Fallout 4 in Noesis

[MrAdults]

In fact, we were pretty sure you were simply working off of our 010 templates, which were publicly pushed to GitHub daily since the 11th, hence extremely easy for you to find with a code search. (You admitted to searching for 'BSTriShape' for example.)

https://www.google.com/?gws_rd=ssl#q=%22BSTriShape%22
Oh, is that so? Sure don't see your article there! And no, I haven't seen any of your templates or any reference material, or I would've happily credited you as I credit everyone in the Noesis readme.txt any time I use any of their material.

I don't know what you're referring to me having trouble with exactly. It seems like you're talking about vertex flags, but I know that you couldn't possibly be. You're not handling them correctly either.

I explicitly noted that the current public release doesn't do this, and that your post, linked above, explicitly mentions you haven't got it figured it out. I don't see how you couldn't have it figured out if you spent more than a day with the format, you simply need to take the time to go through and relate the flags being used to known vertex buffer configurations, and xor'ing flags between known buffers makes this pretty easy.

The vertex flags are not simple OR obvious.

Yes they are.

We've analyzed all the vertex flags with 010 and there are 17 unique vertex flags across all NIF files.

That's not very many, as model formats go.

Anyway, since we found out about this, we've found your pompous attitude very offputting, so I personally don't have a desire to interface with you any further.

I merely pointed out the questionable timing and the fact that your above linked post mentions issues with the things I didn't bother spec'ing out in the script. You're the one being a total asshole here instead of simply posting and saying "Nope, just coincidence - no hard feelings."

Your figuring out *anything* in a "half a day" was highly suspect too, since we'd already published everything by that point.

Oh, horse shit. I've reverse engineered games 1000 times more complex than this. These NIF chunks were incredibly simple to eyeball.

From our point of view, you don't seem to know anything we didn't know a week ago, save for displaying skinned models correctly. And at least for me that's simply because I know nothing about skinning and haven't even bothered to look into it.

Well, congratulations, I guess? I certainly didn't find it by googling.

I think maybe what you're not understanding is that what you've done in "half a day" is fine for Noesis, but not for the actual tools. We need to treat the format rigorously, and it's clear you're not doing so.

I'm glad you've suddenly decided to believe my timeline now. Your release is already pretty half-assed and incomplete, apparently, so I don't see how that's at all "rigorous" either.

I guess before I go I'll leave you with a quiz regarding the vertex format:

This is another load of horse shit. You wouldn't have approached this thread in the way you have if you weren't a complete tool, so I'm not going to answer questions to "prove" anything to you when the catalog of hundreds of more complex games I've reversed speaks pretty well for me.

... and no cheating. Some of this information is already public!

God, you're an idiot.

P.S. Don't flatter yourself, I haven't looked at your source.

Based on your psychotic reaction, either you have severe Asperger's, or you're lying. I'm willing to believe the former, I suppose.

I'm not sure if our IRC logs are available somewhere, but they will attest to that. I'll let you know if our IRC logs are available somewhere and you will be free to look over the past two weeks and see for yourself that you had absolutely zero to do with our work.

I guess at least I can get an ego boost out of multiple people spending two weeks on something that took half a day, then?

[MrAdults]

And if you want to talk about something that should take 2 weeks, this shit should take 2 weeks: http://www.richwhitehouse.com/index.php?postid=68 And it did. In real life time, anyway. I have a job and a family.

A few new, incredibly small, NIF objects? Yeah, no, not 2 fuckin' weeks. So I could accept the possibility that you didn't actually look at the Noesis script, but you're just making it harder with your reaction. But I can also accept that you're just butthurt because you spent a lot of time on it because you're new to this. So that's fine too. I reserve my judgment there, so don't make me hate you any more than I already do.

[throttlekitty]

No mention of Noesis, that's a shame. Have to wonder about the timing, especially given how simple and obvious the vertex flags are, but that's the only thing he seems to be having trouble with and the only thing I didn't bother fleshing out and documenting in the script. Maybe just a coincidence, I suppose.

I hadn't heard of it until a few days ago, pretty neat app you've got, It's something I sure could've used a few years ago.

All Noesis is doing with the bindpose mats is moving the actual bind pose to match the node space. Since it converts the position data up to 32 bits in the process, there's no reason for the bind pose given to exist anymore. This means that the node space pose now equals the bind pose. The ideal bind pose can also easily be recalculated on export by fitting joints into a more ideal precision range for FP16, then transforming geo with (newInvBindPose * fittedPose).

I *think* I get it now, thanks. NifSkope has similar code already for the previous formats, but the inverse was stored directly prior to fo4 and I think we just ran off that.

jon are you being fucking serious right now. I see how his tone is boastful but certainly not threatening. clam yo tits.
Sure, we aim for a more complete/proper tool, but Noesis isn't necessarily that. Just pointing out that we had verts down pretty quick too but ran around trying to figure out what all the off-cases were with the flags.

[MrAdults]

I *think* I get it now, thanks. NifSkope has similar code already for the previous formats, but the inverse was stored directly prior to fo4 and I think we just ran off that.

Yeah, it's a pretty traditional Gamebryo concept (you'll notice the other geo paths in Noesis do similar things too), but it's particularly important in FO4 for the bind pose to be centered at 0 to make the best use of FP16 precision. It seems they decided to capitalize on the cruel Gamebryo legacy of "EVERYTHING HAS A TRANSFORM", which was not a bad idea at all. They could have squeezed even a bit more precision out of it by relying on a non-uniform bind pose in order to quantize to FP16 with maximum precision utilization, then bake that inverse non-uniform transform right into the pose matrices so it would be back in uniform space by the time you concatenated anything to the transform. But they seem to have done pretty well either way, I didn't see any nasty cracks or anything as a result of FP16 positions and it's an elegant way to handle the need for a scale + bias.

jon are you being fucking serious right now. I see how his tone is boastful but certainly not threatening. clam yo tits.
Sure, we aim for a more complete/proper tool, but Noesis isn't necessarily that. Just pointing out that we had verts down pretty quick too but ran around trying to figure out what all the off-cases were with the flags.

Oh, I didn't realize you worked on this as well. That eases my initial suspicion, as you don't seem at all like a flaming dickface.

[throttlekitty]

Yeah, it's a pretty traditional Gamebryo concept (you'll notice the other geo paths in Noesis do similar things too), but it's particularly important in FO4 for the bind pose to be centered at 0 to make the best use of FP16 precision. It seems they decided to capitalize on the cruel Gamebryo legacy of "EVERYTHING HAS A TRANSFORM", which was not a bad idea at all. They could have squeezed even a bit more precision out of it by relying on a non-uniform bind pose in order to quantize to FP16 with maximum precision utilization, then bake that inverse non-uniform transform right into the pose matrices so it would be back in uniform space by the time you concatenated anything to the transform. But they seem to have done pretty well either way, I didn't see any nasty cracks or anything as a result of FP16 positions and it's an elegant way to handle the need for a scale + bias.

I'm used to seeing it for heads because of the prior use of FaceGen, but they aren't really using that now. NetImmerse/Gamebryo files have always been considered 'scene snapshots' rather than 'just a model'. One game had title/main menu/options/char select in a single big .nif, which was a... creative use after seeing most games done modular.

I've spoken with others who complained that their games had big FP seam issues on large meshes, I was surprised to see Bethesda do it, given their style.

Oh, I didn't realize you worked on this as well. That eases my initial suspicion, as you don't seem at all like a flaming dickface.

Honestly I've done little for this round, they had the easy stuff done before I had the time or game. (again with the lack of proper skills for much of this, I'm more an artist, and less protective of my work/discoveries) I'll probably spend the weekend horsing around with shader flags and values.

[Relzaz]

hello, i am new ( and i don't speak english very well) i want to have the armor model but i don't know how to find it .
If someone can help me i will be very happy

[MrAdults]

hello, i am new ( and i don't speak english very well) i want to have the armor model but i don't know how to find it .
If someone can help me i will be very happy

If you're talking about the one in the screenshot, it's just one of the loading screen models.

I'm used to seeing it for heads because of the prior use of FaceGen, but they aren't really using that now. NetImmerse/Gamebryo files have always been considered 'scene snapshots' rather than 'just a model'. One game had title/main menu/options/char select in a single big .nif, which was a... creative use after seeing most games done modular.

The significant downside of (ab)using the bind transform in this way is that it adds another step to the transform process. A lot of older games, and even some newer ones (the popular example is Doom 3) will put skinned verts in local space so that the need to compute a skinning matrix from the inverse bind pose is eliminated, making the "bind" pose is effectively identity. When normals/etc. come along for the ride, this represents a further deterioration if you're just doing a linear transform, though.

1) Do you know where the bitangent is located? (Hint: It's a doozy)

I just noticed this one when this thread got juggled back up. Now, I haven't even looked back at FO4 (I meant it when I said I wasn't going to work on it again until I fnished playing it, and then the Bloodborne DLC came out), but pretty much no one stores binormals/bitangents, and we always store the sign to restore after the CP. Typically in the position, normal, or tangent w, but it can be anywhere you feel like packing it. This behavior is so common that Noesis supports it by default (and has for years), along with a variety of flipping and transform modes, when providing the tan4 type. So I really hope that's what you thought was a "doozy". Additionally, it's also very common to use one of the specialized formats like 10-10-10-2 to maximize precision and store the sign in 2 bits. Noesis also supports a variety of these types natively. The world, shockingly, doesn't consist only of Bethesda games, so, welcome to 3D graphics circa 2006. Medium fish, tiny pond.

Sorry, had to get that off my chest, just glancing through that pile of ignorance-arrogance special-blend again was irritating.

[Relzaz]

I finally understand =
Thanks

[Relzaz]

actually I don't understand...

[devilblades]

So, not all of my textures are loading any idea what is going on, or how to get them all to load? Here is an example, some textures will load, while others wont.

[throttlekitty]

Check your texture pathing? file paths sometimes start with textures\ and sometimes they don't. (ie textures\cat\catbody.dds vs cat\catbody.dds)

[yukitairo]

Thanks! For this, but idk what's the problem here?



That happens with every models when I use the -nifskel thing

[MrAdults]

Regarding no textures, what throttlekitty said, or they're just unhandled material refs. Regarding the skeleton error, try an absolute path.

[yukitairo]

thanx, but..yea I'm kinda new to noesis. lol
absolute path?

[volfin]

I was pleasantly surprised to see Noesis had support for Fallout 4 so quickly. Awesome job.

I did run into a curious problem though. I exported a model from Noesis in FBX format, and tried to load it into Blender using their FBX importer. However, their Importer is complaining about the FBX version. The error says "Version 6100 not supported, Must be 7100 or later."

Is there an easy way to change an FBX's version #, or can Noesis save in newer FBX formats? Thanks