Example of MoTeC data analysis

  • Thread starter ALB123
  • 316 comments
  • 54,131 views
MatskiMonk
Again...

abcdefgh-png.254419


PD didn't say we were going to get a 7 post. (Might be nice if we did though :D)
Click to expand...
Thought they said it at one point. Oops
 
shirakawaa
(see updated message)
You can export interpolated data up to 1000 Hz, by the way.
Click to expand...
It might be hard to get good data out of it if it's all interpolated. Because of the way replays work in GT, in order to get high resolution data of physical parameters, the physics need to be run.

Since suspension tuning relies on frequency analysis and high resolution logging of stroke, wheel hop etc. it's perhaps not surprising this info is not included in the export.

A bespoke suspension tuning tool in the game would be ideal...
 
Last edited:
Ridox2JZGTE
Is suspension data really necessary with the state of the general GT6 tuning methodology in use by tuners here ? :lol: Just use the usual trick used by well known tuners, zero camber, ride height exploit ( ass up or ass down to the extreme ), extreme damper and spring values to get the desired effect ( the usual min/max or flipped front/rear with 1 or 2 click higher or lower ) and near open diff LSD value - oh ballast to the make RR/MR/FF/AWD/FR closer to 50/50 car - killing it's unique handling balance, and when it's easier to drive and quickest in lap times, then claim to be the best tuner :lol: PD needs to have the tire model upgraded if wanted this to cease .... :sly:
Click to expand...
Thank you for teaching me how to tune cars in one post lol.
 
Ridox2JZGTE
Is suspension data really necessary with the state of the general GT6 tuning methodology in use by tuners here ? :lol: Just use the usual trick used by well known tuners, zero camber, ride height exploit ( ass up or ass down to the extreme ), extreme damper and spring values to get the desired effect ( the usual min/max or flipped front/rear with 1 or 2 click higher or lower ) and near open diff LSD value - oh ballast to the make RR/MR/FF/AWD/FR closer to 50/50 car - killing it's unique handling balance, and when it's easier to drive and quickest in lap times, then claim to be the best tuner :lol: PD needs to have the tire model upgraded if wanted this to cease .... :sly:
Click to expand...

What do you mean by flipped suspension? i have heard this term a lot but with the gears as well but couldn't be bothered asking until now sorry it landed on your shoulders to explain :gtpflag:
 
Might be a decent driver training tool, I haven't dug too deep into the functionalities yet, can the track map do a good job on line differentiation and stuff? Also you can do a theorectical best lap, which would give you a target to shoot for.

Also as said this can be use to do a whole session's data as opposed to just fast lap.
 
Johnnypenso
Agreed. Unless there's more to it, it looks like some colourful graphs to please the masses, but for anyone really into tuning and wanting to make their car faster, the vital suspension data is missing and the data really only becomes useful on a trial and error basis which is basically what we already have now, sans the colourful graphs.
Click to expand...

Log Suspension data would easy to find out where have a trap and how to optimize the limit.
Now the MoTeC support such as data logger in game.

Compare where the different from skill is not useful there, since there have Ghost system.
 
Last edited:
Min an max corner speed, numerical G readings, this is incredible. This can aide in testing theories out, lap comparisons, & tuning quite a bit, it even does sector times all on full multiple lap replays. This is impressive.
 
vasiliflame
What do you mean by flipped suspension? i have heard this term a lot but with the gears as well but couldn't be bothered asking until now sorry it landed on your shoulders to explain :gtpflag:
Click to expand...

Dampers values are flipped ( the values can varies, sometimes the front has same comp/ext ), for example front Comp/Ext : 2/4 and rear Comp/Ext : 4/2. or extreme one : front 1/10, rear 10/1, but usually front 5/6, rear 6/5 or the other way around front 5/4, rear 4/5. These are generally used with ride height exploit ( front lower/rear higher for more understeer/reduce rotation, rear lower/front higher for more rotation/less understeer ) Using the ride height difference values at the extremes as a quick fix will give bad side effects on car balance, so spring rate also often needs extra values in difference with the flipped dampers and tendency of high ARB ( roll bar ) values as band aids.

An example for an MR car that's very hard to stabilize ( prone to oversteer ) :
F / R
Ride Height : low high ( can be min / max )
Springs : low high ( can be 1:2 in ratio or more, say 5 10 to balance with the ride height effect )
Damper Comp : high low
Damper Ext : low high ( front extension can use the same value as comp in some cases - on bad behavior cars - understeer on entry due to the extreme ride height difference, 4 to 6 difference is common, eg 2 7 )
ARB : high high ( often 1 value lower at the rear )

The general effect of damper compression at the front / rear : low / high ( tendency to oversteer ) , while high / low ( tendency to understeer )
At the rear, lower comp and higher ext reduce oversteer / more understeer. For the front, understeer on entry : lower comp / higher ext, while understeer on exit : higher front ext with higher rear compression. Setting the front compression / extension at higher range with close values like 6 / 8 or 7 / 8 helps with reducing oversteer, coupled with rear comp / ext at high difference : 1 / 8, helps a lot to reduce rear rotation, but this usually works with ride height used ( lower front / higher rear ) and high ARB ( more than 4 front and rear )
 
Im looking at a datalogue in i2 Pro, Ive got a video recording of the replay I did with my phone and Im trying to figure out how to overlay data on the video and apparently its supposed to auto sync
 
Ridox2JZGTE
Dampers values are flipped ( the values can varies, sometimes the front has same comp/ext ), for example front Comp/Ext : 2/4 and rear Comp/Ext : 4/2. or extreme one : front 1/10, rear 10/1, but usually front 5/6, rear 6/5 or the other way around front 5/4, rear 4/5. These are generally used with ride height exploit ( front lower/rear higher for more understeer/reduce rotation, rear lower/front higher for more rotation/less understeer ) Using the ride height difference values at the extremes as a quick fix will give bad side effects on car balance, so spring rate also often needs extra values in difference with the flipped dampers and tendency of high ARB ( roll bar ) values as band aids.

An example for an MR car that's very hard to stabilize ( prone to oversteer ) :
F / R
Ride Height : low high ( can be min / max )
Springs : low high ( can be 1:2 in ratio or more, say 5 10 to balance with the ride height effect )
Damper Comp : high low
Damper Ext : low high ( front extension can use the same value as comp in some cases - on bad behavior cars - understeer on entry due to the extreme ride height difference, 4 to 6 difference is common, eg 2 7 )
ARB : high high ( often 1 value lower at the rear )

The general effect of damper compression at the front / rear : low / high ( tendency to oversteer ) , while high / low ( tendency to understeer )
At the rear, lower comp and higher ext reduce oversteer / more understeer. For the front, understeer on entry : lower comp / higher ext, while understeer on exit : higher front ext with higher rear compression. Setting the front compression / extension at higher range with close values like 6 / 8 or 7 / 8 helps with reducing oversteer, coupled with rear comp / ext at high difference : 1 / 8, helps a lot to reduce rear rotation, but this usually works with ride height used ( lower front / higher rear ) and high ARB ( more than 4 front and rear )
Click to expand...
Translation for anyone that doesn't want to read the entire post: Tuning in the GT series is a joke.
 
Johnnypenso
Translation for anyone that doesn't want to read the entire post: Tuning in the GT series is a joke.
Click to expand...

:lol: The damper, ARB and springs are okay, the ride height and camber are still funky :lol: I would imagine, a proper tire model would bring it much closer to PC sim level. The ride height and camber are the biggest issue for me as replicas often have lower front or higher front and this makes me have to tune within limitation that puts the car in bad state of handling balance to begin with. So far, most are a bit quicker than in real life, so with a proper camber, ride height, the car would be even quicker ... :eek:
 
diptob79
Can we extract the GPS data from GT6 log file and find out where is the real location of any fictional track?
Click to expand...

Haven't really had time to look, I picked a lap at random and it happened to be at Sierra, so far as I could tell it doesn't use GPS to calculate the track (since it's fictional), it uses the G metre and the corrected distance counter. It then calculates the track 'sections' depending on the length of a straight, or the curvature of a bend. It's not what I'd expected and will require a bit of looking into to properly understand where that may throw up inconsistencies in comparisons.

There is some google maps sync for the GPS data.

It may be different on data from real-world tracks. It's interesting to note that G's are measured not only in latitude and longitude, but also vertically. Basically the Motec system looks like it could generate a track from GPS/G-measurements, complete with X, Y, Z co-ordinates, and then place it on a google earth map. Does't mean anything for GT yet, but if Motec and PD are working together, on the GPS course creator it would make a lot of sense.
 
MatskiMonk
Haven't really had time to look, I picked a lap at random and it happened to be at Sierra, so far as I could tell it doesn't use GPS to calculate the track (since it's fictional), it uses the G metre and the corrected distance counter. It then calculates the track 'sections' depending on the length of a straight, or the curvature of a bend. It's not what I'd expected and will require a bit of looking into to properly understand where that may throw up inconsistencies in comparisons.

There is some google maps sync for the GPS data.

It may be different on data from real-world tracks. It's interesting to note that G's are measured not only in latitude and longitude, but also vertically. Basically the Motec system looks like it could generate a track from GPS/G-measurements, complete with X, Y, Z co-ordinates, and then place it on a google earth map. Does't mean anything for GT yet, but if Motec and PD are working together, on the GPS course creator it would make a lot of sense.
Click to expand...
Exactly was i've been thinking. I guess we'll hear more about motec pretty soon :)
 
What an excellent update from PD doins this! The very nice thing of having Lap Analisys in motec, is that in i2 you can compare your lap with your friends or arrivals lap.

I can see many people getting some improvement with it.
 
@MatskiMonk Nice...I see you've been spending some time in the MoTeC software as well. Pink Floyd fans must truly think alike (nice avatar! :cheers:) Anyway, I was wondering if you know anything about this...

You see your photo that shows the "Track Editor"? I wonder if you can 'modify' what is considered a turn or a straight and save that, so when you bring in data the next time, will it use your new interpretation of what are the straights and what are the turns? I'm not thinking of editing the tracks because I think MoTeC is doing anything incorrectly, but perhaps just to simplify things. Like, let's say I was trying to compare lap times with you. Maybe I'm trying to identify a particular area of the track and I want to divide a particular section up into even smaller pieces, to see where I'm losing time to you in my lap. Will my changes be permanent or just within that editing session? Am I even making sense? :lol:
 
ALB123
@MatskiMonk Nice...I see you've been spending some time in the MoTeC software as well. Pink Floyd fans must truly think alike (nice avatar! :cheers:) Anyway, I was wondering if you know anything about this...

You see your photo that shows the "Track Editor"? I wonder if you can 'modify' what is considered a turn or a straight and save that, so when you bring in data the next time, will it use your new interpretation of what are the straights and what are the turns? I'm not thinking of editing the tracks because I think MoTeC is doing anything incorrectly, but perhaps just to simplify things. Like, let's say I was trying to compare lap times with you. Maybe I'm trying to identify a particular area of the track and I want to divide a particular section up into even smaller pieces, to see where I'm losing time to you in my lap. Will my changes be permanent or just within that editing session? Am I even making sense? :lol:
Click to expand...

I get what you mean, and I've not had chance to experiment too much, but I don't think you can, at least not in the standard i2, maybe pro. The Track itself doesn't seem to actually exist, just a set of data that can be interpreted as a circuit. You can change the start and finish points for each corner and straight, and you can split sections, so you can make each corner an entry and exit section, or split double apexes.. but I think it only applies to that data set.

If you look at the track map of Tsukuba, that's not the track map with an overlay, it's simply my racing line, hence all the corners are smoothed out.

Unlocking the power of this can be tricky, I find it fascinating, but I'm not really ready to let go of my old method for lap time analysis just yet, but this should offer much higher resolution for time analysis - it has the one thing I really wanted from the in game data logger, and that's the ability to show precise time and distance measurements.

I might look in to how much the Pro version of the Software costs.

p.s. Great Album, except for the opening lines of Louder than Words (facepalm)
 
MatskiMonk
Freaking Awesome.
Click to expand...

Could you test 0-300-0 with Veyron (stock) ? I am especially interested in 200-300 time. But all others are interesting too.

Here are the numbers from the real world:

0-100 km/h: 3,1 s
0-200 km/h: 8,5 s
0-300 km/h: 20,7 s

300-0 km/h: 7,1s

And also with Mclaren MP4-12C:

0-100 km/h: 3,3 s
0-200 km/h: 9,7 s
0-300 km/h: 31,0 s

300-0 km/h: 7,1s

More real world test data:

http://www.sportauto.de/marken/news...twagen-beim-extrem-leistungstest-1416753.html

http://www.sportauto.de/marken/news...nell-ist-der-bugatti-veyron-16-4-2747791.html


http://www.sportauto.de/testbericht...claren-mp4-12c-im-highspeed-test-4235189.html
 
Last edited:
YZF
Could you test 0-300-0 with Veyron (stock) ? I am especially interested in 200-300 time. But all others are interesting too.

Here are the numbers from the real world:

0-100 km/h: 3,1 s
0-200 km/h: 8,5 s
0-300 km/h: 20,7 s

300-0 km/h: 7,1s

And also with Mclaren MP4-12C:

0-100 km/h: 3,3 s
0-200 km/h: 9,7 s
0-300 km/h: 31,0 s

300-0 km/h: 7,1s

More real world test data:

http://www.sportauto.de/marken/news...twagen-beim-extrem-leistungstest-1416753.html

http://www.sportauto.de/marken/news...nell-ist-der-bugatti-veyron-16-4-2747791.html


http://www.sportauto.de/testbericht...claren-mp4-12c-im-highspeed-test-4235189.html
Click to expand...

Only got chance to do the Veyron. But stock, on CS tyres, with ABS and TCS at 1 (I figure you can't totally disable aids in the real thing), MT, at SSRX back straight.

0-100 : 3.032 s
0-200 : 7.630 s
0-300 : 16.982 s

200-300 : 9.352 s
Untitled-1.jpg

I then carried on to about 252 mph..

300-0 : 6.596 s (technically this was from 298.5 km/h, that's when I hit the brake pedal on the braking run, as my speedo in game was set to mph)

I don't go for these real world comparos too much since there's so many other variables. It does look like maybe CM tyres would give a closer time.
 
Last edited:
Thanks MatskiMonk.

I actually did these test myself too, and my findings confirm the incorrect aero physics simulation this game has.

So the results are:

Veyron (100% stock, default tyres, Auto)

0-100 kph: 2.4 sec -vs- 3.1 sec (real life)
0-200 kph: 7 sec -vs- 8.5 sec (real life)
0-300 kph: 16.2 sec -vs- 20.7 sec (real life)


Mclaren MP4-12C (100% stock, default tyres, Manual)

0-100 kph: 3.4 sec -vs- 3.3 sec (real life)
0-200 kph: 9.3 sec -vs- 9.7 sec (real life)
0-300 kph: 21.9 sec -vs- 31 sec (real life)

As I mentiond in this thread , above 200kph the broken aero physics of GT6 are really evident. Cars accelerate too fast and they reach much higher top speeds than they do in GT5 and in real life.

Interestingly enough, Veyron has the same top speed in game as in real life (although acceleration is wrong as well), but all (or at least majority) of other street/stock cars are doing 30-50 kph higher top speeds
 
Yeah, the braking and 0-60 times implies too much grip, whilst the high speed accelerations betray the fundamental aero issue the game has in its present state.

EDIT: ha, thought that's what your question was leading to :P
Good idea. :)
 
More tests, all lead to the same findings. As soon as air resistance increases enough to have substantial effect (above 200 km/h or 125 mph), wrong physics calculation is clearly evident:

Mercedes SLS AMG

0-100 kph: 3.8 sec -vs- 3.9 sec (real life)
0-200 kph: 11.3 sec -vs- 11.7 sec (real life)
0-300 kph: 28.0 sec -vs- 42.2 sec (real life)

Lamborghini Murcielago LP 670-4 SV

0-100 kph: 3.5 sec -vs- 3.2 sec (real life)
0-200 kph: 10.4 sec -vs- 10.3 sec (real life)
0-300 kph: 24.3 sec -vs- 29.7 sec (real life)


Corvette ZR1

0-100 kph: 3.9 sec -vs- 4.0 sec (real life)
0-200 kph: 10.3 sec -vs- 11.5 sec (real life)
0-300 kph: 23.8 sec -vs- 35.4 sec (real life)


Lexus LFA

0-100 kph: 4.0 sec -vs- 4.1 sec (real life)
0-200 kph: 11.2 sec -vs- 12.0 sec (real life)
0-300 kph: 26.3 sec -vs- 50.6 sec (real life)


Real world test data sources (at the bottom of each page):

http://www.sportauto.de/marken/news...twagen-beim-extrem-leistungstest-1416753.html

http://www.sportauto.de/marken/news...nell-ist-der-bugatti-veyron-16-4-2747791.html


http://www.sportauto.de/testbericht...claren-mp4-12c-im-highspeed-test-4235189.html


 
Last edited:

Latest Posts

Back