Some small(ish) tweaks that I think would really help immersion
sickboy
Join Date: 2003-09-13 Member: 20804Members, Constellation
Join Date: 2003-09-13 Member: 20804Members, Constellation
You guys may already be planning all this, or chosen not to do it, but this game has potential to be insanely immersive and I want to help in any way possible to achieve that. These are a few things I've been brainstorming:
- When fish (or school of fish) and the player get close to sand, a dust cloud (size and severity based on size and speed of the impacting creature) would show. Temporary sand displacement (even if faked by just spawning lumps) would be even neater for large creatures (and for player 'dusting' of an area). The cloud would stick around for awhile, and be quite difficult to see through if large/severe enough... similar to Lerk gas in (early) NS2.
- When larger fish or the player goes through applicable vegetation (grasses, bushes, etc), it would move. Plenty of games have done this, where the grass/etc. kind of pushes out of the way slightly. A recent Early Access game called Hardland does this in spades, to great effect. The current effect on the purple mushrooms is amazing and a fantastic step in the right direction, but grasses could really use some nudgeability.
- When jumping into the water, a huge bubble cloud would show around the player (and in their face). This would be an actual bubble cloud entity, that the player could quickly swim out of and see when they turn around. An ascending bubble cloud from each breath would be a very nice touch as well.
- A chance for a fish to get 'scared' and move off to another area quickly (though not terribly far unless being chased) when any other creature (even if the same species) is nearby would add for a nice variety in how they interact with one another. In that same vein, a random chance to imply aggression of some sort by chasing another creature could lead to some interesting chain effects... especially if how long something will be chased, direction to flee (and when to change direction), etc. is also randomly decided. Of course, tiny fish charging and scaring a much larger one would have to be disabled for logic's sake.
- Fish moving vertically a bit more often. They tend to stay at the same depth currently, but moving up or down more often would give some variety.
- Fish moving to close-but-different areas every now and then. Most of their current patterns keep them floating around the same area, which is fine. But an eventual 'move 40-50 meters away before continuing to wander around the same area' addition to the pattern would keep the area changing while still allowing for more intentional biome-specific locations. A kind of "graze in the general area, then move to a different patch, then repeat" pattern, I suppose. Every now and then a fish deciding to just go someplace completely different would be nice too (so a temporary usage of the ray's more adventurous AI).
- Removal of the reticule, and object selection becoming UI-less. A hand reaching out, the object reacting in some way, etc. would be more than enough to inform the player the object is able to be used/picked up. The text showing right in the middle of the screen obscures your ability to actually look at stuff up close, making some of those awesome creatures and plants impossible to properly inspect. Quite frankly, I'd prefer to see no UI whatsoever. Everything could be moved to be displayed on game items rather than floating in the player's face.
- Removal of the mask would be delayed a bit, and never be taken off while the player body is in the water (unless manual removal ends up being a feature). This would prevent the constant on/off when bobbing on the surface.
- A player shadow. Right now it doesn't feel like the player is very physical. Since the body can be seen and the fish all give a shadow, one for the player would give a lot of impact to the sense of being part of the world.
- As mentioned elsewhere, harder and darker shadows for environments... also darkening the mask and player body. Day/night cycle, etc... this has already been addressed, I'm just listing it here to keep it all together.
- The light distortion from the waves would show on the tops of fish and the player body, with darker undersides. This is mostly applicable to the larger and/or flatter creatures.
Now, the biggest change (which isn't a small tweak... not that the others were either :P) to really enhance the feel of the fish is their turn pathing and animations. Right now, they turn very game-like, instead of more gradually ("jetpack-in-zero-gravity-like") and with proper bends in their physical structures. This is particularly noticeable with the larger fish, as well as the schools of smaller ones, but pretty much all of them have the same issues to one degree or another.
They do that weird 'statue spinning in place' thing that so many games out there have. Their turns should be matching their body movements almost exactly, with just a bit of floaty feeling unless they're making a hard & fast turn. Fish have amazing control underwater, after all, but their body has to match what they're doing since they're essentially flying without gravity. Animation being off from the movement unfortunately kills the immersion in a split second.
The smaller and faster should feel like a swarm of sparrows or humming birds. The larger and slower like big, laborious zeppelins turning in the sky. Those in between should be like varying degrees of lazily gliding birds until stimuli (or default behavior) has them dart in a different direction. Dart, glide, dart, glide.
The schools of smaller fish have an additional, unique issue in that they don't turn as a group very logically. If a change in direction occurred, there would be new 'leaders' of the swarm, they wouldn't slow down to wait for the previous 'leaders' to come back around. All the fish individually turn on a dime, constantly changing who is in 'front' or 'back'. The exception to this would be when stimuli causes the form of the swarm to sort of turn inside-out (leading to more of a doughnut shape initially, then forming back into a typical swarm).
Unrelated note: that vid's mix of shark + dolphin + penguin attacks were awesome. If you guys end up adding arctic areas, seeing divebombing penguin-like creatures and those kind of bubble trails would be insaaaane!
Anyways, just a few thoughts. Sorry for babbling your ear off.
- When fish (or school of fish) and the player get close to sand, a dust cloud (size and severity based on size and speed of the impacting creature) would show. Temporary sand displacement (even if faked by just spawning lumps) would be even neater for large creatures (and for player 'dusting' of an area). The cloud would stick around for awhile, and be quite difficult to see through if large/severe enough... similar to Lerk gas in (early) NS2.
- When larger fish or the player goes through applicable vegetation (grasses, bushes, etc), it would move. Plenty of games have done this, where the grass/etc. kind of pushes out of the way slightly. A recent Early Access game called Hardland does this in spades, to great effect. The current effect on the purple mushrooms is amazing and a fantastic step in the right direction, but grasses could really use some nudgeability.
- When jumping into the water, a huge bubble cloud would show around the player (and in their face). This would be an actual bubble cloud entity, that the player could quickly swim out of and see when they turn around. An ascending bubble cloud from each breath would be a very nice touch as well.
- A chance for a fish to get 'scared' and move off to another area quickly (though not terribly far unless being chased) when any other creature (even if the same species) is nearby would add for a nice variety in how they interact with one another. In that same vein, a random chance to imply aggression of some sort by chasing another creature could lead to some interesting chain effects... especially if how long something will be chased, direction to flee (and when to change direction), etc. is also randomly decided. Of course, tiny fish charging and scaring a much larger one would have to be disabled for logic's sake.
- Fish moving vertically a bit more often. They tend to stay at the same depth currently, but moving up or down more often would give some variety.
- Fish moving to close-but-different areas every now and then. Most of their current patterns keep them floating around the same area, which is fine. But an eventual 'move 40-50 meters away before continuing to wander around the same area' addition to the pattern would keep the area changing while still allowing for more intentional biome-specific locations. A kind of "graze in the general area, then move to a different patch, then repeat" pattern, I suppose. Every now and then a fish deciding to just go someplace completely different would be nice too (so a temporary usage of the ray's more adventurous AI).
- Removal of the reticule, and object selection becoming UI-less. A hand reaching out, the object reacting in some way, etc. would be more than enough to inform the player the object is able to be used/picked up. The text showing right in the middle of the screen obscures your ability to actually look at stuff up close, making some of those awesome creatures and plants impossible to properly inspect. Quite frankly, I'd prefer to see no UI whatsoever. Everything could be moved to be displayed on game items rather than floating in the player's face.
- Removal of the mask would be delayed a bit, and never be taken off while the player body is in the water (unless manual removal ends up being a feature). This would prevent the constant on/off when bobbing on the surface.
- A player shadow. Right now it doesn't feel like the player is very physical. Since the body can be seen and the fish all give a shadow, one for the player would give a lot of impact to the sense of being part of the world.
- As mentioned elsewhere, harder and darker shadows for environments... also darkening the mask and player body. Day/night cycle, etc... this has already been addressed, I'm just listing it here to keep it all together.
- The light distortion from the waves would show on the tops of fish and the player body, with darker undersides. This is mostly applicable to the larger and/or flatter creatures.
Now, the biggest change (which isn't a small tweak... not that the others were either :P) to really enhance the feel of the fish is their turn pathing and animations. Right now, they turn very game-like, instead of more gradually ("jetpack-in-zero-gravity-like") and with proper bends in their physical structures. This is particularly noticeable with the larger fish, as well as the schools of smaller ones, but pretty much all of them have the same issues to one degree or another.
They do that weird 'statue spinning in place' thing that so many games out there have. Their turns should be matching their body movements almost exactly, with just a bit of floaty feeling unless they're making a hard & fast turn. Fish have amazing control underwater, after all, but their body has to match what they're doing since they're essentially flying without gravity. Animation being off from the movement unfortunately kills the immersion in a split second.
The smaller and faster should feel like a swarm of sparrows or humming birds. The larger and slower like big, laborious zeppelins turning in the sky. Those in between should be like varying degrees of lazily gliding birds until stimuli (or default behavior) has them dart in a different direction. Dart, glide, dart, glide.
The schools of smaller fish have an additional, unique issue in that they don't turn as a group very logically. If a change in direction occurred, there would be new 'leaders' of the swarm, they wouldn't slow down to wait for the previous 'leaders' to come back around. All the fish individually turn on a dime, constantly changing who is in 'front' or 'back'. The exception to this would be when stimuli causes the form of the swarm to sort of turn inside-out (leading to more of a doughnut shape initially, then forming back into a typical swarm).
Unrelated note: that vid's mix of shark + dolphin + penguin attacks were awesome. If you guys end up adding arctic areas, seeing divebombing penguin-like creatures and those kind of bubble trails would be insaaaane!
Anyways, just a few thoughts. Sorry for babbling your ear off.
Comments
Neat idea. Will have to see if performance allows us to do that.
We already have the mushrooms reacting. That's pretty cool I'd say. I don't think we can do much about the bushes for animation reasons and I'm pretty sure we can't do much about grass for technical reasons.
I totally dig the bubble clouds of the birds splashing into the Sardine swarm in that video you linked. The same would be true for player bubbles. I'm pretty sure we can add those easily.
This will be coming for sure. Scared, aggressive, and curious behaviors should add a lot to the game feel.
Not sure what you mean. Swimming up/down head first or slowly floating vertically? We have the former, most obviously seen in Stalker and Crash. Not sure how the latter would improve immersion.
As you have noticed each creature is currently confined to a very limited area. We did this because creatures tended to vanish into the blue before essentially leaving the world empty. That didn't feel good. Perhaps we overshot a bit and now the creatures are too bound. The behaviors mentioned above should loosen things up and a graze behavior would certainly be nice too.
This one is very debatable. We tried having no reticule and it was difficult to pick up stuff, especially moving creatures. A HUD is a very easy and clear way to communicate essential information to the player. That's why it is popular in games. Of course being in earliest access stage there is plenty of room for UI improvement.
Agreed.
Day/night cycle will come.
Probably not for performance reasons.
I totally agree. I want better creature movement. In robots and games we call this locomotion. As I said, I as a person find this very important. But it will only become important for us as developers if people like you lobby for it. Thank you for stating this and explaining in a very clear manner why it matters.
Just examples to illustrate my point. I think this would allow the player to study animal behavior to learn when and where its best to catch certain creatures. But mainly, I feel this would make the world feel really alive.
Oh, also if you are feeling crazy you can make it so they have behavior rules instead of strict scripted behavior. Like, peepers like mushrooms so they hang out near mushrooms. You pick the mushrooms, and the peepers leave. Plant mushrooms, and the peepers show up.
I definitely hope so. I think that a sand cloud trail would really give the fish and player a sense of being part of the world. Plus it could be an interesting way to fake fish eating stuff. Seeing one dart to the ground and a big plume go up would be cool, as well as be a way to mask that nothing is actually happening with the fish aside from a simple pathing waypoint and temporary velocity increase. Implied action and all that. AKA: magic.
I'm not very knowledgeable about animation (was never really my forte), but would a slight angling in the vertical axis of each grass/etc. entity (with the same animations still going) via code triggered by hitboxes not be possible? That is pretty much the extent of what I was suggesting, in case I gave the impression of full physics and individual grass interaction.
Oh and yes, the mushrooms are extremely cool! They are one of my favorite little details in the game currently.
Mostly what I meant is that quite a few of the fish species in-game don't swim up or down head first very often. I know a few do, I suppose I just expected others to do it as well (and more often unless simply traveling from point a to point b). I suppose the exception is bottom-feeders or slower creatures (like whales) that can't quickly go up or down even if they wanted, but even those will take the occasional vertical path.
Completely unrelated note: bottom-feeders would be a very neat addition to the game as well, especially if sand clouds were implemented.
From games I've seen, there definitely has to be some kind of feedback for a UI-less system to work... but it can definitely work and to huge benefit (particularly notable in atmospheric horror/exploration games). Some have a hand move towards the object, some make the object glow, some make it wobble or change color... feedback is definitely necessary, but I really think a game like this doesn't need to have a UI and ultimately suffers from having one. Competitive or data-heavy games, sure (though I'm still of the school of thought that it should be up to the player to figure it all out as part of the learning process)... but not one with so much potential for a focus in exploration.
Yes, a UI can convey a quick and easy message to a player... but it sacrifices massive amounts of the immersion factor (a huge selling point for a game like Subnautica). Even without the UI the player will figure out what is going on extremely quickly so long as there is adequate feedback and logic to the actions. There are some pretty amazing psychology studies out there about non-humans and humans alike being given completely foreign tools/puzzles/etc. and how quickly and naturally they figure them out. If crows can exploit mankind's expansion and technology, and small children can figure out they need to punch trees in Minecraft without the aide of a wiki, surely adults will be able to figure out basic interaction in Subnautica. Just a matter of natural-feeling feedback to nudge some in the right direction.
That's a huge shame if it doesn't work out. I do taxidermy for the local zoo so I've gotten to do some pretty interesting lighting/etc. setups in order to create a particular scene, but when I went to the Denver National History Museum recently the exhibit that blew my freakin' mind was the manatees. They had simple lights to put a moving version of the typical light pattern from waves onto the backs of the animals and it was an amazing effect. It truly looked like they were under water just from a single light.
Is there no way to fake it on the backs of the larger creatures at least? I didn't mean to use actual dynamic lighting in my suggestion, I just meant doing something like popping an overlayed texture of light patterns onto anything not currently in a shadow/cave/lightlessdepth, and having it globally "fixed" (in the sense of CSS 'fixed', or parallax) to assist with the illusion when the creature moves.
Unfortunately, I don't think lobbying will do much to sway a significant enough number in the gaming populace these days. Gaming got popular, and with it came a lot of people that don't really see or care about the magic that the fine details can provide. Those of us that do aren't typically the very vocal or pushy types because our passion for it made us outcasts on our own home turf when the cool kids came into town. (Do I sound like a bitter old man or what?)
Sadly that means the focus ends up elsewhere, though I can't say I fault the logic. A business has to make money, people have to pay the bills, and the most vocal and seemingly popular notions from the public will be that those details don't matter (or flat out aren't even thought about). The rock-and-a-hard-place of a creator is a rough one for sure, I don't envy having to make the decisions you guys do.
Just know that there are plenty of us that love and appreciate those kinds of things, we just aren't very chatty (and when we are, beer was probably involved at some point).
--The mask popping on and off at the surface is just a bug, not intended gameplay feature. It is only intended to come off when the player is on a Submarine / Life pod, not at the surface of the water. This didn't used to happen, it's only a recent issue, so we'll need to look into it. Stuff that gets a bit trickier - does the mask come off when walking on an island, and at what point, and how does it know when you are on land as opposed to just a bit above the water surface causing the same kind of issues with the mask popping on and off?
--Caustics on creatures and view model is definitely something I've been wanting and have been planning on following up with our FX guy about. I think maybe LumpN was thinking of a more complex lighting system that would be bad on performance, but I'm pretty sure we can do this fairly cheaply with just a simple shader effect. But, don't quote me on that
--I'm pretty sure you don't need to lobby for better creature movement, it is bugging us all, and we know it is something we need to address. It just comes down to priorities, and at the moment we are focused on getting in the much higher level gameplay stuff, so that there is a lot more meat to the game, and once we have more of that in place we can start thinking about visual polish like the spinning in place and instant turning and other issues.