The pirate game has not been abandoned! Let me know if you've got any ideas for how to fill in the gaps in the water fog mask between the horizon and the water meshes 👇
Why do you have to use the horizon as a base? Why don't you just render the top of the water plane as white (like you're doing) and then fill everything above that plane with white? Like a flood fill. It seems like the water plane should always make at least a solid boundary between the top and bottom of the screen. So just fill anything above that boundary. In a pixel shader this would just mean that, starting from the top of the image, you would set every pixel in every column (not row) to white until you reached an already white pixel. Since the water is height based there should never be overhangs that have to be filled.
@@joshuafhiggins and you mean have the low LOD versions go all the way to the horizon? I considered that as well, and it could work, but how far those low LOD planes have to extend into the distance would depend on the height of the camera _and_ potentially on the resolution of the display.
@@nathangeorge4017 this was something that crossed my mind, but I honestly just wasn't sure how to approach the flood filling. The fragment/pixel function of a shader operates on each pixel individually, with no information about other pixels. Obviously I could sample the mask texture, but (on 1920x1080 resolution, with about half the screen submerged) that would mean sampling the texture at least ~540 times for each of the ~540 pixels in each of the 1920 columns-unless I'm missing something? That means sampling the fog mask texture ~559 _million_ times every frame, which doesn't sound very efficient to me...
I honestly thing that these reflections are the best improvement you have made visually to the game, aside from maybe the water flattening around the islands. The game is coming along really well now!
Agreed! I think your last statement needs to be corrected though-the _graphics_ are coming along nicely. The _game_ is still as unplayable as it was a year ago :P
I can see from the comments that you're now a servitor to the machine god, 8 hours a day. That's okay man, never think it's 'been too long' to continue your project. You will get used to it, you will upskill in certain areas, and your work, whatever it is, will help you with the game in one way or another. You've got this man! And if your priority is now a career rather than making the game, I wish you good luck on that too. But I really don't think it is. I think you and me are a lot alike, I didn't go to university for the same reasons as you. Well anyway, I've been fucking off on work hours for 15 mins now, so I'll get back to that but just know, you are not alone. There's a lot of people like you, full of passion, working 8 hours a day and MAYBE getting in half an hour every 2 days on their project. It's the consistency that counts. See you at the top.
Thank you for the kind words! These are the kind of comments which make me want to come back to this project, but I'm still trying to figure out where and in what capacity it fits into my life now that I work full time 😅
I think going SSR on reflections would give much better results. Since camera is barely above the reflective surface, it'll be very close to ground truth too. This will also solve (or rather) help with the double lighting issue. With the added bonus of in reflection lighting being completely free.
The problem is that nothing that is off screen can be reflected, which means tops of boats and islands will be cut off in reflections if you don't have them in view. As for the camera barely being above the reflective surface, I'm not really sure what you mean since a boat's crows nest or the top of even a small island is more than "barely" above the water surface, is it not?
@@tomweiland yeah cut off makes sense. I used "Barely above surface" as in SSR only works well when viewing angle of the reflective surface being almost parallel to the camera, going high up and looking directly down breaks the effect. But I don't think that's a realistic thing to happen in this game that's all. Your channel is great find btw, thanks for all
Hope you are doing well, your old C# networking tutorial series has helped me a lot, and your pirate game looks great. Glad to see you're still replying to comments after so many months.
Not sure if you olan to stick the the game or not or just busy but even months later i still enjoy watching your videos! Thank you still! Hope you are taking care and living the good life!
I do want to continue working on it and there will be at least one more devlog in the near future, but beyond that I'm not really sure yet. I don't want to live at home indefinitely and this project/these videos don't come even remotely close to making a livable income. Even if I managed to upload once a month, I would have to average _several million_ views per video just to begin approaching a decent salary, and that is not exactly something I can rely on to happen :P Glad you've been enjoying the videos though!
@@tomweiland completely understandable! It's not easy pulling in the views. Ive seen amazing creators get trampled on by UA-cam no matter how often they upload. its a shame! No problem! It just good to see other people's experiences and creativity ^-^
hey tom, I have just gotten to explore your channel today and I must say, I am in love with it, your videos were so high quality and entertaining I am just hooked. I am a fellow programmer also wanting to see where I want to go and I am inspired by your continuous and hard work for this personal project and I am hoping to start one soon, maybe not game dev after (I am gonna take your word :D). Wishing you the best on your journey in 2024 and onwards!
Simple trick for sea reflections would be to use SSR (screen-space reflections), perhaps you could cross-fade between planar reflections when the sea is calm and SSR when it is agitated. Or you could simply use SSR in every case. Another thing that might possibly be done is to displace the reflected image depending on the camera's angle and how heavily displaced the water plane is, using trigonometry. I'm not quite sure this would result in a good looking result but it feels like it could work with some tuning and adjustments. Keep in mind that if you plan to keep the Planar Reflection for agitated sea, whatever you do, the reflections will be physically inaccurate (that goes for SSR too but it would be less noticeable), since it only replicates what a flat surface would reflect. There would also be an issue with some parts of the terrain OVER the reflection plane and UNDER the sea plane being missing in the reflection. Personally I would advise for using SSR as I don't believe any other kind of hack you can conjure is gonna look good enough to be worth the effort. The next best thing is raytracing but we all know why that's not a great idea :).
The unfortunate thing about SSR is that anything that's not visible on screen also can't be reflected, which in some scenarios can look just as bad/weird as the problems I have right now with the reflections on big waves, so I'm not sure it'd be better. With some tweaking it might be possible to get the current setup looking _decent,_ so I'm not particularly eager to abandon that approach just yet :P
@@tomweiland fair enough. I can appreciate you going for a "higher-tier" physically accurate reflection model. Good luck on creating a reflection setup that fits your project :).
Not sure if you still care, but a decent fix for the fog/mesh gap is to make a slightly fuzzy white band along the pixel curve that's wider than the gap; which looks like a pretty decent foam line and covers the crimes.
This might work if my mesh had a higher vertex resolution but because of the distance between vertices the gap between the mesh and where the fog starts would likely be too big for this to work.
I'm assuming this is a joke lol, but just in case: I have not switched to Godot 😅 I'd like to come back to this project eventually, but I haven't really figured out where it fits in with work and other life stuff.
@@tomweiland Ya I was joking haha😅 Anyways looking forward to whatever comes next, I hope I can speak for a large part of the community if I say that you inspired a lot of us to strive for great projects. We will not forget your greatness. Good luck with anything you do🤞
You r just aweso....i was litterly noob in tNice tutorials...u explained it very simply...thanks i will surely go for next parts of tNice tutorials video.. thanks
The problem is that I don't want it to be masked out if I'm looking at the pixel from under the water...
2 роки тому+1
Oh wait I remember you from somewhere, I think that you used to make devlogs or something. Damn, that water is so impressive I love it! Catch ya next year mate! :)
can you make a tutorial on how to make the water look like yours (i already saw the last water tutorial for the waves but i need multiple waves for my project)
For the reflections, I have some ideas... You can try offsetting that camera to match the highest wave vertex, or you can get the camera texture and offset each pixel according to the highest wave (I hope it isn't too confusing). You can also use a realtime reflection probe which works pretty well but requires more computing power.
The problem is that if you use a second camera to generate the reflections, it inherently assumes a flat reflection plane. Offsetting the camera based on the wave height would make it look more correct in the area that matches that height, but it would then look wrong anywhere where the water is at a different height. Offsetting the camera texture might work to a degree, but around the edges of the screen it would completely break down because you run out of texture to offset (if that makes sense).
An easy way to solve the problem of waves displacing the water plane that you mention in the middle of the video is to have a vertex shader that distorts the reflected vertices in accordance with the wave height, meaning that they always meet the water's surface at the appropriate altitude. It's not perfect, but it at least gets rid of those local discontinuities.
So you mean when rendering the reflection I should use a different shader for everything which moves it in a way that the flat reflection plane ends up producing a reflection that fits onto a not-flat water surface? That's a very interesting idea, but it sounds extremely complicated and time consuming to figure out, and I'm not entirely convinced that it'd even be possible to make it look right 🤔
I was looking at this in sea of thieves, and I think what they've done for a lot of objects like the islands and boats is to just have another low detail object flipped and distorted normally by the shader as it's underwater. Because you can see that the reflection of the island and boat that you see when you're either under water and looking up, or above and looking down, is not the same as the actual object. I think this then gets masked/unmasked by the water shader depending on what side you're looking from! Of course I might be completely wrong, but it makes sense as it provides accurate enough visuals while completely faking it!
_"I think what they've done for a lot of objects like the islands and boats is to just have another low detail object flipped and distorted normally by the shader as it's underwater"_ I think this depends on your graphics settings-lower settings definitely result in lower detail models being used for reflections, but I'm not sure about it higher settings. _"you can see that the reflection of the island and boat that you see when you're either under water and looking up, or above and looking down, is not the same as the actual object"_ Interesting...I'm not sure I've ever noticed that. I might have to go take a look :P Something I want to try out at some point (low priority) is screen space reflections, as I think those would be able to handle the non-flat water surface, though they come with other limitations (only objects visible to the camera can actually be reflected).
For your reflection problem: You might need to vertically offset each object that is on or above the water (boats, debris, player models) by (2 times) the water level at it's position when you render the scene for your reflection texture. Also, on rough water you shouldn't reduce the reflection strength (as you mention at 3:43), but blur the reflections. Water will keep it's full 3%+ reflectivity (source: en.wikipedia.org/wiki/File:Water_reflectivity.jpg) but its surface turns frosted/matte, at least until you start getting foam.
Offsetting everything for the reflection texture seems like it'd be very difficult to get right, if not impossible (since it'd have to be positioned based on how it's going to be reflected, but that's based on how it's positioned). I might blur the reflections, but blurring can get computationally expensive, so I'd still consider reducing the reflection strength a viable option (at least on lower graphics settings), even if it's not entirely realistic.
@@tomweiland I consider there to be two effects that the waves have on the reflections: Firstly, the water is obviously sloped. This is already taken care of by how the reflection texture is applied based on the gradient of the water surface. (I suspect that the deflection angle might be half as much as it should realistically be, but even if that's true, it's not an error anyone would notice.) Secondly, the waves raise and lower the water level. This is causing the problem you showed where the ship doesn't move. To fix this your reflection would just need to move straight up and down. And to calculate the amount, you can just sample the wave height at (below/above) the center of the respective object. In effect, you assume the water to be flat, but at the sampled height instead of at its neutral height. The only problem with this is that a large object might be sitting at a wave crest, while its reflection shows in a valley. But that mainly affects reflections of the top of the object; where the object touches the water (which is where errors are most visible) it will be almost exactly correct.
Idea for the underwater effect: close the water body mesh (add sea floor and walls). After the main pass, render this mesh with reversed face-culling and without depth testing. Add fog manually in fragment shader, based on G buffer depth. This also works for other bodies of water, including waterfalls and floating blobs of liquid.
Walls would be sufficient to make it work, and it's something I considered but didn't implement yet because currently the water mesh is generated at runtime (and giving it walls would make that more complicated).
Man, I got into more complex programming by watching your multiplayer server series. I hope life is treating you well and that whatever you're working on nowadays is running smoothly
That's great to hear! Things are going well, but unfortunately I haven't really figure out where and in what capacity this project fits into my life now that I work full time 😅
Rather you than me... A lot of this stuff seems like a pain in the arse to solve, I admire that you challenging yourself though, it truly is the only way to learn new things and improve!
I've been working on rebuilding the entire project and I do want to keep making videos, but this project is not making me any money (nor will it in the foreseeable future) so a lot of my time is going elsewhere as I'm trying to figure out how to support myself :/
For the surface of the water the solution is simple, make a bubble particles effect that covers up the area where the water meets the surface when the player is half submerged. Many video games over look particles like this and it really shows when you try jumping in bodies of water and there’s no satisfying splash.
Yeah I've seen that in other games, and honestly I'm not sure I like how it looks. Now that I've got this solution essentially working, I'll likely stick with it. There will eventually be splash particles when you jump in the water, but they won't cover your whole screen or something to hide the lack of fog in parts that are underwater.
About the first problem, in SoT when the player is swimming they are either fully submerged or floating above the surface, there is no in between. The camera shifts up/down a little when a transition happens. Having these 2 separate states is also useful to handling the breath/drowning mechanic. I've been thinking about creating a SoT clone myself, let me know if you wanna exchange ideas or maybe turn this project into a coop project.
Yeah, I've played a lot of Sea of Thieves, and the way they handle the transition has always bothered me, which is why I really didn't want to go with a solution like that. It just feels clunky when the game suddenly moves your camera a substantial distance up or down just because you got too close to the surface. For breath I can just check the camera's position against the wave height, so that shouldn't really be an issue. At this point the project honestly isn't even really in a prototype state (there's still no game loop), and I don't plan to involve anyone else at least until that changes.
I can think of a few solutions for the horizon mask problem you have at the moment, 1. you can use the boundaries of the current ocean chunk that is loaded as the mask and use fog to hide the rest of the unloaded water. 2. you showed a black and white image in which the horizon was black and the rest was white, invert it and use that to mask out the blue fog again. hope I could help I never used something like this, just some ideas that came to mind while watching ^-^ the game looks good would love to see dev streams and more content that is Pirate Game related :)🥶
1. Yeah, I'll probably be extending the range of the water mesh. 2. Not entirely sure what you mean by this, or how exactly it'd solve the problem 🤔 I've done streams in the past, so maybe that's something I'll do again at some point, although I'm not quite sure yet and don't want to make any promises 😅
i might overlook sth but is this not only a problem if you are high up in the air? What if you only run your submerged shader once you are close to the water, since in all other cases you won't be submerged anyway? you could reuse your old version of the submerged boolean in some way for that
@@---zh8qn yes, that's what I'll be doing and it'll solve most of the problem, but there can/will still be artefacts when you're close to the water line.
@@tomweiland btw i really appreciate you reading through your comments and answering them, gives a really good feeling to the people watching you :D Keep up the good work
I've just subscribed, so you're now legally obligated to continue development of this game. Thank you. 👍 Also, I found your channel while searching for "should I make my game multiplayer", hoping for people to dissuade me, and you haven't successfully done that yet, so please try again. 😀
Thanks for subscribing! Have you watched this video? ua-cam.com/video/eQt7TwCr6ao/v-deo.html If there's any single one of my videos that will dissuade you from making a multiplayer game, it'll probably be that one :P
well, thanks to someone on unity forums, i've finally found the way to get the world position of screen pixels in standard render pipeline. and now i can sample the same displacement function and have a nice transition between. about the problem the ocean shader vertices are not matching with frag shader, what i did to solve it was, instead of clipping the quad if worldpos.y is below displacement, i've made a gradient. lerping between screen input and underwater color, by 1- saturate ( worldpos.y - displacementY). so that even the vertex displacement is kind of off, it is covered by the gradient over extended. and lerping between a darker color by (1- saturate(abs(worldpos.y - displacementY))), i can have a thick gradient that is darker on the center, which also gives it a nice waterline look.
Sampling the water height at ever pixel is still not ideal because to get the height at a given point with Gerstner waves, you have to sample the wave equation 3+ times. That's a _lot_ of wave height calculations, and I think my masking solution is also more flexible for other things I may want to do in the future.
@@tomweiland Idk why u brought that topic. Thats obvious isnt it. The more complex the calculations are the more performance u gonna sacrifice. But the displacement function has to be same to get the exact point. Or else its not going to match. Which you need to sacrifice some performance. I personally am sampling from a heightmap. Not that heavy. One shouldnt have any permormance issues with some optimizations.
_"The more complex the calculations are the more performance u gonna sacrifice."_ Well yes, obviously, which is why I'd prefer not to have to sample the wave equation 3+ times per pixel. My waves aren't a simple sine wave-Gerstner waves displace vertices horizontally as well as vertically, so you can't simply use the same math to calculate the height at a given point, because it would actually spit out the height for a point that's offset horizontally as well. As shown in the video, sampling the wave height at every pixel is not the only option-the masking solution works too, and in my opinion is better because it's more flexible and will be easier to work with for future things I'll have to do with the water rendering.
@@tomweiland yeah. Thats right. I just forgot about gerstner waves for a second. And you know what? i watched the video again. Realised that i kindof missed out on the last part. Now it makes sense. And the whole conversation was nothing but a waste of time. Sorry for being dumb 🤠 But how do you render your water into a texture? You just say an additional render option or sth like that. What is that?
New devlog!!! Nice to see the water works well now! maybe this sound stupid but for the horizon problem, maybe use the distance between the camera and the fog, if its to fa dont render fog,
I just had a guest speaker from ARM's graphics division at my university speak about graphic rendering. The number one point he hammered in on every slide and said like 50 times is that in computer graphics doing it correctly is ok, but cheating is the best and always preferred. This is because cheating usually means jumping over a lot of steps and making the algorithm/implementation a lot easier to implement. So for your question at 5:31 my thought would be: *how is that a problem?* Yes the fog rendering code doesn't work automatically for all scenarios, but for this case we don't need to care. We can just cheat and turn off the entire fog rendering code whenever the camera is above the water line. Doing it correctly is okey, but cheating is always best ;)
That's what I intend to do, but it still won't guarantee that all artifacts are prevented. The main reason I even showed what it looks like with the camera so far above the water surface was to more clearly demonstrate the problem that also occurs close to the surface (which is the real, more difficult to solve issue), but less noticeably and less frequently. I definitely agree though-cheating is the way whenever possible :P
"Arrr, ye scallywag! A year has sailed by without a whisper from yer ship! The sea of UA-cam be yearnin' for yer return - raise the Jolly Roger once more and let's plunder some views!"
5:50, I wonder if you could just turn off that texture/shader if the camera gets to a certain height, since it's only needed for the under water effect anyways? you could turn it off as soon as the camera is higher than the waves can be
I'll definitely be doing that which will solve most of the problem, but it won't prevent all artefacts so I still need to come up with something else to address those cases.
Lowering the reflection resolution doesn't give it a ripple effect, it just looks worse. I'll probably add a quality setting for it eventually, but that's not a priority right now.
in unreal planar reflection work like separate object, with bounding box and clip plane. With this method you could make separate reflection for ship an attach it to ship. So clip plane would be where bottom of ship is and reflection would apply to meshes in bounding box. Meybe something like this is posible with unity.
The pirate game has not been abandoned! Let me know if you've got any ideas for how to fill in the gaps in the water fog mask between the horizon and the water meshes 👇
You could do what Minecraft does and have a fog effect at the edges. Alternatively you could make a way to seamlessly blend LOD versions of the watter
Why do you have to use the horizon as a base? Why don't you just render the top of the water plane as white (like you're doing) and then fill everything above that plane with white? Like a flood fill. It seems like the water plane should always make at least a solid boundary between the top and bottom of the screen. So just fill anything above that boundary.
In a pixel shader this would just mean that, starting from the top of the image, you would set every pixel in every column (not row) to white until you reached an already white pixel. Since the water is height based there should never be overhangs that have to be filled.
@@joshuafhiggins and you mean have the low LOD versions go all the way to the horizon? I considered that as well, and it could work, but how far those low LOD planes have to extend into the distance would depend on the height of the camera _and_ potentially on the resolution of the display.
Or you could also make everything white and then render the bottom of the plane as black and then fill everything below with black
@@nathangeorge4017 this was something that crossed my mind, but I honestly just wasn't sure how to approach the flood filling. The fragment/pixel function of a shader operates on each pixel individually, with no information about other pixels. Obviously I could sample the mask texture, but (on 1920x1080 resolution, with about half the screen submerged) that would mean sampling the texture at least ~540 times for each of the ~540 pixels in each of the 1920 columns-unless I'm missing something? That means sampling the fog mask texture ~559 _million_ times every frame, which doesn't sound very efficient to me...
After 83 years of my life, I can finally witness another devlog
You didn't have to do me like that 😂
Wow the dedication to watching this man is truly amazing
And you will wait 83 years more
Another Devlog!! But also don’t stress about getting them out fast! The game is looking amazing!
Now if only it was actually playable 😂
@@tomweiland nah dude take your time, ive been making a game for a year too and its still not playable lol
I have binge watched almost all of your dev-logs, I hope the project is going well and its smooth sailing, excitedly awaiting the next dev-log
Glad you've enjoyed the devlogs!
@@tomweilandstill no dev-log? If ur giving up don't I need more vids
I honestly thing that these reflections are the best improvement you have made visually to the game, aside from maybe the water flattening around the islands. The game is coming along really well now!
Agreed! I think your last statement needs to be corrected though-the _graphics_ are coming along nicely. The _game_ is still as unplayable as it was a year ago :P
The island reflection looks amazing. Good progress on the water, hopefully the game gets into a playable state soonish!
🤞
Water is by far one of the most complicated subject once you dive deeper into it. Great video! Would love to see it getting resolved
Yeah it's one of those things that can literally always be improved further, there's no end :P
Never thought i would live to see another pirate game devlog lol
Also the water looks even more amazing now! Great work
Dang, you had really given up hope, huh? :P
I know that working alone on a project can be hard, but I'm happy to see you back :D
Thanks 😅
I can see from the comments that you're now a servitor to the machine god, 8 hours a day. That's okay man, never think it's 'been too long' to continue your project. You will get used to it, you will upskill in certain areas, and your work, whatever it is, will help you with the game in one way or another. You've got this man! And if your priority is now a career rather than making the game, I wish you good luck on that too. But I really don't think it is. I think you and me are a lot alike, I didn't go to university for the same reasons as you.
Well anyway, I've been fucking off on work hours for 15 mins now, so I'll get back to that but just know, you are not alone. There's a lot of people like you, full of passion, working 8 hours a day and MAYBE getting in half an hour every 2 days on their project. It's the consistency that counts. See you at the top.
Thank you for the kind words! These are the kind of comments which make me want to come back to this project, but I'm still trying to figure out where and in what capacity it fits into my life now that I work full time 😅
Awesome video Tom. Very cool
Thanks, I'm glad you enjoyed it :)
I have been waiting so long for this level of quality in a devlog
Glad you liked it :D
I think going SSR on reflections would give much better results. Since camera is barely above the reflective surface, it'll be very close to ground truth too. This will also solve (or rather) help with the double lighting issue. With the added bonus of in reflection lighting being completely free.
The problem is that nothing that is off screen can be reflected, which means tops of boats and islands will be cut off in reflections if you don't have them in view. As for the camera barely being above the reflective surface, I'm not really sure what you mean since a boat's crows nest or the top of even a small island is more than "barely" above the water surface, is it not?
@@tomweiland yeah cut off makes sense. I used "Barely above surface" as in SSR only works well when viewing angle of the reflective surface being almost parallel to the camera, going high up and looking directly down breaks the effect. But I don't think that's a realistic thing to happen in this game that's all. Your channel is great find btw, thanks for all
Code free of charge:
If(aboveWater)
Renderfog = false;
Good luck solving the actual problem tho, glad your back.
Haha thanks :P
Hope you are doing well, your old C# networking tutorial series has helped me a lot, and your pirate game looks great. Glad to see you're still replying to comments after so many months.
Glad the old series was helpful! Have you checked out the newer one?
@@tomweiland not really, just following the pirate game series.😁 looking forward to a new episode.
These are always really cool, good to see another dev log
It was about time 😅
Great to hear from you! I honestly think about your videos every time I work in Unity
That means a lot, glad you like the videos :)
Not sure if you olan to stick the the game or not or just busy but even months later i still enjoy watching your videos! Thank you still! Hope you are taking care and living the good life!
I do want to continue working on it and there will be at least one more devlog in the near future, but beyond that I'm not really sure yet. I don't want to live at home indefinitely and this project/these videos don't come even remotely close to making a livable income. Even if I managed to upload once a month, I would have to average _several million_ views per video just to begin approaching a decent salary, and that is not exactly something I can rely on to happen :P
Glad you've been enjoying the videos though!
@@tomweiland completely understandable! It's not easy pulling in the views. Ive seen amazing creators get trampled on by UA-cam no matter how often they upload. its a shame!
No problem! It just good to see other people's experiences and creativity ^-^
Its been 6 years to the day since you started tjis project! Please continue on. You can do it!!!!!!!!!!
I haven't touched the project at all for a while now due to working full time, but I do want to get back to it at some point 🤞
Glad your still working on this!
Me too :P
You better finish this tom.
Been following it for like two years now and actually NEED to play a finished version.
Well buckle up, it's going to take a while 😂
But yeah, I need to play this too at some point, so I definitely intend to get it to a playable point...
The game looks wonderful so far. I hope the next devlog will be a little sooner but it's okay don't rush it!
It definitely won't be another year until the next one 😂
hey tom, I have just gotten to explore your channel today and I must say, I am in love with it, your videos were so high quality and entertaining I am just hooked. I am a fellow programmer also wanting to see where I want to go and I am inspired by your continuous and hard work for this personal project and I am hoping to start one soon, maybe not game dev after (I am gonna take your word :D). Wishing you the best on your journey in 2024 and onwards!
Thank you for the kind words! Good luck to you too :)
U are ALIVE!
Good ,keep it up!
I'm not dead 🥳
Hey man cant wait till the next devlog have a great day your my new favourite UA-camr ❤
Thanks for the kind words 😅
Simple trick for sea reflections would be to use SSR (screen-space reflections), perhaps you could cross-fade between planar reflections when the sea is calm and SSR when it is agitated. Or you could simply use SSR in every case.
Another thing that might possibly be done is to displace the reflected image depending on the camera's angle and how heavily displaced the water plane is, using trigonometry.
I'm not quite sure this would result in a good looking result but it feels like it could work with some tuning and adjustments.
Keep in mind that if you plan to keep the Planar Reflection for agitated sea, whatever you do, the reflections will be physically inaccurate (that goes for SSR too but it would be less noticeable), since it only replicates what a flat surface would reflect.
There would also be an issue with some parts of the terrain OVER the reflection plane and UNDER the sea plane being missing in the reflection.
Personally I would advise for using SSR as I don't believe any other kind of hack you can conjure is gonna look good enough to be worth the effort.
The next best thing is raytracing but we all know why that's not a great idea :).
The unfortunate thing about SSR is that anything that's not visible on screen also can't be reflected, which in some scenarios can look just as bad/weird as the problems I have right now with the reflections on big waves, so I'm not sure it'd be better. With some tweaking it might be possible to get the current setup looking _decent,_ so I'm not particularly eager to abandon that approach just yet :P
@@tomweiland fair enough.
I can appreciate you going for a "higher-tier" physically accurate reflection model. Good luck on creating a reflection setup that fits your project :).
YAAAAY, haven't seen a new video for a long time, good to see you back
Yeah...it's been _a while_ 😅
Not sure if you still care, but a decent fix for the fog/mesh gap is to make a slightly fuzzy white band along the pixel curve that's wider than the gap; which looks like a pretty decent foam line and covers the crimes.
This might work if my mesh had a higher vertex resolution but because of the distance between vertices the gap between the mesh and where the fog starts would likely be too big for this to work.
I miss you... We all do. I am glad you switched to godot tho and that you are porting this to it. Looking forward to the first devlog in godot.
I'm assuming this is a joke lol, but just in case: I have not switched to Godot 😅
I'd like to come back to this project eventually, but I haven't really figured out where it fits in with work and other life stuff.
@@tomweiland Ya I was joking haha😅
Anyways looking forward to whatever comes next, I hope I can speak for a large part of the community if I say that you inspired a lot of us to strive for great projects.
We will not forget your greatness.
Good luck with anything you do🤞
so glad this project is back keep up the amazing work!
Yeah it feels good to be working on it again :P
Cant wait for next years devlog!
💀
How long have I been waiting for 1 more video >.
Too long 😔
'ooray, now I can look at the reflection o' me face in the water while I trim me beard to perfection! :p
😂
You r just aweso....i was litterly noob in tNice tutorials...u explained it very simply...thanks i will surely go for next parts of tNice tutorials video.. thanks
🤔
glad to see you're uploading again
as for the water fog maybe try using the depth buffer to mask out far away areas
The problem is that I don't want it to be masked out if I'm looking at the pixel from under the water...
Oh wait I remember you from somewhere, I think that you used to make devlogs or something.
Damn, that water is so impressive I love it!
Catch ya next year mate! :)
The next one better not be a whole year from now 😳
High entertainment value on this one. Nice editing. Keep it up.
I put a lot of time into it, so I'm glad you enjoyed it :D
can you make a tutorial on how to make the water look like yours (i already saw the last water tutorial for the waves but i need multiple waves for my project)
Nice to see you Upload again
It was nice to upload again :P
For the reflections, I have some ideas... You can try offsetting that camera to match the highest wave vertex, or you can get the camera texture and offset each pixel according to the highest wave (I hope it isn't too confusing). You can also use a realtime reflection probe which works pretty well but requires more computing power.
The problem is that if you use a second camera to generate the reflections, it inherently assumes a flat reflection plane. Offsetting the camera based on the wave height would make it look more correct in the area that matches that height, but it would then look wrong anywhere where the water is at a different height. Offsetting the camera texture might work to a degree, but around the edges of the screen it would completely break down because you run out of texture to offset (if that makes sense).
the mans is back! I love how the game looks!
Glad you like it :D
Fantastic video bro, legend.
Glad you liked it!
An easy way to solve the problem of waves displacing the water plane that you mention in the middle of the video is to have a vertex shader that distorts the reflected vertices in accordance with the wave height, meaning that they always meet the water's surface at the appropriate altitude. It's not perfect, but it at least gets rid of those local discontinuities.
So you mean when rendering the reflection I should use a different shader for everything which moves it in a way that the flat reflection plane ends up producing a reflection that fits onto a not-flat water surface? That's a very interesting idea, but it sounds extremely complicated and time consuming to figure out, and I'm not entirely convinced that it'd even be possible to make it look right 🤔
Great video as always!
Thanks, glad you liked it!
Was expecting him to say "I'll see you again next year" at the end
Why you gotta do me like that 😭
bro thanks so much, you actually made soft soft easy to understand
🤔
Great devlog! Thanks
Glad you liked it!
eyy Tom, great to see you back ;)
Thanks, it's good to _be_ back!
Thanks, man! Great job! Thanks for help, bcause i`m already finish man game for jam, using a info from your previous devlog
Glad I could help :)
I was looking at this in sea of thieves, and I think what they've done for a lot of objects like the islands and boats is to just have another low detail object flipped and distorted normally by the shader as it's underwater. Because you can see that the reflection of the island and boat that you see when you're either under water and looking up, or above and looking down, is not the same as the actual object. I think this then gets masked/unmasked by the water shader depending on what side you're looking from! Of course I might be completely wrong, but it makes sense as it provides accurate enough visuals while completely faking it!
_"I think what they've done for a lot of objects like the islands and boats is to just have another low detail object flipped and distorted normally by the shader as it's underwater"_
I think this depends on your graphics settings-lower settings definitely result in lower detail models being used for reflections, but I'm not sure about it higher settings.
_"you can see that the reflection of the island and boat that you see when you're either under water and looking up, or above and looking down, is not the same as the actual object"_
Interesting...I'm not sure I've ever noticed that. I might have to go take a look :P
Something I want to try out at some point (low priority) is screen space reflections, as I think those would be able to handle the non-flat water surface, though they come with other limitations (only objects visible to the camera can actually be reflected).
cant wait for the next update of this. I know its been a year but I just seen a comment saying you will have a new video out soon
_Hopefully_ soon 😅
For your reflection problem: You might need to vertically offset each object that is on or above the water (boats, debris, player models) by (2 times) the water level at it's position when you render the scene for your reflection texture.
Also, on rough water you shouldn't reduce the reflection strength (as you mention at 3:43), but blur the reflections. Water will keep it's full 3%+ reflectivity (source: en.wikipedia.org/wiki/File:Water_reflectivity.jpg) but its surface turns frosted/matte, at least until you start getting foam.
Offsetting everything for the reflection texture seems like it'd be very difficult to get right, if not impossible (since it'd have to be positioned based on how it's going to be reflected, but that's based on how it's positioned).
I might blur the reflections, but blurring can get computationally expensive, so I'd still consider reducing the reflection strength a viable option (at least on lower graphics settings), even if it's not entirely realistic.
@@tomweiland I consider there to be two effects that the waves have on the reflections:
Firstly, the water is obviously sloped. This is already taken care of by how the reflection texture is applied based on the gradient of the water surface. (I suspect that the deflection angle might be half as much as it should realistically be, but even if that's true, it's not an error anyone would notice.)
Secondly, the waves raise and lower the water level. This is causing the problem you showed where the ship doesn't move. To fix this your reflection would just need to move straight up and down. And to calculate the amount, you can just sample the wave height at (below/above) the center of the respective object. In effect, you assume the water to be flat, but at the sampled height instead of at its neutral height.
The only problem with this is that a large object might be sitting at a wave crest, while its reflection shows in a valley. But that mainly affects reflections of the top of the object; where the object touches the water (which is where errors are most visible) it will be almost exactly correct.
Just saw a two-year-old video good to see you haven stopped developing 😊
😅
you could name "The polygonal robbery" and add missions on islands like sneaking into another pirate base
8 syllables is a bit of a mouthful and I'd rather avoid using references to the art style in the name. Thanks for the suggestion though!
Keep up the amazing work !!!
Will do :)
Idea for the underwater effect: close the water body mesh (add sea floor and walls). After the main pass, render this mesh with reversed face-culling and without depth testing. Add fog manually in fragment shader, based on G buffer depth. This also works for other bodies of water, including waterfalls and floating blobs of liquid.
Walls would be sufficient to make it work, and it's something I considered but didn't implement yet because currently the water mesh is generated at runtime (and giving it walls would make that more complicated).
Man, I got into more complex programming by watching your multiplayer server series. I hope life is treating you well and that whatever you're working on nowadays is running smoothly
That's great to hear! Things are going well, but unfortunately I haven't really figure out where and in what capacity this project fits into my life now that I work full time 😅
0:46 Minecraft camera in water is the same way
And I don't like it that way :P
This was a fun one, also nice to see that you are still working on it.
Glad you enjoyed it :)
Nice, I love the new water! Good job 👍
Thanks, glad you like it :)
man I missed this kind of tutorials lol. Great work here, thanks!!!
🤔
Lots of love bro
:)
Rather you than me... A lot of this stuff seems like a pain in the arse to solve, I admire that you challenging yourself though, it truly is the only way to learn new things and improve!
Agreed!
Miss this series and your work. Was kinda unique. Hope all is good for you.
Glad you've enjoyed the videos, I will be back soon 👀
imagine making near clip plain of the camera = 0.001...
then under water effect is either on or off.
seems like a reasonable and simple solution
Unity doesn't let you set it lower than 0.01 units-that's the minimum, and I already have it set to that.
@@tomweiland ohh.. okay then :)
Will you still make videos? I really loved them, it's nice to see you still come back to answer some comments tho
I've been working on rebuilding the entire project and I do want to keep making videos, but this project is not making me any money (nor will it in the foreseeable future) so a lot of my time is going elsewhere as I'm trying to figure out how to support myself :/
Looks great! And ofcourse... YAAAAAAAAAAAAAAAAY DEVLOG!!!!!!!!!!!!!!!!!!!
🥳
For the surface of the water the solution is simple, make a bubble particles effect that covers up the area where the water meets the surface when the player is half submerged. Many video games over look particles like this and it really shows when you try jumping in bodies of water and there’s no satisfying splash.
Yeah I've seen that in other games, and honestly I'm not sure I like how it looks. Now that I've got this solution essentially working, I'll likely stick with it. There will eventually be splash particles when you jump in the water, but they won't cover your whole screen or something to hide the lack of fog in parts that are underwater.
Here's for the Riptide 2.0 release!
Thanks :D
To be honest I kinda like how blocky the ocean looks! It gives it a unique style.
Yes I love the low poly art style, assuming it's done decently well!
DEV LOG DEV LOG, I AM HAPPY
Excellent 👌
About the first problem, in SoT when the player is swimming they are either fully submerged or floating above the surface, there is no in between. The camera shifts up/down a little when a transition happens. Having these 2 separate states is also useful to handling the breath/drowning mechanic.
I've been thinking about creating a SoT clone myself, let me know if you wanna exchange ideas or maybe turn this project into a coop project.
Yeah, I've played a lot of Sea of Thieves, and the way they handle the transition has always bothered me, which is why I really didn't want to go with a solution like that. It just feels clunky when the game suddenly moves your camera a substantial distance up or down just because you got too close to the surface. For breath I can just check the camera's position against the wave height, so that shouldn't really be an issue.
At this point the project honestly isn't even really in a prototype state (there's still no game loop), and I don't plan to involve anyone else at least until that changes.
@@tomweiland ok ok
just stumbled here, take your time it will pay off
😅🤞
I think you should share 1080p photo of 2:56. It looks amazing and it will be great as a background image for Windows
If you join the Discord server and ask me there, I can send you one :)
I can think of a few solutions for the horizon mask problem you have at the moment,
1. you can use the boundaries of the current ocean chunk that is loaded as the mask and use fog to hide the rest of the unloaded water.
2. you showed a black and white image in which the horizon was black and the rest was white, invert it and use that to mask out the blue fog again.
hope I could help I never used something like this, just some ideas that came to mind while watching ^-^
the game looks good would love to see dev streams and more content that is Pirate Game related :)🥶
1. Yeah, I'll probably be extending the range of the water mesh.
2. Not entirely sure what you mean by this, or how exactly it'd solve the problem 🤔
I've done streams in the past, so maybe that's something I'll do again at some point, although I'm not quite sure yet and don't want to make any promises 😅
i might overlook sth but is this not only a problem if you are high up in the air? What if you only run your submerged shader once you are close to the water, since in all other cases you won't be submerged anyway? you could reuse your old version of the submerged boolean in some way for that
@@---zh8qn yes, that's what I'll be doing and it'll solve most of the problem, but there can/will still be artefacts when you're close to the water line.
@@tomweiland btw i really appreciate you reading through your comments and answering them, gives a really good feeling to the people watching you :D
Keep up the good work
Omg finally new devlog
Finally 🥳
@@tomweiland yeah
I've just subscribed, so you're now legally obligated to continue development of this game. Thank you. 👍 Also, I found your channel while searching for "should I make my game multiplayer", hoping for people to dissuade me, and you haven't successfully done that yet, so please try again. 😀
Thanks for subscribing! Have you watched this video? ua-cam.com/video/eQt7TwCr6ao/v-deo.html
If there's any single one of my videos that will dissuade you from making a multiplayer game, it'll probably be that one :P
Yeah, that's the one I found first and I'm still here. 😄 Looking forward to the next game update 👍
This one was really well edited. Next time I hope to see beautiful ocean
Thanks, glad you liked it :D
well, thanks to someone on unity forums, i've finally found the way to get the world position of screen pixels in standard render pipeline. and now i can sample the same displacement function and have a nice transition between. about the problem the ocean shader vertices are not matching with frag shader, what i did to solve it was, instead of clipping the quad if worldpos.y is below displacement, i've made a gradient. lerping between screen input and underwater color, by 1- saturate ( worldpos.y - displacementY). so that even the vertex displacement is kind of off, it is covered by the gradient over extended. and lerping between a darker color by (1- saturate(abs(worldpos.y - displacementY))), i can have a thick gradient that is darker on the center, which also gives it a nice waterline look.
Sampling the water height at ever pixel is still not ideal because to get the height at a given point with Gerstner waves, you have to sample the wave equation 3+ times. That's a _lot_ of wave height calculations, and I think my masking solution is also more flexible for other things I may want to do in the future.
@@tomweiland Idk why u brought that topic. Thats obvious isnt it. The more complex the calculations are the more performance u gonna sacrifice. But the displacement function has to be same to get the exact point. Or else its not going to match. Which you need to sacrifice some performance. I personally am sampling from a heightmap. Not that heavy. One shouldnt have any permormance issues with some optimizations.
@@tomweiland and as you are doing the same thing, sampling displacement on every pixel is the only way of doing this.
_"The more complex the calculations are the more performance u gonna sacrifice."_
Well yes, obviously, which is why I'd prefer not to have to sample the wave equation 3+ times per pixel. My waves aren't a simple sine wave-Gerstner waves displace vertices horizontally as well as vertically, so you can't simply use the same math to calculate the height at a given point, because it would actually spit out the height for a point that's offset horizontally as well.
As shown in the video, sampling the wave height at every pixel is not the only option-the masking solution works too, and in my opinion is better because it's more flexible and will be easier to work with for future things I'll have to do with the water rendering.
@@tomweiland yeah. Thats right. I just forgot about gerstner waves for a second. And you know what? i watched the video again. Realised that i kindof missed out on the last part. Now it makes sense. And the whole conversation was nothing but a waste of time. Sorry for being dumb 🤠 But how do you render your water into a texture? You just say an additional render option or sth like that. What is that?
First?
Anyway. Thanks again for the amazing server tutorials, you taught me a lot and I hope your game will thunder the shop when it gets out :3
Damn that was fast 😅
@@tomweiland Literally just opened the app and I saw your video which was uploaded for 8 sec. What were the chances
New devlog!!! Nice to see the water works well now! maybe this sound stupid but for the horizon problem, maybe use the distance between the camera and the fog, if its to fa dont render fog,
The problem with that is that it'd also remove the fog underwater where there aren't any objects being rendered (just skybox) 🤔
2:00 "doesn't click the top right copy icon"
ahhh, a true programmer
😅
Really great work, those reflections are especially impressive! Looking forward to what comes next, just take your time!
Thanks, will do :D
I just had a guest speaker from ARM's graphics division at my university speak about graphic rendering. The number one point he hammered in on every slide and said like 50 times is that in computer graphics doing it correctly is ok, but cheating is the best and always preferred. This is because cheating usually means jumping over a lot of steps and making the algorithm/implementation a lot easier to implement.
So for your question at 5:31 my thought would be: *how is that a problem?*
Yes the fog rendering code doesn't work automatically for all scenarios, but for this case we don't need to care.
We can just cheat and turn off the entire fog rendering code whenever the camera is above the water line.
Doing it correctly is okey, but cheating is always best ;)
That's what I intend to do, but it still won't guarantee that all artifacts are prevented. The main reason I even showed what it looks like with the camera so far above the water surface was to more clearly demonstrate the problem that also occurs close to the surface (which is the real, more difficult to solve issue), but less noticeably and less frequently. I definitely agree though-cheating is the way whenever possible :P
"Arrr, ye scallywag! A year has sailed by without a whisper from yer ship! The sea of UA-cam be yearnin' for yer return - raise the Jolly Roger once more and let's plunder some views!"
It's been almost 2 years 😭
I haven't really touched the project in the last 6 months, though I'd like to get back to it...eventually :/
Great video!
Thanks :)
It is sooo overengineered, I love it
What's overengineered? 😅
I kinda like games that allows you to stay half submerged with no fog. It just feels good
It feels wrong in my opinion :P
5:50, I wonder if you could just turn off that texture/shader if the camera gets to a certain height, since it's only needed for the under water effect anyways?
you could turn it off as soon as the camera is higher than the waves can be
I'll definitely be doing that which will solve most of the problem, but it won't prevent all artefacts so I still need to come up with something else to address those cases.
@@tomweiland i see, good luck! Looking forward to see your solutions in the upcoming devlogs!
This looks so fun to play
Thanks!
Wouldnt a skybox solve it? Im just throwing out ideas though
How exactly would a skybox solve this? Your suggestion is rather vague 😅
A note for the first minute of the video, that bug you mention is in SO many games with water in them.
And it always bothers me :P
Keeps you creating ;)
Yep.
Use a shader that repeats lod copies of the water mesh into the distance. Based on how far you can see, or truncate the water fog
Yeah there's a bunch of different solutions which I'll end up combining, and hopefully together they'll eliminate all artifacts.
YAY ANOTHER DEVLOG!!!!!
About time :P
Yes, to long have I waited for this lol
draw meshes at the edge of the water in the buffer or use floodfill
Yeah that might work.
Lower the resolution of your water reflection in order to give the water ripple effect and to improve performance slightly
Lowering the reflection resolution doesn't give it a ripple effect, it just looks worse. I'll probably add a quality setting for it eventually, but that's not a priority right now.
My game needs this
The reflections? Take a look at Unity's Boat Attack demo if you haven't yet.
The water looks good. I will try tomorrow to create some water in Unity and see how I can handle it.
Good luck!
in unreal planar reflection work like separate object, with bounding box and clip plane. With this method you could make separate reflection for ship an attach it to ship. So clip plane would be where bottom of ship is and reflection would apply to meshes in bounding box. Meybe something like this is posible with unity.
Maybe, but a big part of the problem is that it's a _planar_ reflection, and my water is anything but a flat plane :P