you to make finer adjustments when matching cameras. These additional colors will generally include 4 different representations of common human skin tones (Figure 4, below). These skin tone swatches can be extremely helpful when matching PMW300K1s, as ensuring consistency in skin tones between cameras is of utmost importance. Skin tone points will fall on the line between Yellow and Red on the scope.
Figure 4. Vectorscope skin tones.
The center of the vectorscope represents a lack of color (black, white, and all values of gray contained in an image). As color is introduced, the points on the scope move out from the center towards the colors that are being displayed. The further away from the center that a point travels, the higher the saturation (aka chrominance) for that given color value. When calibrating a camera, the goal is to have each of the color points fall exactly in the center of their respective boxes.