Note: This tool is designed to be used on PL-C and PL-D Pixelink camera models only. For older PL-A and PL-B Pixelink cameras, please use the FFC Wizard tool.
Camera sensors almost always have some pixels that don’t have the full dynamic range that they should. These dead, hot (too bright) and cold (too dark) pixels can be corrected for with the Static Defective Pixel Correction (SDPC) routine. This routine corrects the defective pixel with interpolated values based on neighbouring pixels.
Typical correction methods include averaging with immediate same-color neighbors, or even going out to the second set of neighbors. Correction can span multiple pixels if by some chance many consecutive pixels in a row are defective. Boundary conditions also exist to handle defective pixels at the edges of the image array.
This type of calibration is already done at the factory but you may wish to re-calibrate your camera in the field if you are noticing bad pixels with your specific camera settings. The Pixelink Capture and Pixelink SDK come with the SDPC Wizard application to help you accomplish this.
When to Use
Use the SDPC Wizard when:
- Your exposure rate is very long. Larger exposure times cause a wider variance in charge accumulation per pixel and thus pixel intensity. This can lead to more noticeable differences in pixel intensity.
- You are using a high gain. Increased gain can have a similar effect to large exposure times. It amplifies differences in pixel intensity as well as noise.
- Your operating environment is of a different temperate room temperature. Pixel noise and responsivity can change with temperature. For reference - factory calibration is performed at a camera temperature of 40 degree Celcius.
- You wish to save your factory default or current SDPC values to file.
- You wish to restore a previously saved set of SDPC values or use custom user values.
The procedure is best done in your operating environment after the camera has been running for a period of time. If this is not possible, simulate as close as possible with regards to camera settings, running camera temperature and lighting.
It is recommended you remove the lens from your camera before using the SDPC Wizard. Otherwise, optical effects can cause pixels to be incorrectly flagged as defective.
Once your camera is connected and set to the desired configuration, launch the SDPC Wizard:
- Start Menu => PixeLINK => Tools => SDPC Wizard
After clicking next on the opening dialog, your camera will be detected and you will be brought to the SDPC Configuration page. Check the model and serial number to be sure you are connected to the desired camera.
From this page you can load factory default gain/exposure, load previously saved settings from memory channel 1, or change the exposure/gain manually. Video preview is available for all of these options to check your target image and can be left running during calibration if desired. This will allow you to see the before and after effects of the calibration. The SDPC Preview check box will allow you to manually do this.
By default, the software will save a backup of the calibration data, the new calibration data and a log file to your “Documents” folder. You may disable any of these options or change the save path if desired. You may also load SDPC calibration data from a data file instead of re-calibrating manually.
Clicking “Start SDPC Calibration” will start the calibration progress. A dialog will launch asking you to provide a dark scene to analyze. It is recommended you use the sensor cover that came with your camera to provide the dark scene for calibration. Once this is complete, the software will average a series of frames, looking for pixels behaving poorly and flag any found. The main failures looked for are hot pixels which have a significantly higher value compared to the rest of the scene and noisy pixels which have too much fluctuation in value from frame to frame.
If you enabled Bright Frames SDPC testing from the configuration page then after dark frame analysis is complete you will be asked to provide a uniform bright scene for similar type of bright frame analysis. It may not be possible in all configurations and environments to provide the required bright target scene. In that case, the dialog can be cancelled and calibration will be done using only the dark scene data. The vast majority of marginal pixels are captured in the dark scene analysis. Skipping bright scene calibration should not a major concern for most users.
The final list of marginal pixels is analyzed based on failure mode and weighted to produce a list of defective pixels. If more pixels are found than can be automatically corrected for in hardware, the worst offenders are flagged as defective. All of the analysis data will be saved to file. The output file will flag each type of pixel failure as follows:
Pixel_Good, // This pixel is not known to be defective (default) Pixel_Blinker_Dark, // Too much temporal variation in low light Pixel_Cold_Dark, // Low under low light Pixel_Hot_Dark, // High under low light Pixel_Blinker_Bright,// Too much temporal variation in higher lighting conditions Pixel_Cold_Bright, // Low in higher lighting conditions Pixel_Hot_Bright, // High in higher lighting conditions Pixel_Bad_User, // User flagged defective pixel
This data can be used by the user for manual editing or post capture image processing if desired.