United StatesDue to a bug in the OutRun code, the displayed stage times (with the exception of Stage 5) are inaccurate.
Your actual stage times:
Stage 1 = 57.28 seconds
Stage 2 = 54.87 seconds
Stage 3 = 57.25 seconds (accurate, but only by coincidence)
Stage 4 = 55.53 seconds
Stage 5 = 47.90 seconds
The cumulative difference is more than one full second (1.01 seconds).
All of my Out Run performances are 100% legitimate.
There are several control setups for playing the game: keyboard, mouse, joystick, controller, steering wheel, etc.
I happen to use a keyboard and mouse combination:
Accelerator = Left Shift Key
Steering = Mouse X-Axis
Gear Shift = Mouse Left-Click
Brake Pedal = Mouse Right-Click
I currently use WolfMAME Version 220. My high scores and/or fastest times have been submitted to both Twin Galaxies and MARP (MAME Action Replay Page), where my INP files are readily available for inspection by experts.
I do not own a webcam, nor do I intend to acquire one. If not showing my hands on video is going to be a major issue, then simply reject my submissions. Sorry if my preferred setup leaves a "horrible taste" in anyone's mouth, but I will not be changing my approach.
The videos provided in the original post are excellent examples of the differences between the Arcade (original or emulated) and Saturn platforms. Casual fans might perceive the two versions as being the same, but the differences should be noticeable to experienced players. Regarding the specific turns identified (during the "Alps" and "Cloudy Mountain" stages), I am thoroughly convinced that neither can be navigated at top speed (using MAME) without veering off the main road.
I would be interested to find out what coding changes were implemented by the Saturn programming team to create a more "player friendly" experience: better vehicle handling, slight modifications to the track layout, or perhaps both. In any case, the underlying Saturn code is not identical to Arcade/MAME, and provides a small but distinct advantage.
Thank you for the explanation as well as the list of rejected performances.
Yes, this is not a new discovery. The numerous bugs in the Out Run code have been known for quite some time.
I first reported the stage timer bug to Chris White back in August 2012, when he was working on the Cannonball project.
The short explanation is as follows:
The Stage 5 lap time is correct, but the individual times for Stages 1 through 4 are not. The whole seconds are accurate, but the hundredths of a second are mistakenly pulled from the wrong memory location. They match the current stage 100% of the time, instead of the stage that was just completed. If you are driving at (or near) top speed when passing a checkpoint, have you noticed that the displayed stage time is almost always within the xx.15 to xx.25 range?
I provide a more detailed demonstration in one of my Twin Galaxies submissions from 3 years ago:
I agree that some of the "gear gacha" runs featured on the Replay Burners YouTube channel are questionable. They were likely tool-assisted, as the execution seems a bit "too perfect" for a human.
However, I believe that one submission by AFB should NOT have been removed:
Goal D - No Gear Gacha - AFB - 4:42.12
In my opinion, the run is perfectly valid (see video below):
