A dyno sheet is not proof by itself. The real value is in how the test was run and whether the data makes mechanical sense. A dynamometer measures torque, then calculates horsepower from torque and RPM, so the RPM source, sweep time, gear selection, airflow, and vehicle temperature all matter. If the vehicle is not fully stabilized, if the sweep is too short, or if the run is done in the wrong gear, the numbers can be skewed away from what the truck or car will actually do on the road. We test to replicate real operating conditions, not just chase a peak number. That means proper cooling airflow, stable fluid temperatures, long enough sweeps, and steady-state testing when sustained load matters. Correction factors help compare results from different weather and elevations, but large corrections add uncertainty. That is why observed horsepower still matters—it is the power the vehicle actually made in the conditions where it was tested. If you want to spot inflated claims, start with the baseline, check the ambient data, and make sure the horsepower curve agrees with the torque curve at the stated RPM. A clean dyno sheet should show how the number was earned, not hide the conditions that created it.
The video begins with a simple question: a dyno sheet is often presented as proof that an engine makes a certain amount of power, but how can you tell whether the sheet is trustworthy? To answer that, Banks frames dyno reading as more than just looking at a peak horsepower number. A credible dyno result depends on understanding how the dynamometer works, how the test was run, what correction factor was used, why sweep time matters, and why horsepower and torque always intersect at the same RPM. By the end, the goal is to be able to read any dyno sheet critically and recognize when numbers have been manipulated.
A dynamometer is fundamentally a machine that measures torque. Torque is turning force over a distance, commonly expressed in pound-feet. The example given is a one-pound weight placed at the end of a one-foot lever, which creates one pound-foot of torque at the pivot. When an engine turns on a dyno, it applies torque, and the dyno resists that torque with a measured load. That resistance is then used to calculate the torque value.
Banks distinguishes between two major dyno types. An engine dyno measures torque directly at the flywheel, making it the only way to determine the engine's actual output without drivetrain, accessory, or vehicle cooling variables. This is why manufacturers quote flywheel horsepower when advertising new vehicles. Engine dynos are specialized systems and typically require a separate control room. Banks notes that, unlike most aftermarket companies, it also develops complete engines and therefore uses engine dyno testing in addition to vehicle testing.
For most aftermarket work, however, the engine is already installed in the vehicle, and what matters is the power delivered through the drivetrain. That requires a chassis dyno. Banks uses a Mustang air-cooled heavy-current chassis dynamometer with a dual-roller eddy-current setup. In this system, the power absorber is coupled to the rollers, and eddy-current load is created by an electric core moving across a magnetic field. The resulting electromagnetic braking force resists roller rotation, allowing the dyno operator to apply controlled load and simulate real driving conditions such as towing, climbing a grade, or pushing into a headwind.
Banks emphasizes that a dyno test should mimic the road as closely as possible. That means the vehicle's cooling systems must receive airflow similar to what they would see in actual driving. The company uses its Banks wind machine to provide high-volume, high-velocity airflow, with adjustable positioning to suit anything from a lowered car to a lifted truck. The setup is said to deliver more than 40,000 CFM at up to 60 mph.
This matters because unrealistic cooling can distort power results. Banks specifically criticizes dyno testing done with the hood open and small carpet fans blowing into the engine bay. That setup does not represent on-road airflow or underhood temperature. A cold-air intake tested with the hood open may appear to perform better on the dyno than it will in real driving, because the intake is exposed to cooler, less restrictive air than it would normally receive.
Vehicle temperature stabilization is equally important. Intake air temperature affects power, but so do engine oil, transmission fluid, and differential temperatures. Testing should begin only after the vehicle is thoroughly warmed up, and temperatures should remain within a safe and repeatable range. If power improves from one run to the next without any hardware or tuning changes, Banks argues that the vehicle was not yet in a stable condition. The company also mentions using its iDash system to log up to 100 parameters at 20 samples per second so that conditions can be monitored in real time. Proper normalization takes time, and not every dyno operator is willing to spend that time.
A sweep is a full-throttle dyno pull from a low RPM point to redline. The dyno operator chooses the starting RPM, ending RPM, and sweep time, while the dyno applies the load needed to achieve that rate of acceleration. Sweep time matters because it changes how much of the measured power is consumed by accelerating the drivetrain itself. A short sweep accelerates the rotating mass more quickly, which increases inertial losses. A longer, higher-load sweep slows the rate of acceleration and produces a more accurate measurement of usable power.
Banks compares a long sweep to accelerating up a steep grade at full throttle. The slower acceleration requires less force to increase speed, which reduces the amount of measured power lost to inertia. For that reason, Banks says it runs sweeps as long as possible. To support those longer pulls, the dyno cell uses not only the wind machine for the vehicle but also large blowers that pull cool ambient air through ducts under the floor and up through the chassis dyno. To prevent wheel slip and keep the vehicle planted, the axle is strapped down and truck beds may be loaded with as much as 2,000 pounds of sand.
The company is blunt about short sweeps justified by temperature concerns. If a tuner says a pull cannot last longer than six seconds because of EGT or other thermal limits, Banks argues that this is effectively admitting the vehicle cannot sustain the claimed power. A brief over-fueled burst may produce an impressive peak number, but if the engine cannot hold that output on the road, the number has little practical value.
Gear choice also affects the result. The preferred gear for a sweep is the transmission's direct-drive gear, the gear with a 1:1 ratio between input and output. In a six-speed transmission this is typically fourth gear; in a 10-speed it is usually seventh. In direct drive, the transmission contributes the least parasitic loss because there is no torque multiplication through additional gearing. The goal is to make the transmission behave as much like a solid shaft as possible.
Although sweep tests are common for quick power evaluation, Banks argues that they do not reflect how vehicles are actually used. For development work, the company relies heavily on steady-state testing, in which the dyno holds a fixed RPM and load for extended periods, sometimes for hours. That may sound severe, but Banks points out that maintaining a constant RPM and load for long periods is exactly what happens during freeway driving, towing, or climbing grades.
Steady-state testing is especially important for parts whose behavior changes with temperature, such as intercoolers and tuning devices. Banks gives the example of comparing its Derringer inline tuner for the Duramax L5P against the Edge Pulsar tuner. According to the video, the Pulsar produced more horsepower initially, but only for about five seconds. Roughly 20 seconds into the run, power dropped by about 50 horsepower. Banks attributes this to the factory ECM's EGT limiter, which intervenes at 1472 degrees Fahrenheit, or 800 degrees Centigrade, to protect the engine and turbocharger from excessive exhaust gas temperature.
The broader point is that sustained-load testing reveals whether a claimed power increase is durable or merely a short-lived spike. Banks also notes that steady-state testing can be used to build a power curve by stabilizing the vehicle at successive RPM increments and recording torque once temperatures settle. This approach removes transient variables that are hidden in a conventional sweep. As an example of durability-focused development, Banks references work on the Army's Joint Light Tactical Vehicle, where its D866T engines underwent a NATO-standardized 400-hour durability test.
Accurate dyno graphs require accurate RPM data. On fuel-injected vehicles, RPM can be derived from injector current pulse frequency by attaching a current pulse reader to an injector. Other methods include optical or Hall-effect sensing from a rotating component such as a crank pulley. Banks warns against relying on OBD2 RPM data because it does not sample quickly enough for precise dyno work. Likewise, deriving RPM from chassis dyno roller speed can be compromised by wheel slip and may require substantial correction.
Once torque and RPM are known, environmental conditions still have to be considered. Engine power depends directly on air density, which is affected by ambient temperature, pressure, and humidity. Correction factors are used to standardize dyno results so that tests performed in different weather or at different locations can be compared. Banks cites SAE J1349 as the most common modern standard. Under that standard, the reference conditions are 77 degrees Fahrenheit, 0 percent humidity, and 14.4 PSI, corresponding to an air density of 72.21 pounds per thousand cubic feet of air.
Banks says Southern California often provides near-ideal conditions, but locations such as Denver, Salt Lake City, or Miami may require substantial correction. The problem is that correction factors are estimates, not perfect conversions. The farther actual conditions are from the standard day, the larger the margin of error. Banks offers a practical rule: any positive or negative correction larger than 7 percent should be treated as unreliable. In those cases, observed horsepower, the actual power measured by the dyno under current conditions, may be more meaningful than corrected horsepower.
That distinction matters in the real world. A truck dynoed near sea level in cool San Francisco weather and then tested in Denver at 5,200 feet on a hot day may show similar corrected horsepower, but the observed horsepower on the road in Denver could be dramatically lower. Banks notes that elevation alone can reduce ambient air density by about 14 percent, and hot weather can push the total loss toward 25 percent. For towing in the Rockies, observed horsepower is the number that matters most.
Banks then walks through the anatomy of a dyno sheet using a 2018 Honda Civic Type R example. The notes section should include the time of day, day of year, dyno type, total sweep time, RPM data source, ambient temperature, ambient pressure, ambient humidity, resulting air density, and the gear used for the run. Without that context, it is harder to judge whether the test was performed consistently and under meaningful conditions.
The sheet should also identify the correction standard. SAE usually refers to J1349, while STD refers to the older J607 standard. If no correction factor is listed, the result is likely observed horsepower. Another important item is smoothing. Smoothing reduces noise in the plotted lines. With no smoothing, the graph appears jagged; with more smoothing, the curves become cleaner and easier to read. Most operators apply some smoothing because raw traces can look erratic, but the amount used should be disclosed.
On the graph itself, RPM is plotted on the x-axis, while torque and horsepower are plotted on the y-axis. The torque curve is measured directly by the dyno. The horsepower curve is calculated from torque and RPM, which is why dyno graphs usually begin above idle, often around 1,500 RPM, to avoid lugging the engine during the pull.
The video explains the familiar dyno-graph crossover mathematically. One horsepower equals 550 pound-feet per second, or 33,000 pound-feet per minute. Power is force multiplied by velocity. In rotational terms, that becomes torque multiplied by rotational speed, and after simplifying the constants, the standard equation is horsepower = torque × RPM ÷ 5252.
That equation explains why the torque and horsepower curves always intersect at 5,252 RPM. At that exact engine speed, the RPM term divided by 5,252 equals 1, so horsepower and torque have the same numerical value. Every dyno calculates horsepower from measured torque using this relationship in real time. Banks stresses this point because it becomes one of the easiest ways to spot a falsified dyno sheet.
Banks closes with three practical methods for identifying manipulated dyno claims. The first is the baseline check. A legitimate aftermarket comparison should begin with a stock baseline run. If the stock run is intentionally weakened, the advertised gain from the aftermarket part will look larger than it really is. Banks calls this sandbagging the baseline. The video cites examples involving the Duramax L5P, where several sources measured roughly 403 horsepower, while other companies presented lower baseline figures such as 374, 390, or 378 horsepower before claiming gains from their products. Whether done by inconsistent correction, poor warm-up, or outright fabrication, understating the baseline artificially inflates the improvement.
The second method is to verify the horsepower curve against the torque curve using the horsepower equation. Banks presents a claimed dyno result for a 2022 Ram TRX showing 965.5 horsepower at 6,000 RPM with about 785 lb-ft of torque. Using the equation, 785 × 6000 ÷ 5252 yields about 896.5 horsepower, not 965.5. That discrepancy of roughly 70 horsepower suggests the horsepower trace was altered independently of the torque trace, likely in image editing software.
The third method is to inspect the surrounding test data. Time of day, date, ambient temperature, pressure, and humidity all affect air density and therefore power. A dishonest operator might run the baseline in the hottest part of the day and the modified run in the coolest conditions. Even without a baseline comparison, a run performed on a cold, dry morning with a long sweep will naturally produce better numbers than the same vehicle tested on a hotter day with a shorter sweep. If correction factors are omitted when conditions are better than standard, or exaggerated when conditions do not justify them, the reported power can be inflated well beyond what the vehicle actually makes.
Banks concludes that dyno sheets are useful only when the testing method, environmental conditions, and math all hold up. Understanding how torque is measured, how horsepower is calculated, how correction factors work, and how testing can be manipulated allows a reader to judge whether a dyno sheet is credible or just a polished piece of paper.