[Discussion] Using Magic Data to simulate high-altitude races.

Hi all,
For the track I'm currently working on, I was curious to see how I could simulate a high-altitude race; Monza style speeds with Monaco levels of wing, such as we see at Mexico.
I knew that I could do it via the physics file, but I wanted to try with Magic Data.
Now that I've done it, I'm not sure that I'll actually use it, but I thought I should share what I found out anyway.

Notes:
-------
* Using TSM, so value names are different to CMagic [See Atticus's reply for a good method if you use CMagic]
* Tested with Kedy89's 2002 Mod

Ride Height:
First, I started to decrease the 'Ride Height' value from ~975, until I decided on '600'.
At '600', the 2002 cars will hit 360 km/h+ in the slipstream with the following Wing and Gear values:

FrontWing AI Dry=20
RearWing AI Dry=18
1stGear AI Dry=25
2ndGear AI Dry=32
3rdGear AI Dry=39
4thGear AI Dry=46
5thGear AI Dry=53
6thGear AI Dry=60

However, just changing that absolutely ruined lap times for the player (and the AI, though to a lesser extent) to the point of it being unplayable.

Downforce:
The 'Downforce' value, which started off at around '22', I eventually made '0' where it gives the most grip possible.
This helped, but lap-times were still much slower than before.

Track Grip:
This value has a very large impact on grip levels, and I increased it in intervals of 1000 until I got my desired laptimes (at 24500)....but, at this point, the AI were far too fast.
(Nearly 2 seconds a lap!)

Grip factor Ace:
As of version 4.44 of GPxPatch, the 'Power/Grip factor [Difficulty]' values in Magic Data work properly (i.e, they only affect the AI) and so I was able to adjust this to bring the AI lap times back to a reasonable level.
The 'Power' factor is fine at '495', but I had to lower the 'Grip' factor from '500' to '440' at Ace level. (Obviously this is lower at lower difficulties)

So, long story short:

FrontWing AI Dry=20
RearWing AI Dry=18
1stGear AI Dry=25
2ndGear AI Dry=32
3rdGear AI Dry=39
4thGear AI Dry=46
5thGear AI Dry=53
6thGear AI Dry=60
.......
.......
Track grip=24500
DownForce=0
Ride height=600
......
......
Power factor Ace=495
Grip factor Ace=440

And this will give you what is (I think) a good - and driveable - simulation of a high-altitude circuit.
Lowering the 'Track Grip' value might give a more authentic experience but, for me, I found it too difficult.

---

F1 Graphics Swapper --- Lensdump - Renders and Screenshots --- GP4 Performance Generator



Edited 1 time(s). Last edit at 09/02/2016 12:54PM by AUS_Doug.

Have you actually tried with the air pressure value?

It's below downforce and ride height.

For my Hermanos Rodriguez 1968, I obviously needed a very low number and using an altitude-to-air pressure calculator online, I got a 12986 figure. (I used Monaco's 17055, the lowest among the original 17 tracks, for 30m above sea level - the average altitude of the Circuit de Monaco - as a reference and it's a simple calculus from there.)

It gives you less horsepower, and on the Hermanos Rodriguez you can really feel it if you do a side-by-side comparison without magic data or with a different number, and - I imagine - less grip as well.

Edit: Obviously, you would then need to compensate the horsepower value, but only the horsepower value and not the grip, for a turbo situation, but this could be a clean starting point.

---

My workthread - [www.grandprixgames.org]
Full of classic F1/non-F1 track layouts

My blog about F1 performance analysis - [thef1formbook.wordpress.com]



Edited 4 time(s). Last edit at 09/02/2016 11:12AM by Atticus..

That must be one of the Magic Data values that TSM is missing
I just used CMagic to get the data for Monaco, and compared it to the data that TSM gets for Monaco, and I can't see that 17055 anywhere in the TSM-generated magic data file but I can in the CMagic file.

---

F1 Graphics Swapper --- Lensdump - Renders and Screenshots --- GP4 Performance Generator

AUS_Doug Wrote:
-------------------------------------------------------
> That must be one of the Magic Data values that TSM
> is missing
> I just used CMagic to get the data for Monaco, and
> compared it to the data that TSM gets for Monaco,
> and I can't see that 17055 anywhere in the
> TSM-generated magic data file but I can in the
> CMagic file.


Ouch. I didn't know that. (I don't use TSM.) That's a shame though because, as I've said, it has a huge and very very noticeable effect on performance. Kyalami is another track of mine which is much different with the correct air pressure. (Obviously.)

---

My workthread - [www.grandprixgames.org]
Full of classic F1/non-F1 track layouts

My blog about F1 performance analysis - [thef1formbook.wordpress.com]

I'll certainly be trying the CMagic way tomorrow now, now that I know that's how you did your Mexico!

---

F1 Graphics Swapper --- Lensdump - Renders and Screenshots --- GP4 Performance Generator

What Atticus said . Air pressure in the magic data works wonders for normally-aspirated engines: lower values mean less air (higher altitude), which means less horsepower. I haven't really tested to see if it affects downforce and drag as well, as it does in real life.

Turbo engines don't lose horsepower with altitude, so the higher speeds are a result from less aerodynamic drag. So you probably need to increase the air pressure value instead to get realistic top speeds, and lower the track grip to emulate the lower downforce.


My unfinished tracks: [www.grandprixgames.org]
Send bug reports and track editing questions to f1virtualblog@gmail.com

Can't believe I missed this discussion, I love a good geek-out on some magic data. Funnily enough it was when I was creating some magic data for Mexico 1991 (when we did a round there for the GP4CC) that i found some weirdness around altitude handling in GP4. I'll post a write up later in case anyone's still interested.

=====================================================


Intel NUC 8i3, 8GB RAM, MS Sidewinder Wheel

Sounds good Tom.

---

F1 Graphics Swapper --- Lensdump - Renders and Screenshots --- GP4 Performance Generator

Download: Mexico City 1991 Altitude Testing Pack

Hypothesis: GP4 attempts to simulate the effects of altitude on car handling, and does a s**t job of it.

I noticed something was wrong when I was preparing Mexico City 1991 for Round 2 of the GP4CC Season 3. Prblanco had done some excellent analysis and identified magic data values 46 and 47 seems to correlate with each tracks' altitude. I've reproduced this in the graphs below. Using these formulae, I calculated the values for Mexico's magic data.

While practising for the event, I noticed that I was much slower than the AI through the twisty middle sector. Nothing I tried on car setup worked; I was miles behind! I had lost my touch. My lead in the GP4 Central Championship was sure to be lost!

Eventually, after loading the Magic Data of other tracks into Mexico, I figured out the culprit; the altitude values affect the Player, but have no effect on the AI!

GP4's authors compensate for this (ridiculous, broken) behaviour by changing the AI magic data grip and power values for each track, at each AI level (rookie - ace). This is why the AI grip and power values are so different between Interlagos and Monaco, for example.

Test this for yourself using the download above. In it you'll find the Mexico 1991 track with 2 .dat files, identical except for the Magic Data values 46 and 47. One uses values calculated for an altitude of 0m. The other uses Mexico's correct altitude of 2235m. Compare the difference in your laptimes versus the AI.

I've created a way of fixing this broken behaviour, but that's for a future post...



=====================================================


Intel NUC 8i3, 8GB RAM, MS Sidewinder Wheel

You all should be in Abu Dhabi today as a F1 engineers . Congrats for the discussion

Could this desc46 be the actual air pressure in hPa?

Air temperature also affects the air density, "hot" tracks are below the red line while "cold" tracks above it.


My unfinished tracks: [www.grandprixgames.org]
Send bug reports and track editing questions to f1virtualblog@gmail.com

Nice write up Tom. I'm assuming that this - coupled with the CC-line - is why the AI are just too damn fast around 2001 A1 Ring?
Whoever was in charge of the AI Power/Grip factors didn't do enough to compensate for the desc46 and desc47 there?
(Or maybe I just suck at the AI Ring )

@Prblanco
Not a bad theory that, seems to fit what the graph shows.

---

F1 Graphics Swapper --- Lensdump - Renders and Screenshots --- GP4 Performance Generator

That's weird. I have no difficulties with A1Ring, it is one of the most enjoyable tracks to run and I always do well.

----signature--------------------------------------------------------

RELEASE => Physics (under FIA Technical Regulations)
MagicDatas? Setups? Pit Stop Strategies? Track Specific Performances? Up2Date GP?
Power Torque Curve? Where's this stuff? All download is available at ..

Show off!

=====================================================


Intel NUC 8i3, 8GB RAM, MS Sidewinder Wheel