TDM missions...

[Bikerdude]

Yeah, we should have thrown out the really bad ones before release, now we are stuck with subpar textures forever, because someone might be furious that we replace them.

Personally I wouldn't mind as long as the alignment was exactly the same. But on another note we have only released a handful of maps so far, so isn't it better to fix the crappy textures now rather than later.?

[jdude]

Jdude: Since you are unable participate in this discussion without resorting to insults and strawman arguments ("Excuse me if I don't feel like making the same 'break into some guys house or business' mission 100x over", where the hell did you get that from?!), I will ignore further posts from you in this thread. That is all.

Good.

[Springheel]

And as for me 'going in the right direction' Excuse me if I don't feel like making the same 'break into some guys house or business' mission 100x over I thought maybe people would enjoy some variety in their gameplay.

 

Melan was actually paying you a compliment, which makes your flying off the handle all the stranger.

[jdude]

But so is having artistic consistency... so mixing Doom 3 light and Baked light could also be an artistic disaster ...

 

 

This is why I'm upset. I heard maybe 2 complaints about my mixture of baked and dynamic lights in Saintlucia. Baked and dynamic lights are NOT mutually exclusive. It just seems like there's all these generalizations of what should and should not be done in the level design.

[nbohr1more]

You won't get complaint from me. I think St Lucia looks great!

[Bikerdude]

, it's just that the people complaining haven't taken the initiative to make a level with these things and that those who have are still working on them. Or maybe they don't have the creativity to do so, which is what I've been saying since the start. The impression I'm getting is people here are saying the thief 1 engine looks better than the doom3 engine because of what their capable of which is the real topic at hand here because there seems to be this notion that we can't DO what T1/T2 maps have done with D3 tech.

 

Jdude, WTF...??!!

 

I appreciate your passionate about TDM, but copping an attitude when some of us express ourselfs (be it misdirected) to improve the mod is at best counter productive. Your having a pop at me for starting the thread, a pop at Melan for opening his mouth and questioning mine and anyone else's creative ability which is just plain nasty... And regarding the D3/T2 argument, there are still things T2 does better which is a fact - but that is for people with far more knowledge than me to discuss.

 

If you want less threads like this one then you have 2 options: 1. either ignore them or 2. help people like myself, Melan & others improve their skills.

 

 

Spings, might I suggest moving this thread to the off topic area, which I should have maybe started it to begin with..?

Edited July 7, 2010 by Bikerdude

[Serpentine]

Replacing textures really isn't all that hard or problematic, so long as you keep the aspect ratio between versions, alignment etc will be the same.

I try keep things very uniform and using late stage blending keep quiet a bit of the subtle things from the lower res version.

 

Finding new source to replace weak textures is a bit more challenging, because you need to decide what aspects of the old texture were desirable at some level and try keep all the individual elements (if there are any), takes a whole lot of googling and digging through galleries most of the time.

 

However at the moment most of the work that I'm doing is just balancing and trying to keep textures in some kind of uniform colouring, brightness and such. There are a number of textures which are ideal for some spots, but because the bricks or whatever are the wrong colour/too bright it sticks out like a sore thumb. For example :

 

On the left is the current 1.02 version... it's quite... ugly. Matching the bricks in the second frame in the first and playing around with levels get's things to be a bit 'nicer', then playing with a bunch of random stuff brings the plaster into a more acceptable brightness to use in something like TDM. Nothing finite, still need to check in game, but a decent example of trying to keep things in line with each other.

[nbohr1more]

(again I am speaking from ignorance...)

 

Since we don't have Megatexture tech, to improve the organic variety of textures would it be possible to create "Texture Sets" groups of individual textures that when combined properly will create a complex pattern like a big patch of brickwork being exposed from behind weathered clay but across several texture tiles in a complex fashion?

 

I guess decals could do that too...

 

More decals?

Edited July 7, 2010 by nbohr1more

[Serpentine]

We do need a lot more decals, they really do help to make places look far better than bare brushes.

 

In reality most of the challenge is finding a large enough collection of them to batch process. It's not particularly hard to separate grime etc from the background, but doing one picture is somewhat unsatisfying. Finding ones that are also not distorted, makes it even more challenging - you don't want to take details of the bricks etc that the grime is on, and trying to clone brush and merge things to cover the imprint of larger elements can be insanely hard.

 

And re megatexture etc, it's most likely time to stop looking at things we cant feasibly and ~easily~ do at present.

[nbohr1more]

I was just using that as an extreme example of texture variety... That tech is on the GPL wish list but not really on my wish list

 

 

(unless the megatexture doom 3 mod could be used to store a huge light texture and essential be used as a light map instead of a terrain... but that's my crazy homeless guy type thinking...)

Edited July 7, 2010 by nbohr1more

[Bikerdude]

the one on the far right is awesome Springs

Edited July 7, 2010 by Bikerdude

[Springheel]

There are a number of textures which are ideal for some spots, but because the bricks or whatever are the wrong colour/too bright it sticks out like a sore thumb.

 

Yep, that's a major stumbling block for our textures. Because they were contributed by multiple people from multiple sources over the span of several years, they have little or no coherence. It would be useful to have "sets" that were built to be used with each other, but that requires people who are willing to stick with it and do a lot of texture work, and such people have been sparse.

[Serpentine]

I have no problem with doing it(texture work is quite therapeutic, just zone out), it's just that I have a terrible memory and need people to report them to me, otherwise most of which I find is technical issues, which are good to catch but not particularly exciting/interesting to the end user

 

Thread to report such things - report everything, even if you can give me a vague description

[Fidcal]

Further on my 'add to rather than change' texture view, these are the pros and cons I see:

 

Adding instead of replacing:

increases the download.

increases diskspace.

gives stability to existing textures (mappers feel they can trust what textures they have selected not to be changed.)

New materials can be better grouped; old ones cannot be regrouped without breaking maps (black texture.) Unlike models and entities, the virtual path is stored in the map as part of the texture definition name.

 

Replacing instead of adding:

Clear errors can be fixed (improved bumpmapping, quality, etc.)

Existing maps can be improved without further effort.

 

Note the above two methods are not mutually exclusive and fixing bad textures is still a good thing so long as it does not radically change the mappers' original choice.

 

On the download/disk size we know bandwidth and disk space keep improving so let's see what typical losses might be:

 

Browsing the dds textures I see 171K, 342K, and 683K. The new texture whether replacement or addition balances itself out so only the old size matters. Taking 342K as average (though it may well be most 'bad' textures tend to be smaller) how much download/diskspace might we save? Can there be 100 'bad' textures? Suppose then 100 which might be 34200K or 35MB. And suppose one quarter would be better added to rather than replaced. That would be 9MB extra. Is it worth worring about?

 

Anyway, always worth putting out these ideas before any rampant changes without realizing the consequences.

[Fidcal]

It's funny, I just remembered, I recently changed a badly tiling texture - the fake reflect 02 where there was a clear error in the corner where the moon is (should be in the next update.) That is a good example of replacing a texture rather than adding a new one. Nobody is going to complain about that texture being fixed because all you will see is a bad artifact no longer present.

[Sotha]

Interesting discussion here.

A mapper point of view: if the texture looks horrible, most of the mappers will see and understand this and probably will not use it at all.

 

My point is that really bad textures are probably used relatively rarily.

 

I do not mind slight improvements to textures, but major alterations will probably result in angry mappers when their huge texture alignment efforts go to waste.

 

But yeah, there are some poor textures in TDM and that's why I seem to choose only a few good ones for each materials and stick to them.

[SneaksieDave]

We will get there.

at the moment most of the work that I'm doing is just balancing and trying to keep textures in some kind of uniform colouring, brightness and such... Matching the bricks in the second frame in the first and playing around with levels get's things to be a bit 'nicer', then playing with a bunch of random stuff brings the plaster into a more acceptable brightness to use in something like TDM. Nothing finite, still need to check in game, but a decent example of trying to keep things in line with each other... I have no problem with doing it

 

May the Builder bless you, sirs.

[Tels]

Interesting discussion here.

A mapper point of view: if the texture looks horrible, most of the mappers will see and understand this and probably will not use it at all.

 

My point is that really bad textures are probably used relatively rarily.

 

I do not mind slight improvements to textures, but major alterations will probably result in angry mappers when their huge texture alignment efforts go to waste.

 

But yeah, there are some poor textures in TDM and that's why I seem to choose only a few good ones for each materials and stick to them.

 

I think it goes without saying that a replacement:

 

* MUST have the same size/alignment. Replacing 512x512 with 1024x1024 is a no-go (edit: unless you add "scale 0.5, 0.5" to the material shader, but I think it might still break maps?)

* MUST have about the same colors: replacing yellow bricks with brown ones won't work

 

But apart from that, I think things can be improved. (Just look at all the "hires texture packs for thief 1/2/doom3/unrealwhatever").

[nbohr1more]

Sorry to seem like a suck-up/kiss ass/etc but why do your posts always seem to make so much sense Tels?

I have Swedish blood in my veins is it some kinda Nordic kinship?

 

I vote with for sensible texture replacement ASAP.

 

If FMS authors balk at the changes they can always include the old ones (renamed) in their PK4's right?

[OrbWeaver]

* MUST have the same size/alignment. Replacing 512x512 with 1024x1024 is a no-go (edit: unless you add "scale 0.5, 0.5" to the material shader, but I think it might still break maps?)

 

Changing the resolution of an image shouldn't affect alignment in any existing maps, because the texture coordinates are stored as normalised UV values not absolute pixel positions.

[Melan]

* MUST have the same size/alignment. Replacing 512x512 with 1024x1024 is a no-go (edit: unless you add "scale 0.5, 0.5" to the material shader, but I think it might still break maps?)

This should be confirmed by a dev just to make 1000% sure, but when I replaced several custom textures with higher resolution alternatives, I didn't retexture my testmap but alignment remained correct.

 

[edit]Beaten horribly while I went and checked things.

 

[edit2]Take a look at this swap I just did with two recently upgraded checkerboard tile textures:

 

 

The shots are taken from an identical camera position. The textures in the first image are 512x512, while in the second they are 1024x1024.

Edited July 8, 2010 by Melan

[Serpentine]

* MUST have the same size/alignment. Replacing 512x512 with 1024x1024 is a no-go (edit: unless you add "scale 0.5, 0.5" to the material shader, but I think it might still break maps?)

That's not how textures work, it runs on the relative sizing of the texture, if it didn't then DR editor images would not match up remotely

So long as aspect ratio stays the same it's all good (but whoever made a 512x32px iron bar texture... I'm coming for you, tonight... you.)

 

* MUST have about the same colors: replacing yellow bricks with brown ones won't work

Sure, unless they are crap colours, i.e the pink bricks etc a few posts above.

 

Edit : Beaten even more horribly while I went to make examples!

[New Horizon]

Sure, unless they are crap colours, i.e the pink bricks etc a few posts above.

 

Yup, if they're being changed for the greater good...such as darkening a texture to have the colours better match and make a full set, then that has to be an exception. It may or may not ruffle a feather here and there, but we have to remember that we're still pretty early in development.

 

As far as TDM missions having a way to go, of course they do. Are we even really out of beta yet? I thought we were still considering the 1.0's to be beta. Comparing betas to completely finished products? It's nothing to get worked up about. I'm extremely happy with where TDM is at this stage. It's fantastic, and will only get better.

[Melan]

Well, well, well. I am currently uploading some 186 MB worth of textures to a place Serpentine can download them and decide over which get in a coming update. Some statistics:

 

///textures/darkmod/metal/detailed (8)

///textures/darkmod/metal/grate/ (1)

///textures/darkmod/plaster/ (6)

///textures/darkmod/stone/brick/ (12)

///textures/darkmod/stone/flat/ (6)

///textures/darkmod/stone/cobblestones/ (1) (Only one, but it's going to be popular.)

///textures/darkmod/stone/sculpted/ (44), ...of which: TRIMS (23), PANELS (6), SHIELDS (9), OTHERS (6)

//textures/darkmod/window/ (10)

 

That's a total of 102 not counting some stained glass textures which Serpentine will have to make normals for. As you can see, the biggest category is detailed stonework to spice up walls with. A later, smaller batch will have more windows, but I had to make a cut somewhere.

 

Here are some examples (max 256x256 editor images)

 

(familiar? )

 

 

 

 

 

 

 

 

 

 

 

 

[edit]Let's hope the upload doesn't break now... [/edit]

Edited July 8, 2010 by Melan

[Tels]

Changing the resolution of an image shouldn't affect alignment in any existing maps, because the texture coordinates are stored as normalised UV values not absolute pixel positions.

 

Ah, I wasn't so sure, because you can also scale textures freely, and if the scaling was multiplied by the texture size, increasing the size would indeed misalign it.

 

But good to know it works, so we can even up the resolution of the textures that are too low (a lot of our wall/floor/door textures cover greater areas, but are about twice the size than textures on some models that only cover 1/10 of the area - causing the blurry walls to look even more blurry).