Messages

I don't mind getting the ball rolling.

1) I like it a lot, i think it adds a lot to the immersion factor  I have only tried it with the GPX and i know that rumble is not well implemented on all controllers. Serious riders will want more information from the rumble, casual players will enjoy the fact that rumble has finally been implemented.

2) I have done quite a few hours and i think the problem is that useful info which i am sure is there is masked by the 'Global' nature of the vertical vibration. Even in this first release it is more 'believable' than my experiences with friends console driving games rumble effects.

3) Very low/low filter, high vel sensitivity, low gear period, medium gear amplitude. i have not used 'crash' but i will give it a go.

4) It would be interesting to add options for different parameters to be fed to the rumble. Sliding is probably the most useful as it is not represented well either audibly or visually .

Some of the Y axis rumble is definitely due to how you are applying the power to the rear wheel.

I noticed that for the same piece of track nice smooth power delivery and sensible gear choice give low rumble. More aggressive riding causes almost continual rumble, its as if the rear wheel is chattering on the track surface and this translates to constant velocity changes of the bikes axis. I have mainly been testing with the 990 which has quite harsh power delivery.

I think this might be masking some of the other rumble detail.

I do enjoy the rumble though, now i am used to it switching it off makes the game feel a bit 'broken' 

I think it would be interesting to examine the other axis to see if different details emerge for different bike situations. I am considering coding up my spare USB PIC microcontroller board to drive a linear shaker with 'live' amplitude and frequency control.

The buttkicker gamer2 seems a good package but is in short supply and prices are high at the moment - due to improve in the Autumn - not sure sound driven would work well with GPB as the sound modelling is not very detailed, but driven by a signal(s)  from the telemetry it could be very convincing. Some cheap experiments with an old speaker driver are possibly a way forwardfor now. 

Thanks for linking that; Its mainly tracks i have not downloaded   I will try testing on a few and see which ones are fun to ride with the bike choices available.

Thanks Loopateli, it seems that from a hardware point of view pretty much anything would work. I have some old PC stuff from gaming rigs mainly old Q6600 stuff which is power hungry and noisy compared to modern processors.

Compared to 10 year old stuff it seems a low end AMD A series might be a good contender or even possibly an E series or old laptop.

It would be interesting to see a log of network load when hosting a session. 300/30 is very high, 150/15 is probably top end in the UK and most people i know struggle to get the high end packages to deliver that bandwidth reliably.

The fact that the processing of the data is not loading up an older PC might suggest that not a lot of data is actually moving around and the bulk physics processing is on the clients. A reliable ping might be more important.

One thought, do you need a second license key for the server?

Currently i am on fibre 30M down 3M up, pretty steady connection and decent server ping. The default 1/10th up is a pain though as upping the download significantly has very limited impact on the upload speed.

I was just thinking that i do not see many servers listed, only 3 or 4 generally, 2 of these are Victoria (limited bike selection) and the other 2 are usually race practice sessions - am i missing something?. Be nice to have some 'fun' servers for fooling around on with the MOD packs and a few other bikes circulating.

I am just wondering, if a low power i.e. low noise/power and small 'mini pc' can be used as a dedicated server then i might set one up. Just need some idea of what the basic processing requirements would be. A simple box with no monitor/mouse/keyboard, a single small SSD or uSSD and a basic processor would be cheap, economical and unobtrusive to run.

What would be the best firewall solution?

Something like a 'track day' server and run the bikeMOD pack and some of the more unusual tracks, probably the UK ones for a bit of mindless fun.

If you do not need powerful graphics and the processor load is fairly low then plenty of very low cost options exist. I have decent broadband with low ping so i do not think that would be a problem. Upload bandwidth might be an issue as its typically 1/10th of the download bandwidth with my ISP without a special contract.

Just looking and an AMD FM2 system 'box' could be set up very cheaply  but are you better with high clocks or more cores etc

When head tracking is enabled auto 'tilt' (0-20 degrees) is forced disabled.

The head tracker now supplies the tilt value directly from the head tracker rotation. This is incorrect in simulation, normally the head movement would correct the tilt created naturally by the bike lean. In the current configuration the tilt effect is counter intuitive.

Can tilt be forced ON when using head tracking? Either in the current configuration of a computed maximum tilt and/or as a fully simulated system with the head tracker countering the bike lean angle to compute the correct tilt angle.

Thank you.

Think i will move this issue to bug reports as it seems more suitable.

Just curious, but what are the hardware requirements for a dedicated server? What sort of processor/network/disc/memory loads do you see whilst hosting on a given system.

Seen a lot of discussion of configuration, but not much on hardware.

I guess you do not need any fancy graphics hardware but how much computation load is on the processor or is this off loaded to the clients and just 'managed/routed' by the server or is it having to process the physics data of all clients? This would seem to be the big one for Host performance requirement.

I guess you could run a headless system managed from another PC to reduce costs and clutter.

I guess a 'proper' add on network card is a must have.

Does GPB support multi-core processors? Do they help at all?

Just wondering whether we are talking i7 or NUC type levels of performance

Dont know but would this be any help for grabbing engine sound? Fairly free of backgound noises and got some of that v4 character with throttle open/closed

https://www.youtube.com/v/dhxe2g0FRCQ

It might just be a feature of the Donnington track, on a smoother track the magnitude seems more consistent.

It would be interesting to be able to select any of the 3 axis or a 'combined' output to compare the differences.

Is it a genuine MS pad? lots of copies about and the quality is 'variable'. Cant go wrong with the GPX designed to give better control for racing games.

If you want to persevere you need to try something like this to limit the travel but you might find the pad is still marginal for GPB which is very sensitive to poor joystick outputs;

Brilliant, cant wait to give it a try - have you had it in game yet?

That's what happens when you read the manual first  . Fortunately i have stayed with the range of integer values reflected by the menu options that i have now found 

Loving this plug in 

Not had a game crash scenario yet, maybe MaX can trap can track a 'watchdog' to tell when the plugin in no longer receiving fresh data and exit gracefully shutting the rumble motors down on the way out.

But i have tried a smoother track now (Mallory) and with my current rumble ini settings i am surprised how much 'feedback' appears on the rumble. It is an insight into just how much processing is going on with the bike modelling, every little fluctuation in surface and wheel traction seems to feed through. Does not seem to have as much variation in magnitude with the bike leaned over either.

Had to try it  'threshold=0' brilliant ! Constant vibes from low level to 'the shakes'

 

Awesome !

I have read the manual and tweaked the gearshift vibration and duration down a bit, increased the velocity and reduced the thresh hold and it feels great (very convincing) on the bumpy donnington track. I think we all obsess with force feedback, but a lot of the riding experience is the bumps and vibrations that normally feed back through the bars and foot pegs, the feeding of the acceleration data continually to the rumble motors seems to capture this quite well.

I have not tried it on a smooth track yet, having too much fun with the bumps. If we ever get a really good road race track this will add a lot to the experience.

If anyone wants the rumble it pays to experiment with the .ini file values to get the best experience with your particular controller, just altering the values by very small increments made a big difference to the feel.

Rumble motors in the GPX controller are great and you can sense a lot of detail in magnitude and frequency and they do not just clatter randomly like my friends console playstation rumble controllers. Probably the best implementation of rumble i have experienced.    Even my rumble lighting works (not very useful but fun) 

Observations so far;
Strange rumble when bike resets to the edge of the track, maybe 0 mph = no velocity rumble would fix this

Magnitude seems higher when the bike is more upright, bumps in corners with high lean do not seem to have comparable rumble magnitude. Is this a function of world/bike axis interpolation ? Maybe all 3 axis need to be resolved before being sent to the rumble filter.

Might be nice to mix in a bit of steer torque value when the bars are flapping about and the values become more 'dynamic'.

Utterly excellent and really adds a lot to the riding experience.

Thanks MaX.