OSW Mmos to PS5

[gamechanger]

If you are doing a new build, would it be possible to remove the "g29_correction: axis address..." debug output or make the log level configurable maybe? It makes it difficult to see the steering rotation changes with this debug output. Thanks again for your efforts here! I'd still like to buy you a coffee.

As luck would have it, that is the reason for making a new build! I want to take a snapshot of the FFB levels for a reference car+track lap with the G29 so I can try and make sense of what I'd be looking for from the T300RS.

actually I have a question, if the game ( like GT7 ) detected you are using different type of wheel, is the output effect will be different ?
so what the FFB data received from G29 / T300 / Fanatec DD pro will be different too ?

or actually they are receiving the same FFB data but just interpret differently on different brand of Wheel?

[GeekyDeaks]

actually I have a question, if the game ( like GT7 ) detected you are using different type of wheel, is the output effect will be different ?
so what the FFB data received from G29 / T300 / Fanatec DD pro will be different too ?

or actually they are receiving the same FFB data but just interpret differently on different brand of Wheel?

From what I have seen so far, it seems that there is no standard for encoding the FFB effects and they can differ even amongst wheels from the same manufacturer. After reading between the lines in this thread (https://gimx.fr/forum/viewtopic.php?p=7293#p7293), I suspect the reason work on the T300RS stopped and the focus was moved to the G2X was simply because logitech were gracious enough to publish the details and it took out the guesswork. https://opensource.logitech.com/wiki/Te ... formation/

I'm not holding out much hope that the T300RS FFB is going to fix things, but I'll keep plugging away whilst it holds my interest as it's fun learning about this stuff

EDIT: I should probably also point out that I'm learning loads in the process and one of the reasons for playing with the build again is to take a closer look at the FFB effects received for the G29 and refer back to the documentation to see if there is a better approach to this than my original 'hammer it down' idea

[gamechanger]

actually I have a question, if the game ( like GT7 ) detected you are using different type of wheel, is the output effect will be different ?
so what the FFB data received from G29 / T300 / Fanatec DD pro will be different too ?

or actually they are receiving the same FFB data but just interpret differently on different brand of Wheel?

From what I have seen so far, it seems that there is no standard for encoding the FFB effects and they can differ even amongst wheels from the same manufacturer. After reading between the lines in this thread (https://gimx.fr/forum/viewtopic.php?p=7293#p7293), I suspect the reason work on the T300RS stopped and the focus was moved to the G2X was simply because logitech were gracious enough to publish the details and it took out the guesswork. https://opensource.logitech.com/wiki/Te ... formation/

I'm not holding out much hope that the T300RS FFB is going to fix things, but I'll keep plugging away whilst it holds my interest as it's fun learning about this stuff

EDIT: I should probably also point out that I'm learning loads in the process and one of the reasons for playing with the build again is to take a closer look at the FFB effects received for the G29 and refer back to the documentation to see if there is a better approach to this than my original 'hammer it down' idea

As I don’t have a real G29 or T300rs on my hand, I cannot compare the FFB with my Simucube after converted by GIMX

Just curious how the FFB Quality of GT7 can be provided

For my SC2 pro with GIMX now, can feel the curb with like vibration , but for sausage curb ( like the Monza T1 / T2 ) , you don’t really feel much different and just the same vibration ( for ACC / iracing you can feel the car is jumping )

[GeekyDeaks]

I copied the file over into my GIMX install directory. When I launch I seem to be getting libgimxudp.dll missing error? Any ideas?

Pushed a new build here https://github.com/GeekyDeaks/GIMX/rele ... 8.0-g29c.2 - this should contain everything so you don't have to copy .exe files around to switch between the hack and vanilla versions. It also keeps the logging behind the haptic debug flag

[gamechanger]

I copied the file over into my GIMX install directory. When I launch I seem to be getting libgimxudp.dll missing error? Any ideas?

Pushed a new build here https://github.com/GeekyDeaks/GIMX/rele ... 8.0-g29c.2 - this should contain everything so you don't have to copy .exe files around to switch between the hack and vanilla versions. It also keeps the logging behind the haptic debug flag

fantastic!!

[Killamunch]

I copied the file over into my GIMX install directory. When I launch I seem to be getting libgimxudp.dll missing error? Any ideas?

Pushed a new build here https://github.com/GeekyDeaks/GIMX/rele ... 8.0-g29c.2 - this should contain everything so you don't have to copy .exe files around to switch between the hack and vanilla versions. It also keeps the logging behind the haptic debug flag

Both yourself, and gamechanger have been absolute legends. I really hope you guys can chip away at the task at hand so we can get a better experience. I think at this point the T300 route sounds challenging but possible.

[GeekyDeaks]

After a bit of logging and looking at what was going on with the FFB constant force in the vanilla build, I noticed that there were huge sudden changes in direction just prior to the wheel going over the 0 point. I suspected this was the 'kick' feeling you experience in the wheel, so I decided to only apply the gain when the direction of the constant force changed sign (effectively we mute one sample of the FFB instead of all of them in the range). So far I'm a lot happier with the results as the little 'kick' around the mid point feels more like in the vanilla build when it occurs, but it doesn't result in positive feedback oscillations. I managed to increase the `zero_gain` and widened the `axis_range` to help compensate for the oscillations occurring far off-centre and didn't notice any real change in the FFB strength overall. I thought it might affect the kerb rumble, but they seem very inconsistent in GT7 anyway, so I found it tricky to really tell any difference with or without the mod.

Anyway, if someone wants to give it a try, I posted the build on github again: https://github.com/GeekyDeaks/GIMX/rele ... 8.0-g29c.3

You might want to change the settings a bit to widen the range and increase the gain e.g.

Code: Select all

<g29_correction zero_gain="50" axis_range="2000"/>

[gamechanger]

After a bit of logging and looking at what was going on with the FFB constant force in the vanilla build, I noticed that there were huge sudden changes in direction just prior to the wheel going over the 0 point. I suspected this was the 'kick' feeling you experience in the wheel, so I decided to only apply the gain when the direction of the constant force changed sign (effectively we mute one sample of the FFB instead of all of them in the range). So far I'm a lot happier with the results as the little 'kick' around the mid point feels more like in the vanilla build when it occurs, but it doesn't result in positive feedback oscillations. I managed to increase the `zero_gain` and widened the `axis_range` to help compensate for the oscillations occurring far off-centre and didn't notice any real change in the FFB strength overall. I thought it might affect the kerb rumble, but they seem very inconsistent in GT7 anyway, so I found it tricky to really tell any difference with or without the mod.

Anyway, if someone wants to give it a try, I posted the build on github again: https://github.com/GeekyDeaks/GIMX/rele ... 8.0-g29c.3

You might want to change the settings a bit to widen the range and increase the gain e.g.

Code: Select all

<g29_correction zero_gain="50" axis_range="2000"/>

Thanks, Let me have a try !

[dragn]

will try aswell. im still using the g27emu since i dont have a ds4. a g29 firmware without the requirement for it would be nice since i use a drivehub for authentication anyway

[GeekyDeaks]

I have one other idea to try out also. Here is a chart of the axis values and the constant force being sent to the G29 when moving the wheel back and forth about 35 degrees either side of the centre (the blue line is the wheel position, red is the force):



You can see the raw values here:

https://docs.google.com/spreadsheets/d/ ... sp=sharing

The thing I was looking at last time was the wild swing between the direction of the force, but I noticed this morning that it looks like it's occurring when you turn past the point at which the wheels are pointing and it's probably not something associated with the 0 point on the axis. You can see this effect in game by stopping the car and then turning the wheel far off to one side, then letting the FFB bring the wheel back until you are no longer trying to overcome the elasticity of the tire - if you let go at this point it starts to oscillate.

Looking at the numbers closer, I realised that the wild swing seems to be because GT7 is not kicking out FFB values less than about 1800, so it goes from a value >+1800 to <-1800 in one FFB report. I'm now thinking another approach might be to logarithmical map the values below a configurable point, say 5000, to smooth out the curve when it starts to change direction.