How to Read a Dyno Graph, Correction Factors and What the Numbers Mean

Two customers bring the same model BMW 320d (N47) in for a before/after dyno pull after a remap, a week apart. The power figures come out 8bhp different even though the file is identical. The customer wants to know if he’s been shortchanged. He hasn’t; he’s just seeing how many variables sit between the engine and the number on the printout.

What the graph is actually showing

A dyno graph plots power and torque against engine RPM, usually as two curves rising and falling across the rev range. Torque tells you how hard the engine is pulling at a given RPM; power is a calculated value from torque and RPM, so the power curve always crosses the torque curve at 5252rpm when measured in lb-ft and hp. If those curves don’t cross there, something in the units or the calculation is off, worth knowing if a printout looks unusual.

Correction factors matter more than people think

Raw dyno numbers are affected by air temperature, humidity, altitude and barometric pressure, because denser air means more oxygen and more power. That’s why results get run through a correction standard, commonly SAE J1349 or DIN 70020, to normalise the figure to a reference condition. Two runs on the same car, same dyno, same day, can still differ by several percent if a different correction factor is applied to each, or if one run used a different standard entirely. A shop comparing before/after figures should use the same correction method for both pulls, on the same dyno, ideally the same day.

Why two runs on the same car can differ

1. Engine and oil temperature: a cold run and a fully warm run can differ by several percent
2. Tyre pressure and rolling resistance on a rolling road, if the car isn’t strapped down identically
3. Fuel used: a tank of poor-quality or lower-octane fuel changes ignition timing behaviour and knock retard
4. Operator technique: gear selected, how quickly the throttle is applied, whether the run starts from the same RPM
5. Ambient conditions on the day, even with correction applied

What to check before and after a remap on the dyno

Before the pull: coolant and oil at normal operating temperature, tyre pressures correct, fuel level and quality noted, no active fault codes that might already be limiting power, since a lazy MAF or a boost leak will flatter the “after” run for the wrong reasons. After the pull: check for knock retard in the logged data, not just the peak number, and check EGT if the car has an exhaust temperature sensor logged. A file that gains 15bhp but shows knock correction pulling timing back under load isn’t a clean gain, it’s a map that needs revisiting.

Reading past the headline number

The peak power figure is what goes on the sticker, but the area under the torque curve, how much pull the engine has through the mid-range, usually matters more for how a car actually feels on the road. A file that adds 10bhp at the top but fills in a flat spot at 2000-3000rpm will feel like a bigger improvement than the peak number suggests.

Hub dyno vs rolling road, and why it matters for comparisons

A hub dyno, which bolts directly to the wheel hubs, removes tyre slip and drivetrain flex from the measurement and tends to read a few percent higher than a rolling road, where the tyres sit on rollers and some power is lost to tyre deformation and roller friction. Neither figure is “wrong”, they’re just measuring at different points in the drivetrain. The number that actually matters is the comparison between two runs done on the same machine in the same way, not how that figure stacks up against a printout a customer saw from a different shop on a different type of dyno entirely.

For background on the file types behind a remap, see original vs virgin vs FRF ECU file, and the ECU model pages for EDC17C46 and MED17.5.5 for platform specifics.