Example of MoTeC data analysis
- Thread starter ALB123
- 316 comments
- 54,138 views
- 891
- Rio
- LeoStrop
Sorry for not being very clear, this explanation is at the training video(part 21/151).
The values for the axis of the graph are in minutes, that is, it shows your progression based on time. If you press F9, the graph will be displayed in meters, which is helpful when you overlay laps because you will see the comparison at the exact same place of the track, for both or more laps.👍
Last edited:
- 5,161
- Massachusetts
Oh, very cool! I saw the button with a little clock and I know I've pressed it before, but I think I did so with only 1 lap loaded and no second lap overlayed. So, I don't think I noticed the X-Axis (horizontal axis) changing. Thank you for pointing that out to me. It helps make comparing lap times MUCH easier now.
Before, I was having to use the track map to figure out where I was on the track and the go back to the Driver tab to compare the exact differences. I really need to make a custom Tab with the exact things I need on it. Unfortunately, since this is just the free version of the software (MoTeC i2 Standard, not Pro) you can only have 1 Track Report in the entire project file. So, I would have to kill the Tab marked Track Report if I wanted to make a custom Tab that had the Track Report AND a Time/Distance Graph, for instance.
- 891
- Rio
- LeoStrop
Why do you have the standard? I have the pro, download from here: http://www.motec.com.au/i2/i2downloads/
Does it expire with time? I haven't use any license or activation code....
Well i'm glad i helped, here is how my custom tab looks like, but it needs more improvement.
Attachments
- 5,161
- Massachusetts
Thank you @LeoStrop! I think that the link provided when the MoTeC integration was announced was for the Standard version of the i2 software. Well, I removed the Standard version and installed the Pro version today. What a wonderful difference that makes. Just being able to have more than 1 overlay is really nice. Then add in the fact that I can now create my own Workbooks and Worksheets, designed exactly as I would prefer them to be -- it is such a blessing. I don't even know why they bother keeping two separate versions.
- 891
- Rio
- LeoStrop
You're welcome!
Yeah, i didn't even noticed there was a standard version. It's very easy to copy the items you like and paste them into your personal tab. It's just hard to decide which ones to use, the track report needs a lot of space to be visible, i may let it alone in one tab and use the track position on the custom.
Anyway, have fun!
Yeah, i didn't even noticed there was a standard version. It's very easy to copy the items you like and paste them into your personal tab. It's just hard to decide which ones to use, the track report needs a lot of space to be visible, i may let it alone in one tab and use the track position on the custom.
Anyway, have fun!
- 5,161
- Massachusetts
Yes. Copy & Pasting is very easy. See, on the Standard Version you were only allowed like 5 Distance/Time Graphs in the entire Project file. Since the preset tabs already contained some of those graphs, you'd either have to delete one from a different tab, if you wanted to put one on your own custom tab and other features had limits of less than 5. Like I said, the track map thing has a limit of 1 in the Standard version. You're right, you do need to have it blown up pretty big to be very useful though. Anyway, I just think this is a great tool. Earlier I was doing some analysis of my laps around Brands Hatch and it's remarkable to see the differences (I was focusing on Turn 2) between taking an earlier apex vs. later apex...Sometimes the results aren't exactly what you'd expect. There were several instances where I thought I was carrying more speed than I really was. Anyway, you already know all that stuff... haha It's a fun tool. No doubt.
- 30
- Wichita Falls
- Salvaged1Gdsm
- Salvaged1Gdsm
I finally caught wind of this feature and I've been messing around with it for about 4 hours now reviewing an old free run best lap of mine, this MoTeC tool is pretty b.a. heres some screenshots from the i2 Pro software.
Tsukuba Circuit
'91 Acura NSX
500pp
Sport Softs
357hp
264.9ft-lb
1250kg
Tsukuba Circuit
'91 Acura NSX
500pp
Sport Softs
357hp
264.9ft-lb
1250kg
Attachments
- 5,161
- Massachusetts
Yeah, it really is an awesome tool to use with Gran Turismo 6. If you have the patience you can edit the Track Layout to more precisely define what is considered a corner and what is considered a straight. I've found a couple of times where the MoTeC software was a bit imprecise in that regard. But, if you're truly trying to determine your top speed in a straight, or a maximum speed through a particular corner, that information is extremely important!
Anyway, I'm very glad to see another GTPlanet member who is taking advantage of MoTeC i2 Pro. If you come up with any interesting uses or tricks or tips, please feel free to leave them in this thread. As much as I love this software, I have a feeling I am probably not using it to its potential and there are probably easier ways for me to do things that I just don't know about yet.
On a side note, there are a handful, or two of people who are conducting a study into the effects of running negative camber on your vehicles in GT6. We all pretty much know how it works in the real world at what it's supposed to do. Well, there is no question that real world does not equal GT6 world. Anyhow, in that thread, which you can find at the link below, is making great use of MoTeC for us to analyze our laps Speed, G Forces both Lat & Long, etc.. Check it out if you are interested.
https://www.gtplanet.net/forum/thre...t-drivers-wanted.320272/page-10#post-10347940
Anyway, I'm very glad to see another GTPlanet member who is taking advantage of MoTeC i2 Pro. If you come up with any interesting uses or tricks or tips, please feel free to leave them in this thread. As much as I love this software, I have a feeling I am probably not using it to its potential and there are probably easier ways for me to do things that I just don't know about yet.
On a side note, there are a handful, or two of people who are conducting a study into the effects of running negative camber on your vehicles in GT6. We all pretty much know how it works in the real world at what it's supposed to do. Well, there is no question that real world does not equal GT6 world. Anyhow, in that thread, which you can find at the link below, is making great use of MoTeC for us to analyze our laps Speed, G Forces both Lat & Long, etc.. Check it out if you are interested.
https://www.gtplanet.net/forum/thre...t-drivers-wanted.320272/page-10#post-10347940
- 1,175
- In front of you.
- F1_Racer68
Been using MoTeC i2 Pro for years with other Sims. Had been asking for this capability since GT5. So glad to finally have it. Now if only PD would allow the suspension and other telemetry data to be included. This is far superior to the in game data logger.
- 30
- Wichita Falls
- Salvaged1Gdsm
- Salvaged1Gdsm
- 5,161
- Massachusetts
Overlays are great. That feature alone is why you want to make sure you have the Pro version installed and not the Standard. Otherwise, you can only select ONE additional lap and while it might not be such a big deal when looking at graphs of speed or G forces or whatever, it most certainly IS a big deal when looking at the channel reports and wanting to see 5 laps data next to each other.
How do you make the background black in MoTeC? I am not in front of my laptop right now and I can't remember seeing an option for background color...unless I am blind.
How do you make the background black in MoTeC? I am not in front of my laptop right now and I can't remember seeing an option for background color...unless I am blind.
- 891
- Rio
- LeoStrop
Overlays are great. That feature alone is why you want to make sure you have the Pro version installed and not the Standard. Otherwise, you can only select ONE additional lap and while it might not be such a big deal when looking at graphs of speed or G forces or whatever, it most certainly IS a big deal when looking at the channel reports and wanting to see 5 laps data next to each other.
How do you make the background black in MoTeC? I am not in front of my laptop right now and I can't remember seeing an option for background color...unless I am blind.
I see that you still didn't watched the entire training video.
Tools> Options > Colours > Select Scheme.
- 5,432
- Codemasters (EA)
- PJTierney
Forgive my ignorance as I haven't tried out the MoTeC applications yet, but do they provide anything along the lines of suspension/camber data or notes on fuel depletion at all?
Looks like a good tool regardless which I might start using when tuning cars.
Looks like a good tool regardless which I might start using when tuning cars.
- 891
- Rio
- LeoStrop
- 5,161
- Massachusetts
The application itself does, yes... Unfortunately, you can't take advantage of that with Gran Turismo 6 because it doesn't export that information (suspension/camber data). I'm not 100% certain about fuel depletion since I haven't exported any replays files that were run with fuel depletion turned on. Obviously, we are all hoping that PD expands on this incredible feature and updates it in the future to include the data that they are currently excluding.
What I have been doing, in the mean time... If I am getting very similar lap times with two different suspension setups, I just export replays of both (keeping track of which log file corresponds to which suspension setup) and then I overlay the two runs. Then I can examine min, max & avg speed, G forces (Longitudinal & Latitudinal) and overall time through particular segments of the track.
I'm sure people who have used this tool with real Racing Sims will tell you that MoTeC is much more useful with those Sims than it is GT6 and I'm sure they're right because GT6 doesn't export nearly as much data, but I think once you start using it you'll see it can be extremely useful nonetheless.
EDIT: I see that @LeoStrop had posted an imagine showing exactly what information is exported by GT6 into MoTeC i2 Pro. 👍
- 8,698
You should be able to use gforce as a marker for fuel consumption. Using average figures and assuming m constant over a lap, F=ma, W=Fd, P=W/t = Fv = mav. Power P is directly proportional to fuel consumption for a given car, track, driving style etc. So acceleration figures will work as a suitable analogue.
Separate the forces into additive (e.g. positive longitudinal and negative vertical) and subtractive (magnitudes of the remaining forces), there should be an imbalance between the two. The magnitude of that imbalance will be proportional to the energy input required, and thus (ideal) fuel consumption.
It would only likely be comparable for a given car with given power parts installed, and it neglects aero (although the average of the square of the speed multiplied by the throttle position will serve as a useful marker for that, again only for comparison).
So the idea would be to find the line and driving style that minimises the energy imbalance and, potentially, the average of the square of the speed, whilst maximising lap time (i.e. the average speed). That latter condition implies holding a steady speed instead of speeding up and slowing down (whilst achieving the same average speed as in the steady case).
Test at different fuel loads to test the sensitivity of the constant mass assumption at the start, for longer laps and / or lighter cars.
Another method would be to "obtain" the torque curve of the car's engine, multiply the rpm by the torque by the throttle position, "integrated" over the whole lap. This method would be comparable across any cars you have torque data for, but even without knowledge of the absolute torque values, it can be used to compare approximate energy requirements for a given car and state of tune (assuming linear power delivery, i.e. constant torque, and ideal fueling - the latter being reasonable for a game).
Comparing these approximate engine power figures to the inertial and aero approximations above should also be interesting.
Separate the forces into additive (e.g. positive longitudinal and negative vertical) and subtractive (magnitudes of the remaining forces), there should be an imbalance between the two. The magnitude of that imbalance will be proportional to the energy input required, and thus (ideal) fuel consumption.
It would only likely be comparable for a given car with given power parts installed, and it neglects aero (although the average of the square of the speed multiplied by the throttle position will serve as a useful marker for that, again only for comparison).
So the idea would be to find the line and driving style that minimises the energy imbalance and, potentially, the average of the square of the speed, whilst maximising lap time (i.e. the average speed). That latter condition implies holding a steady speed instead of speeding up and slowing down (whilst achieving the same average speed as in the steady case).
Test at different fuel loads to test the sensitivity of the constant mass assumption at the start, for longer laps and / or lighter cars.
Another method would be to "obtain" the torque curve of the car's engine, multiply the rpm by the torque by the throttle position, "integrated" over the whole lap. This method would be comparable across any cars you have torque data for, but even without knowledge of the absolute torque values, it can be used to compare approximate energy requirements for a given car and state of tune (assuming linear power delivery, i.e. constant torque, and ideal fueling - the latter being reasonable for a game).
Comparing these approximate engine power figures to the inertial and aero approximations above should also be interesting.
I ran a HSV-010 '12 on SSRX, exported to MoTeC and then to CSV to load in R for statistical analysis. My (first) goal is to produce the engine power and torque curve as well as friction and drag coefficients from an acceleration test. I go full throttle in each gear from low RPM up to the limiter. There are two mysteries that keep me from proudly sharing that work:
When accelerating full throttle at about 300 km/h, the rear wheel speed is 10 km/h slower. The front wheels turn at the vehicle speed. I understand the concept of slip ratio, but then the rears should turn faster? Please correct me if I am wrong, but this looks like a bug to me. Any hints would be welcome.
To get to numbers that resemble the peak power of the GT6 GUI, I have to make drag go quadratic with vehicle speed. In that case the max. power is within 1.5% of the stated value and also the shape of the power curve resembles the one in the car settings GUI. But, AFAIK, drag force is quadratic with speed, and corresponding power is cubic with speed. If I model that, then the shape of the power curve is OK, but the max. power is 25% too low, with similar results for other variations with multiple parameters. Am I doing something stupid, or could this be another bug in GT6?
My approach is to solve a system of linear equations: engine_power(rpm) - drag*vehicle_speed^p = weight x acceleration x vehicle_speed
Here according to me p should be 3, but 2 gives much better results. Acceleration is based on "G Force Long", Weight is from the GT6 car settings GUI, vehicle speed is also from the MoTeC log-file. I filtered all samples to keep only those with full throttle and without wheelspin, shifting, limiter, etc.
When accelerating full throttle at about 300 km/h, the rear wheel speed is 10 km/h slower. The front wheels turn at the vehicle speed. I understand the concept of slip ratio, but then the rears should turn faster? Please correct me if I am wrong, but this looks like a bug to me. Any hints would be welcome.
To get to numbers that resemble the peak power of the GT6 GUI, I have to make drag go quadratic with vehicle speed. In that case the max. power is within 1.5% of the stated value and also the shape of the power curve resembles the one in the car settings GUI. But, AFAIK, drag force is quadratic with speed, and corresponding power is cubic with speed. If I model that, then the shape of the power curve is OK, but the max. power is 25% too low, with similar results for other variations with multiple parameters. Am I doing something stupid, or could this be another bug in GT6?
My approach is to solve a system of linear equations: engine_power(rpm) - drag*vehicle_speed^p = weight x acceleration x vehicle_speed
Here according to me p should be 3, but 2 gives much better results. Acceleration is based on "G Force Long", Weight is from the GT6 car settings GUI, vehicle speed is also from the MoTeC log-file. I filtered all samples to keep only those with full throttle and without wheelspin, shifting, limiter, etc.
- 16,432
- United Kingdom
Not enitrely sure I followed all that, but in this... "My approach is to solve a system of linear equations: engine_power(rpm) - drag*vehicle_speed^p = weight x acceleration x vehicle_speed"
What you term drag already includes the vehicle speed^2, (and yes power required to overcome that is ^3), then you are multiplying by (at least) the square of the vehicle speed again... I'm not sure because I'm not familiar with the rest of what you are doing, but perhaps this is why you get better results when p=2.
edit: Also, I see similar things happening with the wheel speeds. It's most noticeable in RWD cars that the Rear wheels turn too slow. I just figured that they calculate how fast the wheel should be turning, calculate what the speed differential would be from slip on the driven wheels, then knock that off the wheelspeeds... or add it on to the vehicle speed... where as they should knock it off the vehicle speeds or add it on to the wheelspeeds... ? I'm just guessing now
Last edited:
- 5,161
- Massachusetts
I don't mean to hijack your interesting conversation, but something you were talking about reminded me of a question I've had for some time now. Do you know how how Corr Speed is determined in MoTeC i2? The value is always different than Vehicle Speed. What's the lo-down on that?
I understand the confusion ALB123. With drag I meant "drag coefficient", e.g. P_drag = drag_coef vehicle_speed^2. So drag_coef is a single number that does not depend on vehicle speed that I estimate to be about 0.6 in SI units.Not enitrely sure I followed all that, but in this... "My approach is to solve a system of linear equations: engine_power(rpm) - drag*vehicle_speed^p = weight x acceleration x vehicle_speed"
What you term drag already includes the vehicle speed^2, (and yes power required to overcome that is ^3), then you are multiplying by (at least) the square of the vehicle speed again... I'm not sure because I'm not familiar with the rest of what you are doing, but perhaps this is why you get better results when p=2.
edit: Also, I see similar things happening with the wheel speeds. It's most noticeable in RWD cars that the Rear wheels turn too slow. I just figured that they calculate how fast the wheel should be turning, calculate what the speed differential would be from slip on the driven wheels, then knock that off the wheelspeeds... or add it on to the vehicle speed... where as they should knock it off the vehicle speeds or add it on to the wheelspeeds... ? I'm just guessing now
- 5,161
- Massachusetts
Sorry MatskiMonk. Thanks fort noticing ALB123.
- 16,432
- United Kingdom
Ah okay, as I say, I'm not sure of the equation that you are using, but it's my uderstanding that to come up with the drag force on the car the equation would 0.5 × 1.204 (std. density of air) × v² × Co-efficient of drag × Frontal Area of vehicle. Cd×A might well be about 0.6 (e.g. 0.3×2m²).
Attached is the first cleaned version of my drag analysis (input, script, output).
I did a run on SSRX with a HSV-010 '12, accelerating from low RPM in each gear.
The first part of the script is to filter out everything except full throttle, no wheel spin, no clutching, no bridge. This gives:
I should have run the test with closely packed long gears because half the samples are gone due to wheel spin.
The second part of the script estimates engine power and drag. I used two formulas (p=2 and p=3). The first one (figure below) matches better with the game, but the second one I believe is physically correct.
The estimated values are:
Car data from GT6 car settings GUI:
The problems mentioned in the previous post still exist: Slip < 0, p = 2.
I did a run on SSRX with a HSV-010 '12, accelerating from low RPM in each gear.
The first part of the script is to filter out everything except full throttle, no wheel spin, no clutching, no bridge. This gives:
I should have run the test with closely packed long gears because half the samples are gone due to wheel spin.
The second part of the script estimates engine power and drag. I used two formulas (p=2 and p=3). The first one (figure below) matches better with the game, but the second one I believe is physically correct.
The estimated values are:
Code:
Slip : -3.232999 %
Wheel.Radius : 358.8659 mm
p : 2
Drag.Param : -52.18265
Engine.RPM.nodes : 2349 2960 3571 4182 4793 5404 6015 6625 7236 7847 8458 9069 9680 10291
Engine.Power [HP] : 67 81 135 185 244 314 384 452 512 559 554 549 530 537
Engine.Torque [Nm] : 204 194 269 316 363 413 454 486 503 507 467 431 390 372
p : 3
Drag.Param : -0.528427
Engine.RPM.nodes : 2349 2960 3571 4182 4793 5404 6015 6625 7236 7847 8458 9069 9680 10291
Engine.Power [HP] : 69 50 100 140 187 244 300 361 416 461 458 457 459 456
Engine.Torque [Nm] : 208 120 199 238 278 322 355 388 409 418 386 359 337 316Car data from GT6 car settings GUI:
Code:
Honda RAYBRIG HSV-010 '12
600 PP
Max. power 575 HP @ wide RPM range
Max. torque 54.2 kgfm (= 532 Nm)
Weight 1100 kg
Tires RH/RH
Brake balance 5/7
Ride height 60/60
Spring rate 15.09/16.14
Dampers 4/4/4/4
Roll bars 3/3
Toe angle -0.15/0.15
Camber 0.1/0.1
Gears 4.314, 2.985, 2.169, 1.664, 1.280, 1.000
Final gear 4.307
Power limiter 81.1%
High RPM Turbo
Downforce: 300/550The problems mentioned in the previous post still exist: Slip < 0, p = 2.
Attachments
- 1,231
- VS & 74 (France)
- GT-Alex74
Hey guys, I don't really understand "Corr Speed" compared to "Vehicle Speed". From what I saw, it looks like the latest is the speed shown on the gauge / in the game while running, whereas the first is some sort of corrected speed which would try to compensate for wheelspin, but it looks weird with GT6 data. Basically, a run with a tuned Scirocco R gives me Cspeed = fastest wheel speed (front), and Vspeed = non-powered wheel speed (rear). It's on SSRX, so I haven't looked at that in corners.
I'm thinking the game already corrects the speed and shows your real travelling speed already, so the Corr Speed here would be off. So it seems to me using "Vehicle Speed" would be more appropriate to check out 0-100 or stuff like that in GT6. Any confirmation / infirmation ?
I'm thinking the game already corrects the speed and shows your real travelling speed already, so the Corr Speed here would be off. So it seems to me using "Vehicle Speed" would be more appropriate to check out 0-100 or stuff like that in GT6. Any confirmation / infirmation ?
