Reaction to Ermin's Video

Sometimes I do wonder if people actually read the update release notes or the developer notes. I mean, it isn’t like Reiza developers don’t know that certain things are not working properly. With pretty much every update there are changes to physics, differential or ffb. I own AMS 2 since early access and I remember very well how it was a year ago to drive the 2019/20 Stock Car. Since then the Stock Car has come a long way which shows me that there is some serious work going on.
I actually managed to watch the GM rebuttal video for 13 minutes before I had to turn it off because he just is to nauseating for me. He might be a good guy but his content just doesn't add anything to my daily life.
Said all that, we all have different opinions for different reasons and that is ok.

 

99% YTbers don't care to read about development updates (they would know what was fixed and what still under development) but they are quite fast asking for keys to Devs when a new SIM is released...
Problem about those kind of YT videos (referring to JB, GM, etc) are more on the entertainment side while others (like Erwin) are more acting they really know something while clearly not.
I prefer Niels or Steelcast videos overall since those try to explain how physics works and how you can setup your car...
Just look at each YTber comments section and you will really understand which is their audience target...

 

In RaceRoom, "100% closed diff" doesn't mean 0% wheel rps differentiation, it means "100% of the available potential locking is applied" (confirmed by their physics guy). So in practice many LSD cars in that sim always have slightly open diffs for realistic preload values. Here is an example of the Audi R8 EVO going through Brussels in R3E:



Probably explains some of the differences in lift-off response between sims.

 

I have to say that I absolutely get where GM is coming from. I think he makes valid arguments in a well mannered fashion and with a positive attitude. Unlike Ermin.

Regardless of what is realistic behaviour (I'm in no position to judge that, I haven't ever entered a corner irl anywhere near the speed I do in simracing) I do feel a bit the same as GM.

• Lift off oversteer / lift off less understeer. I don't really care about the dfinition used but where I can manipulate throttle input easily in other sims to make small entry adjustments I find it much more difficult in AMS2 to get a similar controlled response. Dunno if it's more or less realistic but I do think that for the majority of simracers it feels good to have that control so having a similar repsonse will probably add to more satisfied users.
• Setups. I understand that with the gazillion track/car combo's it's a heck of a job to catr for specific well balanced setups for everything. At the same time, the majority of the simracers (I think) do not want to spent a gazillion hours in setups either. Sure, I tweak setups every now and then. But mainly I either use a stock setup that is actually very decent (ACC) or use someone else's setup that gives a really nice baseline. The amount of cars/tracks in AMS is awesome but it is also a bit overwhelming knowing that for each combo I have to actually spent quite a bit of time in setups to get the "sort of optimal" experience. Personally I feel that it is worth the time/money investment from a development pov to get this done. You have the very fanatic simracers but what I see around me (including myself) is that the vast majority of simracers are lazy and don't want to spent much time in setups. A.k.a. it's understandable but it's still a major drawback that stock setups are sub-optimal.
• Braking. I love that there's a huge difference between cold and warm tyres, as there should be. I don't have massive complaints about the braking tbh but yes it is a bit more fiddly to get right compared to other sims.

Overall I love AMS2. I get the impression GM does aswell. Erwin loved it, then hated it, then loved it then...yeah well he's Ermin. In the end I'm one of those simracers that loves AMS2 but hardly play it. That's mainly due to a) being lazy (read: limited time) not wanting to spent to much time in setups and b) multiplayer being lackluster atm. I have a feeling I'm not the only one like that, I'll just patiently wait

 

Jesus christ so much drama over youtube ... i mean ... fortunately everyone is quite well behaved but still, 1 hour of talking about other people talking, I couldnt do more than 5mins of each video.

Regarding the LOOS.
In order to be able to "feel" the ramp angles work in GT3 for example, you need to reduce the viscous locking and also preload quite a bit, in AMG, thats at least 30 preload and max 50 viscous to get some kind of balanced lock/unlock and then play with the angles and you can get a nice LOOS if you want that, it works but preload and viscous intereferes. I also dont get why we get the option to even adjust viscous locking on modern cars with plate LSD.

Problem comes in for Retro and vintage formula cars, they have LSD but you cant adjust viscous lock or there is some other issue with preload maybe. Basically, in retro McLaren for example, even the most extreme "open" LSD setting doesnt open the diff unless you get below 80kmh or so, which is just weird and according to Maser and his friend, it doesnt behave that way, not to mention the actual setting, even at much higher preload, more plates lower coast ramp angle, the cars didnt have a problem to turn in during the race. So unless you go real slow, you understeer into oblivion in these cars. And no other games have this issue, would you say AMS2 is the only game/sim in the world that is realistic? Especially when people who used to drive these cars say the opposite?

 

@steelreserv

The obvious way to resolve this disagreement: youtuber boxing match

 

Well some says no but some others says yes (specially the ones driving on BR championships).
Just some guessing...
AC, 1 contact point tyre model (car spinning like on/off, kerbs of death, etc)
ACC moving from 1 point contact to 5 for the same issues as above (still not refined as should, etc)
AMS1 based on ISI Rfactor engine, still good but same problems as AC when you feel loosing the car it is too late (but better ffb of the 3)
rF2 best tires model so far even if you can exploit it easily (GT3 drifting all over the place as example)
AMS2 ideally on par with rF2 but differential is a pain in the ass for Reiza but they are fixing it somehow...
So basically every sim is different and behave in a different manner, some can be fixed working on same setups parameters others need a different kind of tweak so...
YES GM you need to adapt based on the SIM the simple fact 3 of 4 works Inna similar way doesn't mean it is correct because as you can see every software is different and simulate things in a different way, some in a similar way but others with a different approach....but this doesn't mean is wrong overall....

 

I meant specifically Retro-formula cars, like I said GT3 feel more on par after setup tuning.

 

No no, just quoting you as incipit of my post

 

You forgot to mention Ermin's bigotry

 

You were too polite and didn't provide a link. The conversation above carried on as though this didn't exist...

Nice succinct presentation, also

 

We shouldn't.

But remember the starting point was PC 2's "provide every diff option for every car" as a kludge to avoid doing the hard work that Reiza has been undertaking for a year plus to sort out all these diff and diff-connected issues. Most people just used the same favourite diff on every car, even if it was unreal.

I believe that as things progress and settle the remaining unrealistic or unavailable in real life options will get snipped, just as many have already been eliminated over time (recall we used to be able to use a spool on open diff cars if we wanted to!). In the meantime, they can be helpful to play with to diagnose some behaviours.

 

'Problems' oddities yes are still there but diminished thankfully but track specific still ie long sweepers as apposed to tight turns. Preload changes +others, over recent updates have improved things as I have mentioned and others elsewhere, but not eliminated fully imo.

REF BT44
**Ethan Shippert the real life engineer and race driver of the Bt44 has to be regarded as a valid source of feedback. Thanks for the reminder.

Agreed and ongoing, I would say 'power and transmission line inc lag' issues as the seemingly % per car specifics are tweeked. I am and remain very optimistic, and enjoying also.
Sidenote.
Thankfully the tubers chose gt3's not Retros!.
Seen enough inc bits of todays live stream stuff. gm.
New Month very soon

 

In RST you do, its more or less the "Wheel Spin Graphs" only difference vs a wheel velocity graph is wheel spin normalizes for the vehicle speed aka long. slip ratio. As far as the graphs he posted, those graphs are very clearly not the driven axle which doesn't have a diff anyways.

A locked diff will always cause both inside and outside tires to spin at the same velocity. The difference in 100% locking is in the definition that the sim uses. You can have 100% locking on the diff and still see a torque difference that causes the differential to unlock, but you wouldn't say that diff is fully locked you would say that it's open, when the diff is locked you will see both the inside and outside tires spinning at the same velocity.

 

Which graphs are you talking about? If the one from R3E I posted, I can ensure you that is the rear wheel rps difference in %. The front wheel rps of course is not controlled by a diff and would exhibit much larger differences.

While I would agree, my point was that some other sims do not model it in this way. Indeed, RaceRoom appears to always allow the diff to be slightly open. This can be seen even in the one car that has a spool in that game. What I wanted was to show that broad statements based on comparisons such as "the diff in AMS2 doesn't behave like every other sim, therefore the physics are wrong" don't make sense because those other sims could be (and often are) wrong too.

Every sim cuts corners somewhere, then fiddles with the parameters no one has information about to make the cars drive nicely. This was the magic wand used by Niels Heusinkveld on AMS1. The difference is, AMS2 doesn't have a lead physics guy making 40-minute videos where he hypes the features of the sim that exist more or less in every other competent sim while glossing over the limitations.

 

While I would agree, my point was that some other sims do not model it in this way. Indeed, RaceRoom appears to always allow the diff to be slightly open. This can be seen even in the one car that has a spool in that game. What I wanted was to show that broad statements based on comparisons such as "the diff in AMS2 doesn't behave like every other sim, therefore the physics are wrong" don't make sense because those other sims could be (and often are) wrong too.

Every sim cuts corners somewhere, then fiddles with the parameters no one has information about to make the cars drive nicely. This was the magic wand used by Niels Heusinkveld on AMS1. The difference is, AMS2 doesn't have a lead physics guy making 40-minute videos where he hypes the features of the sim that exist more or less in every other competent sim while glossing over the limitations.[/QUOTE]

Yep fully agree with that statement, and what I was trying to elude to in my post.

 

Anybody care to react to the new video?