peter_spy

Lanczos or Mitchell, IIRC.

I think the footstep sounds question is rather tricky, I haven't checked the raw samples, but I bet they don't have similar compression / normalization levels, so they can be tweaked in relation to each other in sound shaders. As for realism, I'd ditch it in favor of familiarity and playability on the first possible occasion So the whole " why thief with hard soles" discussion is irrelevant to me, I have no problem with tile being loud, stone and wood being a bit more quiet, because that's a "surface difficulty level" convention known from Thief, etc., etc.

Never had any problems with one-time games, really. If it fits the purpose, like the story, type of gameplay, or what authors wanted to say in general, then it's fine for me. I gladly pay for 4-hour games that know exactly what they want to say in given amount of time, than titles that add replayability options or some ingame chores just for the sake of the length (hello Ubisoft).

I used this example mostly because Ubisoft tries so desperately not to make any political statements with their shooters. They were really defensive about that around when Division came out. They were also criticized for Far Cry 5 I believe, for having a sort of portrayal of modern South American gun nuts, but without doing anything specific about with it. I think there was even an article titled something like "Far Cry 5 and the desperate art of saying nothing" And as far as "murder simulator" part goes, with all the HUD indicators, damage numbers, health bars, etc. and pretty "industry standard" character animation, this doesn't give me that queasy feeling The Last of Us 2 does.

But they will attract politicians again, because at this level of fidelity they're very close to snuff films and "murderer simulators". Games have been on that path several times already while looking way more "innocent" than they do in this example. Hell, we're at the point where Mortal Kombat animators got PTSD from YT research they had to do for all the breaking bones or people being hanged. If you don't want to say anything political, you definitely don't do a game like that. You do a bland Ubisoft game like Far Cry whatevernumberitisnow.

Yeah, that was one of my thoughts as well, why do you perfect violence to such insane degree, while something as normal as two people making love looks like smashing plastic toys against each other? And I'm not buying a "serious adult story" bit that the game director is pitching. Maybe companies like Naughty Dog or Rockstar have driven themselves into a corner, because all they ever did was perfecting graphics and animation techniques, while the only means of interactivity they know is shooting people in the face. At this point they just have nothing else.

Sorry for double post but I guess it deserves it: While I cringe at this level of violence and meticulous animation that portrays it, I admit Naughty Dog have some of the best lighting artists out there. Especially since majority of lighting in the game is ambient.

The character is definitely last gen in comparison. And also it's not the same thing to have a 30-million poly static mesh (statue), and a 30-million poly animated character. I don't think it even would be possible to rig such complex model at this point. Simpler models, below 10 million poly, would probably crash a modeling app during rigging or export.

Substance Painter is for authoring materials, baking additional textures, painting over models, etc. The whole Substance suite is geared towards 3d artists, so they can make better materials, and it has been an industry standard for years. And there's nothing lazy about it. Making materials for games is awfully time consuming, and Substance makes it a little bit easier and faster. Again, no idea how's that connected with "making an engine look good". Good Substance materials will make models look good on any engine.

In general, FMs that are currently in beta should not be used as example to troubleshoot new version of the engine. There are too many variables to take into account, and usually bugs are result of mapper's work in progress. Released missions that have been thoroughly tested and updated by mappers post-launch are much more reliable source material, and even those can have bugs that were not caught around the time of release.

Not sure what you're talking about here. What tools, other than industry standard ones?

Intels still beat AMDs in games, but they are more expensive too. AMDs seem great at rendering, encoding and all that kind of stuff. This is where intel doesn't seem to keep up.

Meanwhile, Intel releases 10600K and 10600K chips that are still 14 nm

Small tangent here, but using suffixes for diffuse textures would probably make things easier to grasp, especially for new mappers. I use simple _d, _n, _s, _em (emissive mask) notation in all my materials, never had any problems with it.

I always thought megatexture was bit too extreme approach. I know that engines prefer as few textures as possible, and there's nothing wrong with making texture atlases for several models per UV space, I do that myself as often as possible. But I wouldn't want to work with megatextures in the long run, the balancing act most engines use today is an okay compromise IMO.

And that's what will become of 3d artists. Slider pushers will get peanuts (although 3d artists are not exactly gods even now). But IMO what's more important, if the majority of such "artist" community will not have basic artistic / technical skills, they will have no say on what direction the tools or the tech could or should take. They will just passively get on to whatever new thing is given to them. And that might ultimately be what's this all about. To make as many people as possible disposable, and lower the expertise to make them passive, so you're in the full control of what technology is being pushed and when. Btw. level designers are next, there are already neural networks learning how to furnish rooms, houses, public spaces, etc.

By the way: https://youtu.be/K6y9PJipfpk I forgot about that grading system in MGSV, and that's interesting because it does let you play more aggressively while it also rewards ghosting approach.

Idea: it could actually be more fun if the system was somewhat modifiable by mappers, or maybe rewritten entirely. Like having a number of points set by mapper (e.g. 100 or 1000) where detections would deduct from the point total. Although I don't like the fact that even suspicion lvl 2 counts towards that penalty. Searches I can agree with. Or, maybe focus on positive reinforcement, like in latest Hitman games. Get points for beating a level faster, getting more loot, getting special loot items, performing special actions and finding more secrets. That would encourage replayability, give more stuff to do for more hardcore / competitive players or even speedrunners.

With Tencent's money, they're trying to cover all fronts, but megascans content does run like shit even on decent modern machines. And that auto LODing system that will come with UE5 will require some serious hard drive throughput. There are rumors that PS5 has some next-gen SSD that allows it, so that would make sense. But I doubt indie studios will want to force highest hardware requirements for their games. Either that or they'll be spending most of their time optimizing someone else's models. Not to mention that there are quite many indie studios specializing in stylized art (pixel, voxel, handpainted), so no jobs for slider pushers there I wouldn't be surprised though, if a UE5/Megascans combo will be heavily used in VFX and Cinematography. Prototyping lighting and camerawork will be much faster than doing tests on location only. Side note: I know that I'm getting grumpy, but the whole thing kind of makes me like idTech4 even more. Its limitations (no GI, old shaders, etc.) are opportunity for you to grow as 3d artist, and you will see that improvement in the scenes you make. If you have no idea about how light works, how surfaces react to light, how objects are made, and you don't really care about all that, it will be reflected in the quality of your maps and models. But once you start learning all this, or if you have some experience in drawing, photography, etc., the quality of your work will become better and better. It's a joyous process and no amount of sliders and picklists will give you that

Btw. I've seen that Rebirth cinematic: https://youtu.be/9fC20NWhx4s IMO it's a very good representation of what AAA industry is right now: style over substance lvl 1000. Ultimately, the tech and resources mean nothing to the audience, when the script is at this level of shallow and cringe. Although the tech nerds in comments under this video are getting orgasms one by one.

A bit more grounded discussion on UE5 demo: https://www.youtube.com/watch?v=5KgWJ1RxDhw It also aligns with what I've been thinking, e.g. that the quality of the character model is still last gen. Or that these 30-million polygon models are just impossible to have in a production environment. I've also checked the latest Megascans video: And you know what, I thought I had a pretty beefy rig. This scene runs at 20 fps for me, in a viewport. In fullscreen mode is more like 5 fps. And it really doesn't look much better than e.g. last Uncharted games. But right now I kind of understand why they've been pushing this "no-polygon limit" approach – so they can bind you to use Megascans and their hugely unoptimised assets. This scene is also not possible to have in a current production environment (i.e. a full game with AI logic, etc.). But AAA companies will jump at this as fast as they can, because the workflow they propose in this video is hugely beneficial to them. Once Megascans libraries are comprehensive enough, the role of "3d artist" will be to search through them, find an asset, choose from available dropdowns, push some sliders on a material level and insert asset into a level. In a sense, there will be no need for a 3d artist, as level designers could easily do that job themselves, or there will be a highly replaceable "asset finder" post, where any monkey with a week of training would do. But that's AAA business I guess, indie games will still hire modelers and 3d artists with a classic skill set.

IMO it's rather cumbersome and time-consuming. If your time is limited, you'd probably want to spend it on better level design or other aspects of the actual mission.

We're talking about streamlining and simplifying the asset workflow, so using FBX just to add another conversion step defeats the whole purpose.

This is pivot point, it works like that in all modeling software. ASE uses 0,0,0 point (scene origin) as an absolute pivot point for any model you want to export. Yes, and that gives you a unified format for both static and animated meshes, so you don't have to maintain separate plugins for md5 (which are outdated or hard to find as well). The same will eventually happen with any user-maintained plugins as well. And IIUC, you don't have to support all the options e.g. lights. Never seen any options on video clips though. You can embed media, but that means texture files, not videos. It is proprietary, although Autodesk provides C++ SDK with format description and header files for readers and writers. It's not true that it's only binary, you can save FBX either as binary or as ASCII.

Wonder how you're going to approach it though. Main advantage of FBX is that you can select any model in your scene and export it – modelling software and game engines will recognize its pivot point placement and rotation. Its material assignment will also be maintained when the model is reexported after some edits later. Main problem with ASE is that it actually does what it stands for: ASCII scene export. It exports all the information about all materials used in the scene and puts it in the model. It also requires the model pivot to be at 0,0,0 point in your scene. And you have to reedit the material path manually in notepad, each time you reexport the model. Assuming it's even possible to make ASE work more like FBX, how do you want to achieve it? People use different modelling packages, will you make tools for 3dsmax, Maya and Blender, to cover the most popular ones? Will you maintain the Blender add-on after each Blender update? Edit: I didn't even mention animation, as with FBX you get unified format for both static and animated models, which resolves the problem with ancient plugins for md5.