Discussion in 'Automobilista 2 - General Discussion' started by Karsten Hvidberg, May 30, 2020.
Ginetta P1 at Bathurst mainly. I have also tried GT3 and Stockcar. Same result for me.
Thank you Karsten, looking forward to trying this out.
I spent a few hours playing around with .31 yesterday. Only in the Porsche Cup and only at Nurburgring. I'm starting to really like it now though. I just bumped up really obvious stuff like kerb scale, general road noise, wheel tightening under braking and also increasing front_grip_slip_feel_coeff helped a lot. Clearly I like things on the more exagerated side.
I need to spend more time dialing it in and I haven't A\B tested it against default. But overall it's great, I have a good feeling of where grip is, good feeling under braking, a nice amount of feedback fx wise without it being over the top to the point where effects are grabbing at the wheel and still get that smooth transition around the centre that just isn't possible on default without a ton of damping.
Anyway looking forward to comparing .46 when I have .31 where I'd like it to be.
Will be very interesting!
I'm having a lot of trouble with .46 and not sure why. I'm always testing with Formula Reiza on Interlagos with SC2 (same SC settings as Karsten recommended) as my baseline so perhaps
this is downforce related differences from what you all are reporting.
I'm consistently getting clipping alert from my SC2 wheel mid high speed corners and it feels more like turning against a soft bumpstop than typical cornering forces. I tried increasing
and decreasing center_full and LFB slider value to get more center but its so far off not sure what to try.
The one positive thing I can say is that the new acceleration effect has made my starts off the line faster and enjoying this effect.
The formula reiza is rather extreme & I did not try it.
But I think you can lower the velocity_scale considerably and that should make it work.
Something to note, I've experienced drastic differences in FFB gain (per-car) in certain cases, perhaps related to switching between default / Custom ffb profiles(?) I had some cars set to 120% and others at 50% gain.
Exiting AMS2 and staring a new session may solve the issue, it did in my case.
Yeah that would explain it. reducing velocity_scale helped a lot but its still very far from the mark. I think a number of other parameters need tweaking to work well with Formula Rieza. Give it a try and see what you think.
Currently my favorite is version .43
I'm testing the .46 file on G920 ... I hope tonight to have a couple of hours free to be able to deepen it and be ready to adapt to the future .47 of Karsten, but summarily it seems to me already a good goal that obviously needs to be a refinement and no doubt will be focused on in future updates by Karsten & Company. For now I have tested the .46 LOW version with a configuration of values equal to Gain 65, LFB 50, FX 50-60 and Damping 0 (or 50) ... compared to the 45 (which is the first version with the yaw scale, the velocity_scale and with the progressivity depending on the downforce of the car) it must be said that it is certainly improved, it is still to be defined, but the direction and the sensations go in the direction of the versions .43 and .44 (my favorite on Logitech G920) ...
It is also true that the .31 file is a reference point among all the others (indeed according to many the best). Having done this, I think we should take the granitity of .31, the elasticity and innovations of .43-.44, and then apply the speed, downforce and scalability corrections of this version .46 and of future versions, all without creating anomalies, empty and/or inhomogeneous areas ...
I think it is a long and difficult job, especially if the feelings of each of us are subjective, especially if we have many different steering wheels and above all because no one is right because this is a simulation.
How to help Karsten & Company ??
- Not quickly comparing the various versions with the same values found and used with our favorite, because each file has its own poem to interpret and therefore each file could be ugly because it is badly set ... the gain of .31 is different from the gain of .43 and .44 and is again different on .46 and will be different on * 47, same situation happens with LFB (between the various versions), with FX and Damping ... if we then add, that each of us, can vary the values, the scales and forces of 3 masters (the rack, the road and the effect) the problem is amplified;
- Use the new files to test for a while before defining it (and/or returning to the favorite), as the brain and eye/hand sync is adjusted to your favorite file that you use every day ... a quick test (without making getting used to the brain and manual tacticity) falsifies our report card with incorrect feedback that can lead even those who compose the files astray ...
In short, we could risk going crazy .... but I think you all love Karsten & Company , so we communicate results and feelings only after extensive and long test sessions.
See you tonight, hoping to be able to give useful information on this .46 file .... good job to all and thanks ...
When I was giving my inputs and recommendations I was testing high DF cars as well to make sure there is nothing weird going on and also the ratio of forces applied to GT cars for example, and in the latest file Karsten made this pretty good so imo in most cars, you shouldnt need to adjust the FFB multiplier.
Anyway, can you post your settings please? Ingame + SC2 force.
I tested F Reiza just now on DD1 and it seems great. I get about 11Nm in very high speed downforce corners (300 kmh), I dont run my DD1 super tough as I like to keep the steering forces relatively realistic + a bit more to simulate G forces. Now after driving couple of race cars and a formula (F3 equivalent) on a closed track, I can say the G forces arent that big to warrant a 20Nm cornering force, not to mention the steering wheel is relatively light all the time. Of course the faster formula and prototype cars with huge amount of downforce will be a bit different. But its impossible to replicate the exact feeling in game. The most noticeable sensation in any car on the track is probably the braking and thats one of the reasons for the dec2_scale and exaggarate brake value, the increased centering torque during heavy braking is to imitate that feeling somewhat, while acel2 is to provide a bit of ligthness during heavy acceleration when the front tyres get unloaded.
Try also decreasing by 0.1 both master_road & master_effects. With the formula reiza i got as well some occasional clipping beeps in mid-high corners but i was either close to touching the curbs or i was on a bumpy road. In game settings if i remember correctly i had 32-34 gain, lfb 5, fx 55 and damping 10 or 15.
To be honest i was shocked at how this update made my Ts300 come to life.
Especially balance changes are really felt now, where i didnt feel anything before.
Maybe its a bit overreacting at times, but for this wheel this is the way to go i think.
I had my gain to 90 before, which suddenly felt too strong and i now tuned to 75
in a first step. What i find most astonishing now is the shift of loads of the car
(did only drive a few rounds in gt3s and porsche cup). i can somehow feel now
how the car is leaning to the sides.
for me this update is the best till now (not that i know how .31 felt )
Great job Karsten and team!
At first I thought v.46 was too weak/vague at the centre, and some of the effects, especially scrub, were a bit high. And I felt I was missing some of the acceleration/deceleration effects.
I did a few quick changes, and for me this improved the feeling quite a bit:
I'd say, for me, this is as good as v.44, with the added features from v.46 on top.
(HighEnd file, DD1, McLaren GT3 @ Nordschleife)
Edit: After some more practice I improved my personal best @ Nordschleife with both the McLaren GT3 and the Porsche GT1. That's usually a sign of improved FFB
.46 (low-end for G29) is my favourite so far.
Feels particularly good with some of the RWD cars - German Group A, AR Camaro. Feels a bit weaker with a couple of the FWD cars I tried but I'll trying upping the main gain for them. And the braking feel I liked from default is at least as good in this version.
I'll be intrigued to see if this version can be improved upon and even if not, I'll be happy to stick with this one.
At a preliminary stage of testing. I thought it could not get any better from ver 45. I was wrong. 46 high end, tx wheel. the added accel and decel effects are good, it feels faster and more depth in detail with responsiveness.
I would not say this ver 46 has any inherent weakness. Its only up from here.
Though its a big benefit all the new effects, I am going to do an A/B thing (not a full blind obviously style test) with the settings from torsteinvh.
While I do not feel as though its missing too much, its possible a tweak to scrub for example, is in the works.
I use 100 gain for main, 50, 50, 50 and dial it down between 0 and 30 at most on a per car basis.
At default I get the most minor of red bar in the telemetry, but its not present during a run.
At imola I can see clearly the potential for a faster time, its way more granular than the files in their 30's. Really appreciating the added depth, complexity and texture these files give to the experience, how they scale and all that stuff.
And naturally having said something of the less than stellar take offs in the past in AMS2, this file goes a long way to improving that, while retaining the smoothness of something like raceroom.
Pertaining to that, its nice now that the ffb forces on either side of center seem so much more real these past couple of months - this level of fidelity definitely broke new ground in FFB - so well done.
To test some of the things I like to do is take the mp4 round jerez like a maniac, bathhurst in older f1 types and newer, the v8s as well, and gt3s round brands, era appropriate stuff at imola... I go through a system each time, and maybe I am losing out with a 50,50,50 I don't know but as a reference point this 46 version across a wide variety of cars is more on point than ever. The gt1s I like to go on spa, the german ones at hockenheim or nord (but porshe on the schlief and not in testing so much).
It feels a lot more instinctive in the 40s than the 30s. I like the ffb to step out the way but be present as needed and I think a good balance is being found.
Usually like to stick to things I can compare a little across titles. Some of the most enlightening stuff has been fiddling with gen1 f1s and tuning them just enough to be wild enough, then driving them, and in cases making them more compliant. The ffb is handling the task of all these things since the 43 update to my mind. Then the ginetta p1, and of course the sigma p1 car at santa cruz. So the ffb is even helping dial in setups so the fidelity is a reality now which is just so good.
edit: its hard to quantify but in the effort of fidelity I think at jerez in mp4 for example its not as 'hardcore' fast as it was previously. Maybe I need to up the gain to 110 I dont know. Its definitely more fidelity but less of a sledghammer effect in 46. Just an impression I will now race this one. Noted the same kind of thing at brands. I keep all the files its not a problem.
So I will now check the quoted settings. after I test that all in a jerez short race to see if the intensity is there.
edit2 so one can see my mistakes/working out haha well i did try 100, but 96 is what I previously had it at (tx wheel, normal 100, jerez in mp4), and this allows for more intense or better driving. Also take offs definitely way better. Things happen a bit faster in this ffb and I am getting used to it. For example spins and saving spins has never felt this good. Or throwing out the back and then arresting a slide.
edit 3: this ffb is definitely way more race ready. You can fit a lot more into your inputs. cars with good downforce and grip, for example, bathurst, you can pretend you're a fully fledged professional driver with the way you dart down the mountain.
Is it any wonder the quoted poster found his time improved? The ffb gives a lot more clarity than even the [now admittedly 'obtuse'] force of 45 or 44. Going down the mountain however I did note he is probably right about his accel and decel thing, and also the progressive loading on a gen2 did not seem to be as much there. But I will go round again and use more brakes, I was darting and weaving. --- so its just that the car, the gen 2 v10 has so much grip, but a gt3 probably will show it better, though the v12 gen2 did actually load up on slip loss (or load down?).
What I mean if no one reached the conlcusion that popped into my mind (trying an unbiased approaced for myself which I can be) it could be more pronounced that, further giving a sense of speed above the lateral and visual aspects. if thats an accurate description/and prescription
because in a gt3 and even others the front feels vague when I may expect more weight going downhill. the telemetry geforce meter is forward/back so the force is there. will now try with extra brake and a brands hatch setup on bathurst. hehe.
so with the front end loaded up more, and honestly crashing has never felt so good - the effect exists but is not as pronounced I would think. I will try the things I do with the extra settings. it may not solve that if it needs solving but it may do other things instead
I will probably look into the file myself
Was that a typo or did you actually change (dec_boost_feel) from .07 up to .2? That's a large jump. Haven't got to try it yet, so thought I would ask.
hehe i did it and maybe thats the setting, decent improvement of sorts, maybe more longitude weigth or some such is required for my taste, but it was over powering in a sense to what that particular setting changes.
thus I am using 45 again to get a good feel of the differences. its still my prefered one overall, notwisstanding most likely that 46 is superior in many tangible and beneficial ways.
this is because at first I thought 46 was doing 'too much' if that makes sense. a re-frame of my thought but probably the more correct assessment in terms of holistically what it does for the sim.
it must be stated that my driving skill today is not at its height. Its possible that with much better driving the thing is way better, but I dont think its going to work in regards to the weights and front vagueness.
I dont mean its horrible - only that its not as clear in terms of input delineation (what you need to do for 1+1=2, I think the scaling gets in the way a little) as 45..
edit: This is confirmed for me with 45 again on. I know 46 is technically superior, but in 45 it just does things so sublime and direct. I think a 45.5 would not whittle that away too much, that aspect.
You know in race room lol sometimes it feels more on-rails than the other games, but I thought of 45 that this aspect was more organic but with the same effect. I feel 45 helps you out whereas 46 prevents you - just straight off the bat kind of thing. Yes I could learn that, but no it seems counter intuitive if that makes sense.
While rr is prob more fake in some ways than a madness engine game of AMS2's nature, that driving experience is something to aspire to. I believe that fully.
but 45 is god-level in my view. so its prob unfair on 46 maybe
to give an idea - on the mcL f1 gt1 it felt great, but on open wheelers and gt3s it didn't compare to the direct nature at the core, so to speak, of 45
well thats the aspects that would prevent me from downgrading to 45 again. the fidelity was just way more in 46. I still have them and will not rule out an interim adjustment to check things out, just one or two values.
however, its best I redo partly some car setups. this is not an easy version to back away from, this 46. its very good.
Of course this had me think: what could be holding back this game in some ways, all games, but this one in tandem with the engine - is per-car FFB settings/files.
Honestly just think of that for a moment. Is this what the game needs for next-level type things. Or an internal offset to accommodate such things, as set by the devs, or others.
My change of heart comes from user error/user not wishing to adjust so much. I set, for example the main one I was concerned with, the m12/4 thingo... clutch plates 6 at jerez and 80 power, slight higher coast...than previous values because default will work. helps one push it even harder. for the more fidelity this version has, and the way jerez is laid out, 6 clutch plates for my wild driving there (I will try to spin out on pit exit/startup.) blah blah this gives a good sense and feel.
If you drive right you can crash hard and its interesting to see how it behaves. 46 is a good version and its allowing better overall driving, to be fair plus I was then able to put the gain per car here on the mp4/12 back to 100. I am not at this point changing from 50 50 50, though I will just see what thats like.
^ Pffft- its pretty amazing, again, think it was version 43 where I turned it to that. 75 fx seems pretty sweet. damping down to 20, and lfb at 50 since its got too much to it now for me to want to mess with as its in a decent spot; but will test 60 too later. cooool.....
edit but the big vibrations and all the bumps is why I kept things at 50 50 50 haha - do that for a while and its not feasible. I am changing one setting at a time.
Bravo: not. set it to 100 gain, 60, 75, 50.
The slightest of adjustments up, more pronounced if you lower dampening. still get it all. then set the clutch plates back to 4 - was a hasty setup change.
This means I set screen bounce legacy and at 40, then, so its bumpy on screen and bumpy in the hand. The reason why I seem to turn it down (then forget) is that the core of the ffb file is so well done, it just blends in. Thus I am not limiting all of it, just getting immersed. missed those engine rev vibrations being prominent though; before they were present but not full.
And what that means for anyone following and experiencing it themselves is the core and the road feel combine very well. Regardless of that, should be enjoyable either way. Probably will not adjust individual settings in-file now.
So I know have 3 (three) full posts going over bits n bobs about this.
But I have just made the rather difficult - actually - choice of going back to version 45. For those of us whos driving skill oscillates between really, really, really good, and just really good lol... well version 45 is simply better as a game and in gameplay.
Its that simple. For technical marvel 46 is the goods but in actual gameplay I find the more sublime (50,50,50 and core file - 'on rails') feel of the 45 version to be sufficient and game play-helping. The version 45 high end, at least on a tx, is simply too good to pass up. Its not about driving skill - the file just does all it needs to do and gets out the way - doubly so now the road noise is lowered on the tracks.
I said that in the spirit of feed back its not a preference as I do think that in order to push through all the bumps and whatnot, the vibrations (which I never liked on v twins in the main), it is required to punch through that in order to deliver the input - which is why I prefer 45...
but also it probably interferes in 46 with the new pitch angles etc of the latest patch, I am not sure, but it does not seem like its in line with how smooth 45 is. some people did not like 45 because it was not enough. I rather think the sublime aspect of it outweighed the other aspects.
That was intentional
I like to make quite big changes on first edit to notice the difference clearly. Then I can always dial it back if it is too much.
Not concluded yet, though. Need to drive a bit more first.
back on 46 for a little bit - no other game does bumps as much as AMS2 it seems to be a feature. This ffb can be good when you dial most of that out, but its still has not got the directedness and 1:1 type thing from input to game like in 45. I still feel as though 46 gets in the way of itself a little too much.
i do have a slightly sore shoulder atm, and also if I lower the per car ffb right down to 60, its possible if I try it again it could be the goods. however the steering is simply not there from what I know of it in terms of trackability. its like the steering ratio is changing or something inside the file and probably needs further working.
edit got to admit its a lot better, with high file, 63 per car gain on 100 gain setting, 60, 75, 50. could lower damping a bit. this is stock car 2019 at santacruz.
To explain I think its out, yes, it moves like a free flowing water jet at the first chicane, but then it turns like a lead bolloon in any of those corners....kinda... so I think its 20% out of wack or something, hard to say.
The difference between chunky and finesse... but then ams2 is a chunky game. except version 45 removes a lot of the superfluous stuff and focusses better I think. its more direct steering and movement is maybe not as 'alive' as this version but as a game plays better
Separate names with a comma.