A converted MP4 broke the speed calculation: variable frame rate, duplicated frames, and 1.8 seconds that didn't exist
The situation
In a vehicular-homicide case a roadside business gave police a proprietary surveillance export. A detective couldn't play it on a standard computer, converted it with free online software, saved the result as an MP4, and deleted the proprietary export from the working folder. The converted video appeared to show the defendant's vehicle crossing the center line.
What the other side said the evidence showed
The prosecution's accident-reconstruction expert calculated vehicle speed from distance traveled across frames in the converted file, and interpreted a bright area near the tire as the tire crossing the painted line.
A file that plays is not a file that was preserved. Convenience and forensic accuracy are different goals.
The distinction that decides this case is container versus codec. A container is the file structure — MP4, AVI, MOV. A codec is the method used to encode the stream — H.264, H.265, MJPEG. Changing the container can preserve the encoded stream untouched. Transcoding replaces the stream itself, and transcoding is what convenience tools do by default.
Surveillance systems frequently record at a variable frame rate driven by motion detection, meaning the interval between frames is not constant and the native timing data is part of the evidence. Force that into a fixed 30 frames per second and the software must invent the difference: some frames get duplicated, some intervals get compressed, motion looks smoother, and the timing relationship between frames — which is what a speed calculation depends on — no longer reflects reality.
That is why an expert calculation performed on a converted file can be confidently wrong. The arithmetic is fine. The input assumed constant frame timing that the file no longer had.
What the examination involved
Established what the conversion actually did
Confirmed the detective's software transcoded rather than remuxed — replacing the encoded stream, not merely rewrapping it in a playable container.
Compared native and converted frame structure
The native recording used a motion-driven variable frame rate. The converted MP4 had been forced to a constant 30 fps, duplicating frames and compressing intervals.
Recovered the proprietary export from the police evidence archive
The native file still existed despite having been deleted from the working folder, which made a direct comparison possible.
Measured the difference between the two files
412 unique frames in the native file against 521 displayed frames in the conversion, many of them duplicates, with apparent event duration differing by 1.8 seconds. The native file also carried system timing data absent from the MP4.
Re-examined the visual interpretation at native quality
The conversion's heavier compression produced blocking around headlights and lane markings. At native quality the bright area interpreted as a tire crossing the line moved independently of the tire edge — it was a reflection.
What the examination found
- The offered video was a transcoded derivative, not a forensic copy
- Native variable frame timing had been replaced with a forced constant frame rate
- Frames were duplicated and intervals compressed, changing apparent motion and duration by 1.8 seconds
- The speed calculation rested on an assumption of constant frame timing the file no longer satisfied
- Compression artifacts around headlights and lane markings affected visual interpretation
- The native export existed in the evidence archive and had not initially been produced
Outcome
The native file remained usable, but any analysis derived from the converted copy required revision or withdrawal, and the conversion itself needed to be disclosed and documented. The reflection interpretation did not survive review at native quality.
Takeaways for counsel
- Ask whether the video you received was transcoded, and with what. "I converted it so it would play" is a common and entirely undocumented step.
- Any expert calculation from video — speed, distance, timing — depends on frame timing. Ask whether the analysis was performed on the native file.
- Variable frame rate is standard on motion-triggered surveillance systems. Forcing it to a constant rate corrupts exactly the data a reconstruction needs.
- Compression artifacts create apparent objects and edges. An interpretation that only appears at high compression deserves review at native quality.
- Ask whether the proprietary original still exists somewhere. It frequently does, even when it's gone from the working folder.
Related questions
Because measurements taken from it can change. Transcoding replaces the encoded stream and can alter frame timing, duplicate or drop frames, and change duration — none of which is visible on casual playback, all of which affects any calculation of speed, distance, or interval. It also strips system timing data the native file carried.
Services involved
Have a case that turns on digital evidence?
Tell us the case, the device, and your deadline. The scoping call is free, and we'll tell you honestly whether an examination is worth it.
Appointed counsel: see our Colorado public defender program.
More case studies
Deleted files in an employee-theft prosecution: what the artifacts proved, and what they couldn't
Deleted texts and edited screenshots: the one-hour timestamp error that moved a message before the order took effect
The video played at trial wasn't the video in discovery: PowerPoint, transcoding, and normalized audio
Related reading
- Digital evidence terminology for attorneys
What each forensic term means, what it changes in a case, and the question to ask.
- Forensic case studies
Worked examples of how digital evidence is examined, challenged, and presented in court.
- For Colorado public defenders
Independent examination at indigent-defense rates, with estimates written for funding requests.
This case study is illustrative. It is written to show how digital evidence is examined, challenged, and presented in court; it is not an account of any client matter, and any resemblance to a specific case is unintended. The techniques and failure modes described are real. Nothing on this page is legal advice, and no result is a prediction or guarantee of the outcome of any other matter. Inquiries are governed by our Digital Forensics Engagement Terms; contacting us does not create a forensic engagement, chain of custody, or preservation duty.