Just played through this on PS5 on a decent OLED. It does not look like this video.
It's impressive, but I'm skeptical this video is real-time on a PS5, certainly I doubt this is at 4K - the jaggies when I ran this were everywhere.
It was worse in the city demo after the action sequence. Certainly a lot going on, but in terms of the trade-off, it was visually quite obvious where the costs are side-by-side with Miles Morales on it's PRT mode and for a real game the later would almost certainly be better overall.
Hopeful people will get more out of it in future, but wanted to temper expectations here from the video.
EDIT: Just found the signpost to enable real-time night mode with Lumen. Wow. It's clearly still struggling to run on a PS5 but it's not that far away from where it would need to be. And when it's not struggling with jaggies/framerate it's pretty jaw-dropping.
I also just played this through on a PS5 and an OLED and was blown away. Don’t think I’ve ever been so impressed with a virtual world before.
The moment that drove it home for me was flying up into the sky, striking out in a random direction, flying to the roof of a building and zooming right up to a cluster of pipes and seeing individual rivets accurately modelled. All of this, for a 10 minute demo.
This is a staggering achievement, by all accounts made possible by some pretty amazing new tooling. Genuinely baffled that anyone could fail to be impressed by this.
Edit: don’t forget to open up the menu and play around with the lighting, crowd, AI settings, etc.
Those rivets probably aren't modelled, they'll be reconstructed from volume information in the texture. Which is still impressive and a great way of dealing with that type of geometric detail, but it has limitations, and the engine isn't processing actual models at that level of detail.
You can still do stuff like that in UE4. Have a look at what's possible with Quixel Mixer, you can create detail like that surprisingly quickly. I'd tentatively argue that modelling tools are the real MVP when it comes to increased geometric LOD in modern engines. They allow you to add that kind of detail quickly enough that it becomes economical to actually detail rivets.
Either way, it's very cool. The crowd simulation stuff is going to be useful, current tooling there absolutely sucks outside of a few very expensive dedicated products.
Yes. Nanite is very clever. A mesh representation that can be shown at a huge range of levels of detail is what makes it go. Much of the work is moved to asset preprocessing. The actual rendering is simplified. There are about as many triangles being drawn as there are pixels, regardless of scene complexity. If a triangle is bigger than a pixel, it's time to zoom in to a higher LOD for that part of the mesh. If a triangle is much smaller than a pixel, a higher LOD can be used. This is a reduction from O(N) to O(1). So draw time is constant regardless of scene complexity.
Watch the SIGGRAPH video to get the feelings of: that's impossible - oh, I kind of see how that works - one pixel triangles? - how do they get that mesh representation set up right? - oh, graph theory - that data format has to be a pain to generate with all those constraints - they're rendering mostly in the CPU? - that's all that needs to be done at render time? - GPUs need to be redesigned to be a better match to this - how to stream this stuff? - that compression scheme is clever.
It's very cool. I bet Digital Foundry takes an in-depth look - can't wait.
After reading your comment I went back in to the demo to have another look at the rivets up close. In the options menu, you can switch the viewport so that it shows you the triangles that make up object geometry under the "Nanite" system. Would be interested to hear from anyone with access to the demo and the right background what they make of this.
Edit: I took a couple of screenshots of default view vs Nanite triangles view but not sure what the convention is on HN for where to host them. Happy to put them up if someone can clue me in.
I'm impressed but I think Spider-Man/Miles Morales on the PS5 handle it better. What I gather from this demo is that UE has made these AAA features easily accessible to more developers, and that Nanite means that artists will need to worry less about LOD for the assets they create.
Some downsides, the "metahuman" face modeling isn't great, it seems very uncanny valley compared to titles like The Last of Us II. There's some obvious framerate issues during the car chase scene where I wouldn't be surprised if it dipped into the 20s.
> Some downsides, the "metahuman" face modeling isn't great, it seems very uncanny valley compared to titles like The Last of Us II.
I think each character in last of us 2 is a custom scan and artist creation. Metahumans is made from an amalgam of scans and lets you parameterize the face (adjust everything with sliders) so it is a much harder problem.
>Genuinely baffled that anyone could fail to be impressed by this.
I was much more impressed by Battlefield 2042 graphics when I first saw it. So while this looks "good", may be I was expecting much more. But then again we are limited by Hardware so I guess I have too high of an expectation.
It's always been like that for demos and promo videos. Sometime they straight up run the demo on a PC with crazy specs while pretending it's running a console.
I played through it myself and tbh I think the video doesn't do it justice. I haven't been this blown away in years. Until I got to take over the controls I was utterly convinced there had to be pre-rendered video thrown into the mix.
If you have a compatible console I'd really recommend checking it out!
I played with it for about an hour on an LG C8 55" OLED and was impressed.
I think it is pretty clear there is a level of detail difference between the 'white room' scenes with Neo and the 'city'.
I would be shocked if the 'white room' scenes where not fully motion captured and then transposed into 3D models to run in engine. A bit unfair as while it may be real time rendered it is fully scripted. The player has no control. It is essentially a movie converted into a 3D environment then optimised to hell and back with pixel corrections where needed.
However the 'city' environment is quite stunning with incredible level of detail and you can free roam either walking/driving around the street or fly by. Really quite something to play about with the real time rendering pipeline to see the triangle make up, etc of every object or adjust the position of the sun in the sky, etc.
A lot of fun. Reminded me a lot of when I first got to play Spider-man on the PS4 and just swing around the city. Will be great to see where this leads in a year or two once devs have really had time to get a feel for how everything works.
I wasn't aware that was common practice. I assume this is for ease of playtesting a game before any optimisation phases have happened? I wonder whether it has a setting to convert down to actual PS5 power levels.
Agreed. I was really hoping they'd release this sort of product on PCs with a "here's what we want in 5 years" level of power. Only high end machines would even be able to run it, but i want to see their vision for future fidelity.
The textures looked super flat to me in this, the world looked decent, but all together i hoped the PS5 was holding their tech down.. because i expected more. Nanite (codename for high model density rendering) examples previously were _much_ better in my view. Eg: https://youtu.be/U-dRBjqAGgI
Glad you got the settings to work to improve your experience, now also imagine waiting 5 years for developers to really optimize and innovate how they use this engine and the PS5's resources.
Going to be funny listening to elitist PC gamers talk about how inferior everything still looks by then, even though they just play nostalgic pixel art games on Steam all day.
Considering the diff between a pc and ps5 is fairly minimal other than the storage i/o. PCs are already 1-2 gens ahead in CPU and GPU. Any sort of optimizations/tweaks they make for the PS5 are going to be on the PC very quickly as the archs are basically the same. Now I own a lot of different consoles and have in the past built my own PC. However, these days the difference between gaming on a console and PC is so 'same' I have not bothered to buy this generation of consoles (I usually get 30-50 games for each console and generation). The switch is probably the most innovative for this generation. With the xbox and ps5 being hardware locked PC's and some interesting exclusives. But it looks like I am going to be buying uncharted 4 again :)
I'm pretty sure the value proposition is that for $500 you can play state of the art games, compared to $1500 for a pc to do the same. Also not sure what unique benefits the PS5 ssd i/o will have. This is currently not something you can do with consumer pc components
A PS5 and a PC of the same specs is comparable in price. Also a PS5 is currently 1000-1200 depending on bundle. The I/O bit with ps5 is interesting for loadtimes and open world games, though it is getting tougher to buy a PC without nvme. The current xbox looks interesting, but MS is basically pushing 'also release on PC'. Like I said I have owned many generations of consoles. However, this gen I am finding it tough to justify also getting one in addition to my PC. Even the PS4 before it I bought only because of one silly unique game that was not on PC but on every console. Something like the n64 when it came out vs the PC it was a no brainer and 'here take my money'. I am not getting that with this gen of consoles. If you want a console over a PC go for it, seriously. I personally am finding the distinction hard to justify with the current gen of consoles.
"A PS5 and a PC of the same specs is comparable in price."
What?. No. A PlayStation 5 digital edition retails for $399.99. What Celeron powered hunk of junk are you able to put together for $399.99?
Obviously use the pc if you have it. My point is entirely for the case of having neither. It is patently untrue you can get a full pc capable of running ps5 quality games in the price range of a ps5
Pci-e 4 + directsorage hardware decompression can do similar rates. What you can't get is integrated CPU/GPU memory on GPUs this fast, but you can get way faster GPUs and CPUs.
Unfortunately if you watch it on something other than the rendering device you are very likely to be seeing through a video compression/decompression cycle. Providers like YouTube are notorious for rescaling the compression to match the current qualities of the channel. Just a big caveat when looking for rendering artifacts because they may be indistinguishable from compression artifacts.
Note that "playing it on a PS5" is meaningless unless you had demo software. The PS5 has a YouTube player and it's not representative of your in-game experience.
There was a Half-Life mod called The Specialists that I played ages ago. It was a Matrix-inspired arena shooter. There was only one map, but it was well done and I got a tremendous amount of gameplay out of it. The bullet timing mechanic was wild. At the time I'd never seen something like that in a multiplayer game.
I was kind of hoping this demo would include something similar, but this looks more like a cinematic than anything. Then again I'm too old to play games like that now. Civilization and Railway Empire are more my speed.
EDIT: Here's a gameplay video if you're interested.
There eventually were many maps in the game. The lobby one might be the best for role playing The Matrix, but the game also had Castor Troy as a player model so the theme was a bit more broad.
This was one of my favorite games for a while. I was part of a “clan” and played tournament games — even made my own website for the whole thing. It sucked up a lot of my time for a summer or two.
They will feel nostalgia but I don't think playing the specialists, a niche mod for a pastime at that time, is the same as playing fortnite, an heavily marketed locked-up game with a world audience.
Oh wow that brought back some memories. Overclocking my Pentium 4 Northwood paired with a 6800gt then running that benchmark, followed by Super PI, and then making sure it's stable with Prime 95. Spent so much time running those benchmarks.
Al that work to be realistic, and then these ridiculous one-linere like "I thought you'd never ask" and "you boys'll get a kick out of this". You'd think with all the money they're spending on visuals they could afford a screenwriter or two.
I'm not sure why they don't just hire voice actors and ask them to naturally respond to the situation while watching the video, with a set of target phrases.
"Say something to the effect of this, when this thing happens on screen, but phrase it however you like."
Visually stunning, but no matter how realistic it is, you're killing the chance of enjoying it because you go "Nobody would say that", and it also doesn't sound natural.
This is why I enjoyed the Fallout games so much. Even the earlier ones like 3 and New Vegas that weren't realistic looking had fantastic, believable dialogue.
This is incredible. It's actually better than some released games, I've been driving around the city for a while now and it's so real!
What I would like not just for this but games in general is more interaction with the environment, you see a building, you can go into it. You see a bench you can sit down etc.
The realism only lasts until you try to do something out of bounds then it's just a reminder that it's all pretty superficial.
But yeah this will be a day one buy when the full game/experience is released!
At this point the main thing I see missing to achieve more realism are better physics, global illumination and better facial animations. Models/textures are already there.
At some point we're gonna need some type of non-profit, open source company that develops massive asset packs for game engines in order to move the industry forward significantly. Companies are wasting a lot of money re-inventing assets from the ground up for every triple A game.
It exists already. UE5 is integrated with Quixel MegaScans. Quixel is a company Epic acquired that does photogrammetry i.e. scans of real objects to film-quality assets. Like everything UE you can use it for free. You only pay once you have >$1M of revenue.
I think this idiot is saying artists should work for free so game publishers can make more money in the same way that thousands (millions?) of devs donate their time to make VCs richer by working on open source projects. No thanks. Need to eat. If you want to use my assets so you can make money selling your game then you can pay me.
Facial animations seemed fine for me. What got me off was surprisingly the body animations. When Keanu and Carrie-Anne were walking and doing hand gestures, it felt really strange. Too smooth.
A little later in the video where another character is walking, the butt motion is really weird. It's like someone decided that "her butt will wiggle, regardless of realism".
It's the small imperfections that give things away. Before, it was textures that were too perfectly smooth. These days we can make things look a bit more realistically imperfect.
Now, it's animations that are too perfect, as you've pointed out. Even a perfectly trained athlete or ballet dancer still moves in slightly irregular and imperfect ways.
People gliding over the ground is a huge giveaway as well, when you walk past someone else your feet don't slide to the right while you are still walking straight ahead, your movement changes.
In this demo, most still frames could often easily be confused with a still frame from one of the films. Blown away, there. You're so right about the lighting, models and textures being virtually indistinguishable.
However - in motion, the effect is completely broken - only every few seconds - but enough to push it into uncanny valley territory.
I notice Nirobi (? sic, the Jada-Pinkett Smith's character) in the backseat seemed to be the most offensive in terms of unnatural animation.
I'm guessing - I'm hoping - ML models could be of tremendous use in terms of improving animation in video games? The problem is always that it is too unnaturally smooth. People just don't move like that.
Just as we got texture detail to the point of perfection of the unnatural - pores on the face, etc - we need to get animation to the perfection of the unnatural.
Even freaking same with the camera. After getting used to VR, the weirdly static, linear, unnatural camera movements in regular console games have aged like milk. Not sure why we are still treating cameras in 3D like it's 2001.
One thing I find missing with facial animations often is blood flow under the skin. Skin often looks plastic because it doesn't change hue with pressure changes. An example of this would be firmly pressing your lips together - they'll go a shade lighter due to the pressure causing the blood to flow out of the lips. These things are almost imperceptible, but they go a long way to making something look alive.
There is something really weird and psychotic about, here's this technology that lets you render anything you can imagine in any way you can imagine, and what we keep trying to do is copy the way we already perceive the world as closely as possible. I wonder if this urge to copy reality is borne out of our biological urge to reproduce. Otherwise why would we find something as boring and unimaginative as photorealism so pleasing?
It's about pushing boundaries. Creative people see a boundary - they want to push it. Up until very recently it was about "how much polygons and textures you can put on screen". Now it's about "how real can you make it". Once we routinely get to the other side of uncannny valley and everything is real enough, we will see another boundary and creative people will push that instead.
even more psychotic: the real world but you get to kill anyone you want, and violence and death is basically the main focus of everything that happens in the world.
I totally agree. Any recommendations of games with great visuals that don’t focus on killing? I’d love to experience amazing virtual worlds without gameplay involving weapons.
All I can think of are sports games like FIFA, or side quests in Red Dead Redemption. Any other suggestions?
The Forgotten City, My Summer Car, Quantic Dream's games, Supermassive's games, Cloudpunks, Mind Scanners, Sable, Obduction, Superliminal, Eastward, Beautiful Desolation, Dropsy, If On A Winter's Night Four Travelers, Desert Child, Diaries of a Spaceport Janitor, Legend of Hand, Draugen, Felix The Reaper, Firewatch, The Talos Principle, The Witness
I am interepreting great visuals broadly to include great art direction, but there are some very nice looking realistic 3D games in that list
Some other ones that I don't particularly care for but that many people love: Disco Elysium, Outer Wilds, Return of the Obra Dinn
There are also many great games that feature combat as a core mechanic but are no more "about it" than The Odyssey or The Seven Samurai are. For example: Yakuza 0, Kenshi, INSOMNIA: The Ark, Vampyr, Valakas Story, Pathologic 2, Ghost of a Tale, A Plague Tale
If they are trying to make a game based off that universe, with the same actors/actresses, why would they not aim for the most realistic possible portrayal possible?
Your statement also ignores phenomenal games like Flower, Journey, NiGHTS Into Dreams - your statement even excludes the Mario series. Even 'Halo' has imaginative worlds that can benefit from photorealism.
Photorealism does not remove creativity. It's also a standard, a boundary to which digital artists have aspired to since the beginning of the medium.
You might as well ask why realist painting is a thing?
We don't - Fortnite and Minecraft are good examples of hit games that don't look photo real deliberately.
The Matrix demo is set in a city because (a) that's where the Matrix is set, and (b) it's relatively "easy" to do photo-realism in a city because you can use actual photos, and because other than the people the objects are reflective shiny boxes or surfaces.
Indeed, that is the answer. People need something that they can relate to, otherwise they will just shrug and move on. This is why in all sci-fi movies about the far future, that future in many ways always looks suspiciously like today; it is not for the lack of imagination on the part of the author, the goal here is precisely to leave some points of reference and prevent the audience from losing interest. (Hidden advertisement and product placement plays not a small role here as well.)
Is this all rendered models? I honestly thought much of it was just filmed humans. I mean, once they get in the car, it's pretty obvious it's a game, but before that, there were a lot of shots that looked filmed.
> The demo has two parts. The first is a narrative sequence that doesn't respond to button taps, and it opens with a digital facsimile of actor Keanu Reeves—not playing a character, just Keanu being Keanu—remarking on The Matrix. He strolls through moments from the first film while pondering the original trilogy's questions about what is and isn't real. (Every time he passes a mirror, the bearded, long-haired Reeves who appears in it is video footage; otherwise, he's all digital. This article doesn't include any real-life Keanu.) Soon after, a digital facsimile of his Matrix co-star Carrie-Anne Moss appears—she only appears as a digital avatar, never as filmed footage—and their conversation eventually warps into a car chase.
I'm fairly sure that quote is not factual. Many early appearances of Keanu are clearly a form of video footage; look at the later gameplay scenes to see the actual digital rendering of Keanu, with skin rendering not even matching ten year old games like LA Noire. Even when one Keanu is walking past another in the Morpheus scene, you can see that one is computer generated, with slightly better skin rendering than later, and one is video. Same goes for when Keanu walks through the bullet-time visual effects - the characters involved in the effects are clearly much lower fidelity than the filmed Keanu.
In the sequences prior to the car I think it's most obvious by looking at the eyes. Especially the wrinkles have an odd "glow" to them. I think they may have recreated all the sequences that look like they were taken out of the movie but it's hard to tell because many of them are shown on screens with a filter.
Honestly, after reading that supposedly some of the shots of current age Keanu are not rendered but real, I'm wondering if that's even true though. Maybe the odd appearance of the eyes is an artefact of the lighting of the all white room.
Probably not, the body animations from 1:40 look very unnatural to me, the scene at 2:05 looks worse than in the original movie and more like a ps4 game
Says who? There is place out there for both Minecraft and for extremely high detail one-off experiences, we aren't destined to all move into the Oasis.
It's the natural evolution of the internet as a medium. Technology is always born without content and leeches off older technologies content.
Theater -> Books
Music -> Radio
Books + Radio -> Movies
Movies + Books -> TV
Movies + TV -> Games
So far we have used books and movies as content, but the internets native content is an open action 3D MMO, it just takes time for new mediums to develop their own content type.
Why have a linear hardcoded storyline when the users can live out their own stories in the game?
I know it might be hard to see now, but we don't have the energy to not build it.
Also kids lead the way, and TV does not exist in their world, they call that YouTube and Twitch.
And while movies might be fun for them, the time spent in Movies vs. Minecraft + Roblox is an approximation error.
>>Why have a linear hardcoded storyline when the users can live out their own stories in the game?
Because a lot of people seem to enjoy exactly this - 10-30 hours of a very linear, hand crafted experience, not sandbox "you can do anything" kind of experience.
>>I know it might be hard to see now, but we don't have the energy to not build it.
I don't even understand what that sentence means, to be honest with you. I work for one of the largest video games companies in the world and we somehow still build both sandbox as well as single-player focused games.
>>Also kids lead the way, and TV does not exist in their world, they call that YouTube and Twitch.
Kids have zero purchasing power, and it's the money which decides what to make. Your "average" gamer, by industry's own stats, is a 20-30 year old male, who owns a console, and buys exactly 3 games a year - latest Fifa, latest CoD, and one other game(and companies spend hundreds of millions of dollars on marketing to occupy that 3rd spot).
Minecraft and Roblox are as big as they are not only because they are good games, but also because they are fantastic at extracting money from parents through children. Trying to figure out the future of gaming by looking at those two is like saying that future of movie making and holywood is animated cartoon short stories, since Peppa The Pig is unbelivably popular with kids.
I have seen the ‘gamers buy 3 games’ stated by various people before. Do you happen to know off hand what percentage of gamers that applies too. I am trying to figure out if I am just a rounding error when publishers and devs are trying to pitch and develop new games.
1) it's an industry average, so some people will only buy a console to play Fifa with their mates and literally never buy anything else, some people buy loads of games, but the average across the the industry is 3 games per year, and those games are extremely likely to be the ones listed above(in fact the 3rd one is usually, more often than not, a copy of GTA 5, as crazy as this is)
2) It only applies to people spending money. So if you just play WoW and nothing else, and never buy any new games, you aren't even counted.
3) Not all gamers are equally interested in all games. So of course a developer of a small indie game probably isn't aiming for the same crowd of Fifa/CoD buyers. Just like a small film director isn't making a decision on whether to make a movie based on whether they can compete with Marvel - they usually have different crowds. So yes, if you are predominantly interested in short, single player experiences in the £20 range, then probably no one at EA or Activision is pitching games for your crowd - but there will be other developers who do, and for whom you are the #1 target audience.
This is obviously absurd, or else there wouldn't be billions of dollars in advertising toys and kids games every Christmas season. They might not directly have purchasing power but they have it indirectly by convincing parents to buy stuff.
Yes, and these two aren't the same. (Young) kids have zero purchasing power, but like I said, Minecraft and Roblox are the behemoths they are because they are good at making kids convince their parents to spend money - that's the exact same point you just made, right?
>>The PS5 is basically a paper weight with these electricity prices.
Again, not sure what you mean - PS5 uses very very little power compared to a gaming PC, about 200W under load - my own GPU uses about 300W and that's without the rest of the system. An hour of gameplay on the PS5 costs about £0.05, and that's if you hit the electricity cap and are not on a fixed tariff. At my prices it's more like a two pence per hour. If these expenses turn your PS5 into a paperweight, I have to question how were you able to afford one in the first place.
And in addition - not sure what PS5 has to do with this?
>>So what happens when these kids grow up and don't have to nag their parents?
They buy whatever they like, not sure what you are trying to suggest here.
I disagree: video games do not have to converge on a single genre. Sometimes I feel energetic and ready for a brain challenge and I fire up Factorio, sometimes I want to play super-chess and I launch a Gary Grigsby game; sometimes I want to just immerse myself in another world and/or culture and I play Blackbook/Witcher/ACS, sometimes I'm ready for some fun with friends and it's CS:GO or R6, or Il-2 when I want the flying version of it; finally, when I want to just enjoy a good story, then it's time for Witcher or Metro.
From Everquest to PUBG which are the two only adult evolvers.
You can tell by looking at what they changed by how copied they became.
And Minecraft and Roblox which are the young evolvers.
Nobody has managed to copy Minecraft and Roblox successfully yet and the AAA studios are stuck in a grind making meaningless sequels.
But then if you look at the technology, you see that theres an obvious combination that is missing: Everquest - RPG + physical action gameplay like mario/zelda + modability of Roblox/Minecraft.
Single player cinematic games (without much freedom, not open world) are lovely experiences. Sure, I enjoy unwinding in Minecraft, but you seem weirdly dismissive of single player games with good writing. Some such games are cartoonist and wonderful, some are photorealistic and wonderful, and some are going beyond realism into something completely new (also wonderful).
As long as we still enjoy reading good books, we probably will continue enjoying watching good movies and playing good single player games.
My 3D MMO engine can now have 2000 non-instanced animated characters each with an interchangeable item equipped at 60 FPS without a world on a 1050Ti (80W GPU).
My target is to keep that number above 1000 when the world + physics is in.
On Jetson Nano (1/2 Nintendo Switch at 5W GPU) that number is 300 and I hope 150 on release!
The server can handle 10.000 action players (no spraying/headshot but mario/zelda mechanics/physics) at an avererage of 50 players in view on a large AWS machine.
Nobody has ever played an action MMO (with fun gameplay and physics) that plays like mario but scales to 10.000 players yet. When they do, they won't play anything else ever again (or do anything else in meatspace) is my prediction.
You are spot on; gameplay is more important than surface.
Are you doing your own engine or what scaffolding are you using? I'm assuming something that is pretty close to end vulkan? This is a topic close to my heart, lots of questions about how to scale, where to put the seams (node based loading has issues, have to be careful about loading screens, etc), how to keep player state backend access super fast once attributes bloat, how to keep low-lag player interaction when you want to force as close to a pure server-dominant state decision layer.
I think the tech itself is there just about, but there are some real technical limitations to why it hasn't been done (well) before. I'd love to hear how you are addressing these things.
As for the conversation about the matrix and the metaverse, my conclusion a long time ago was that the real matrix/metaverse will be FOSS. Thats why all these big companies (Meta, Epic, etc) dipping their toe into this admittedly juicy arena are going to fail in their vision. (also one of the reasons why I still have hope for a hacker culture of freedom in the future)
be warned though: with all this censorship they are coming for your game worlds next
I stopped at OpenGL (ES) 3 that has VAO (the last meaningful feature, internal improvements omitted, those can come later behind the scenes), Vulkan only scales the embarrasingly parallel stuff which does not impact gameplay.
(Mostly visual, zero motion-to-photon improvements have been made and again the meaningful practical internal improvements can be added retroactively in the OpenGL drivers.)
The last engine has to be modifyable and resellable, so the ownership model should probably avoid corporate models seen the legal/bureaucratic problems that brings. Right now I'm keeping all rights for myself and my son will probably inherit those.
I'm still experimenting with the licence pricing that you can find on my github sponsors page.
The b2c product should cost ~$10/year and gives you the right to mod and redistribute for free independently of me, something both Minecraft and Roblox lack.
Unfortunately the client will initially only be partially open-source; keeping the core engine .exe closed, and the game specific .dll open-source as an implementation example.
Also the cl.exe compiler is not redistributable, so I'm looking at solutions (use another compiler for the game .dll) for that until people move to linux:
They can't censor TCP/IP, or atleast not in a way I can't work around with my own encryption. The internet removes that possibility permanently, it's binary: either you have global connections and then the world is open or you don't and then we're all screwed back to the stone age.
Cool, but the problem is people ( marketing department included ) extrapolate this tech demo into something more, and we all know how that ends..
Maybe the most recent one is Cyberpunk 2077 and there were many reasons for that disaster other than the "tech", but it always ends with a couple million angry people robed and a few hundred burned-out employees with long lasting mental issues.
I might have been lucky (decent system) but I got through Cyberpunk 2077 fine, was not bug free but I enjoyed it, put like 100hrs into it over Dec 2020.
I'm not sure I should watch it. I've been actively avoiding trailers and any kind of spoiler of the movie. Does this contain spoilers? Or is it more like the Animatrix was to the sequels (e.g. provide valuable context before watching the movie)?
There are no spoilers of any kind. It is essentially completely disconnected from any of the actual films other than featuring avatars for two of the lead characters and revealing agent behavior that was made clear before the middle of the first Matrix film.
It's a video that shows the characters talking about the future of 3D reality and then a bunch of guns shooting out the tires of agents chasing them in cars.
While the technology is extremely impressive, I often find we over doing it to the point of un-canniness. For example, the dithering in the explosion/bullet effects appears to be confound with the over-sharping. As well, many of the effects are over done with the intensity of the particles and size of them. The gloss of the cars are extremely aggressive to show off the ray-tracing technology but it make the cars look goofy. I can see the de-noising failing with some surfaces because they are using ray-tracing as many places as possible.
I feel we already close to reality, but are in-human desire to over-crank things to justify the huge performance hits or showing off some these features is hurting us. Games that I see that don't jump largely from medium to ultra are often because they softly layer the effects to the point where it's not noticeable at first glance, but as time goes on it builds up immersion.
Detail is amazing, the technology is insane, and the attention to detail is jaw-dropping, I don't want to undermine the impressive work both from a technology as well as hard work implementing it.
Graphics look great, but gameplay looks dull. Same tired old stuff, shooting at prettier targets (none of which can shoot straight other than when the plot calls for it).
Probably also makes for a more predictable experience; you don’t want to show off your shiny new engine and have a subset of people getting a poor experience because of their varied hardware (which I guess is also a marketing reason)
Epic is worth almost $30b. They make the largest 3d engine and own Fortnite, a franchise worth billions on its own. They have their own game marketplace. Their legal team is squaring against Apple. They've got more resources than a tiny indy shop these days.
I think the first part with Long Haired Keanu is all filmed. He has all the details on his face and his face wobbles the way it does in real life. Same goes for the wrinkles on his shirt when he is touching the sofa.
The neo that wakes up in front of computer didn't look real. Also when he transformed to his young self, that immediately looked bad and definitely CGI.
The statement at start of video is not fair (a lie).
Even if it's a render, it's probably just 3d capture played back. You can see how unrealistic the young Keanu is versus the current one. The jaw specially moves in a weirdly inhuman way both in game play and in intro when he suddenly shape shifts.
This article is also speculation and I don't believe them. If it was all a render, all faces should have maintained the same detail and realism. But only Keanu's face is real enough. You can even see his individual hair which are out of shape. That's way too real. Those hair can't possibly be 3d captured like that.
As someone else pointed out, he does change to 3d render at 1:39 in a far view and you can easily see how unrealistic that one is.
Because it's a technically impressive "dragons lair". Ultimately, the game is selecting which circle you're shooting at and pressing the button. I'm not saying that this is the only game you can write with this engine, but as art costs go up, so the experience necessarily becomes more prescribed.
It's impressive, but I'm skeptical this video is real-time on a PS5, certainly I doubt this is at 4K - the jaggies when I ran this were everywhere.
It was worse in the city demo after the action sequence. Certainly a lot going on, but in terms of the trade-off, it was visually quite obvious where the costs are side-by-side with Miles Morales on it's PRT mode and for a real game the later would almost certainly be better overall.
Hopeful people will get more out of it in future, but wanted to temper expectations here from the video.
EDIT: Just found the signpost to enable real-time night mode with Lumen. Wow. It's clearly still struggling to run on a PS5 but it's not that far away from where it would need to be. And when it's not struggling with jaggies/framerate it's pretty jaw-dropping.