203 replies to this topic

[gliebzeit]

Mc Wolf, do you have this entry in your XXseason.ini file?

[EventXX]
name=I. Großer Preis von Berlin
shortname=Berlin GP
trackDirectory=avus50s
trackShortName=AVUS 1954
flag=gerflg
month=9
day=19
numberOfLaps=60
startingGrid=10

[Mc Wolf]

Thank you, Greg......now it works. My entry parameter were wrong.

[Alter]

Thank you Saiph for fixing the bug!
Alter

[Gartner]

282 and 256 ? hmm... how are your gearbox settings for SL and Connaught? My topspeed is only 270 (SL) and 251 km/h (Connaught)... (dancing the slowfox)

[Bob Simpson]

Maybe your straight is too short?

I used a final ratio of 3.938 at 8,550 rpm

Edited by Bob Simpson, Aug 07 2017 - 11:43 AM.

[Gartner]

Bob Simpson, on Aug 07 2017 - 11:41 AM, said:

Maybe your straight is too short?

I used a final ratio of 3.938 at 8,550 rpm

Thank you, Bob. That's it. Now moving in the right direction...

[Lord]

Thanks for the great work, mates! It's really head-spinning, congrats!

However, I have to report a couple issues with the mod:

1. The Lancia is a bit too long in the tail, the nose is a bit too straight and it's too squared, when compared to the REVS Institute photos: I don't claim to be right, but this is what I've managed to model for my rFactor mod (it's just an add-on, actually) with those photos.I've attached the link below to my  google album. If somebody can take some time in teaching me the art of modding GPL, I'd love to give you guys a new model, should you want it (I hate to be all bark and no bite ^^; ). Nonetheless, great work!

https://photos.app.g...9X9Zu1lRIcAlue2

2. Two days ago I had to format my Win10 PC and because of that I had to reinstall GPL: after installing GEM+2 2.5.0.32, I've installed 1965, 1966, 1967Historical, 1969X and the GT, all went well. I've tried to install 1955 v1.0.3 (the latest available on the official website) but GEM crashed every time I tried to activate it. I've also tried Saiph's fix, but the problem remained. I've attached the debug file here, only edited out my name. Stange thing is, the mod was working well before the format.

EDIT: To problem nr. 2, the issue was Avast recognizing the gplc55 exe as a virus (a trojan, in fact): adding it to the exclusions solved the problem. I wonder why that happened, since I know for sure that nobody here is dickish enough to put malware on purpose. A false positive seems to be the most reasonable explanation...

Attached Files

•  Debug.txt   88.04K   1 downloads

Edited by Lord, Oct 01 2017 - 07:29 AM.

[hagapito40]

I found this video...
on 1:20 Moss say that the 250f reach nearly 180 mph at Reims. The images shows a V12 lightweight model but he is probably talking about the 6L 2.5 liters raced by him on 1956.
I still think that the top speeds are wrong...even if the lap times are right.

https://www.dailymoti...m/video/x2umlhe

Edited by hagapito40, Oct 11 2017 - 04:49 PM.

[gliebzeit]

Moss drove the Maserati 250F that he is speaking about in 1956.  By that time the Maser was tuned to approximately 270 bhp at 8000 rpm, some had a lightweight chassis, there was usually a wrap-around windscreen and the nose was profiled to be narrower.  It may have reached nearly 180 mph on the Reims straight.

Edited by gliebzeit, Oct 11 2017 - 05:43 PM.

[Border Reiver]

What was the average lap time difference from '55 to '56? Presumably most of any change would have come from travelling down the straights quicker since the majority of corners are either taken flat out, or are very slow and so I doubt much lap time change in only a year would be due to better brakes or traction over just one year, so it must come from more power and less drag. Basically then, would the difference in '55 to '56 laptimes be in agreement with the change in reported top speed between '55 and '56 since that is where the time must be getting saved? Obviously an increase in engine power and a reduction in weight and drag would easily account for higher top speeds in '56 compared to '55.

Rob

[Pavel]

I also questioned top speed that 55 mod cars shows. All sources over internet says that Merceds W196 could reach 300 km/h, and Maserati 250F - 290 km/h. In mod their speed is less 25 km/h. On most short tracks cars from mod shows lap times close to real one. But for some reason at Spa cars from mod are too fast - 20s difference it's too much. Yes, track was little differ in 1955 than in 1967. But not as much.

I feel, that grip-power balance is bit wrong. With less grip and more power Spa will not be track with flatout corners. Top speed will be the same but speed in corners will drop and lap time will increase. Just my thoughts after driving this mod for 6 months

[gliebzeit]

Rob, there were no races at Reims in 1955 ... not even non-championship ones.  So, a comparison between 1954 and 1956 is only possible.  Fangio did a 2:29.4 in the Merc W196s in 1954.  He did a 2:23.3 in the Ferrari D50 in 1956.

[Border Reiver]

Thanks Greg. After I posted I remembered that there was no 55 race, but anyway, yes, I think that still works as a comparison.

So, the track was also the same for both '54 and '56 I believe and the track length was 5.158 miles long, and the difference in times over two years was ~6 seconds. As it turns out this equates to a change in average lap speed of just under 6 mph then. Since the car won't simply have been 6 mph faster at every point around the lap, it also makes sense that the car could easily have been around 12 to 15 mph faster in absolute top speed over two years in order to account for this change in lap time, especially because if it were attaining higher top speeds on the straights then the driver would also have had to start braking earlier too and for longer to get rid of this additional momentum the car will now have had. As I said above, I don't believe that the lap time can have been getting gained through the slower turns and so it must be reduction in time taken along each straight which was making the difference.

Rob

[Lord]

gliebzeit, on Oct 11 2017 - 05:04 PM, said:

Moss drove the Maserati 250F that he is speaking about in 1956.  By that time the Maser was tuned to approximately 270 bhp at 8000 rpm, some had a lightweight chassis, there was usually a wrap-around windscreen and the nose was profiled to be narrower.  It may have reached nearly 180 mph on the Reims straight.

Moss is referring to the 250F T1M, which was a modified vestion of the T1: longer and more tapered nose, longer back (still rounded in shape), revised and slightly lighter chassis. Think of it as an hybrid between the classic T1 '54 and the T2 '57. Also, 290 kph (180 mph) isn't that bad, is comparable to the Vanwall VW5-9 in 1958.

Border Reiver, on Oct 12 2017 - 05:50 AM, said:

What was the average lap time difference from '55 to '56? Presumably most of any change would have come from travelling down the straights quicker since the majority of corners are either taken flat out, or are very slow and so I doubt much lap time change in only a year would be due to better brakes or traction over just one year, so it must come from more power and less drag. Basically then, would the difference in '55 to '56 laptimes be in agreement with the change in reported top speed between '55 and '56 since that is where the time must be getting saved? Obviously an increase in engine power and a reduction in weight and drag would easily account for higher top speeds in '56 compared to '55.

Rob

Pavel, on Oct 12 2017 - 07:31 AM, said:

I also questioned top speed that 55 mod cars shows. All sources over internet says that Merceds W196 could reach 300 km/h, and Maserati 250F - 290 km/h. In mod their speed is less 25 km/h. On most short tracks cars from mod shows lap times close to real one. But for some reason at Spa cars from mod are too fast - 20s difference it's too much. Yes, track was little differ in 1955 than in 1967. But not as much.

I feel, that grip-power balance is bit wrong. With less grip and more power Spa will not be track with flatout corners. Top speed will be the same but speed in corners will drop and lap time will increase. Just my thoughts after driving this mod for 6 months

That might be the age-old question "should I follow the top speed or the laptimes?" I always face with other modders in rFactor. It always comes a time when, while you try to push outside the box of what you know it's working. For example, in rFactor the AI files (AIW and RCD) are optimized for a rear-engined, fully independent suspensions, high grip car. Once you start trying to emulate (let's say) a '50s GP car, it simply goes nuts: the AI slips everywhere, engines start blowing up, brake zones are all sorts of :angry33:ed up. So you start adding "helping hands" to the AI in form of torque stabs, grip multipliers and so on, but the underlying problem is thatill that the AIW/RCD system in itself isn't designed to do that and you're forcing it to at least come close to the behaviour you want.
Just as an example, I'm making a 1958 mod for rFactor: I've made the cars as close as possible to the real deal (bought books about them, did research, the usual stuff a conscious modder would do, the same stuff the GPL teams do and did time and time again). The result is that the cars are spot on in Monza and Monaco, go faster in some circuits (like Reims and Silverstone) and slower in others (Spa). After modifying a copy of the mod, it occurred that the results varied wildly but always following the AIW files.
This only means that, for a determined style of car, a new AIW file is needed, optimized for the cars you want to work with. This will leave the other cars almost useless though, since those aren't designed for the new AIW.

I honestly think that something like this is occurring with the 1955 mod also. Personally, I'd concentrate on the cars rather than the laptimes, but that's just my way to go about things.

[Border Reiver]

I think I would also add here that there is an underlying assumption of some basic "truth" here. (By which I mean we are assuming some datum which is absolutely correct against which everything is judged and corrected).

In this case the big assumption is that the tracks themselves are correct so for the correct input car parameters that they will yield correct lap times and correct top speeds etc. This actually then becomes a circular argument since from the outset we know that the 67 cars originally shipped are something like 4% quicker per lap than they should be, but then also the tracks as delivered by Papy are also not correct and are generally too wide, allowing faster cornering.

With the tracks that were then built as add on tracks, we know the laptimes that were set in reality and so for an "average" driver in a given car we can aim to match the track to give a real lap time for that same car, but, this assumes the car being used for this comparison is also "true", and then also neglects a lot of other points like the virtual driver cannot be hurt by pushing the limits, the track in GPL is uniformly smooth and in good repair with a clearly defined edge that you can run right up to etc. Some addon tracks are also based on better or worse original data and can have been optimised for speed by using different ability drivers as their baseline and indeed different cars and mods to determine which target times to use for their "calibration". To drive the track you wouldn't really feel something like a 2 degree camber change, but this will have a huge effect on the lap times produced, and this sort of detail is generally not detailed in track plans, even if the track is based off of a 100% correct shape from an aerial photo.

This becomes even more circular as an argument when further mods have relied on these derived addon tracks to set their performance, and then further tracks have used these later mods to set their lap speeds. The potential and cumulative error, or uncertainty if you prefer, simply cascades down the further you get away from original game content that has been used as the datum reference.

With the mods, in order to keep things correct and logical, there has also been a tradition to maintain this ~4% over performance from the original Papyrus game. This means that the step up in performance over two years from 65 mod back to the original 67 cars is in relative terms correct, with the 66 mod sitting as it should between these two, but in a truly global sense they are all too quick. An approach that has been used is to find tracks which were used in both seasons in question, e.g. Silverstone, Spa, Zandvoort, Monza, Monaco are common candidates. Using the same test driver, or ideally multiple test drivers to take an average, you set times in an existing mod at each track as a baseline and then again in the new mod under development. From historical data you find that in e.g. 3 years for the same real driver or car model there was a 3% decrease in times at Silverstone, a 2% change at Monza, and 4.2% change at Monaco, or whatever, and then you look to adjust the mod physics to yield a similar change there too. Using tracks of different types helps with this as it shows if the change is down to handling, engine power etc., but this also removes the absolute need for the tracks themselves to be correct and perfectly accurate since we are keeping the tracks constant in both tests and looking for a relative performance change across all tracks, but weighted by each of the specific track characteristics. For sure, it is not perfect and it does make some assumptions, but it is also probably about as good as can be achieved given the available resources.

Rob