Blackbox workflow
The viewer is being archived. The format is not. Your logs are ordinary files, and the measurements inside them are still the same measurements. What changes is which app you open them in, and how much you should rely on it for future flights.
Betaflight's standalone Blackbox Explorer is being frozen and archived on 1 December 2026. The project's own announcement is the source to read: the Blackbox Explorer repository and the Betaflight release notes that link to it. Archived means the code is kept public and read-only: no new releases, no bug fixes on the main branch, and no new log fields added by future firmware.
It does not mean the app you already installed stops opening. It means the version you have is the version you keep. For old footage that is fine. For a quad you are still tuning, it means the viewer stops tracking the firmware.
A blackbox log is a file on your SD card or disk. It is written by the flight controller, not by the viewer. Betaflight still writes blackbox data, and the field definitions still live in the firmware source. So:
The practical advice is boring and effective: keep the original log, name it with the date and the tune that produced it, and keep a note of what you changed afterwards. A screenshot of a chart cannot be recomputed. The log can.
| Tool | Good at | Watch out for |
|---|---|---|
| Blackbox Explorer (your installed copy) | Frame-by-frame inspection, raw traces, exporting CSV | Frozen. No fixes and no awareness of future header fields. |
| Betaflight Configurator | Downloading a log from the FC, and basic blackbox viewing | It is a configuration tool first; it does not produce a tuning decision. |
| PIDtoolbox | Offline spectrum and step-response analysis with full control | You still have to read the charts and decide what to change. |
| Easyquadtune | Measuring the log, scoring whether it can support a tune, and writing a before/after CLI | It is not a frame-by-frame trace viewer. Use it for the decision, not the footage. |
None of these replace the others cleanly. A sensible split is: keep a frozen viewer for raw inspection, and use a measuring tool for the tune. The mistake is treating a viewer as a tuner, which it never was.
A trace viewer shows you the gyro line. It does not tell you whether that line overshot its own settled level, whether the log had a clean step to measure, or whether the noise peak you are looking at is frame resonance or prop noise. Those are measurements, and they are what a tune is built from.
So the useful question after the archive announcement is not "which viewer replaces Blackbox Explorer". It is "what measures the log". Easyquadtune does that in the browser: it scores the log before tuning from it, labels the step response Chart-ready, Partial or Not safe, and refuses to raise gain on a chart that cannot show tracking. If the log is not good enough, it says so instead of producing a confident script.
No. The logs are your files. Archiving the viewer does not touch them.
No. An archived project is frozen, not switched off. Your installed copy keeps running; it just stops receiving fixes.
There is no such announcement. The viewer and the file format are separate things. The risk is an unrecognised new header field, not an unreadable file.
The original log, then the PID and filter values that produced it. A screenshot is the least useful artefact because it cannot be recomputed.
Upload a Betaflight .bbl or .csv and get the
spectrum, the step response per axis, a log-quality verdict and a
before/after CLI. The first log is free per browser.
Analyse your log Or read a full sample report first
Questions about your own log? Ask in the Discord — a small, manual community where logs get a second pair of eyes.
More guides: how to read a blackbox log, wobble on hard stops, overshoot vs droop, frame resonance vs prop noise, I-term relax, dips in fast turns, when a log is not good enough.
Easyquadtune is not affiliated with the Betaflight project. Any example here is illustrative of the tool's output, not a tune for your aircraft. Always fly responsibly and perform an initial hover test in a safe environment after any tuning change.