Magic Data Discussion
Posted by TomMK
Dunno if useful, but from looking at these I'm guessing that the faster circuits need to have less difference in fuel consumption between the player and AI.
At least at Monza, Spa, Hockenheim and Sepang player's fuel consumption is only 180-966 higher. Dunno why the difference at Magny-Cours and Silverstone is higher than at Hungaroring though.
Maybe this difference is connected to how the game is operating with the car acceleration. I guess it is discretive - like "1" (button "A" or whatever else you use for acceleration pressed) and "0" (button "A" not pressed) for the player, while cpu have their cars accelerated more realistic. Or maybe I'm saying rubbish.
___________________________________________________________________________
For a list of EVERY download for GP4, look here:[docs.google.com]
Edited 1 time(s). Last edit at 09/11/2019 12:17PM by Soutsen.
At least at Monza, Spa, Hockenheim and Sepang player's fuel consumption is only 180-966 higher. Dunno why the difference at Magny-Cours and Silverstone is higher than at Hungaroring though.
Maybe this difference is connected to how the game is operating with the car acceleration. I guess it is discretive - like "1" (button "A" or whatever else you use for acceleration pressed) and "0" (button "A" not pressed) for the player, while cpu have their cars accelerated more realistic. Or maybe I'm saying rubbish.
___________________________________________________________________________
For a list of EVERY download for GP4, look here:[docs.google.com]
Edited 1 time(s). Last edit at 09/11/2019 12:17PM by Soutsen.
Rear wing setting is slightly randomized for cc-cars, but I have no idea how exactly GP4 does that, or if we can influence that in any way.
My unfinished tracks: [www.grandprixgames.org]
Send bug reports and track editing questions to f1virtualblog@gmail.com
My unfinished tracks: [www.grandprixgames.org]
Send bug reports and track editing questions to f1virtualblog@gmail.com
What I meant is that GP4 randomizes the rear wing values from the magic data, sorry if I wasn't clear. So even if the magic data says 1, some cars may have 2 or 3 for example.
My unfinished tracks: [www.grandprixgames.org]
Send bug reports and track editing questions to f1virtualblog@gmail.com
My unfinished tracks: [www.grandprixgames.org]
Send bug reports and track editing questions to f1virtualblog@gmail.com
More findings from my GP4.exe investigation.
First, about rear wing setting randomizer. Rear wing is randomized for each team, at the start of each race weekend. For each team, there is a 50% chance rear wing setting will be exactly as set in magic data, 25% chance both cars will use one step lower and 25% chance both cars will use one step higher. It is possible to rework the code to eliminate this randomness.
Second, about values desc45 and desc46 in magic data, previously called "downforce and ride height", and later "drag and player downforce". These values are, most likely, air temperature in degrees Celsius and air pressure in hectopascals. desc45 in original tracks ranges from 17 (Imola, Silverstone and Spa) to 31 (Hungaroring). desc46 from 926 (Interlagos - highest altitude in 2001 calendar) to 1022 (Monaco - sea level).
While loading magic data, the game code calculates a value with the formula ((298 * desc46) * 16384) / ((desc45 + 273) * 990). 273 being the conversion from Celsius to Kelvin, 16384 just a scaling factor. The formula is very similar to the one for density of dry air, and this result is used in a number of places in the code. How exactly this affects downforce and drag, as other people noticed, remains unknown for now. This is the only piece of code where desc45 and desc46 are directly referenced (indirect references may still exist).
Desc47 meanwhile, which I previously called air density, is really only used as a scaling factor for engine power. But it still correlates quite well with the results from the in-game formula, meaning it is indeed used to compensate how normally-aspirated engines have less power at higher altitudes.
Thanks for 32BOBO32 for helping confirm what the formula is about.
My unfinished tracks: [www.grandprixgames.org]
Send bug reports and track editing questions to f1virtualblog@gmail.com
Edited 2 time(s). Last edit at 08/23/2021 07:29PM by Prblanco.
First, about rear wing setting randomizer. Rear wing is randomized for each team, at the start of each race weekend. For each team, there is a 50% chance rear wing setting will be exactly as set in magic data, 25% chance both cars will use one step lower and 25% chance both cars will use one step higher. It is possible to rework the code to eliminate this randomness.
Second, about values desc45 and desc46 in magic data, previously called "downforce and ride height", and later "drag and player downforce". These values are, most likely, air temperature in degrees Celsius and air pressure in hectopascals. desc45 in original tracks ranges from 17 (Imola, Silverstone and Spa) to 31 (Hungaroring). desc46 from 926 (Interlagos - highest altitude in 2001 calendar) to 1022 (Monaco - sea level).
While loading magic data, the game code calculates a value with the formula ((298 * desc46) * 16384) / ((desc45 + 273) * 990). 273 being the conversion from Celsius to Kelvin, 16384 just a scaling factor. The formula is very similar to the one for density of dry air, and this result is used in a number of places in the code. How exactly this affects downforce and drag, as other people noticed, remains unknown for now. This is the only piece of code where desc45 and desc46 are directly referenced (indirect references may still exist).
Desc47 meanwhile, which I previously called air density, is really only used as a scaling factor for engine power. But it still correlates quite well with the results from the in-game formula, meaning it is indeed used to compensate how normally-aspirated engines have less power at higher altitudes.
Thanks for 32BOBO32 for helping confirm what the formula is about.
My unfinished tracks: [www.grandprixgames.org]
Send bug reports and track editing questions to f1virtualblog@gmail.com
Edited 2 time(s). Last edit at 08/23/2021 07:29PM by Prblanco.
Thanks for sharing this knowledge. 
My Grand Prix 4 Files
I'm a total dick. How many people can say that?
The problem with desc46 is it affects the player car differently to the AI so as that value varies between different Magic Data, your performance will vary against the AI for no good reason.
So I keep desc46 constant in all my MD, and only vary desc47 to simulate air pressure differences (i.e. cars go faster in a straight line at high altitude, Mexico City being the best example in recent years).
=====================================================
Intel NUC 8i3, 8GB RAM, MS Sidewinder Wheel
So I keep desc46 constant in all my MD, and only vary desc47 to simulate air pressure differences (i.e. cars go faster in a straight line at high altitude, Mexico City being the best example in recent years).
=====================================================
Intel NUC 8i3, 8GB RAM, MS Sidewinder Wheel
Mexico is a nice example because whether you're modding for naturally aspirated engines or turbocharged engines makes a massive difference. From the F1 website (https://www.formula1.com/en/latest/features/2015/10/all-time-high---the-rigours-of-racing-at-altitude.html)
So, based on recent findings, what I believe to be the best approach is:
Desc45 and desc46 are temperature and pressure, so they should be realistic values regardless of season. Air pressure for Mexico City should be only around 780 hPa, which is a huge contrast to other tracks.
Desc47, being only a multiplier for engine power output, should be set differently whether simulating a turbo season or not. For non-turbos, it should be somewhere in the 12000-13000 range because of the reduced air intake; for turbos, the value should be very close to 16384 (equivalent to 100%), as power doesn't change with altitude in this case.
What makes Mexico one of the tracks with fastest top speeds in the turbo era is the low drag from less dense air. The reduction in air pressure from desc46 should decrease drag enough to significantly raise straight line speeds, at least for player car. And if that creates a discrepancy between player and AI, then we can later use the cc power factors (desc53 and later) to compensate.
Of course, this is only conjecture, I haven't tested these values to see if what I said really holds. But I hope that as we continue to improve our knowledge of GP4 we'll be able to make more conscious choices in modding.
My unfinished tracks: [www.grandprixgames.org]
Send bug reports and track editing questions to f1virtualblog@gmail.com
Quote
Typically, for every 100m increase in altitude, engine output dropped by about one percent. Back in 1992, when F1 racing last visited Mexico, engine power would be down by around 22 percent compared with, for example, sea-level Suzuka.
But, this doesn’t happen with the turbocharged engines. With the turbo forcing air into the engine at much the same pressure it would at lower altitudes, the power remains the same.
So, based on recent findings, what I believe to be the best approach is:
Desc45 and desc46 are temperature and pressure, so they should be realistic values regardless of season. Air pressure for Mexico City should be only around 780 hPa, which is a huge contrast to other tracks.
Desc47, being only a multiplier for engine power output, should be set differently whether simulating a turbo season or not. For non-turbos, it should be somewhere in the 12000-13000 range because of the reduced air intake; for turbos, the value should be very close to 16384 (equivalent to 100%), as power doesn't change with altitude in this case.
What makes Mexico one of the tracks with fastest top speeds in the turbo era is the low drag from less dense air. The reduction in air pressure from desc46 should decrease drag enough to significantly raise straight line speeds, at least for player car. And if that creates a discrepancy between player and AI, then we can later use the cc power factors (desc53 and later) to compensate.
Of course, this is only conjecture, I haven't tested these values to see if what I said really holds. But I hope that as we continue to improve our knowledge of GP4 we'll be able to make more conscious choices in modding.
My unfinished tracks: [www.grandprixgames.org]
Send bug reports and track editing questions to f1virtualblog@gmail.com
Hmm the answer might be a little complicated, since we don't fully know how GP4 works.
Individual driver performance comes from the performance file, which is separate from magic data. The CC power and grip factors are multipliers (500 = 100%) which are applied to all cars equally (except player), depending on which difficulty level you're running.
Speed itself depends from different factors such as engine power and aerodynamic drag (and by extension, car setups). There are strong indications that calculations for AI are much more simplified than for players but again, we don't fully know the details.
If you ask a question like "how much % in power/grip should give a 1% difference in laptimes?", there is no formula that can give you a definitive answer. Each track is different, each season mod is different. A very crude approximation valid for the unmodded game is that, on average, every 5% power equals 1% laptime, and every 3% grip equals 1% laptime.
For gear ratios yes, I think they're equal to all AI cars regardless of engine power/wing setup. The 7th gear calculations are done as follows: 6th from magic data becomes 7th, 1st remains 1st. All other ratios are linearly interpolated between those two extremes, meaning 2nd to 5th gear ratios in magic data are ignored for cars with 7 gears.
Edit: by the way, many values in the magic data (and performance file) are multipliers, especially those whose values are around the 15000-17000 range, where 16384 (2^14) equals 100%.
My unfinished tracks: [www.grandprixgames.org]
Send bug reports and track editing questions to f1virtualblog@gmail.com
Edited 1 time(s). Last edit at 08/30/2021 01:22PM by Prblanco.
Individual driver performance comes from the performance file, which is separate from magic data. The CC power and grip factors are multipliers (500 = 100%) which are applied to all cars equally (except player), depending on which difficulty level you're running.
Speed itself depends from different factors such as engine power and aerodynamic drag (and by extension, car setups). There are strong indications that calculations for AI are much more simplified than for players but again, we don't fully know the details.
If you ask a question like "how much % in power/grip should give a 1% difference in laptimes?", there is no formula that can give you a definitive answer. Each track is different, each season mod is different. A very crude approximation valid for the unmodded game is that, on average, every 5% power equals 1% laptime, and every 3% grip equals 1% laptime.
For gear ratios yes, I think they're equal to all AI cars regardless of engine power/wing setup. The 7th gear calculations are done as follows: 6th from magic data becomes 7th, 1st remains 1st. All other ratios are linearly interpolated between those two extremes, meaning 2nd to 5th gear ratios in magic data are ignored for cars with 7 gears.
Edit: by the way, many values in the magic data (and performance file) are multipliers, especially those whose values are around the 15000-17000 range, where 16384 (2^14) equals 100%.
My unfinished tracks: [www.grandprixgames.org]
Send bug reports and track editing questions to f1virtualblog@gmail.com
Edited 1 time(s). Last edit at 08/30/2021 01:22PM by Prblanco.
Quote
klausfeldmann
So, is anybody able to give us a current overview of all magic data description
This is my Descriptions.txt
; desc1= Front wing [CC dry setup] ; desc2= Rear wing ; desc3= 1st gear ; desc4= 2nd ; desc5= 3rd ; desc6= 4th ; desc7= 5th ; desc8= 6th ; desc9= Front wing [CC wet setup] ; desc10= Rear wing ; desc11= 1st gear ; desc12= 2nd ; desc13= 3rd ; desc14= 4th ; desc15= 5th ; desc16= 6th ; desc17= Front wing [Player dry setup] ; desc18= Rear wing ; desc19= 1st gear ; desc20= 2nd ; desc21= 3rd ; desc22= 4th ; desc23= 5th ; desc24= 6th ; desc25= Dry brake balance ; desc26= Front wing [Player wet setup] ; desc27= Rear wing ; desc28= 1st gear ; desc29= 2nd ; desc30= 3rd ; desc31= 4th ; desc32= 5th ; desc33= 6th ; desc34= Wet brake balance ; desc35= Softer tyre [52 Hard, 53 Medium, 54 Soft, 55 Supersoft] ; desc36= Harder tyre [52 Hard, 53 Medium, 54 Soft, 55 Supersoft] ; desc37= >= 50 AI chooses softer tyre, otherwise harder ; desc38= ; desc39= ; desc40= ; desc41= ; desc42= Track grip ; desc43= ; desc44= ; desc45= Temperature (deg C) - top speed only affected, minimal effect in corners / acceleration? (17 - 31) ; desc46= Air Pressure (hectopascals). AFFECTS PLAYER-ONLY downforce, and drag ; desc47= Engine power output - e.g. altitude simulation? Affects acceleration ; desc48= Fuel per lap - see also desc70/71 ; desc49= Slipstream? A distance in feet that is subtracted from the distance between 2 cars, then compared to the speed(?) of one of the behind car, and if the speed is less, something is triggered. Player only. ; desc50= Player tyre wear factor - see also desc72 ; desc51= CC aggressiveness (aka Braking Range) min - now configurable in GPxPatch 4.50 (added to car.field_108. related to cc grip factor, 32768 low, 65536 & 0 mid, 32767 high) ; desc52= CC aggressiveness (aka Braking Range) max - now configurable in GPxPatch 4.50 ; desc53= [Ace] CC power factor ; desc54= [Ace] CC grip factor ; desc55= [Pro] CC power factor ; desc56= [Pro] CC grip factor pro ; desc57= [Semi-Pro] CC power factor ; desc58= [Semi-Pro] CC grip factor ; desc59= [Amateur] CC power factor ; desc60= [Amateur] CC grip factor ; desc61= [Rookie] CC power factor ; desc62= [Rookie] CC grip factor ; desc63= CC random performance range min ; desc64= CC random performance range max ; desc65= CC error chance - now configurable in GPxPatch 4.52 ; desc66= CC recovery sectors ; desc67= Sectors to pit-in begin ; desc68= Sectors to pit-out end ; desc69= Pre-pit speed limit ; desc70= Fuel consumption Player ; desc71= Fuel consumption CC ; desc72= Tyre wear ; desc73= Track sector after which AI cars stop being cautious on 1st lap ; desc74= Hotseat turn duration - e.g. 70800 equals 1:10.800. Try to match to the real-life pole laptime if possible. ; desc75= Real-time factor, speeds up or slows down the in-game time (+/-1000 equals +/- 1.0) ; desc76= Helps decide if a tyre set needs changing? (dry to wet / wet to dry) - higher values = quicker reaction ; desc77= Same as above? (dry to wet, wet to dry) ; desc78= Rain chance ; desc79= Segment nr (start of some range) related to pit in/out. Related to determining surface type under car. ; desc80= Segment nr (end of some range) related to pit in/out. Number of segments from the end of the pits (so higher means TOWARDS the start of the pits!). Related to determining surface type under car. ; desc81= CC race grip (always 256) ; desc82= A time duration (ms), used when the car is not in the pits, though it is related to the pits. ; desc83= Segment where the player car starts in quicklaps mode ; desc84= Black Flag penalty (ms) (must be between 10000 - 30000) ; desc85= Black Flag severity (1024=80kph) ; desc86= (Wet) CC engine mapping (higher, more detuned) ; desc87= (Wet) CC tyre wear and grip factor (higher, more) ; desc88= (Wet) CC tyre wear factor (lower, more) ; desc89= (Wet) CC grip factor (lower, more) ; desc90= "Handbrake". Stop car moving in garage? Seems to be a lateral offset from the pitbox. ; desc91= Car depth in garage (Player car) ; desc92= Car depth in garage (AI cars) ; desc93= Car orientation in garage ; desc94= Pitstop stall depth from pitlane (Player + AI) ; desc95= Pitstop stall depth from pitlane (Player + AI) finetune ; desc96= These seem to be tyre wear factors for each tyre type (always 16384?) ; desc97= These seem to be tyre wear factors for each tyre type (always 16384?) ; desc98= These seem to be tyre wear factors for each tyre type (always 16384?) ; desc99= These seem to be tyre wear factors for each tyre type (always 16384?) ; desc100= These seem to be tyre wear factors for each tyre type (always 16384?) ; desc101= These seem to be tyre wear factors for each tyre type (always 16384?) ; desc102= pitstop group 1 % ; desc103= stop 1 ; desc104= pit window 1 ; desc105= ; desc106= ; desc107= ; desc108= ; desc109= ; desc110= pitstop group 2 % ; desc111= stop 1 ; desc112= pit window 1 ; desc113= stop 2 ; desc114= pit window 2 ; desc115= ; desc116= ; desc117= ; desc118= pitstop group 3 % ; desc119= stop 1 ; desc120= pit window 1 ; desc121= stop 2 ; desc122= pit window 2 ; desc123= stop 3 ; desc124= pit window 3 ; desc125= ; desc126= failure chance : suspension (The total sum should not go beyond 126040. Note: 630 is added to each, total becomes 131080 after 8*630 is added) ; desc127= failure chance : loose wheel ; desc128= failure chance : puncture ; desc129= failure chance : engine ; desc130= failure chance : transmission ; desc131= failure chance : oil leak / water leak ; desc132= failure chance : throttle / brake ; desc133= failure chance : electrics ; desc134= ; desc135= ; desc136= ; desc137= ; desc138= bump factor ; desc139= bump shift
=====================================================
Intel NUC 8i3, 8GB RAM, MS Sidewinder Wheel
Sorry, only registered users may post in this forum.