Asset Update to Ogre3D Discussion Thread

[AeroPhonicsOfficial]

Side notes:

All assets should be originals!

We can try recruiting artist on https://www.tesrenewal.com

Original style with edits for realism and expansion and making to more than 50 polys

I plan on re-learning C++ and Ogre3D so that I can implement the advanced features post 1.0 and now

Carpe Diem!

[Okulo]

Our last contact with a Skywind asset creator was one that accused us of stealing, I think. I'm not sure the people working at Skywind entirely understand OpenMW or the kind of IP-culture that OpenMW fans like you and me represent... I'd gladly see some of them prove me wrong, though.

[Berandas]

Side notes:
All assets should be originals!

This is what I was discussing with Deonsion before.
If the goal of this project is to make models as vanilla-like as possible, just in higher resolution, it's unnecessary to create everything from scratch. We can simply use dozens of meshes from various replacement mods if the authors agrees, then do the rest by ourselves.
As mentioned on the first page, creating new meshes from scratch that are pure copy of vanilla asset is still a derivate work, so that solves nothing, it's just an extra work.
When we make new meshes that are not just like vanilla, it's simply "an another replacer" and completely loses the point of this project. Also the new meshes would have to be as much vanilla-shaped as possible to prevent various errors ingame (floating, bleeding, caspering, inaccessible areas, etc.). And basically this is something similar to what SkyWind project is doing already.

[AeroPhonicsOfficial]

Not a pure copy and a new textures, most replacers deviate from the originals. I'll work and asset and show you what I mean

[Tarius]

I am away for a while and this gets started.

First, as people have already suggested, look through whats already been done, no sense in doing whats already finished and acceptable when you have on the order of several thousand assets to do.

Second, some models(maybe all) will need to be matched almost perfectly or you WILL end up with bleeding and various visual placement issues.

Third, I am one of the people who has made replacement models from scratch including some new textures for them. I matched them almost perfectly, but mine are alot smoother.(my mod was listed in that list posted earlier) Its one thing to make a model how you want, but its more work to make sure it matches the original geometry really well. This is very important and the importance can not be understated.

Fourth, there are many many very good texture replacers out there. No, not all of them much the original style but many are quite good. This may be a long shot idea, but if someone could find and persuade connary to release his texture compendium to us, that would basically finish most of the textures.


With the above said, I would suggest that bodies be the first thing completed as that would be not only the most visually noticable work, but is also one of the things that is needed most dire.

[Tolchock]

Hiya -- I'm afraid I am no modeler (as of yet!), but I would really like to contribute to this project. Perhaps I could help by working on tedious admin tasks, so the artists can make the best use of their time. I could help organise the claim system, and document what models need to be done. I could take before and after screenshots of *all* the models, so there is a way to record the project's progress (also, it would be really great for interior/exterior designers and texture artists to have a good idea of what is available).

I agree with AeroPhonicsOfficial's view on the workflow, but I think it might also be a good idea to consider reference count, as well as difficulty, when replacing the models (so the most common and easiest items are done first).

Here are some examples of how claims could be organised (having a google docs spreadsheet would be great):

Velothi Religious furniture

Spoiler: Show

f\Furn_velothi_altar_01.nif --1251 refs
f\Furn_velothi_prayer_stool_01.nif -- 780 refs
http://1drv.ms/1vePTFN

Claimed by:
Complete (y/n):

Vines

Spoiler: Show

f/Flora_BC_Vine_01.nif -- 962 refs
f/Flora_BC_Vine_02.nif -- 887 refs
f/Flora_BC_Vine_03.nif -- 1059 refs
f/Flora_BC_Vine_04.nif -- 502 refs
f/Flora_BC_Vine_05.nif -- 214 refs
f/Flora_BC_Vine_06.nif -- 192 refs
f/Flora_BC_Vine_07.nif -- 399 refs
http://1drv.ms/1pJdcqR

Claimed by:
Complete (y/n):

(refcount is for Vanilla Morrowind only)

[AeroPhonicsOfficial]

Or start with the hardest, that way it is mostly done when the first phase is done

P.s. Refcount is a definite must take into account, good idea!

[MacKom]

Fourth, there are many many very good texture replacers out there. No, not all of them much the original style but many are quite good. This may be a long shot idea, but if someone could find and persuade connary to release his texture compendium to us, that would basically finish most of the textures.

I`m not so sure this would be a good idea. True, there are some nice texture replacers out there, but the majority of them is just upscaled and sharpened version of vanilla textures, and it shows. Besides, if we were to make new models as replacers, it would take a lot of time and effort to make sure they match up to a specific replacer perfectly. Sometimes it`s just better to start over, doing the sculpt, baking out maps, etc. The result ends up looking much cleaner.

[Berandas]

As for the textures, I would go use the same procedure as for meshes - look foor what's already done. But don't be hasty, we don't even have meshes in process, so let's wait with textures for now.

Everyone talking about hipoly sculpts, but as I mentioned before, it's quite unnecessary to create hi-poly sculpt for most meshes. Objects like tables and pots are not like heads, they do not use single texture. You would have to bake new texture with completely new UV unwrap for every object. That would mean less detail in total and incompatibility with other already made texture replacers.

[Tarius]

Fourth, there are many many very good texture replacers out there. No, not all of them much the original style but many are quite good. This may be a long shot idea, but if someone could find and persuade connary to release his texture compendium to us, that would basically finish most of the textures.

I`m not so sure this would be a good idea. True, there are some nice texture replacers out there, but the majority of them is just upscaled and sharpened version of vanilla textures, and it shows. Besides, if we were to make new models as replacers, it would take a lot of time and effort to make sure they match up to a specific replacer perfectly. Sometimes it`s just better to start over, doing the sculpt, baking out maps, etc. The result ends up looking much cleaner.

That isnt at all what I meant. And it would take more time to make a new texture than it does to to shove textures in a folder and get rid of the ones that dont match very well.

but the majority of them is just upscaled and sharpened version of vanilla textures,

And this is completely and utterly not true.

You know how I know what I know about MW texture replacers? I went through several thousand textures by hand to see how well they match up to the originals taking only the closest and the best for my game. Am I that crazy? Yes, I might be.

That would mean less detail in total and incompatibility with other already made texture replacers.

This is a very good point, if you are going to redo the models, another reason to keep them as close as possible is because of the textures. Thats the single most time consuming thing in many cases, the effort needed for UV mapping.