45 réponses à ce sujet

[Shadooow]

Hey, I would like to propose a project like improving default tilesets in NWN. It could fit into the CTP project, although its dead, but I could help.

Problem is, lets say I will look for forest tileset on vault. I find dozen of forest tileset combos (and dozens reskins) and most of them are very similar just one extra terrain or tile over other ones etc. Tried most of them myself and afterall made my own compilation. I dont want to force anyone to use my own compilation, but could we agreed on what should it contain and what shouldn't and make a community approved compilation?

Speaking of forest, there are lots of options:
- henesua or DLA canopies
- elven tree houses
- water terrain (high/low shore)
- trees terrain (rural-like, in several different texture variants)
- high/low terrain (classic rural style and TNO style too)

Other tilesets don't have so many custom tiles created but I was able to create some tile variants simply by removing some polygons from tile. Like in rural tileset there are two pillars in "road to forest exit" tile. Ive created a copy without those pillars and voila - new tile.

Maybe even CEP would be interested? AFAIK CEP is adding some tilesets but from what Ive seen and tried they are simple reskins without any single custom tile, not really sure whats point of that but nvm.

[Bannor Bloodfist]

CTP was working on a huge forest combo pack. With a lot of different terrains and crossers.

However, I need to warn you, raise/lower typically ads dozens of tiles, add more than one crosser that will interact with the raised section and the number jumps again.

Making two different terrains to be able to work with each other adds roughly 30 or so tiles on top of the two terrains.

In other words, tilesets get huge REALLY fast. Our version had 9 terrains and 9 crossers (none of which were fully complete yet) and totaled over 755 tiles. The incompleteness stems from missing tiles for specific locations where two terrains or crossers interact. Likely the set would have grown by at least 100 more tiles.

That is a basic set though, IE one color variation on foliage etc... to add additional variations, you either duplicate nearly every tile to allow for the variation or restrict what sections/groups/features are to be duplicated for the next variant.

Texture wise, you are talking a huge undertaking as well. Many textures can be easily replaced, but there are also many that can't simply be replaced without re-uvw mapping the corresponding tiles.

Basically any large tileset will require a fairly skilled and fairly large team of folks. Folks willing to spend an exorbitant amount of time creating/modifying and fixing tiles.

That team also needs to plan ahead and actually AGREE on stylistic issues and plans before you even get started.... then stop the never-ending tendency to bloat with "this new idea" "this new terrain" etc. Believe me, that last bit about stopping bloat is a killer in more ways than one.


Believe me, I am not attempting to talk you out of attempting such a project, I am just warning you that the entire amount of work involved is exponentially larger than anyone who has not attempted it, realizes.

Modifié par Bannor Bloodfist, 07 juillet 2012 - 08:57 .

[NWN_baba yaga]

I agree with Bannor about the problem large tilesets seems to have. The work realy explodes when you have to make every terrain/ crosser compatible with each other and it´s a neverending process that just tedious imo. I learned a lesson and will only do smaller tilesets in scope in the future.

[henesua]

FYI - ShaDoOoW, I didn't release any art for NWN. I think you mean Helvene. I don't want to receive credit for other's work. I'm so rusty at CC that I couldn't even complete a couple trees for the October CCC.

Anyway, I like the idea of a refresh for all the standard tilesets. But defer to the better judgement of the tileset experts.

[Shadooow]

henesua: yes typoed the name, I indeed meant Helvene

Bannor: yes i know the caveats. And I dont even say that these features must be there. My own compilation in fact doesnt include elven trees, neither raise/lower, neither airships. I had my reasons not to include them.

Shame that CTP didnt finished it, the work CTP has done with tilesets is really amazing. Tilesets in CTP meet all standards I could think of. Not big, yet offering many tiles and variations, textures matching default nwn etc. The tilesets I found on vault don't meed these often.

At least, perhaps we could make a list of custom tiles and features and their authors.

[Bannor Bloodfist]

I thought I could maybe help with the list bit... maybe but I just checked my two partitions that had all the original content broken out and documented... both sets of folders have nothing but sub folders, no actual files.

I will send out some emails to see if anyone else from the prior team still has that documentation though, we might get lucky or I might have a cd/dvd with some backups. I sincerely hope so. I am normally very anal when it comes to keeping multiple copies of things.

Edit: I do still have a copy of the ctp_forest at whatever stage it was left in, but I am missing all the documentation of where the source files came from.

Modifié par Bannor Bloodfist, 08 juillet 2012 - 04:11 .

[Master Jax]

I've actually been working on this for a couple of years now. Hope your idea gets momentum, though! I am presently working on the TNO tileset and Rural one back to back, because they look very similar due to my use of resources. The issue in my case, as I have mentioned before, is I use other tilesets as resource, and because of this, they are a mix of models and textures from both original content and custom content authors. I can't release it publicly because of this, and because the whole thing is just too heavy to upload. Even if we would split things so as to upload them, it would still be a little on the heavy side, and this without adding new tiles or anything, just overriding the existing ones!

If some of those issues were solved though, I would be more than happy to share these once I am done! It's the least I could do for a community that has given so much over the years!

[Rolo Kipp]

<putting his own...>

I also like the *concept*.
Respecting original authors is crucial, but every tileset out there could be improved, bug-fixed or tweaked.

One idea might be to offer a centralized "Community Tileset Reference" project on the Vault and collect the links to existing tilesets that have restricted rights (notably Babylon) and updates/tweaks for derivative works.

It could also be a "clearing house" for announcing/previewing new tilesets in the single most stable (if old and creaky) repository we have.

Perhaps starting with a transcription of the excellent tileset list by Skunkeen, if permission can be obtained.

<...shoulder to the wheel>

[OldTimeRadio]

Bannor Bloodfist wrote...
However, I need to warn you, raise/lower typically ads dozens of tiles, add more than one crosser that will interact with the raised section and the number jumps again.

Making two different terrains to be able to work with each other adds roughly 30 or so tiles on top of the two terrains.

In other words, tilesets get huge REALLY fast.

Bannor, do you happen to know of a place where this sort of breakdown of the number of tiles which are required might be covered in more detail?  I think it's extremely useful information.  I'm still new at doing tilesets and, for what I was trying to do, I was absolutely happy with just a single level of tiles w/out doing the raise/lower thing.  I'd love to see more actual numbers of how these sorts of choices affect how big a tileset gets.  Do you know if someone ever broke the math of it down?

Thanks!

Modifié par OldTimeRadio, 10 juillet 2012 - 03:35 .

[Zwerkules]

OldTimeRadio wrote...

Bannor, do you happen to know of a place where this sort of breakdown of the number of tiles which are required might be covered in more detail?  I think it's extremely useful information.  I'm still new at doing tilesets and, for what I was trying to do, I was absolutely happy with just a single level of tiles w/out doing the raise/lower thing.  I'd love to see more actual numbers of how these sorts of choices affect how big a tileset gets.  Do you know if someone ever broke the math of it down?

Thanks!

I once found that information by chance when I was looking for something else in the NWN omnibus. When I really needed that information and started looking for it, I never found it again.
There was a list of which tiles were needed if you combined two, three or four terrains (without raised terrain).
I don't remember who wrote that list. That would have made searching for it a lot easier.
Maybe someone has seen the list too, knows where to find it or who wrote it.

[NWN_baba yaga]

@OTR:
Thats a pretty good tileset. I realy like your style and textures are awesome. Nice details on everything. Any plans to release it some time soon?

[Shadooow]

Rolo Kipp wrote...

Respecting original authors is crucial, but every tileset out there could be improved, bug-fixed or tweaked.

IMO each tile should have tilefade (unless its walkable on higher level) and use default textures from that tileset (unless using new one ofc). Then we could use the concept of animloops, toggling the various splatches on the ground etc.. off.

Using the concept of raise/lower in dungeons (not sure when Ive seen it now) allowing to paint corridors on the edge of raised terrain to connect both levels should be also essential. Ceilings and probably compatibility with NWNCQ should be also guaranted.

[OldTimeRadio]

@Zwerkules  - I hate that, heh heh!  Just knowing that it was spotted swimming in the Omnibus is a good sign that it can be tracked down again, though. 

@NWN_baba yaga - Thank you- but models and textures are unashamedly taken from Titan Quest.  We have a copy for each of our computers here but I'm pretty sure I got the actual textures from a pack on Garrysmod.org.  There's like a zillion people who play Half-Life2/Garrysmod so pretty much every game's models and textures are ported over there for the last several years.  Mostly, the models and textures I use are usually either from some other game or from the web.  I tend to actually spend most of my time doing more MaxScripty, environment-mappy, skin-weight-y, walk-mesh-y, pheno-type-y, animation-y sort of things.

[NWN_baba yaga]

hehe yeah ok OTR. I tend to use source assets from various places too as a reference. But when I´m actually done with the concept I rework the needed stuff;) I might want to check out some titan quest... lol:D

[Bannor Bloodfist]

Actually OTR I think the absolute minimum is 3, a left, right and center tile. Plus a 4th for edge, actually 3 for edges, so 6.

That is not giving any variations. Most tilesets have at least 2 if not 3 or 4 variations for each individual tile location. so, 3 tiles, with 3 variants each, and 3 edge tiles, would give you 12.

That is without ANY crossers. If you have crossers (streams, bridges, roads) etc, then the number goes up again, but not quite as high. You won't need crossers on the left and right corners as they don't have the ability to use a crosser. But you need variants to allow the crosser to dead end at top, dead end at bottom, and connect through top/bottom... so 3 more tiles, and more with variants. Most crosser sets have limited number of variants on raised bits though, typically only one, sometimes two.

Now, if you have more than one base terrain, that connects to another base terrain, and you wish to be able to raise/lower both terrains, the number goes up again. As you need the original 12 per terrain, plus variants that allow both types of terrain at top and bottom, or meeting in center top and bottom. Say sand/grass... you need all the variants for sand alone, and all the variants for grass alone, then the additional variants where the top half is sand and bottom is grass, top half is grass and bottom is sand, and variants where the top is split sand/grass and bottom is split sand/grass. plus top split grass/sand and bottom split grass/sand.  And to be honest, to be fully workable, you need variants on the corners too.  Top sand, bottom grass (all 3 edges) Top sand, bottom grass on two edges, and Top sand, bottom grass on ONE edge.  As well as the oposite having Top grass and bottom sand for each of the variants.

NONE of this is all that difficult to create, but it DOES take time. Time to create and time to test.

Modifié par Bannor Bloodfist, 10 juillet 2012 - 10:49 .

[OldTimeRadio]

Thank you very much, Bannor!  If I said I could really encompass that all on one reading, I'd be lying.  Bookmarked to hammer into my skull over time.

[Michael DarkAngel]

If I may add something to this that might possibly help.

Bare minimums needed for tileset creation:

• Single Terrain -- 2 tiles (1 terrain, 1 edge) Tile #1 from the image below will serve both purposes.
• Single Terrain w/Raise-Lower -- 6 tiles (4 terrain, 2 edge) Tiles #1 thru #4 with #1 and #4 dual-purposing as edge tiles.
• Single Terrain w/Crosser -- 5 tiles (1 terrain, 2 crosser, 2 edge) Tiles #1, #5 and #6 with #1 and #5 dual-purposing as edge tiles.
• Single Terrain w/Crosser and Raise-Lower -- 13 Tiles (4 terrain, 6 crosser, 3 edge) Tiles #1 thru #10 with #1, #4 and #5 dual-purposing as edge tiles.

To this, feel free to add as many variations, tile features, tile groups as you see fit.  As you add more terrains or crossers, you add exponentially to the amount of minimum tiles required.
[EDIT] Find a better tile representation of the above image -->> Gallery  -->> Discussion

 MDA

Modifié par Michael DarkAngel, 12 juillet 2012 - 02:25 .

[_six]

MDA, tile 9 is exactly the same as tile 10, at least from the perspective of a .set entry. However you are missing the r T and X shaped crosser junctions from your diagram, which the vast majority of crossers will always have. 
It also doesn't consider placing a crosser across a raise edge (which admittedly is a fairly specialist use but can be seen done quite effectively in TNO or Worms' sets). 

Of course, the absolute bare minimum to place a crosser is just tile 6 alone, so I'm just trying to point that crossers (and to a lesser extent terrains) always depend on precisely what they're intended to be used for, for what is necessary in terms of tiles. Ergo it's rather academic and misleading to try and throw numbers out there for what's needed without knowing precisely what is being implemented.

Two terrains and one crosser might warrant 30 tiles to implement if you require the crosser to paint around all possible orientations of the terrains with just a couple of those tiles having variations. But if its just a straight bridge across a pit, you might only need 3 tiles.



A slightly more on-topic point... I'm not entirely sure there's a huge amount of point these days to adding to the base tilesets more than has already been done so. It sounds the OP is talking about cherry picking existing add-ons to the tilesets, which to be honest seems more pointless to me, as you'd be making new areas anyway so why not just use new 'instances' of the tilesets like the CTP did? I'd be more interested in improvements to whats already there, a la what neverroofers did with the caves tileset override.

Modifié par _six, 11 juillet 2012 - 07:18 .

[OldTimeRadio]

Nice, thanks!

[NWN_baba yaga]

I dont think it´s wise to improve the default tilesets anymore and the time could be spent into newer ones instead. The sewer (now we have the PQ/_six tileset or DLA anyway), demonic dungeon (ugh) and especially the city exterior just looks so dull to me. If we create an alternative tileset list for the default ones and add/ enhance these I´m in.

[Pstemarie]

Well...what do we have for replacement tilesets out there.

1. Rural = Castle Exterior Rural (with Q mods)
2. Forest = Baba's Project or Tom Banjo's Forest or Six's Wildwoods
3. Dungeon = Maxam's classic Dungeon or Versatile Dungeon, Six's Catacombs
4. City = Zwerkules Medieval City or Six's Wildlands
5. City Interior = Project Q City Interior or City Interior 2 (with Q mods)
6. Castle Interior = Project Q Castle Interior or Castle Interior 2 (with Q mods)
7. Underdark = Baba's Underdark or Sen's Underdark
8. Mines = Never Roofer's Mines

And that's just the ones I use. In reality I don't think we need a project either that refurbishes the default Bioware tilesets. The Community has already replaced most of them with much better sets. What would be better IMO is if someone started a project that looked at all these high quality tilesets and what was needed to make them work nice together.Basically you'd be looking at a tophak project that merged conflicting 2das.

[Krevett]

Indeed a tophak merging all those fantastic sets would be sooooo nice!

[_six]

We'd have to make sure every time someone updates a tileset hakpak, we'd be able to update the merger. Wait, I have an idea that makes lexical sense...

*hollers* ROLO!

[Rolo Kipp]

<collecting...>

And the sticking point continues to be embedded data in the mod and intelligent parsing/modification thereof.

Specifically, the tiles referenced in the mod *must* be correctly defined in the set.

However, I still think we can create a wrapper that assembles mod resources at *run-time* (skipping if it's already built) and enable asynchronous content generation. It is simply absurd that you have to massage an entire tileset when you add a single tile or even a feature that re-uses already present tiles.

The rules for .set files are known. We *can* automate this.

*sigh*

I guess I really should move my Ruby lessons up a burner or two... We really do need this :-/

Edit: In the meantime, Bannor, MDA... who owns the Community Tileset Project now?
What challenges would we have to revive the organization with a slightly different mission - to coordinate the existing, 3rd party tilesets (and new *cough* Babayaga *cough* ones that come along) into, as Paul points out, a comprehensive replacement superset?

<...a few more gems>

Modifié par Rolo Kipp, 11 juillet 2012 - 03:29 .

[NWN_baba yaga]

My offer is to create requested or specific features/ groups to these tilesets and make sure that my own ones work with all the other sets listed above. 2da and itp editing is realy not my thing guys:D