Jump to content


Photo

The player has no animations or shadow


  • Please log in to reply
48 replies to this topic

#26 chakkman

chakkman

    Advanced Member

  • Member
  • PipPipPip
  • 932 posts

Posted 02 January 2018 - 03:34 PM

I get Springheel's 360 degree vision concern, but, imagine the bodily awareness in VR. I know that you still won't move yourself, but, that also applies to any other game, racing games e.g., and still you see the driver's body and arms on the steering wheel, and, IMO, it definitely adds to the immersion. And, you will have a whole other kind of awareness with VR, so it may add to the gameplay. 



#27 cabalistic

cabalistic

    Member

  • Development Role
  • PipPip
  • 339 posts

Posted 02 January 2018 - 03:48 PM

Well, VR and shadows can definitely work together: ;)


  • Anderson likes this

#28 kano

kano

    Member

  • Member
  • PipPip
  • 313 posts

Posted 03 January 2018 - 11:02 AM

I would love to see our thief get an animation upgrade, so that he doesn't look like he's catatonic whenever I walk in front of the mirror. But I am not skilled with character animation, sadly.


  • Aosys likes this

#29 kano

kano

    Member

  • Member
  • PipPip
  • 313 posts

Posted 03 January 2018 - 11:04 AM

There's really no sense in having elaborate player model and animations. Just look at Thief 2014. Even with all the AAA manpower, the animations still look stiff.

Riddick did this really, really well.


  • Anderson likes this

#30 Springheel

Springheel

    Creative Director (retired)

  • Admin
  • 37246 posts

Posted 03 January 2018 - 11:11 AM

Riddick did this really, really well.

 

But the game still took control of your character whenever you climbed a ladder or used a healing station, in order to make the animations line up properly.


  • Judith likes this
TDM Missions:   A Score to Settle   *   A Reputation to Uphold   *   A New Job   *    A Matter of Hours
 
Video Series:   Springheel's Modules   *   Speedbuild Challenge   *   New Mappers Workshop  *   Building Traps

#31 Judith

Judith

    Advanced Member

  • Member
  • PipPipPip
  • 1322 posts

Posted 03 January 2018 - 03:46 PM

Riddick did this really, really well.

 

As I said, the problem isn't with animation. As perfect as it can be, its weakest link will always be player input. Silly key presses or thumbstick tilting will always break the immersion this way. The only solution I know of would be kind of "animation queue", as in Dark Souls, where you will be punished if you randomly spam movement keys or attacks. But that's very specific kind of gameplay too.



#32 stumpy

stumpy

    Advanced Member

  • Member
  • PipPipPip
  • 1798 posts

Posted 03 January 2018 - 10:20 PM

the shadows on those guards in the video are just unrealistically wrong, must be some type of projected stencil, that doesn't work with distances, the guard holding the torch his head's shadow would have burred to an unrecognisable shape on the wall behind him across the street that's if the torch has a 360 bulb, but it looks as if it doesn't so there wouldn't be a shadow behind him. the other guard doesn't have a torch else it would bleach out his shadow on the wall, nor is one visible in his waving hands.



#33 Abusimplea

Abusimplea

    Member

  • Member
  • PipPip
  • 399 posts

Posted 03 January 2018 - 11:42 PM

Oooh, and doesn't it bother you, that you can stand in the shadows and if there's like a fireplace or something bright behind you, the AI can't see you? I mean it's unrealistic too, right?

It bothers me. And as AI already has to somehow do a line of sight check against the player, the cast ray might get continued through the player until it hits some geoemetry. AI could take into account the difference in brightness and color between player and occluded surface when calculating player's visibility.

 

 

I'm not sure most players would find it "immersive" to be caught because of situations that they can't control.

Players control light occlusion and therefore shadow-casting just as they control surface noisyness: They see the threat and act accordingly. Lights obviously are the most visible thing in the game. Light occlusion and shadowcasting are well-known things. People experience that in the meat space all the time. They just never have the need to avoid occlusing lights and casting suspicious shadows there. But if they know, that they have to consider that in the game, they should not have too much problems doing that. As any difficulty-altering thing, it would be optional anyway. So casual players would not get griefed by the slightly increased difficulty.

I am pretty sure, ghosters would love it.



#34 Anderson

Anderson

    Advanced Member

  • Member
  • PipPipPip
  • 1015 posts

Posted 04 January 2018 - 12:04 AM

 

As I said, the problem isn't with animation. As perfect as it can be, its weakest link will always be player input. Silly key presses or thumbstick tilting will always break the immersion this way. The only solution I know of would be kind of "animation queue", as in Dark Souls, where you will be punished if you randomly spam movement keys or attacks. But that's very specific kind of gameplay too.

 

I would phrase it otherwise. There just need to be more animations for ladder/rope climbing depending on the positioning of the PC (player character). It's usually expensive as in the Batman Arkham Asylum/City games.

That's the bottom line. Animators that have the time/skill/will to take it on. Otherwise no point to create player shadows and 3'rd person view.

 

But really, you don't need many. It's quality over quantity. I liked the attempt made here in regards to combat, even if it's in first person, just for the sake of demonstration. The demo is available for free on Steam by the way: https://youtu.be/i3feRkMZMRA?t=15m15s


Edited by Anderson, 04 January 2018 - 12:21 AM.

  • Judith likes this

 "I really perceive that vanity about which most men merely prate — the vanity of the human or temporal life. I live continually in a reverie of the future. I have no faith in human perfectibility. I think that human exertion will have no appreciable effect upon humanity. Man is now only more active — not more happy — nor more wise, than he was 6000 years ago. The result will never vary — and to suppose that it will, is to suppose that the foregone man has lived in vain — that the foregone time is but the rudiment of the future — that the myriads who have perished have not been upon equal footing with ourselves — nor are we with our posterity. I cannot agree to lose sight of man the individual, in man the mass."...

 

 

- 2 July 1844 letter to James Russell Lowell from Edgar Allan Poe.

 


#35 Judith

Judith

    Advanced Member

  • Member
  • PipPipPip
  • 1322 posts

Posted 04 January 2018 - 02:52 AM

 

It bothers me. And as AI already has to somehow do a line of sight check against the player, the cast ray might get continued through the player until it hits some geoemetry. AI could take into account the difference in brightness and color between player and occluded surface when calculating player's visibility.

 

Now that was sarcastic, and I didn't think anyone would take this idea seriously, as it would make the game literally unplayable. With effective FOV of 80 to 90, player has no means to be aware what's behind him all the time, and he would be punished for something he didn't know he did. But that's also what illustrates general issue with ideas: everybody's having them. But knowing what would translate into meaningful gameplay is something that requires some more thought, or just testing experience. The core mechanics doesn't have to be super intricate or complicated to be interesting. Simple elegance and consistency works much, much better, and people will suspend their disbelief anyway. We're talking about a guy who's invisible in the shadows, what's realistic is unimportant. What's interesting and meaningful is.


Edited by Judith, 04 January 2018 - 03:28 AM.


#36 Abusimplea

Abusimplea

    Member

  • Member
  • PipPip
  • 399 posts

Posted 04 January 2018 - 05:24 AM

Now that was sarcastic, and I didn't think anyone would take this idea seriously, as it would make the game literally unplayable. With effective FOV of 80 to 90, player has no means to be aware what's behind him all the time, and he would be punished for something he didn't know he did. But that's also what illustrates general issue with ideas: everybody's having them. But knowing what would translate into meaningful gameplay is something that requires some more thought, or just testing experience. The core mechanics doesn't have to be super intricate or complicated to be interesting. Simple elegance and consistency works much, much better, and people will suspend their disbelief anyway. We're talking about a guy who's invisible in the shadows, what's realistic is unimportant. What's interesting and meaningful is.

I don't think, it would make the game unplayable at all. It would make most missions only slightly harder to play. A few missions would indeed become a nightmare. But most would keep beeing pretty playable. I am in no way a ghost and not trying to brag or something like that. But i completed most TDM missions on hardest and most of them without using consumeables. There is some room for more environmental-awareness-based difficulties in TDM.

 

FOV is not that much of an issue because you indeed can "turn your head" while walking using mouse and keybord. I always crouch and go sideways or backwards as the situation calls for it - it works pretty well in TDM missions with their mostly surprise-free environments. Not that far a way from scanning locations for their shadows to also look for high contrast surfaces to avoid while at it.

And such a mechanic would obviously be off (or tuned down to irrelevancy) in easy mode anyway. So it would not scare away any new players.

 

You might be sarcastic about it or not. You might belittle me for beeing a mere player since 25 years.

But evaluating the contrast between player and occluded surfaces for adjusting the effective visibility of the player is a pretty feasible idea gameplay-wise, most likely easier to implement than detection of cast shadows, and would spice up the hide in shadows mechanic a bit (yes, TDM already does that better than any other current stealth game out there - but there still is room for improvements).

Sadly, we probably don't get any of the two anyway (because dev time is scarce too).



#37 Judith

Judith

    Advanced Member

  • Member
  • PipPipPip
  • 1322 posts

Posted 04 January 2018 - 05:38 AM

FOV is not that much of an issue because you indeed can "turn your head" while walking using mouse and keybord.

 

The question is whether turning your head back evey few seconds like in a paranoid mental state is in any way fun. It isn't. Does it contribute to player character or the world in any menaingful way? No it doesn't. Main character is (typically) a master thief, burglar, infiltrator, spy, what have you. He comes prepared, he is in control, he has tools and abilities to do the job. Stealth gameplay is about gaining control in hostile territory.

 

You might belittle me for beeing a mere player since 25 years.

 

I'm not trying to belittle you, and I'm a player since early 90's. But I'm also interested in design, technical side of things, and video game language evolution. It really helps to sort out better ideas from worse, and to think about other players and play styles than yours (which is typically very limited).



#38 Abusimplea

Abusimplea

    Member

  • Member
  • PipPip
  • 399 posts

Posted 04 January 2018 - 08:21 AM

The question is whether turning your head back evey few seconds like in a paranoid mental state is in any way fun. It isn't. Does it contribute to player character or the world in any menaingful way? No it doesn't. Main character is (typically) a master thief, burglar, infiltrator, spy, what have you. He comes prepared, he is in control, he has tools and abilities to do the job. Stealth gameplay is about gaining control in hostile territory.

Don't forgetm, that in most situations we also get sound cues from our environment. even though that sound cues are a bit broken in a lot of missions due to room sealing issues, you still get the rough distance of nearby AI. And if you are in a noisy factory with multiple entrances, you really should be paranoid - especially as a master thief.
It is a bit like with sniper mechanics in other games. In some you just point at the head and score the headshot. Other games (or same on harder dificulty levels) require you to consider distance and wind for scoring the headshot. I get that most players don't want the more sophisticated simulations - but there are tons of games for them out there already.
Sure, more sophisticated simulation of lighting and detectuion is a niche - as is TDM already.
 
As already mentioned: Tools are not even needed for getting through most missions. Player experience matters a lot. But after so many years of TDM, you know where to hide, how AI reacts to what situations, what patrol patterns are most often used and how to douse a guard's torch without him noticing...

 

But i see, neither of us will convince the other:
I will not convince you, that an optional more sophisticated and slightly more difficult to fully master visibility-calculation would be a good thing.
And you surely will not convince me of the opposite.
 

It really helps to sort out better ideas from worse, and to think about other players and play styles than yours (which is typically very limited).

Yeah, surely i could not have thought about the impact on the player base. And of course i am just spilling out ideas as they come without any sort of prefiltering, because only you would be able to do that. Really no belittling there at all...



#39 Judith

Judith

    Advanced Member

  • Member
  • PipPipPip
  • 1322 posts

Posted 04 January 2018 - 08:39 AM

 

Yeah, surely i could not have thought about the impact on the player base. And of course i am just spilling out ideas as they come without any sort of prefiltering, because only you would be able to do that. Really no belittling there at all...

 

Maybe there's something wrong with my tone, but I don't mean that at all.

 

But it's true that I'm somewhat tired of niche or nitpicky, peculiar kind of ideas, that will make a niche game even more obscure, and will make level design even harder (both options do). Nobody says TDM has to follow fit-for-all Dishonored path, but missions with really elegant and conscious gameplay design are still a rarity (btw. thank the Builder for folks like Kingsal).



#40 HMart

HMart

    Advanced Member

  • Member
  • PipPipPip
  • 639 posts

Posted 04 January 2018 - 09:20 AM

Abusimple Judith is right, that idea of "back lighting" was already discussed before by the TDM team and was decided that it was more problematic then its worth, it wouldn't only cause a nightmare to mission designers trying to balance difficulty on their missions, it would demand from players an unrealistic amount of attention to their environment, causing many to be frustrated by constantly being caught without knowing why.  TDM is not a real life robbery simulator is a game pure and simple, the TDM team has to have in consideration fun (and being constantly caught and repeating is not fun) and for that some reality has to go through the door. In the end it would be one more AI feature taking value CPU time and performance down for little gain on gameplay and increase on frustration potential. 


  • Judith and CarltonTroisi like this

#41 Springheel

Springheel

    Creative Director (retired)

  • Admin
  • 37246 posts

Posted 04 January 2018 - 12:42 PM

Now that was sarcastic, and I didn't think anyone would take this idea seriously

 

 

 

 

It's actually a pretty commonly-requested feature.  One of the earlier discussions on the subject:

 

http://forums.thedar...of-see-ability/

 

and a more recent one:

 

http://forums.thedar...earby-movement/

 

My opinion is summed up thusly:

The technical hurdles in implementing such a system aside, I have no idea what this would achieve, other than forcing the player to become aware of his surroundings in a fashion that is unrealistically awkward in an artificial 3d environment.

You've walked down a long hallway, slipping from shadow to shadow, when a guard at the end of the hallway turns to look your direction. Are you safe or do you have to move?

Our system: Check your lightgem--oh crap, you're on the edge of detection, better crouch.

This system: Check your lightgem, then wonder whether somewhere behind you down the hall is a lit area that happens to be directly behind your head. Swing the mouse to turn and look around behind you--there are a few bright spots, but now you can't see exactly where the guard is standing, so swing the mouse to look the other way again to try and line it up as he walks your direction....will crouching line your head up with a bright spot on the floor behind you? Better turn around and check.....etc.

 


  • Judith and HMart like this
TDM Missions:   A Score to Settle   *   A Reputation to Uphold   *   A New Job   *    A Matter of Hours
 
Video Series:   Springheel's Modules   *   Speedbuild Challenge   *   New Mappers Workshop  *   Building Traps

#42 Abusimplea

Abusimplea

    Member

  • Member
  • PipPip
  • 399 posts

Posted 05 January 2018 - 11:27 AM

The lightgem's brightness would have to represent maximum current player visibility for any AI instead of brightness of the players surface only. It could fall back to the current behaviour of showing the player's illumination if no AI has a line of sight and its vision cone pointed in the general direction of the player.

As the AI-specific visibilities would have to be calculated anyway, it would not be hard to use the maximum of them and the general player illumination for the lightgem's brightness.

 

So i think, the lightgem is a strawman. The needed adjustments on the lightgem would be easy and cheap.


  • Anderson likes this

#43 Anderson

Anderson

    Advanced Member

  • Member
  • PipPipPip
  • 1015 posts

Posted 05 January 2018 - 11:54 AM

The lightgem's brightness would have to represent maximum current player visibility for any AI instead of brightness of the players surface only. It could fall back to the current behaviour of showing the player's illumination if no AI has a line of sight and its vision cone pointed in the general direction of the player.

As the AI-specific visibilities would have to be calculated anyway, it would not be hard to use the maximum of them and the general player illumination for the lightgem's brightness.

 

So i think, the lightgem is a strawman. The needed adjustments on the lightgem would be easy and cheap.

 

With all that in consideration, without an animator, none of this has any rationality to be implemented ad hoc. Shadows reflecting sluggish animations is not an option.


 "I really perceive that vanity about which most men merely prate — the vanity of the human or temporal life. I live continually in a reverie of the future. I have no faith in human perfectibility. I think that human exertion will have no appreciable effect upon humanity. Man is now only more active — not more happy — nor more wise, than he was 6000 years ago. The result will never vary — and to suppose that it will, is to suppose that the foregone man has lived in vain — that the foregone time is but the rudiment of the future — that the myriads who have perished have not been upon equal footing with ourselves — nor are we with our posterity. I cannot agree to lose sight of man the individual, in man the mass."...

 

 

- 2 July 1844 letter to James Russell Lowell from Edgar Allan Poe.

 


#44 Springheel

Springheel

    Creative Director (retired)

  • Admin
  • 37246 posts

Posted 05 January 2018 - 11:55 AM


 

So i think, the lightgem is a strawman.

 

Am I misrepresenting someone's position?
 

 

The needed adjustments on the lightgem would be easy and cheap.

 

 

So the lightgem will change brightness moment by moment even if the player isn't doing anything?  An AI turns his head and suddenly the lightgem spikes.  He turns his head a different way and it plunges into darkness.  That would completely destroy the player's ability to know whether they are hidden or not, or why, or whether they should do anything about it. 

 

Oh phew, your lightgem is dark, you must be safe.  Oops, an AI just turned his head and now it's bright!  Do you move?  Crouch?  Wait for the AI to move?

 

The lightgem because useless at that point.  Everything becomes trial and error.

 

it would not be hard to use the maximum of them and the general player illumination for the lightgem's brightness.

 

 

It's fun to watch people make claims about what wouldn't be hard to do.

 

Just for fun, would you like to lay out exactly how this would work?  How do you propose to get a numerical value from all the different AI in the room and apply it to the lightgem?


TDM Missions:   A Score to Settle   *   A Reputation to Uphold   *   A New Job   *    A Matter of Hours
 
Video Series:   Springheel's Modules   *   Speedbuild Challenge   *   New Mappers Workshop  *   Building Traps

#45 Anderson

Anderson

    Advanced Member

  • Member
  • PipPipPip
  • 1015 posts

Posted 05 January 2018 - 12:12 PM

 

 

 

So the lightgem will change brightness moment by moment even if the player isn't doing anything? 

He didn't say that.


 "I really perceive that vanity about which most men merely prate — the vanity of the human or temporal life. I live continually in a reverie of the future. I have no faith in human perfectibility. I think that human exertion will have no appreciable effect upon humanity. Man is now only more active — not more happy — nor more wise, than he was 6000 years ago. The result will never vary — and to suppose that it will, is to suppose that the foregone man has lived in vain — that the foregone time is but the rudiment of the future — that the myriads who have perished have not been upon equal footing with ourselves — nor are we with our posterity. I cannot agree to lose sight of man the individual, in man the mass."...

 

 

- 2 July 1844 letter to James Russell Lowell from Edgar Allan Poe.

 


#46 Springheel

Springheel

    Creative Director (retired)

  • Admin
  • 37246 posts

Posted 05 January 2018 - 01:00 PM

He didn't say that.

 

That is the clear consequence of "The lightgem's brightness would have to represent maximum current player visibility for any AI".


TDM Missions:   A Score to Settle   *   A Reputation to Uphold   *   A New Job   *    A Matter of Hours
 
Video Series:   Springheel's Modules   *   Speedbuild Challenge   *   New Mappers Workshop  *   Building Traps

#47 Abusimplea

Abusimplea

    Member

  • Member
  • PipPip
  • 399 posts

Posted 05 January 2018 - 01:45 PM

With all that in consideration, without an animator, none of this has any rationality to be implemented ad hoc. Shadows reflecting sluggish animations is not an option.

Player-occluded surfaces affecting player visibility would not require player-cast shadows. So animations or the lack of them do not matter for that.
 

So the lightgem will change brightness moment by moment even if the player isn't doing anything?  An AI turns his head and suddenly the lightgem spikes.  He turns his head a different way and it plunges into darkness.  That would completely destroy the player's ability to know whether they are hidden or not, or why, or whether they should do anything about it.

Yes, it would change its brightness depending on AI locations. As AI position would affect visibility and the lightgem indicates visibility, it would have to change brightness depending on AI locations. The vision cones should probably not taken into account though. I erred on that one.
 

The lightgem because useless at that point.  Everything becomes trial and error.

The lightgem would still give you a correct estimation of how long it takes until the AI starts believing, that it could have seen something (needs player experience to take distance into account with and without the new feature).
 

It's fun to watch people make claims about what wouldn't be hard to do.

Just to be clear: Calculating the player visibilities for each AI would probably involve casting rays. So that would not be cheap and probably not easy to implement. I was talking about using the max of the resulting values that would have been stored in the AI state just like all the other stuff (like alertstate for example).
 

Just for fun, would you like to lay out exactly how this would work?  How do you propose to get a numerical value from all the different AI in the room and apply it to the lightgem?

 
Pseudocode:

lightGemBrightness = player illumination // that is what we have now.
foreach ai in listOfAis: // There surely is some list of AI in the engine already.
  aiVisibility = getPlayerVisibilityFromAiState(ai)
  lightGemBrightness = max(lightGemBrightness, aiVisibility)


#48 wesp5

wesp5

    Advanced Member

  • Member
  • PipPipPip
  • 620 posts

Posted 05 January 2018 - 01:50 PM

 

That is the clear consequence of "The lightgem's brightness would have to represent maximum current player visibility for any AI".

I know it probably has the much inferior stealth system overall, but VTM: Bloodlines has two stealth indicators: A bar representing the darkness of the player location and a number on top showing the probabilty of enemies to detect the player. Judging from left-over icon graphics in the game files the probability number was added later in development on top of the brightness indicator, probably because Troika noticed that a simple combination won't work. On that I agree with Springheel, you'll need a halo around the lightgem or similar for a seperate AI indicator!


Edited by wesp5, 05 January 2018 - 02:01 PM.

  • Anderson likes this

#49 Springheel

Springheel

    Creative Director (retired)

  • Admin
  • 37246 posts

Posted 05 January 2018 - 04:06 PM

Just to be clear: Calculating the player visibilities for each AI would probably involve casting rays. So that would not be cheap and probably not easy to implement. I was talking about using the max of the resulting values that would have been stored in the AI state just like all the other stuff (like alertstate for example).
 

 
Pseudocode:

lightGemBrightness = player illumination // that is what we have now.
foreach ai in listOfAis: // There surely is some list of AI in the engine already.
  aiVisibility = getPlayerVisibilityFromAiState(ai)
  lightGemBrightness = max(lightGemBrightness, aiVisibility)

 

That glosses over how it would actually work.  How do you establish whether the player is silhouetted against a bright surface?  How much of the player has to occlude a bright surface?  How bright does the surface have to be?  How many traces would you use and what numerical values would you give to the difference between the player and the surface behind them?  How far does the trace go and is there any falloff over that distance?  How would you translate this info into a numerical value, and how would it relate to the numerical value already used by the lightgem?

 

For every one of those questions there are several hurdles to clear in making it work correctly, and that's not even considering the performance implications.


  • Anderson likes this
TDM Missions:   A Score to Settle   *   A Reputation to Uphold   *   A New Job   *    A Matter of Hours
 
Video Series:   Springheel's Modules   *   Speedbuild Challenge   *   New Mappers Workshop  *   Building Traps




0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users