denis74 escribió:Evidentemente un hardware de hace 7 años no puede tirar como un pc de hoy en día, si le pones todo al máximo como es en este juego. Bajón gráfico sufrirá para tener unas físicas, clima dinámico, tiempo del día cambiante, público con animaciones.. En cuanto a las físicas es algo que se está discutiendo allí. Por suerte son muchos los usuarios que vienen del iracing, rfactor, gtr, live for speed. Así que caña se les está dando en relación a las físicas.
Una vez salga el juego habrá DLC de coches, eventos y pruebas de Rally.
Dutch escribió:¿Este juego se supone que va a ser un simulador mas parecido a iracing, a race pro o a shift2 (este no es un simulador ni de lejos pero bueno...)? me refiero en cuanto a las fisicas.
Saludos.
Dutch escribió:Si decis que se puede jugar con pad eso lo dice todo, no? iracing por ejemplo te dicen que necesitas un volante. Es absurdo hacer un juego y programar el control para que vaya bien con el pad y llamarlo simulador.
Saludos.
KoX escribió:Dutch escribió:Si decis que se puede jugar con pad eso lo dice todo, no? iracing por ejemplo te dicen que necesitas un volante. Es absurdo hacer un juego y programar el control para que vaya bien con el pad y llamarlo simulador.
Saludos.
hombre, los que no tenemos volante tenemos derecho a jugar no? xD
Yo entiendo que no es ni por asomo parecido, y realmente yo me conformaria con unas fisicas como las del PGR4, más arcade pero que "simula"/"ambienta" bastante bien movimientos de cámara, frenadas etc...pero con un Modo Historia más "realista"
Si este juego tiene las dos cosas, pero yo por usar mando me pierdo la parte más simulador, pues mira es problema mio.
Además que sin lugar a dudas habrá dos métodos de control, volante y pad, ambos balanceados por separado.
Adjusted volumes of turbo, tranny, backfires and splutters
Added subwave to exterior, added comp to rumbles interior and exterior; lowered delay mix for focus; boosted various levels, add subtel distortion dsp to on and off throttle for sat effect.
Ok, I'm a few days into thread hell now so thought it was about time I posted an update
I've created a separate high-priority thread that's now running nothing but input device state reads @ 250Hz. This ensures that the latest state for a device is always available (to within 4ms), and I've modified various areas of code to allow for on-demand updates. This means that when systems decide to interrogate the inputs they can request that the internal representation of the device is updated with the latest hardware state.
Although this is a key component in the fight against the lag, no improvement will be visible until the physics threading code has received the required updates too. This is typical of threading solutions in complex systems, only when the last piece of the puzzle clicks into place do we then finally see the fruits of our labour!
As I'm sure you can understand, the physics guys are seriously busy right now, but they will be getting round to this threading work as soon as possible. Until then, I've got plenty of stuff to be getting on with :-
• On-demand FFB updates. So far I've only dealt with reading the device inputs on a thread, so I need to give the same treatment to the updating of the FFB.
• Add some tweakers. The ability to both set the master rate, and also provide device-specific step-downs so that certain devices don't need to be sampled at every tick of the master update.
• Prediction. I'm not a big fan of Input prediction as I find it can sometimes create as many problems as it solves. I've got something implemented though which I'll most likely set to off by default, but it might be worth trying to activate on lower-end systems to help combat the effects of lag.
• Reflect the changes across all other devices/platforms. Currenty I've been concentrating on the PC-specific DirectInput device class but the same work will also need to be done for XInput, Keyboard, Mouse and console versions of all the above (about a dozen different device classes altogether).
Another few days trudging around the codebase, and it's time for another update
The good news, as hopefully most people have seen now, is that we've had the first big win in the fight against the lag. The render guys identified an area of the code that was basically causing an entire frame's worth of lag. From what I've seen myself, and read around the forum, I'd say this was contributing up to about 50% of the overall lag. There was a worry that there might be a performance hit with the render changes, but the feedback from you guys suggest that hasn't been an issue - which is a relief
Back to my own personal hell now, and I've been busy in a few areas :-
• Visualisation. Lag is one of those problems that is extremely difficult to diagnose, because it's not really possible to accurately quantify or record using the normal techniques, such as logging. Although you can record timestamps at the point where the device data is read, or even used, you can't record timestamps for when the user sees the visual representation of that data (monitor's don't tend to reply when you ask them such questions ). What is possible to quantify though, is the suspected specific causes of the lag - in this case the irregularities between the input and physics thread updates. To this end I've implemented some graph drawing code which tracks the physics update timestamps against the input read timestamps, which should be a big help for the physics guys when they get round to revamping the physics thread.
• Tweakers. I've added both a master thread update rate (number of milliseconds between updates), and device-specific step-down rates to allow certain devices to be sampled at a lower rate than others. Both of these can quite happily be tweaked in real-time, and I'll be exposing them via the dev menu.
• Prediction. I've implemented input prediction, and created graphing code to plot the predicted values against the real values. As I suspected, it's not working so great, but should get better when the physics updates are working as planned.
I'm now working on the FFB update thread and I'll report back when I have the results from that. Cheers!
All the tyre sounds are going to be replaced. We'll also have enough variations to avoid hearing the same old sound every time you lose grip
Just a note that I've been working on the physics over the break. I decided to take the opportunity with the lack of distraction to really try to focus on what we're discussing here. I've started with the LMP car and went a new direction with the tyres (the chassis is pretty clear cut so just balance changes). The basics of it are as follows. Previously i used the base rubber grip split( front vs rear) to be large part of the balance and then leave the load sensitivities of each tyre the same, and now I've done the opposite with the base rubber grip the same front/rear and split the load sensitivity values to create balance. Also closed the gap between the cornering stiffness's of the front/rear tyres to remove as much of the "resistance to turn at speed" as possible (too far and straight line stability really suffers). Reduced some of the aero side forces too. Tried to mix everything in to get a more convincing drive per above. In general a free-er feeling car. There's good power oversteer now in general and the cars are reasonably catch-able. The overall grip is lower too. I had to re-balance the chassis' with less rear roll stiffness and some shock work to bring it back to neutral on trailing throttle, yet leave in some throttle oversteer.. Also the Formula B got new front suspension geometry as the game didn't like the high roll center from the upward inclination of the a-arms (no change visually), occasionally flipping at Bologna's Alta chicane with heavy curb use.
Enjoy and I'll continue on with this direction if there's agreement.
denis74 escribió:No te lo vas a creer, Din-A4. He visto el video y en efecto gira demasiado. Me ha extrañado bastante. Me lo acabo de poner y no gira así. Yo creo que debe ser su configuración de volante aunque como sabes,,,yo juego con pad.
Si quieres grabo un pequeño gameplay y te lo pongo.
I posted in the other Physics thread about using higher sliding grip I'm experimenting with, i agree with your observation there. Again this is nothing new for me, but it's been nice being able to focus once again on true sim work. The current tyre model has never been used for pure sim, as it has always been a part of the Shift series where lots of smothing over was done to make it drivable with a game pad. So now i can really flesh out the details more without all the hacks.
The braking locking/unlocking can be tuned, but it does effect the on power longitudinal which is my main focus. The braking then falls where it will. There no seperation in the parameters between the two unfortunately.
Physics / Handling Build 124
Noted Changes
1. It feels like there’s more control over positioning the car (LMP) anywhere you like in terms of line. IE steering it more precisely where you want to go in the straights.
2. Driveability on power feels better on the LMP.
3. It is possible to draft other cars – very cool!
Main issues remain:
1. Lack of turn in – especially high speed corners. It should be possible to make the car oversteer with a heavy handed / faster turn in. (The Ariel Mugen permits this and I re-created a similar feeling to reality by trying a Scandinavian flick into some of the high speed corners)
2. Lack of turning power in long slow corners - it should be possible to wind on more steering at a slightly lower speed (than that which causes understeer) to make the front end bite and balance the car into oversteer
3. Ability to lock / unlock wheels under braking – this is a strength of the CARS sim but if it could be improved at all it would incredible. This is about processing speed for the system to recognize that the driver has altered brake pressure and react accordingly by restoring braking grip to the tyres.
4. The AI cars are complete nutters!
Kart
Love the engine noise and RPM
Main issues:
1. Under 100% braking the Kart should spin – it will always lock a rear wheel in the end, especially karts with only rear brake discs which is the common set up. (regarding this note, karts regularly pitch slightly sideways under braking and the grip drop off with a wide slip angle is MINIMAL. Would be great to re-create this as it one of the most exciting aspect of 2-stroke karting)
2. On power – great feedback through the throttle to get the kart drifting. The problem is that the grip drop off is too great and most of the drift are irrecoverable as a result. Mostly this is caused by the return snap caused as the rear re-aligns – this happens even when the steering is straight whereas in real life you just leave the steering straight and the oversteer just bleeds off. It’s a big part of the charm of these karts, you drive them on opposite lock all the time. This snapping return oversteer where it kicks back is very evident if you take a trip through the gravel, as it picks the whole kart up and sends it in a new direction
3. Turn in – like the LMP it should be possible to hurl the kart into the corner with a slight drift on and try to balance it all the way through the corner. Which takes me to the Mugen
The characteristic of the real karts is violent, twitchy oversteer – requiring much faster steering responses but in the fact the grip level even when sideways is fairly high.
Ariel Mugen
Very impressive look and feel. What I like most about this model is that it seems to be the most pliable and so far is the only machine I have been able to four wheel drift into a corner. This is because it has TURN IN. This is the car that responded well to the Scandinavian flick, wheread the kart for example would interpret the switch in weight across the platform as a signal to take all grip away and spin.
There seems to be a good level of control on power out of corners and the sim lets you drift it out and balance the car on the throttle fairly well. More of this please if its possible.
Ideas for improvement
1. Even more throttle control
2. More grip when you make a transition slide between corners (it does bite too hard if you switch directions whilst sliding)
3. Locking wheels. Ariels tend to lock front wheels and you get a puff of smoke through the wheel-arch or a glimpse of a frozen wheel. Would be great to replicate that. They have very sharp brakes and due to the lack of downforce you can lock a wheel if you jab the brake too hard – however they respond quickly if you release the brake pressure, the wheel turns and you can re-apply the brake.
Caterham 500
Look and feel is superb – the way the long bonnet tracks ahead of you is just like the real thing. Great work. I barrel rolled it at the first corner though - no pain, no gain.
Main issues:
1. The grip level as the speed increases becomes extremely low, to the extent that it feels like a wet track. The tyres start spinning over bumps when you reach around 85mph and if you apply some steering the rears want to slide. Not sure what is causing this as it doesn’t feature in the other models – but it doesn’t reflect reality
2. The overall grip level is too low. When you accelerate its great to be able to break traction so easily but I found the tail of the car far too slippery on the way into medium to fast corners. It wants to spin all the time.
3. On power there is not quite enough grip to hold a powerslide through a corner. When you get it wrong by drifting too sideways you might lift off the gas to recover the slide. It would awesome if the sim could recognize this and judge that the friction of sliding (with no throttle) is reducing the cars speed, and raise the grip level so that you can catch the slide, feather the throttle (re-applying on and off) and drift out of the turn.
4. Braking - this could be more aggressive and like the Atom you get some really fun moments locking the front wheels (our model seemed to favour locking rear but i believe that was a physics/grip thing) and looks like this video, where you can also hear the on/off throttle feathering as the car drifts: http://www.youtube.com/watch?v=6tN6_xUL8j4
With the Mugen and the Caterham you find occasional corners where it picks up so much understeer that you can’t overcome it. If there is any way to close the relationship between the front and rear ends so that you (the driver) can adjust the dynamic handling by adding some steering, or a little power it would really help.
Slip Angles
Just to clarify as slip angles are a minefield of misinterpretation – I am advocating MORE grip (ie less drop off) at the point when the optimal slip angle is achieved, AS WELL AS beyond it. At the moment when the thing starts to slide we are dropping grip away too soon when in fact the tyre should be providing its best grip at that point. Would be great to chat through this sometime so I can tune in to the dev changes at your end!
Loving your work guys!! And Happy New Year!!!
He chequeado un nuevo campo de display que aparece un panel con tu actual vs vuelta personal. El juego muestra tu "potencial", mejor vuelta y calcula tu mejor progreso en la pista con unos segmentos dinamicos (mas que sectores).
El tiempo de la vuelta potencial es calculado en tiempo real cuando el juego observa que eres más rápido que el tiempo de tu vuelta potencial, el tiempo mostrará una mejora, pero si haces un error en la próxima curva, el juego se hecha atrás en la predicción..
Tambien hay un valor delta en vuvo que muestra, y este es el tiempo delta para poner un punto preciso relativo a tu mejor progreso grabado anteriormente en ese punto. (nota: no tiene nada que ver con su actual o en el mejor tiempo de vuelta en sí mismo).
Ir por fuera de la pista (mas de dos ruedas), impactos fuertes, ir en sentido contrario (mas de 90º), y mas cosas, invalidan tu vuelta y necesitarás correr unas curvas limpias para que el calculador se ponga otra vez en marcha.
Hola
Hemos estado toqueteando el Force Feedback, y esta es la lista de cambios (este trabajo es separado del imput lag):
Lista de cambios:
1. Eliminado la vibracion irreal al pasar por algunas curvas
2. Las fuerzas del volante se corresponderán a la suspension
3. Centro de la fuerza ajustado a la velocidad y frenado.
4. cuando el coche esta estacionado, el volante no vuelve al centro milagrosamente, tendremos que girarlo nosotros.
5. Suspension delantera y fuerzas G influyen en la fuerza del volante. El volante se vuelve mas duro cuando hay mucho grip frenando y acelerando.
6. Sentiremos mas el volante cuando pasemos por pianos y zonas bacheadas.
7. Sacudidas del volante cuando nos choquemos.
8. Fuerza aplicado en subviraje
Primero se esta haciendo esto en volantes logitech
¿Como compararías Project CARS a títulos como GTR2 e IRacing?
La mayoría del equipo trabajó en el GTR2 original y compite regularmente en iRacing pero Project CARS representa el next-gen de simulación de carreras. Mediante la inclusión de los chicos que aman esos juegos (y otros) estamos recibiendo algunos de gran comprensión que son los temas actuales que realmente resuenan con jugadores (ejemplo, telemetría extensiva, setups, "sentir el finde semana de carrera&caracteristicas) y mezclado con nuestros propios conceptos para conectividad social, hora del día y clima dinámicos,,etc.... Sentimos Project CARS que está en una fuerte posición que tiene a ambos unidos bajo el mismo techo de la comunidad de simulación de carreras y compite directamente con la competición.
¿Habeis conseguido la licencia oficial para varias disciplinas que vais a ofrecer?
Hay negociaciones en todo el tiempo pero necesitan un tiempo para conseguir el fin legal. Tan pronto como se deciden las cosas, lo diremos a nuestra comunidad (ver previos anuncios del Caterham, Motorsport Vision track,etc...) así que mantén los ojos bien abiertos así la próxima gran licencia podría estar a la vuelta de la esquina.
¿Puedes darnos más detalles sobre tu descrito método para Project CARS?
Eso relata directamente al modo Carrera en Project CARS que idolatra un sentimiento de progresión a través de tu carrera desde junior o rookie comenzando por una pasión bruta y un talento por las carreras, consiguiendo el reconocimiento por un equipo semi-profesional, encontrando tu medida y especialización en una disciplina que tu ames (BTCC, DTM, Rally, Indycar, cualquiera) y entonces iendo a ganar las más prestitigiosas títulos/trofeos en la categoría y consiguiendo ser reconocido en el mundo del automovilismo. Eso puede llevar un año in-game o temporada o puede ser un objetivo a largo plazo, pero en última instancia, la analogía simula la progresión y las oportunidades de un piloto de carreras real tendría como oposición a un más juego como "rutina de dinero en efectivo o XP '. Este tipo de progresión es comúnmente llamado un modo de Franquicia en juegos deportivos, pero es que realmente no ha sido explorado por completo en un título de carreras antes. Otros juegos han bajado el dedo del pie en el agua aquí (GRID, F1), pero definitivamente hay espacio para impulsar aún más.
¿Como vais a incorporar DLC con Project CARS?
Los planes no han finalizado todavía, pero nuevo contenido será lanzado regularmente para el juego una vez lanzado. y se vinculará a temporadas de deportes de motor, disciplinas, fabricantes, eventos específicos, épocas (por ejemplo. histórico o coches de los años 80), lugares (por ejemplo,. pistas alemanas) y muchos más. Por lo que es más como una continua entrega estilo MMO que una seria de packs estilo consola.
¿Cual es el punto fuerte del motor MADNESS cuando se trata del desarrollo de un juego como Project CARS?
El motor MADNESS fué construido con la previsión de ser un motor multi-threaded y multi-processor que puede ser escalado fácilmente dependiendo del tipo de juego. Por lo que ya ha visto un montón de mejoras desde su inicio por primera vez en Need for Speed ??SHIFT para permitir completas sombras dinámicas proyectando luces para verdaderas terroríficas carreras nocturnas, grandes mejoras de la IA, físicas, modelos de neumáticos, input, etc. Cualquier nuevas características de juego se pueden considerar como módulos para conectarse en el motor que ya está capacitado y preparado para ello.
¿Ha sido un reto implementar el sistema de clima y noche/día dinámicos?
Trabajar en estos sistemas se debe comenzar pronto y la clave es siempre considerando la respuesta emocional del jugador si se trata "Wow, eso luce real" o "Vaya para que va a llover, mejor entrar al pit en la próxima vuelta". Recibir esta respuesta emocional es el mayor reto.,de lo contrario, sólo estás haciendo bastantes gráficos que no tienen ninguna función o algo que funciona, pero no se ven muy bien. Técnicamente, por supuesto, hay algunos obstáculos en el lado del render y algunos cambios en la fuente en el lado del arte, pero es algo que hemos pensado haciendo proyectos anteriores y por lo tanto está bastante preparado para ello ahora.
¿La Gestion del Equipo está limitado al single-player, o será algo que puedas organizar en clanes?
La Gestión del equipo en la Carrera (single player) significa la contratación de compañeros de equipo de la IA como tu segundo piloto comunicando con tu equipo via radio del coche en las carreras, y gestionando tu carrera a través del calendario (escogiendo que eventos para asistir, equilibrando tiempos de prácticas con eventos de exhibición, etc..). En los modos de juego conectado (online), la Gestión del Equipo significa (jugar con amigos" similar a jugar en clan con diferente gente haciendo diferentes roles, contratación, comunicación, sistemas de seguimiento de estadísticas. Así la gente puede socializarse y jugar juntos de una manera co-operativa y competitiva. Cuando se combina la red basada en nube puedes imaginar tu teléfono dando una notificación que un compañero de equipo ha cambiado el setup de tu coche, que tu corres en ese fin de semana mientras otra ha aplicado tus nuevos logos de sponsor en el capó frontal. Puedes contestar agradeciéndolo, y organizar sesiones de testeo esa noche para probar los nuevos ajustes con uno de tu equipo que actúa como jefe de equipo viendo tu telemetría.
¿En que plataformas va a aparecer la red social "nube", y como difiere en varias plataformas?
Siempre que sea posible, los datos serán almacenados en la nube permitiendo ser accesibles tanto en una web, aplicaciones de móbil o en cualquier lugar donde puedas acceer a tu cuenta. Así que la única diferencia sería probablemente en la navegación de todos esos datos dependiendo del dispositivo que se esté utilizando.
http://www.thereticule.com/world-of-mas ... ew-part-2/
A brief update - we're still not quite ready to enter full design, planning and production but we have come up with a short term priority list based on member feedback and available team resources. In very rough priority order (but all will be started soon if they haven't been already) this is:
Input lag (new threading system)
New FFB system
Car specific steering lock support
Fix low speed grip issue
Trees improved
Improved car reflections
Car and track updates as usual
Tire deformation
Open face driver added for 67 car
AI parameters exposed in UI
Night lighting
"High detail player car" option added to UI
Vehicle livery selection in UI
1st person driver animations
GUI usability improvements (including track select menu)
Kart physics fixes (low speed grip fix will help a lot)
Leaderboards (specifically client-server instant updates, ghost support)
Multiplayer
Pit-to-car radio
Damage fixes/improvements, including wheel to wheel contacts
Dynamic time/weather system
Driving school
Slow motion replays
Camera tool for use by team members
Web leaderboards
Helmet cam with reflections
denis74 escribió:* En el día de ayer se han conseguido grandes progresos del FFB una vez aplicado el parche. (ayer salió a primera hora de la tarde el build pero hacia las 6 de la tarde sacaron un parche para mejorar aun más el FFB. Tengo que decir que la gente que juega con volante está encantada, según voy leyendo, con el resultado conseguido. Buenas impresiones.
* Lás físicas han dado otro paso más para algunos coches y también cambios en general. Por ello han reseteado las tablas de tiempos. Doug Arnao (de GTR2) sigue con los trabajos en este campo.
* Esta es la web de estadísticas del juego que poco a poco van mejorando:
http://cars-stats.wmdportal.com/
* Licencia del coche del 2011 BAC Mono
* Ya tenemos Silverstone en el juego
* De Enero-Abril vamos a poder probar en builds lo siguiente:A brief update - we're still not quite ready to enter full design, planning and production but we have come up with a short term priority list based on member feedback and available team resources. In very rough priority order (but all will be started soon if they haven't been already) this is:
Input lag (new threading system)
New FFB system
Car specific steering lock support
Fix low speed grip issue
Trees improved
Improved car reflections
Car and track updates as usual
Tire deformation
Open face driver added for 67 car
AI parameters exposed in UI
Night lighting
"High detail player car" option added to UI
Vehicle livery selection in UI
1st person driver animations
GUI usability improvements (including track select menu)
Kart physics fixes (low speed grip fix will help a lot)
Leaderboards (specifically client-server instant updates, ghost support)
Multiplayer
Pit-to-car radio
Damage fixes/improvements, including wheel to wheel contacts
Dynamic time/weather system
Driving school
Slow motion replays
Camera tool for use by team members
Web leaderboards
Helmet cam with reflections
*Trailer no oficial hecho por el miembro HDTanel (le recordareis por sus videos de pc del F1 2011). Incluso gametrailers ayer publicó dicho video:
http://www.youtube.com/watch?v=4NSjfaSw4_M&hd=1
Los responsables de Need for Speed: Shift, los chicos de Slightly Mad Studios, han anunciado que su próximo simulador de velocidad, el conocido como Project Cars, se lanzará también en Wii U.
El videojuego, recordemos, fue anunciado en un comienzo para PC, Xbox 360 y PlayStation 3. De momento no hay fecha de lanzamiento para este juego desarrollado por la comunidad, un título en el que el feedback de la beta abierta es fundamental para los resultados de éste.
Gorkexo escribió:¿Pero y este pepino saldrá en consola? Raro me parece... Esto me tira más a un rFactor... Este juego luce demasiado bien para verse así en consola, además de que en Pc meterán muchísimos Mods como sucede en el rFactor por poner un ejemplo.
started with proper drivers integration today:
- it's just a start, lot of details in animations and driver behaviour are missing (will be sorted during this week) and also driver materials and skinning will be polished
- new driver meshes exported into game - primarily for GT cars, but scripts and game are also ready for specific F1 driver which will be added later (soonish) and KART dirver (later)
- animations and cockpit are setup works for ASANO LM11 only for now. More cars should be ready (at least basically without detailed animations) during next week once we will sort out all details for new driver.
- 3rd person driver + 1st person driver works together, which means we have driver arms visible from 1st person for this car and we can start with helmet-view once we will have some more time.
- setup car files to fit driver into cockpit - pedals and paddles aren't working yet (should be done later this week)
- initial cockpit camera setup done ... defaults will be adjusted back and forth this week.
ToDo:
- alcantra/ fresnel are too strong and making hands volume visually bigger than acceptable (especially in shadows) + character is too bright in comparison with the rest of cockpit
- even thought I changed all emaps to "inside", all reflections are still too strong (making helmet/visor too bright) . However this should be caused by alcantra/fresnel settings
- add fabric structure detail to normal maps (maybe as additional texture pass)
- tweak skinning + add full animation set
- work with Clarky to get each driver type to specific folder to help organize data better
- coop with coders to prepare system which will allow to load custom skins/helmets paints for all team members
Ian Bell: And we'll have the best 1st person driver in soon.
New code checked in; tyre slip force feedback has changed allowing the under and over steer to be felt more easily.
Gain can be used to change the weight of the constant force strength (mostly when yaw occurs).
Average Weight default value has changed to 0.8 and its function completely changed since previous builds. It now represents the default weight of the centre spring during normal standard zero accel/decel/cornering driving.
Spring Coefficient is default 1.0 and controls the max strength of the centre spring.
Tested on G27 today with the Formula B at Connecticut Hill Short with testing across other wheel types tomorrow.
Players MUST delete save game profile otherwise the new controller defaults will not work.