Preserve Originals and Processing History
A team overwrites its originals after changing contrast.
Read the full lesson and open its transcript below. A narrated video will be added after production.
What you will learn
A team overwrites its originals after changing contrast.
A practical question
Someone adjusts contrast on every archived photograph and accidentally overwrites the raw files. Months later, a new inspection method needs the original shadows, but those values are gone. A reproducible image pipeline begins by protecting the captured master.
The useful idea
Store the source image with its capture metadata and calibration. For each derived result, record the processing steps, parameters, and software version. Keep the output linked to those inputs. A display-ready export can be useful, but it doesn't replace the raw evidence that produced it.
Work through the example
Pick one published inspection result. Starting only with the stored master and settings, try to recreate it. If you can't tell which crop, threshold, or scale was used, the process is missing a necessary record. If a manual adjustment was made, name it rather than presenting the output as automatic. A processing history should let another person start with the raw capture and recreate the final result. Write down the order of operations, parameters, software version, calibration, and any manual decision. Save a representative intermediate image where a consequential change occurs. If a later result differs, these records help locate the step responsible. A final image alone cannot show whether an apparent scratch was captured, sharpened, or introduced along the way. Use this case to test your understanding: Recreate one published result from a raw file and recorded parameters. Here is the limitation to watch for: A processing history that omits software and data versions may not be reproducible. This supports a careful conclusion: Version the source image, processing settings, calibration, and output together. Ask a second person to recreate one published output from the raw capture and recorded settings. The first missing parameter they encounter shows exactly what the processing history still needs.
Limits and next step
Keep the history in a form another person can follow. Next, we'll compare two versions pixel by pixel and learn what a simple error number can and cannot tell us.
Worked example
Recreate one published result from a raw file and recorded parameters.
Quick check: test your understanding
Try it yourself: Consider this case: Recreate one published result from a raw file and recorded parameters. Explain how you would check the result. In your answer, address this limitation: A processing history that omits software and data versions may not be reproducible.
Check the reasoning
Version the source image, processing settings, calibration, and output together. A processing history that omits software and data versions may not be reproducible.
What is the main idea behind preserve originals and processing history?
Version the source image, processing settings, calibration, and output together.
How is it used in the example?
Recreate one published result from a raw file and recorded parameters.
What limitation should be checked?
A processing history that omits software and data versions may not be reproducible.
Further reading
Digital Image Processing, 4th edition — Rafael C. Gonzalez and Richard E. Woods.
Original ToolsInVision teaching examples. Editorial approach · Suggest a correction
Accessible referenceLesson 72 transcript317 words · Narration ready for the upcoming videoOpen transcript
Someone adjusts contrast on every archived photograph and accidentally overwrites the raw files. Months later, a new inspection method needs the original shadows, but those values are gone. A reproducible image pipeline begins by protecting the captured master.
Store the source image with its capture metadata and calibration. For each derived result, record the processing steps, parameters, and software version. Keep the output linked to those inputs. A display-ready export can be useful, but it doesn't replace the raw evidence that produced it.
Pick one published inspection result. Starting only with the stored master and settings, try to recreate it. If you can't tell which crop, threshold, or scale was used, the process is missing a necessary record. If a manual adjustment was made, name it rather than presenting the output as automatic. A processing history should let another person start with the raw capture and recreate the final result. Write down the order of operations, parameters, software version, calibration, and any manual decision. Save a representative intermediate image where a consequential change occurs. If a later result differs, these records help locate the step responsible. A final image alone cannot show whether an apparent scratch was captured, sharpened, or introduced along the way. Use this case to test your understanding: Recreate one published result from a raw file and recorded parameters. Here is the limitation to watch for: A processing history that omits software and data versions may not be reproducible. This supports a careful conclusion: Version the source image, processing settings, calibration, and output together. Ask a second person to recreate one published output from the raw capture and recorded settings. The first missing parameter they encounter shows exactly what the processing history still needs.
Keep the history in a form another person can follow. Next, we'll compare two versions pixel by pixel and learn what a simple error number can and cannot tell us.