Analyze what the ECU actually did, not just what one graph looks like.
Import multiple ME7Logger, ECUx, VCDS-style or generic delimited logs. ME7 Lab recognizes common channels, scores data quality, segments pulls, marks events, compares revisions and now groups unlike units into readable tuner-focused plot lanes.
Is this log actually worth comparing?
ME7 Lab scores parse quality, time continuity, recognized channels and usable engine-speed data. It also identifies what kinds of analysis are possible from the channels present.
Read the pull without fighting the graph.
Requested and actual values share a scale when they should be compared. Unrelated units get separate lanes automatically, so lambda, boost and duty cycle no longer flatten each other.
Tip: hover anywhere on the graph for exact values. Click a legend item to hide a noisy channel. Requested traces are dashed so actual-vs-requested tracking is obvious at a glance.
Mark the interesting parts automatically.
Events are evidence markers—not tuning instructions. Thresholds are deliberately conservative and each event shows the measured values behind it.
Separate comparable acceleration runs.
Pulls use RPM slope and throttle when available. If throttle is missing, ME7 Lab falls back to sustained increasing RPM and marks the confidence lower.
Compare two logs with the same metrics.
Use this for revision-to-revision testing. The comparison reports measurable differences and leaves the interpretation to the tuner.
Baseline A
Revision B
See what ME7 Lab recognized.
Alias matching handles many common ME7Logger/ECUx/VCDS-style labels. Unrecognized numeric columns still remain available for plotting and derived calculations.
Build a local calculation.
Create A − B, A + B, A ÷ B or A × B. The derived signal becomes available in the plotter and exports for this browser session.
Map-aware analysis is the next major step.
The next layer can connect BIN/XDF table axes to log operating points so ME7 Lab can show which calibration regions were actually exercised and compare cell-level performance across revisions.