AMS 2 | Release Delayed to March 2020 - Jerez Confirmed

Paul Jeffrey

Premium
Reiza Studios have released their October 'Development Roadmap' for Automobilista 2... and it contains plenty of interesting news, a new laser-scanned circuit and a delayed release window.

With the successor to fan favourite Automobilista well into development prior to hitting public release, these Reiza roadmap posts keep getting juicier and juicier each month - and the October edition is certainly no exception!

Although packed full of interesting news, images and videos, one bit of information that will undoubtedly cause some disappointment amongst the community is confirmation that the release window of December 2019 has slipped to early 2020, with March floated as a provisional new release period for Automobilista 2.

With that news aside, you can check out the full October roadmap below:

Hello again folks!

October has come and gone - another month of intense work for us as we enter the latter stage of Automobilista 2 development before its initial release.

After months grinding away producing, updating and exporting a large number of cars and tracks, figuring the ins and outs of the new game engine and developing the required additional resources, things are now beginning to take form and potential is starting to turn into reality as the focus shifts from assembling the meat and bones of the sim to fine tuning the game into a cohesive whole.

There´s a lot of work still to be done to achieve that, so let´s go over where we are
:)


AMS2 Release Schedule Update

Beginning with what may not be so good news - although things have been progressing nicely and more or less within the planned schedule, with most major issues already cleared and a large content base already in place, we have decided to push release a bit from December 2019 to March 2020.

This is not a decision we have taken lightly - we understand many people are anxious for AMS2 to come out (not least ourselves), and even if we might have been able to just get to a releasable state by December, the content, features and polish we can add in these 3 extra months more than justifies the delay and will ultimately lead to a stronger launch for Automobilista 2.

For Early Backers the wait will be a little shorter as Automobilista 2 Beta should become available in February.

In the next two months we will be fleshing out our plans for Automobilista 2 in more detail, which hopefully will make the extra wait more understandable as well as more bearable
:)


For now, here is what we have been up to this past month:

Progress, Progress Everywhere

AMS2 super v8.jpg


On top of many cool cars and tracks continuing to make their way into the game (like this sweet new and revamped Super V8), we covered a lot of groundwork in various essential fronts of the sim. Some of the highlights:

Physics & FFB - We have made a couple of small but significant adjustments to the physics code, crucially fixing a bug in unsprung inertia calculation (which was already fixed in AMS1 but not yet in this version of the physics engine).

We are also working on a new FFB system which basically works very similarly to Realfeel system used in AMS1 (basically converting the forces coming through the steering arm into the FFB).

The combination of bridging these gaps with the ongoing development of the physics derived from AMS1 and a growing understanding of the SETA Tyre Model is starting to net some very exciting results - Over the next two months I´ll try to expand a bit more on where these improvements are coming from, but for now I can say we´re growing more confident that the overall physics and resulting driving experience in AMS2 is going to deliver the step forward we hoped to achieve.

AI - A lot of attention was given to AI this month. One of the main differences vs AMS2 AI is that while the player physics run the SMS SETA Tyre Model, the AI runs on a Brush Model similar to what we had in AMS1 which was by and large the same as the player´s.

This means both models need to be developed for each tyre type so that the AI has tyres that match the characteristics and performance of the player be that on straights, in and out of corners, running cold, hot, under or over pressure, come rain or sunshine and over a wide range of difficulty and aggression settings - a lot of variables to fine tune.

Luckily we do have quite a bit of experience with a similar AI code and this fine tuning process, and elsewhere in the code SMS has been very thorough in making AI functions fully parametrized, making it an easier (if still time consuming) process to experiment and fine tune the AI to get them to perform, behave and battle as one would like in all conditions.

There´s still some ways to go to make things work smoothly accross the board, but the good news hopefully demonstrated in the previews below is that we have the resources we need to get the AI as good if not better than AMS1 already for initial release, and then build on that with new features to try make AI racing more satisfying than ever before.


AMS2 Dedicated Tool - Another important milestone has been putting together the AMS2 Dedicated Tool so servers can host multiplayer sessions not necessarily through the game itself - this was one of the main requests we received from leagues and luckily here too we were able to rely on SMS providing us with something that was already over halfway there and we are now pushing accross to hopefully a releaseable state in v1.0

Weather & Climates - Similarly we´ve been doing some work on various weather and climate profiles to achieve more accurate rain volume, lighting and temperature ranges for the various locations we have in the sim - significant even within a country of continental proportions like Brazil where climate varies a lot from the southern tracks to the ones further north.

These weather and climate profiles also affect how much rain you may get in different places and the temperatures will affect how quickly or not it dries up.

It also helps recreating the kind of warm summer sunset you´d experience at a seaside italian kart track as can be seen here in intense VR glory:


3D Animations

One new front we have started tackling this month are 3d animations, as even though we have been able re-use some of the assets that came with the engine for things like driver animation, animated crowds and pit crews, there is a lot more that can be done elsewhere. For cars specifically, the suspension assembly is now fully articulated; then there are wipers, vibrating components, damage, switches and a bunch of different things we can add to make the environment both inside the cockpit and out more lively.

To help us out making quicker progress on this front we have partnered up with a South African Studio whose expertise is already proving invaluable getting us up to speed on this front. Below you can see a demo reel of various such animations being added to the Caterham:


Jerez Coming for Automobilista 2

AMS2 Jerez.jpg


We are thrilled to confirm another traditional circuit is coming for Automobilista 2: Jerez de la Frontera is a challenging Spanish track of rich heritage, having hosted several GPs in the 80s and 90s and continuing as popular testing venue well into recent years.

Jerez was stage to several historical moments in F1, from the classic 1986 photo-finish between Senna and Mansell, to Senna´s record-breaking 50th Pole Position in 1990 following Martin Donnely´s massive accident, to the infamous Villeneuve-Schumacher title deciding collision in 1997.

As with other recent tracks, Jerez has been modelled for Automobilista 2 with the aid of accurate LIDAR data, and will be included in the initial release of the sim. Also as per usual when going for circuits of such history, we plan to include the historical layout used through the 80s, although most likely after AMS2 release.

Jerez has been one of the tracks we have managed to fast-track licensing and production thanks to the influx from the AMS2 Early Backing Campaign, so thanks to everyone who has already opted in and helped us move this along!

Automobilista 2 Community Skins

Last month we announced that those who joined the AMS2 Early Backing Campaign would also have the opportunity to become part of Automobilista 2 by creating their own livery for any of the various fictional or semi-fictional series in the sim.

Some great results are already coming in and several drivers have already earned their place in the AMS2 grid, such as this beauttiful F-V10 by @Bruno Mendes

AMS 2 Community Skins.jpg


If you also want to be a part of AMS2 there is still time - check out the AMS2 Early Backing Campaign!

That is all for October! Even though release may not be quite around the corner anymore we are confident there is still plenty to look forward to before the end of the year as we outline our 2020 roadmap for Automobilista 2 - plenty of exciting news on the way still so stay tuned

AMS 2 Cars.jpg



Automobilista 2 is expected to release March 2020 for PC.

Keep an eye out on the AMS 2 sub forum here at RaceDepartment for all the latest news and discussion about this exciting upcoming new game.

Like what we do at RaceDepartment? Spread some love and give us a like/follow/share on Social Media!

Podcast fan? Don't forget to catch up with the latest happenings on your Podcast platform of choice! Just search 'RaceDepartment'.

Sim Racing is awesome!
 
Releasing AMS2 in the run-up to Christmas would have been a bad move anyway IMO. Hopefully this way the Reiza Devs can have a break at Christmas with their families instead of working flat-out on post release hotfixes.

Also, I think it's a mistake to release a game before it's fully ready. Reviews are usually written about the game when it's first released and bad reviews can have a lasting effect even if later patches fix the problems.
 
Physics & FFB - We have made a couple of small but significant adjustments to the physics code, 1./ crucially fixing a bug in unsprung inertia calculation (which was already fixed in AMS1 but not yet in this version of the physics engine).

2./ We are also working on a new FFB system which basically works very similarly to Realfeel system used in AMS1 (basically converting the forces coming through the steering arm into the FFB).


3./ The combination of bridging these gaps with the ongoing development of the physics derived from AMS1 and a growing understanding of the SETA Tyre Model is starting to net some very exciting results - Over the next two months I´ll try to expand a bit more on where these improvements are coming from, but for now I can say we´re growing more confident that the overall physics and resulting driving experience in AMS2 is going to deliver the step forward we hoped to achieve.

Ad 1./
- Can you shed some light on what this bug was and how it affected car behaviour? If you've already done so internally on the AMS2 forums, where might I find more info on this?

Ad 2./
- What was wrong with the PC2 FFB system from your perspective? Are you ditching the "mixer pipeline" concept which is controlled by a lisp-like syntax? Or are you re-purposing the existing code to build a less flexible, more "one size fits all" system which is easier to calibrate for you?

Ad 3./
- Sounds very interesting! Looking forward to hear more about it!
 
Ad 2./
- What was wrong with the PC2 FFB system from your perspective?

Other than 99.9% of people couldn't make head nor tails of how to adjust it and the corresponding almost universal negative reviews about that aspect PC 2? Based on your in-depth knowledge, you might be one of the 0.1%.

The rest of the world prefers a properly defined and calibrated FFB system with no adjustments beyond overall strength required (as in AMS). There is only one optimal "simulation" combination of settings for FFB--why allow people to adjust it away from optimal or go around in circles perpetually trying to get the best settings. Exposing all those settings (in the UI at least) is compensation for a lazy developer who didn't bother to calibrate all the physics of their cars properly (relative to each other) requiring the adjustment of settings to get particular vehicles to feel "right" because they need different settings.

The only reason I am excited about AMS 2 is because I know Reiza will apply their design philosophy to the outrageously frustrating FFB from PC 2. No more stupid incomprehensible settings (incomprehensible because unless every car uses the same "base" physics, you have no way to know how one setting will affect the vehicle and are stuck in trial and error mode, so please don't reply with some smart people understand the fundamentals of the settings--I know that). Every car will feel appropriate relative to every other car. All basic responses (progressive loss of grip, the effect of down-force, etc.) will feel "natural" and plausible based on the particular characteristics of the tires and wings and so forth of the car. It won't be a roulette wheel spin of randomness based on which cars got paid attention to by which people on the team. There will be the usual Reiza impeccable quality control :)
 
Other than 99.9% of people couldn't make head nor tails of how to adjust it and the corresponding almost universal negative reviews about that aspect PC 2? Based on your in-depth knowledge, you might be one of the 0.1%.

I think you are confusing PC1 and PC2. PC1 FFB was indeed confusing but PC2 FFB is much simpler and the menu explains what each "flavour" is and what each slider does. The nice feature they failed to communicate very well is the self adjusting nature of the immersive and informative flavours to minimise clipping. Nor are reviews almost universally negative on the force feedback.

The rest of the world prefers a properly defined and calibrated FFB system with no adjustments beyond overall strength required (as in AMS).

This is rubbish. AC, ACC and Raceroom have multiple adjustments including "enhanced" effects. AC, PC2, ACC and RR are much more popular than AMS1 by an order of magnitude.

There is only one optimal "simulation" combination of settings for FFB--why allow people to adjust it away from optimal or go around in circles perpetually trying to get the best settings. Exposing all those settings (in the UI at least) is compensation for a lazy developer who didn't bother to calibrate all the physics of their cars properly (relative to each other) requiring the adjustment of settings to get particular vehicles to feel "right" because they need different settings.

Again this is rubbish. Steering wheel force feedback is the only force feedback most of us get from our sim, while racing drivers get feedback through the seat of the pants and vibration of the car through the wheel which isn't a steering rack force. People have varying preferences on how much if any of this addition feedback they want to feel through the wheel.
 
I think you are confusing PC1 and PC2. PC1 FFB was indeed confusing but PC2 FFB is much simpler and the menu explains what each "flavour" is and what each slider does. The nice feature they failed to communicate very well is the self adjusting nature of the immersive and informative flavours to minimise clipping. Nor are reviews almost universally negative on the force feedback.

This is rubbish. AC, ACC and Raceroom have multiple adjustments including "enhanced" effects. AC, PC2, ACC and RR are much more popular than AMS1 by an order of magnitude.

Again this is rubbish. Steering wheel force feedback is the only force feedback most of us get from our sim, while racing drivers get feedback through the seat of the pants and vibration of the car through the wheel which isn't a steering rack force. People have varying preferences on how much if any of this addition feedback they want to feel through the wheel.

I disagree that it is rubbish. First, canned effects don't count in what I am talking about, which is pure/proper physics-based FFB. The sliders in PC 2 are useless if you prefer realistic feedback, but I agree that you can customize the characteristics of the feedback to suit your tastes if you prefer something else. However, the fact that you need to tweak by vehicle to get the best result (see dozens of threads on this topic) proves my overall point that the adjustments are a superfluous distraction from what should have been a cohesively programmed combination set of FFB and vehicles. Even iRacing, with its multiple tire model versions and varying generations of vehicles, suffers from unnerving inconsistency from one vehicle to the next. rF2 is one of the worst--teasing with the most fantastic FFB in one vehicle and then wonky in the next (same issue of generational change with the latest tire modelling and FFB standards not applied across the vehicle portfolio).

So we can agree to disagree--your arguments were totally unconvincing. AMS, for whatever its other and overall strengths and weaknesses are, proves that FFB does not require any user adjustments to be excellent, plausible, natural-feeling, user effort-free and maximally-compatible with the widest set of wheel hardware. The almost universal praise of the FFB, even by all the people slamming the title while they complain about the graphics or other outdated aspects, proves this. Stripping all the silliness out of it (sliders and non-physics-based "enhanced features") is how and why it works so well. Consistency in vehicle physics modelling quality is what adds the icing to the cake. Just drive some hacked mods that do not respect this approach to reveal the results for yourself.
 
@Renato Simioni, I went to the Endurance Brasil race at Santa Cruz do Sul yesterday and was able to take a closer look on how the AJRs were updated, like push to pass, DRS, updates on aero in general, etc.

Will these changes in the AJR be carried to Automobilista 2 or will the car stay in its current virtual spec?

AJR aero has already been updated to better match real world performance (vs theorical CFD figures the original was based on) and it will have DRS and P2P.

Ad 1./
- Can you shed some light on what this bug was and how it affected car behaviour? If you've already done so internally on the AMS2 forums, where might I find more info on this?

Ad 2./
- What was wrong with the PC2 FFB system from your perspective? Are you ditching the "mixer pipeline" concept which is controlled by a lisp-like syntax? Or are you re-purposing the existing code to build a less flexible, more "one size fits all" system which is easier to calibrate for you?

Ad 3./
- Sounds very interesting! Looking forward to hear more about it!

1- Niels described the bug well in this post - it was fixed by ISI at some point over rF1 development but it was still present in the older code the Madness engine is based on. You could somewhat get around it by plugging inertia values that accounted for the wrong formula, but our car physics had inertia values already set to the correct formula so this was an important bug to fix.

2- Changing the FFB system was less about what was wrong with the native ffb and more about what was right with the simpler Realfeel-like system we´re implementing (basically converting the forces coming through the steering rack combined with a limited number of "hacks" for special circumstances and less room for misconfiguration). In time we will provide more documentation explaining these changes in more detail.
 
I disagree that it is rubbish. First, canned effects don't count in what I am talking about, which is pure/proper physics-based FFB.

I am loath to engage in arguments with people who appear not to be open to changing their minds (i.e. learning new facts and consequently reorganising their current understanding in light of newer, more accurate knowledge), but I'm going to make a decent effort anyway, because I feel the developers behind the PC2 FFB subsystem thoroughly deserve it:

To my knowledge, the only effect in the PC2 FFB pipeline that isn't pure/proper physics-based is road-noise from non-laser-scanned tracks, which is instead based on a combination of pseudo-random noise and some other physics signals. And this particular effect is quite good for what it is.

There are other effects that sample/record physics data from other spots in the car than the steering rack (to simulate e.g. missing inner ear / butt accelerometer clues), but to the best of my knowledge, these are NOT canned -- they are just derived from other physics data (lateral acceleration at the rear wheels for instance).

The sliders in PC 2 are useless if you prefer realistic feedback

This is factually incorrect.

The following is how to achieve realistic 'pure rack' FFB (note that this requires a high end wheel such as a CSW or DD with ample dynamic range for best results):

- Select the RAW Flavour
- Leave Tone at the default of 50 and never touch it again. This represents 'pure forces as they affect the steering rack through the suspension geometry'.
- Gain determines how much of your FFB wheel's maximum torque capability you want to be able to use. Setting this at 100 is similar to dialing a stereo power amplifier to 0 dB attenuation (i.e. full power). Some wheels need to be set at less than 100 to feel 'smooth' in their responses. Once you've found a proper Gain value that your wheel likes (as close to 100 as possible), leave it be and never touch it again.
- Set FX to 0 and never touch it again. 0 means 'the road noise that naturally comes through the simulated tyres and suspension links'. Settings above 0 essentially amplify road noise. As an example, my G27 wheel prefers settings in the range of 5-10. If I go above that, FX 'overshadows' other physics signals. A high-quality wheel doesn't need FX > 0 because it will already be powerful enough to convey subtle road surface effects.

The final setting is Volume. With the above settings now being 'locked in', Volume corresponds to the single 'FFB strength' slider you requested and will need to be dialed in for the cars that weren't adjusted optimally during development. The goal here is to reach a setting that doesn't clip during steady state cornering at maximum lateral load, which is easy enough to do:

Turn on the physics HUD and ensure that the FFB histogram graph has large 2nd (20-40% force) and 3rd (40-60% force) columns during cornering and that the 1st column (0-20% force) is largest when exiting turns and unwinding the wheel. Bumps and kerbs should show up in the 4th (60-80% force) and 5th (80-100% force) columns and the 6th column (100+% force or 'clipping') should be mostly zero during normal driving.

And that's basically it. Notice that it only took around 20 lines of explanation. That's hardly rocket science?

However, the fact that you need to tweak by vehicle to get the best result (see dozens of threads on this topic) proves my overall point that the adjustments are a superfluous distraction from what should have been a cohesively programmed combination set of FFB and vehicles. Even iRacing, with its multiple tire model versions and varying generations of vehicles, suffers from unnerving inconsistency from one vehicle to the next. rF2 is one of the worst--teasing with the most fantastic FFB in one vehicle and then wonky in the next (same issue of generational change with the latest tire modelling and FFB standards not applied across the vehicle portfolio).

This is because dialing in FFB per wheel and per car is time consuming (it hasn't got anything to do with 'programming' per se -- the word you're looking for is 'calibration'). There's also the not-so-small fact that suspension and SETA tyre design/construction parameters directly affect the steering rack forces, meaning that some cars by dint of their suspension layout alone don't produce as much FFB as others.

AMS, for whatever its other and overall strengths and weaknesses are, proves that FFB does not require any user adjustments to be excellent, plausible, natural-feeling, user effort-free and maximally-compatible with the widest set of wheel hardware. The almost universal praise of the FFB, even by all the people slamming the title while they complain about the graphics or other outdated aspects, proves this. Stripping all the silliness out of it (sliders and non-physics-based "enhanced features") is how and why it works so well. Consistency in vehicle physics modelling quality is what adds the icing to the cake. Just drive some hacked mods that do not respect this approach to reveal the results for yourself.

Both rF2 and PC2 use more advanced tyre models than AMS, which in contrast uses a simplified empirical model which I believe is incredibly well understood by now. I applaud Reiza for taking the time to refine the FFB for each and every car they create, but the fact of the matter is that development time is limited and SMS needed to split said time between the development and polish of the SETA tyre model, the FFB system and the per car calibration. In the end, SMS simply ran out of time and resources. That doesn't imply that the FFB system is poor, all it implies is that per-car FFB calibration required more development man-hours than was ultimately available.

Think of it as a lack of polish, not necessarily a lack of underlying quality and note how Reiza has pushed back the release of AMS2 for similar reasons: They, too, found themselves needing more man hours for polish. In the end, available dev time per car simply matters more than non-developers might ever grasp.

That said, I understand why Reiza might want to stick with the FFB approach they've already mastered and thus limit the amount of new tech they will need to get their heads around.
EDIT: As shown below, Reiza is aiming to make the FFB harder to mis-configure. They will in effect repurpose the existing FFB engine and programming language and implement a new UI and default FFB flavour that will work for people who were not trained as Mechanical Engineers, while still allowing tinkerers to create custom FFB flavours. Turns out Reiza will allow us to have our cake and eat it!

In closing, you are free to believe whatever you want to believe. The above are the facts as I understand them from following the PC1 and PC2 tyre and FFB development with keen interest over a number of years. But that certainly doesn't mean that I'm deluded enough to think I know everything there is to know about the subject (one of the reasons being that I don't have source code access).

As an aside, another PC2 FFB tinkerer recently discovered that adding back a set of forces (which were initially left out of the FFB configuration due to their small magnitude) by using the excellent under-the-hood FFB 'programming language' tends to make people feel that the FFB makes slightly more sense in many cases.
 
Last edited:
1- Niels described the bug well in this post - it was fixed by ISI at some point over rF1 development but it was still present in the older code the Madness engine is based on. You could somewhat get around it by plugging inertia values that accounted for the wrong formula, but our car physics had inertia values already set to the correct formula so this was an important bug to fix.

Thank you!

Here's hoping this will put to rest the persistent complaints about cars feeling off. I know a certain modder who's modded the MADNESS engine since the days of NFS Shift and he has always maintained that SMS use too small unsprung inertia values. Similarly, people were on SMS' case during PC1 and PC2 development, repeatedly asking whether the inertia values were correct. I guess this is such a foundational (and simple) issue that no-one thought to double check the deepest layer of the code.

2- Changing the FFB system was less about what was wrong with the native ffb and more about what was right with the simpler Realfeel-like system we´re implementing (basically converting the forces coming through the steering rack combined with a limited number of "hacks" for special circumstances and less room for misconfiguration). In time we will provide more documentation explaining these changes in more detail.

Heh -- that's a very diplomatic answer. Where might I read more about the "hacks" for special circumstances? I'm a AMS1 & 2 Reiza backer if that matters any.

I like the idea of "less room for misconfiguration". Part of me obviously worries that this will limit tinkering substantially, but if the end result is easier to configure FFB for 80+% of the user-base, I can learn to live with it I guess.
 
The FFB options in PCars 2 can be described as »It’s not a bug, it’s a feature.«

In fairness, the UI was put in relatively late in the development cycle and had to also allow for adjustability on PS4 and XBox (where the FFB system works a little differently than on PC). IMHO, Gain really shouldn't be visible to the user on PC -- it should have been locked at 100 to minimise potential confusion.

Heck, even the part of the physics overlay HUD that shows the FFB graph and histogram was put in *very* late by someone who wasn't even on the UI team. I'm just happy that we ended up having this functionality at all!
 
I am loath to engage in arguments with people who appear not to be open to changing their minds (i.e. learning new facts and consequently reorganising their current understanding in light of newer, more accurate knowledge), but I'm going to make a decent effort anyway, because I feel the developers behind the PC2 FFB subsystem thoroughly deserve it:

To my knowledge, the only effect in the PC2 FFB pipeline that isn't pure/proper physics-based is road-noise from non-laser-scanned tracks, which is instead based on a combination of pseudo-random noise and some other physics signals. And this particular effect is quite good for what it is.

There are other effects that sample/record physics data from other spots in the car than the steering rack (to simulate e.g. missing inner ear / butt accelerometer clues), but to the best of my knowledge, these are NOT canned -- they are just derived from other physics data (lateral acceleration at the rear wheels for instance).



This is factually incorrect.

The following is how to achieve realistic 'pure rack' FFB (note that this requires a high end wheel such as a CSW or DD with ample dynamic range for best results):

- Select the RAW Flavour
- Leave Tone at the default of 50 and never touch it again. This represents 'pure forces as they affect the steering rack through the suspension geometry'.
- Gain determines how much of your FFB wheel's maximum torque capability you want to be able to use. Setting this at 100 is similar to dialing a stereo power amplifier to 0 dB attenuation (i.e. full power). Some wheels need to be set at less than 100 to feel 'smooth' in their responses. Once you've found a proper Gain value that your wheel likes (as close to 100 as possible), leave it be and never touch it again.
- Set FX to 0 and never touch it again. 0 means 'the road noise that naturally comes through the simulated tyres and suspension links'. Settings above 0 essentially amplify road noise. As an example, my G27 wheel prefers settings in the range of 5-10. If I go above that, FX 'overshadows' other physics signals. A high-quality wheel doesn't need FX > 0 because it will already be powerful enough to convey subtle road surface effects.

The final setting is Volume. With the above settings now being 'locked in', Volume corresponds to the single 'FFB strength' slider you requested and will need to be dialed in for the cars that weren't adjusted optimally during development. The goal here is to reach a setting that doesn't clip during steady state cornering at maximum lateral load, which is easy enough to do:

Turn on the physics HUD and ensure that the FFB histogram graph has large 2nd (20-40% force) and 3rd (40-60% force) columns during cornering and that the 1st column (0-20% force) is largest when exiting turns and unwinding the wheel. Bumps and kerbs should show up in the 4th (60-80% force) and 5th (80-100% force) columns and the 6th column (100+% force or 'clipping') should be mostly zero during normal driving.

And that's basically it. Notice that it only took around 20 lines of explanation. That's hardly rocket science?



This is because dialing in FFB per wheel and per car is time consuming (it hasn't got anything to do with 'programming' per se -- the word you're looking for is 'calibration'). There's also the not-so-small fact that suspension and SETA tyre design/construction parameters directly affect the steering rack forces, meaning that some cars by dint of their suspension layout alone don't produce as much FFB as others.



Both rF2 and PC2 use more advanced tyre models than AMS, which in contrast uses a simplified empirical model which I believe is incredibly well understood by now. I applaud Reiza for taking the time to refine the FFB for each and every car they create, but the fact of the matter is that development time is limited and SMS needed to split said time between the development and polish of the SETA tyre model, the FFB system and the per car calibration. In the end, SMS simply ran out of time and resources. That doesn't imply that the FFB system is poor, all it implies is that per-car FFB calibration required more development man-hours than was ultimately available.

Think of it as a lack of polish, not necessarily a lack of underlying quality and note how Reiza has pushed back the release of AMS2 for similar reasons: They, too, found themselves needing more man hours for polish. In the end, available dev time per car simply matters more than non-developers might ever grasp.

That said, I understand why Reiza might want to stick with the FFB approach they've already mastered and thus limit the amount of new tech they will need to get their heads around.
EDIT: As shown below, Reiza is aiming to make the FFB harder to mis-configure. They will in effect repurpose the existing FFB engine and programming language and implement a new UI and default FFB flavour that will work for people who were not trained as Mechanical Engineers, while still allowing tinkerers to create custom FFB flavours. Turns out Reiza will allow us to have our cake and eat it!

In closing, you are free to believe whatever you want to believe. The above are the facts as I understand them from following the PC1 and PC2 tyre and FFB development with keen interest over a number of years. But that certainly doesn't mean that I'm deluded enough to think I know everything there is to know about the subject (one of the reasons being that I don't have source code access).

As an aside, another PC2 FFB tinkerer recently discovered that adding back a set of forces (which were initially left out of the FFB configuration due to their small magnitude) by using the excellent under-the-hood FFB 'programming language' tends to make people feel that the FFB makes slightly more sense in many cases.

First off, your reply reconfirmed most of my perspective, despite your attempt to convince me/others/yourself otherwise.

Secondly, thank you for supplying the intelligence on how to properly set the FFB in PC 2. Despite asking multiple times myself and having read others asking the same question I have never seen an explanation. SMS should have simply had a default that corresponds to "pure rack" (called something user-friendly) and we could all skip the 20 lines of explanation that as I mentioned, you are one of an extremely tiny minority who are in the know.

Thirdly, and sadly, even if I go adjust my PC 2 to these new proper settings, I still won't be able to hop in any car and drive what a dev thought was the best possible presentation of their "product"--I still have to fiddle with each car not having the slightest clue as to what might be sub-optimal with that particular vehicle.

So, my conclusion still stands and if you re-read what I wrote, I was never suggesting that the underlying engine was defective or that canned effects were required use in PC 2. The implementation of it is extremely poor and requires far too much science experimentation from even the most die-hard sim racer. And, the lack of polish (for whatever reasons) was the nail in the coffin.
 
First off, your reply reconfirmed most of my perspective, despite your attempt to convince me/others/yourself otherwise.

Secondly, thank you for supplying the intelligence on how to properly set the FFB in PC 2. Despite asking multiple times myself and having read others asking the same question I have never seen an explanation. SMS should have simply had a default that corresponds to "pure rack" (called something user-friendly) and we could all skip the 20 lines of explanation that as I mentioned, you are one of an extremely tiny minority who are in the know.

Thirdly, and sadly, even if I go adjust my PC 2 to these new proper settings, I still won't be able to hop in any car and drive what a dev thought was the best possible presentation of their "product"--I still have to fiddle with each car not having the slightest clue as to what might be sub-optimal with that particular vehicle.

So, my conclusion still stands and if you re-read what I wrote, I was never suggesting that the underlying engine was defective or that canned effects were required use in PC 2. The implementation of it is extremely poor and requires far too much science experimentation from even the most die-hard sim racer. And, the lack of polish (for whatever reasons) was the nail in the coffin.
I can add a bit more to this. You certainly don't need to adjust it on a per car basis. It is more a per suspension geometry basis. To put it simply, cars of a given class will behave similarly for the same ffb settings meaning you can set one up and you can freely change cars within class.

Now, one of my biggest criticisms if this game is the inability to save individual settings for each car both in ffb as well as in controls. But it is certainly not the nightmare you seem to paint.
 
thank you for supplying the intelligence on how to properly set the FFB in PC 2. Despite asking multiple times myself and having read others asking the same question I have never seen an explanation. SMS should have simply had a default that corresponds to "pure rack" (called something user-friendly) and we could all skip the 20 lines of explanation that as I mentioned, you are one of an extremely tiny minority who are in the know.

You're welcome.

And FWIW, I agree that the 'User Experience' could have been better. That's not a FFB system defect, that's a User Experience (UX) defect.

Thirdly, and sadly, even if I go adjust my PC 2 to these new proper settings, I still won't be able to hop in any car and drive what a dev thought was the best possible presentation of their "product"--I still have to fiddle with each car not having the slightest clue as to what might be sub-optimal with that particular vehicle.

It would have been nice to have had the ability to save this setting per vehicle so you'd only need to do this every time there was an update to each car, yes.

Again, a UX defect.

So, my conclusion still stands and if you re-read what I wrote, I was never suggesting that the underlying engine was defective or that canned effects were required use in PC 2. The implementation of it is extremely poor and requires far too much science experimentation from even the most die-hard sim racer. And, the lack of polish (for whatever reasons) was the nail in the coffin.

In software development 'implementation' is a narrowly defined term that directly refers to the underlying code. I disagree that the FFB implementation is 'extremely poor'.

I do however agree that the UX (the User Interface, the presentation plus the accompanying documentation) perhaps leaves something to be desired for more casual users. Reiza appear to have recognised this and will ship a simpler UX without compromising the underlying FFB system's quality and flexibility.

I will also admit to being triggered by your assertions that the SMS devs are/were 'lazy'. They aren't and weren't (just as Reiza's developers surely aren't): During WMD development I witnessed a group of passionate developers who on numerous occasions submitted code late at night during the weekends and who worked their backsides off to meet tight deadlines.

I was also triggered by your use of certain terms, but perhaps I read more into said terms than was warranted. And as always, you are of course entirely entitled to your opinion and your frustrations!

Cheers :)
 
And before I forget:

I'm super excited about AMS2 -- I have every faith in Reiza delivering on the potential of the MADNESS engine, now that the underlying new tech has had a few years to mature and 'gel'.

The notes about the small fixes and adjustments (which will likely have a big impact on the user experience and driving experience) is just the icing on the cake for me. :thumbsup:
 

Latest News

Online or Offline racing?

  • 100% online racing

    Votes: 87 7.4%
  • 75% online 25% offline

    Votes: 123 10.5%
  • 50% online 50% offline

    Votes: 170 14.5%
  • 25% online 75% offline

    Votes: 329 28.0%
  • 100% offline racing

    Votes: 460 39.2%
  • Something else, explain in comment

    Votes: 4 0.3%
Back
Top