Practice data only. Every value, number and event in this lab is fictional. We built gauntlet.db to practice identifying timestamp formats. It is not an extraction from a real device, but every format in it is one you will meet in casework.
The scenario
Overnight on November 7–8, 2026, a unit at a self-storage facility was broken into. An extraction of a person of interest's phone, plus a computer and an external drive seized with it, produced the twelve raw values in gauntlet.db. The detective wants a timeline by morning.
One problem: the tool that parsed these values guessed at the formats. It reported one event in 1995 and another in 2092. Before anything goes on a timeline, work out what each number really is. The devices were set to US Eastern time.
Download
Download the lab file (ZIP, 1 KB) Or open it in Lab Tools
| File | SHA-256 |
|---|---|
gauntlet.db | f95b0593e3845b21d9cbad5d5ff2994de60eacfe8886d2a9f2bb58e6659c52a5 |
validate-this-02.zip | a7897584b032fc5be2f9db69eeb8934a6a15c2118a287f0f67bb4b5a903f82d5 |
Nothing to install: in the Lab Tools SQLite viewer, click Load the Validate This #02 file, open evidence_values, and click any raw_value. The viewer lists every common format that would put that number between 1990 and 2100. Picking the right one is your job. The Number & Time Converter does the same for a value you paste.
The challenge
- For each of the 12 items, name the format and convert it to Eastern local time, to the second. Two items are not timestamps at all.
- Items 3 and 6 are both 18 digits long. How do you tell Mac Absolute nanoseconds from Windows FILETIME?
- Item 10 decodes to January 1, 2001 in one format. What is it really, and why did a tool offer a date?
- Item 5 is the only event before November 1. Which UTC offset applies to it, and what goes wrong if you use the same offset as the rest?
- Seven events happened on the evening of November 7, Eastern time. What date would a report that lists only UTC give them?
- Which two values did the tool report as 1995 and 2092, and which format did it wrongly assume for each?
- Put the real timestamps in order. Write the timeline in Eastern time.
- Bonus: item 9 is stored as a decimal. What do the digits after the point tell you, and should they appear in your report?
Check your answers
Pick each format and give the Eastern local time, 24-hour, to the second. Leave the time blank for a value that is not a timestamp. A row is marked correct only when both parts are right. Answers are checked in your browser against hashed values, so the page source won't give them away. Your entries are saved on this device only.
17941111641794110310611134385877753000000179411286509000000017934921407250008157978222040000003876992411815797925815803572.5171847134385812493881148035550123Hints
Hint 1: count the digits
For dates in 2026, each format has a typical length and leading digits. A 10-digit value starting with 17 is very different from a 10-digit value starting with 38. Write down each value's length before you convert anything.
Hint 2: use the case window
A converter can make almost any number look like a date. Every real event here falls between late October and November 8, 2026. A format that lands in 1995, 2005, 2056 or 2092 is a wrong format, not a surprising fact.
Hint 3: daylight saving time
In 2026, US daylight saving time ended at 2:00 AM on Sunday, November 1. Before that the Eastern offset is UTC−4 (EDT); after it, UTC−5 (EST).
Walkthrough
Show the full solution
The digit ladder
Each format counts different units from a different starting point (its epoch), so values from the same year have very different sizes. For dates in 2026:
| Digits | Starts with | Likely format | Counts |
|---|---|---|---|
| 9 | 8 | Mac Absolute seconds | seconds since 2001-01-01 UTC |
| 10 | 17 | Unix seconds | seconds since 1970-01-01 UTC |
| 10 | 38 | HFS+ seconds | seconds since 1904-01-01 UTC |
| 13 | 17 | Unix milliseconds | ms since 1970 |
| 16 | 17 | Unix microseconds | µs since 1970 |
| 17 | 13 | WebKit / Chrome time | µs since 1601-01-01 UTC |
| 18 | 8 | Mac Absolute nanoseconds | ns since 2001 |
| 18 | 13 | Windows FILETIME | 100-ns intervals since 1601 |
| 19 | 17 | Unix nanoseconds | ns since 1970 |
The ladder gives you a hypothesis, not a finding. Confirm it with the app's documented schema, a known event, or your own test data before it goes in a report.
1. The answers
| # | Source | Raw value | Format | UTC | Eastern |
|---|---|---|---|---|---|
| 1 | Wi-Fi log: Disconnected from home network | 1794111164 | Unix seconds | 2026-11-08 04:12:44 | 2026-11-07 23:12:44 EST |
| 2 | Call log: Outgoing call started | 1794110310611 | Unix milliseconds | 2026-11-08 03:58:30 | 2026-11-07 22:58:30 EST |
| 3 | Desktop sync log: Phone backup copied to a PC | 134385877753000000 | Windows FILETIME | 2026-11-08 05:02:55 | 2026-11-08 00:02:55 EST |
| 4 | Fitness app: Walking session started | 1794112865090000000 | Unix nanoseconds | 2026-11-08 04:41:05 | 2026-11-07 23:41:05 EST |
| 5 | Rideshare app: Trip receipt issued | 1793492140725000 | Unix microseconds | 2026-11-01 00:15:40 | 2026-10-31 20:15:40 EDT |
| 6 | Messaging app: Message sent: "leaving now" | 815797822204000000 | Mac Absolute nanoseconds | 2026-11-08 02:30:22 | 2026-11-07 21:30:22 EST |
| 7 | External drive (Mac): Photo file created on the drive | 3876992411 | HFS+ seconds (since 1904) | 2026-11-08 14:20:11 | 2026-11-08 09:20:11 EST |
| 8 | Messaging app: Message read | 815797925 | Mac Absolute seconds | 2026-11-08 02:32:05 | 2026-11-07 21:32:05 EST |
| 9 | Photos: Photo taken (stored as a decimal) | 815803572.517 | Mac Absolute seconds | 2026-11-08 04:06:12 | 2026-11-07 23:06:12 EST |
| 10 | Call log: Duration of that call | 1847 | Not a timestamp | – | – |
| 11 | Web browser history: Visit: self-storage access hours page | 13438581249388114 | WebKit / Chrome microseconds (since 1601) | 2026-11-08 03:14:09 | 2026-11-07 22:14:09 EST |
| 12 | Contacts: Number saved for "Unit Mgr" | 8035550123 | Not a timestamp | – | – |
2. Two 18-digit values
Look at the leading digits. In 2026 a Mac Absolute nanosecond value starts with 8 (item 6: 815…), and a FILETIME value starts with 13 (item 3: 134…). Each one, read in the other's format, lands far outside the case: item 3 read as Mac nanoseconds gives April 2005, and item 6 read as FILETIME gives a year after 4000.
3. The call duration
Item 10 is 1,847: the length of the call in seconds (30 minutes, 47 seconds). Read as Mac Absolute seconds it becomes 2001-01-01 00:30:47 UTC, because Mac time counts from 2001 and 1,847 seconds is half an hour past that point. A converter offers a date for any number that fits a format. That doesn't make the number a timestamp. Item 12, 8035550123, is the same lesson: it is a saved phone number, and every format puts it far outside the case window.
4. The daylight saving trap
Item 5 decodes to 2026-11-01 00:15:40 UTC. On October 31, Eastern time was still EDT (UTC−4), so the receipt was issued at 8:15:40 PM EDT on October 31. Apply EST (UTC−5) like the rest and you get 7:15:40 PM, an hour early. Convert each value with the offset in effect on its date, not the date of the case.
5. UTC and the calendar date
Every event between 7:00 PM and midnight EST on November 7 falls on November 8 in UTC. A report or tool export that shows only UTC will put the evening's activity on the wrong day for anyone reading it as local time. State the time zone with every time you report.
6. 1995 and 2092
1995 is item 8 (815797925), Mac Absolute seconds read as Unix seconds. Mac time starts 978,307,200 seconds (31 years) after Unix time, so the misread lands 31 years early. 2092 is item 7 (3876992411), HFS+ seconds read as Unix seconds. HFS+ starts in 1904, 66 years before Unix time, so the misread lands 66 years late. Both are 10 digits or fewer and look like ordinary Unix values. Only the leading digits and the case window give them away.
7. The timeline (Eastern)
- Oct 31, 8:15:40 PM EDT: Trip receipt issued (item 5, rideshare app)
- Nov 7, 9:30:22 PM EST: Message sent: "leaving now" (item 6, messaging app)
- Nov 7, 9:32:05 PM EST: Message read (item 8, messaging app)
- Nov 7, 10:14:09 PM EST: Visit: self-storage access hours page (item 11, web browser history)
- Nov 7, 10:58:30 PM EST: Outgoing call started (item 2, call log)
- Nov 7, 11:06:12 PM EST: Photo taken (stored as a decimal) (item 9, photos)
- Nov 7, 11:12:44 PM EST: Disconnected from home network (item 1, wi-fi log)
- Nov 7, 11:41:05 PM EST: Walking session started (item 4, fitness app)
- Nov 8, 12:02:55 AM EST: Phone backup copied to a PC (item 3, desktop sync log)
- Nov 8, 9:20:11 AM EST: Photo file created on the drive (item 7, external drive (mac))
Item 5 is a week before the case window. It goes on the timeline, but in context: it shows a rideshare trip, not activity on the night.
Bonus: the decimal
Item 9 is 815803572.517: Mac Absolute seconds with milliseconds after the point. Report the time to the second (11:06:12 PM EST) unless the precision matters. Truncate rather than round: 11:06:12, not 11:06:13. That matches how the converters on this site display it and keeps your times consistent across sources.
Report wording
The value 815797822204000000, stored for item 6 in the messaging application's database, was interpreted as Mac Absolute Time in nanoseconds (nanoseconds since 2001-01-01 00:00:00 UTC). It converts to 2026-11-08 02:30:22 UTC, which is 9:30:22 PM EST on November 7, 2026. The conversion was performed with [tool and version] and confirmed by manual calculation.
Notice that the wording names the format, the epoch and the method. If the format were challenged, another examiner could reproduce the result from the raw value alone.
Go further
Timestamp formats, epochs and time zone validation are covered in depth, with graded labs, in our examiner course, Mobile Forensics: A Bit of Knowledge. For quick review, drill the timestamp terms in the glossary study mode.