Hard Light Productions Forums
Community Projects => The FreeSpace Upgrade Project => Topic started by: FreeSpaceFreak on November 01, 2010, 01:46:57 pm
-
It's now done!
Plug & Play pack (http://www.mediafire.com/?1kz54fk8sty7dl7) (7MB) - contains game-ready POF and diffuse, glow, shine and normal map at 2048^2 resolution DDS.
Map pack (http://www.mediafire.com/?9j5vg64qvzigutv) (68.5MB) - contains all possible maps in 4096^2 resolution PNG.
Enjoy!
[ORIGINAL POST]
After the Akrotiri and several other high-poly Ancients, this one deserves a makeover as well - after all, it's a canon one. I know Aardwolf started this too a while ago, but he told me he doesn't intend to finish it anymore, hence why I take a shot at it.
It's currently at ~13.6k tris, the curves take up quite some. I was wondering whether to keep the lines clean, as they are now, or add some greebles to them? I'm also not too sure about the 'tail' of the large part, the original model is... inconclusive in that area.
(http://img831.imageshack.us/img831/5262/knossoswip01.jpg)
(http://img594.imageshack.us/img594/4648/knossoswip02.jpg)
(http://img828.imageshack.us/img828/113/knossoswip03.jpg)
(http://img263.imageshack.us/img263/9606/knossoswip04.jpg)
(http://img178.imageshack.us/img178/251/knossoswip05.jpg)
(http://img202.imageshack.us/img202/6186/knossoswip06.jpg)
-
Hmmm... maybe I should wait to create any more FS2 ANIs until this is finished?
-
Honestly, I feel like what you have there is way too over-smoothed, to the point that the poly expenditure doesn't garner any real visual effect at all. I'd much rather see the Knossos go the greeble route, but maybe that's just me.
-
I agree.
-
Honestly, I feel like what you have there is way too over-smoothed, to the point that the poly expenditure doesn't garner any real visual effect at all. I'd much rather see the Knossos go the greeble route, but maybe that's just me.
This, basically.
-
I'd actually go for a balance of both. Tone down the smoothing a bit to allow actual detail on the model.
-
WAY oversmoothed. One or maybe two iterations of the MeshSmooth function would serve you better. Then you should work on modeling in the macroscopic parts of the texture, like the canyons that have the energy lines or the dark techy parts under overlapping scales.
-
Right, well, this is how it looks at 4.6k. But if polycount is important, I probably shouldn't be modelling in wavy recesses for the glowy lines and techy parts...
(http://img33.imageshack.us/img33/7746/knossoswip07.jpg)
(http://img227.imageshack.us/img227/8748/knossoswip08.jpg)
Original textures, for reference:
(http://i93.photobucket.com/albums/l77/Aardwolf001/Knossos%20Upgrade/original_texture_1.png) (http://i93.photobucket.com/albums/l77/Aardwolf001/Knossos%20Upgrade/original_texture_2.png) (http://i93.photobucket.com/albums/l77/Aardwolf001/Knossos%20Upgrade/original_texture_3.png)
-
Is greebling really important on this model? The parts are flying at such a speed that you can hardly make out the small details after all.
Or is tehre a way to deactivate the rotation?
-
There, that looks much better. Polycount isn't really an issue, but it's always good practice to be efficient.
You may want to do your modeling on the low-poly mesh first, then smooth it, as it will be much easier and probably look better than trying to model on the new high-poly mesh. And as Norbert has said, the thing's going to be in motion most of the time.
Go for the really big details that affect the silhouette and the lighting; you can clearly see on the original textures how there's plates on top, and then an unarmored surface. Keep in mind what details will be communicated with the normal maps. Don't feel restricted by the textures' positions on the original model, either. So long as it looks great, close is good enough--like the Cain and Hatshepsut, for example.
You should also try and get the geometry on the individual parts to line up as they rotate. And remember that the whole thing is going to eventually be UV mapped and baked.
-
Right, well, this is how it looks at 4.6k. But if polycount is important, I probably shouldn't be modelling in wavy recesses for the glowy lines and techy parts...
It's not that we don't want high poly counts; it's that we don't want unjustified poly counts. Please go ahead and model the details if you want, since those are visible usages of polies and will greatly improve the look of your model.
-
Don't feel restricted by the textures' positions on the original model, either. So long as it looks great, close is good enough--like the Cain and Hatshepsut, for example.
Hmm, well what's "close"? :drevil:
(http://img207.imageshack.us/img207/4503/knossoswipr01.jpg)
(http://img155.imageshack.us/img155/5256/knossoswipr02.jpg)
-
Sure. That looks pretty good. However, that's looking almost Protoss-y, and I don't exactly see Ancient technology with those delicate protrusions (two front spikes) and such smooth curves. You'd think that a portal that's been around for millenia would look a bit more ruggedly built.
I do like the general direction you're taking it though. Just keep in mind those textures when you're making the model. I'd personally like to see some more waviness in things like the grooves.
Something that I think might be closer would be a blend of (http://images2.wikia.nocookie.net/__cb20090805034153/crysis/images/thumb/5/50/Alien_Scout.png/830px-Alien_Scout.png)
and covenant technology. That's just my preference though, of course.
-
lol the original knossos textures has bits of Orion tile on it.
-
Sure. That looks pretty good. However, that's looking almost Protoss-y, and I don't exactly see Ancient technology with those delicate protrusions (two front spikes) and such smooth curves. You'd think that a portal that's been around for millenia would look a bit more ruggedly built.
What the ****. What natural phenomenon in space is going to wear down a construction such as this?
Disregard that message, it looks good like it is and I think you should keep going in that direction.
-
What the ****. What natural phenomenon in space is going to wear down a construction such as this?
Micro-meteor impacts is all I can think of at the moment.
I dunno. I doubt cosmic rays or solar wind cause any real damage.
-
What the ****. What natural phenomenon in space is going to wear down a construction such as this?
Micro-meteor impacts is all I can think of at the moment.
I dunno. I doubt cosmic rays or solar wind cause any real damage.
Gamma Draconis didn't have any planets, stars, etc in it, so there likely weren't any asteroids/meteors either.
-
What the ****. What natural phenomenon in space is going to wear down a construction such as this?
Micro-meteor impacts is all I can think of at the moment.
I dunno. I doubt cosmic rays or solar wind cause any real damage.
Gamma Draconis didn't have any planets, stars, etc in it, so there likely weren't any asteroids/meteors either.
Didn't...have...any...stars?
What about Gamma Draconis?
-
Beat me to it.
-
What the ****. What natural phenomenon in space is going to wear down a construction such as this?
Micro-meteor impacts is all I can think of at the moment.
I dunno. I doubt cosmic rays or solar wind cause any real damage.
Gamma Draconis didn't have any planets, stars, etc in it, so there likely weren't any asteroids/meteors either.
Didn't...have...any...stars?
What about Gamma Draconis?
That wasn't a star... it was a Had--!
-
Derp, didn't mean to put stars but instead meant to put moons.
-
Odds are it had plenty of **** tumbling around, though.
-
Yeah but the chances of any of them hitting and doing any damage to the Knossos, even in 1000 years, is very, very small. And yes, I know the Knossos wouldn't be moving for most of that seeing how it wasn't powered up, etc
-
Yeah but the chances of any of them hitting and doing any damage to the Knossos, even in 1000 years, is very, very small. And yes, I know the Knossos wouldn't be moving for most of that seeing how it wasn't powered up, etc
Man homie are you kidding with me yo, the Knossos is going to be hurtling around the system at tens of thousands of kilometers per hour or some ungodly huge number like that, crashing into swarms of every imaginable kind of space detritus from simple solar photons to high-energy cosmic rays to dust and gas to honest-to-Buddha micrometeorites.
-
Yeah but the chances of any of them hitting and doing any damage to the Knossos, even in 1000 years, is very, very small. And yes, I know the Knossos wouldn't be moving for most of that seeing how it wasn't powered up, etc
Man homie are you kidding with me yo, the Knossos is going to be hurtling around the system at tens of thousands of kilometers per hour or some ungodly huge number like that, crashing into swarms of every imaginable kind of space detritus from simple solar photons to high-energy cosmic rays to dust and gas to honest-to-Buddha micrometeorites.
:(
-
Yeah but the chances of any of them hitting and doing any damage to the Knossos, even in 1000 years, is very, very small. And yes, I know the Knossos wouldn't be moving for most of that seeing how it wasn't powered up, etc
Man homie are you kidding with me yo, the Knossos is going to be hurtling around the system at tens of thousands of kilometers per hour or some ungodly huge number like that, crashing into swarms of every imaginable kind of space detritus from simple solar photons to high-energy cosmic rays to dust and gas to honest-to-Buddha micrometeorites.
:(
orbits, *****es!
-
On that note, if the device was deactivated for all those years, how come its individual parts didn't separate due to gravity (or rather, lack of momentum)?
-
because it's Ancient (unknown, mysterious) technology.
-
High Technobabblon particle flux. :nervous:
I think that is momentum though, or lack of acceleration.
-
Yeah, but the individual segments weren't rotating until Bosch activated it.
-
This mysterious phenomenon called "gameplay" held it together.
-
Yeah, but the individual segments weren't rotating until Bosch activated it.
Oh. Didn't know/remember that.
-
This mysterious phenomenon called "gameplay" held it together.
This is the best explanation I've heard! I vote it becomes canon.
-
They were asymptotically bound.
-
...we should probably stop derailing this any further. Let's get moar pics/suggestions!
-
Yeah, agreed. Greebles? :3
-
No greebles till the mesh is improved.
-
Some work on the front, making those odd protrusions look a little less fragile, also making sure the other rotating parts fit underneath.
(http://img99.imageshack.us/img99/6271/knossoswipr03.jpg)
Yeah, I'll be adding some volume to the tails.
(http://img510.imageshack.us/img510/4003/knossoswipr04.jpg)
-
*fanwank*
-
I'm still not sure I like how the front spikes looked. Maybe it's just because they're quite different from Vanilla. /personal preference.
Otherwise, sexy :D
-
I'm not liking the fangs, but otherwise :yes:
-
These double tentacles may or may not be a good idea...
(http://img573.imageshack.us/img573/8372/knossoswipr05.jpg)
As for the tails, I went with thin blades instead of thick ones, since the original looks kinda fragile there - the thin middle section, then the heavy tail volume behind it.
(http://img408.imageshack.us/img408/235/knossoswipr06.jpg)
-
Hmm. I'm thinking that it's still looking too streamlined, especially with the second set of spikes that contour the front (what happened to the blunt "nose" of the the profile view?), and the thinner rear section. I still think of the Knossos as a more chunky construction, akin to SUV rather than a sports car.
If that's what you're going for, then great. Especially since there are no other canon Ancient constructions, you do have a fair bit of free reign here, as long as it doesn't "feel" wrong.
-
I get the feeling that the model's got no volume or that's too thin in some areas.
-
Am I the only one who see's a Vasudan in that mesh?
-
Am I the only one who see's a Vasudan in that mesh?
****ING DAMMIT
CANNOT UNSEE
-
Still don't like the sharpness of the fangs...
-
Am I the only one who see's a Vasudan in that mesh?
****ING DAMMIT
CANNOT UNSEE
Ditto.
-
I feel it too far a departure from the cutscene anims Knossos.
/$.02
-
I feel it too far a departure from the cutscene anims Knossos.
/$.02
So many people said the same thing about c914's Cain / Lilith and everyone loved it in the end. :P
-
Nice work, I encourage you to get the model done as quickly as possible. :) One, nearly irrilevant critique: IMHO, there should be more spikes on the top.
-
Well, since the general opinion seems to not like the sharp things at the back, I redid them, adding some more volum, also closer to the original now. This better?
(http://img19.imageshack.us/img19/9056/knossoswipr07.jpg)
(http://img716.imageshack.us/img716/1151/knossoswipr08.jpg)
-
Yeah, looks nice. What's the spiky things between them though?
-
Decoration?
-
[insert technobabble here].
And that's why it's needed for [phlebotinium] to work. :)
-
Cool. I think it's kinda weird that the fins now face each other though. I mean, the thing is supposed to beam phlebotinum at the center of the portal, not into its own components. It also means that the top profile now looks somewhat more boring.
An idea is maybe make it so that the spikes look like they interact with the inner rotating bit somehow. Could look pretty cool.
(Are your screenshots isometric or perspective? I'm asking because I'm having a hard time telling. I'm guessing it's iso though)
-
The spiky things are the supports of those fins. Also, what Dragon said :)
Yeah, the shots are isometric.
-
Noted. That probably explains why I'm getting a feeling of "something about it looks wrong/feels different from vanilla" :P
-
Yeah, that one varies so much from the original I'm not really sure I'm down with it.
-
Could we get side-by-sides with the original?
-
(http://img834.imageshack.us/img834/49/knossoswipr11.jpg)
The pink line outlines the new one...
(http://img203.imageshack.us/img203/4723/knossoswipr10.jpg)
A perspective shot:
(http://img190.imageshack.us/img190/3454/knossoswipr09.jpg)
-
Well done. :yes:
-
Sweet. A perspective view screenie. I can now say that I love the head.
(Tail spikes/tentacles make it look like a squiddy now though)
-
I feel like the back needs to remain solid.
-
I feel like the back needs to remain solid.
I would agree with this. The back looks a little fragile. But the head looks really good, much better than the original!
-
You know how the fangs on the original kind of converge towards the center of the model to form a V-shaped trench? Might be better to try and do something with that idea or it might not. (Also, I think you could think about making the fangs longer, just to make it seem like there is greater interaction between inner/outer components).
-
I have to say this:
squid attack!!
I would say it's better if the back part is modeled as a single body and not that way.
I realize however the amount of work on the back part might have been quite big, so I'll just go along with what you decide about it.
Aside from that, the overall model looks good :yes:
-
Yeah, the back part probably needs to be more solid (and a bit wider too).
squid attack!!
(http://img.photobucket.com/albums/v423/kc1991/1305.jpg)
pic related.
-
Perspective shot looks great, keep it up. :yes:
-
looks good, but i have to agree with others here, the back should be solid.
-
I think it looks pretty squitastic, de gesu
-
Okay gotcha, the back will remain solid. How's this? The recesses on the rear fins will be those techy bits on the texture.
(Pink line outlines the new one again, I scaled up the head a little)
(http://img241.imageshack.us/img241/3312/knossoswipr12.jpg)
(http://img221.imageshack.us/img221/9013/knossoswipr13.jpg)
-
Oooh! That is good, do more of that! :yes:
-
That's more like it!
-
Get rid of the wavy fins at the bottom, make it rather one single arc instead of three. Other then that, looks good. Just my 2 cents.
-
That's more like it!
+1
+9001
-
Much better! Much closer to the original, more potential there indeed.
Get rid of the wavy fins at the bottom, make it rather one single arc instead of three.
I agree; don't slave yourself to the texture mapping on those particular polygons, make that area a nice smooth arc.
-
I think the earlier, spikier version looks more than a bit Shivan. Might be interresting if someone could make fitting textures to turn that into a Shivan version of the portal.
-
Good, very good.... I like how this is turning, keep it up FSF! :yes:
-
I think the earlier, spikier version looks more than a bit Shivan. Might be interresting if someone could make fitting textures to turn that into a Shivan version of the portal.
That might be useful if anyone updates Derelict.
-
I think the earlier, spikier version looks more than a bit Shivan. Might be interresting if someone could make fitting textures to turn that into a Shivan version of the portal.
That might be useful if anyone updates Derelict.
Too bad, I've already overwritten that :nervous:
-
Oh well. It's not important.
-
One model at a time people...
-
FSF, could we get a wireframe?
-
So, uhh, like this then?
(http://img149.imageshack.us/img149/6093/knossoswipr14.jpg)
Wireframe: Yeah, I'll be doing some optimization.
(http://img337.imageshack.us/img337/5291/knossoswipwire01.png)
-
Niiiice. :D
-
Very nice. This is coming along awesomely.
-
Right, is the big part pretty much done then?
On to the small one:
(http://img140.imageshack.us/img140/2694/knossoswipr15jpg.png)
(http://img52.imageshack.us/img52/4029/knossoswipr16jpg.png)
-
Wow, the enthusiasm... I guess it's time to wrap this up?
(http://img146.imageshack.us/img146/1773/knossoswipr17.png)
(http://img252.imageshack.us/img252/2500/knossoswipr18.png)
(http://img840.imageshack.us/img840/9262/knossoswipr19.png)
Blendfile attached for backupping purposes.
[attachment deleted by admin]
-
Looks good to me :)
-
:yes:
-
Needs a few more polygons around the edges to smooth things out. Otherwise good :nod:
-
K. I guess we want this neatly unwrapped onto a single texture?
-
I think they need some less smooth areas, with some detail in them. I recall that in the original knossos for instance, one of the modules has 3 circles\cylinders in the front. Those are good to contrast with the smoothness in the rest of the design
-
I see what you mean, but wouldn't the textures take care of that?
(http://img24.imageshack.us/img24/7514/knossoswipr20.png)
-
Huh, I never really noticed them having so many circles.
-
If you could make the edges between the panels and the recessed bits wavy like the texture, that would be many orders of magnitude more badass.
-
If you could make the edges between the panels and the recessed bits wavy like the texture, that would be many orders of magnitude more badass.
So it would, most definitely. Unfortunately it would also be many orders of magnitude more polies...
-
Well, how many are we at right now? I'm not really digging all the straight lines that have shown up, so I'm wondering what could be done. (POM anyone? jks)
Also, could you slide the circles on the second half of the head so that they're facing up a bit more? It seems to me they're the components that interact with the outer ring with phlebotinium radiation, and should therefore should point up a bit more.
Again, I don't have any experience with modeling, so if what I'm asking for is unfeasible/stupid, please ignore.
-
Hmm, well right now it's at about 15.7k polies. With some optimization, I may be able to get it down to 14-14.5k. In order to get all the wavy edges modelled in, that polycount would be at least doubled, so count on 28-29k. Hmm, that's actually not that bad... Would it be worth the extra polycount?
-
That would be pretty neat.. at least for some of the major texture details!
-
IMHO, no. I don't think the Knossos would really benefit from that many polies.
-
I'm not sure that your wirefram shot is telling the whole story, as that poly count seems oddly high. The Hatty is near the 30K mark, and it's surprising that the Knossos would approach those numbers. Am I missing something in regards to where all of those polies are going?
-
I think he means that that's the total for all 8 subunits, unless I'm missing something here.
-
It sounds a heap, but 32k polys total would mean a poly budget of 4k per component, which isn't really all that much compared to most HTL ships of comparable size of an individual component. As such, detail is going to have to be somewhat limited I guess. :\
-
That doesn't really matter a massive amount in the long run, I think, since the components are just about always in motion, so you won't be able to pick out fine details anyway.
-
That all makes sense. I would also say that the knossos being in motion will make it hard to pick out any details.
-
Okay, so no modelled-in waviness then? Like Kolgena and VA said, it's the grand total polycount I'm talking about.
-
Knossos being in motion is one thing. There's also Knossos in stills.
And since this is a very beautiful (and big) design i think that the polycount is not that important.
Afterall it's less than 10 high-poly fighters.
-
Knossos being in motion is one thing. There's also Knossos in stills.
And since this is a very beautiful (and big) design i think that the polycount is not that important.
Afterall it's less than 10 high-poly fighters.
The poly count is important, especially on objects like the Knossos, where you can't see the detail.
-
The poly count over all is less important than the texture count.
And normal maps can only go so far.
And how many of these are in any one mission?
I'd say, keep the existing working item. Make a duplicate where the "ripples" are modeled in. We'll see how much it increases poly count during smoothing and conversion and which impacts performance the most. If it's negligible, we'll put in the "ripple" version.
Of course, I'm only speaking as someone who can barely manage to import a DAE, size and smooth it, export it and have it turn into a POF in Blender, so I realize that this is something I can't really do and I appreciate the amount of skill and work it takes to do this, and I think it would look really sexy. And I think the engine can afford it, but it would be better to have one that isn't to fall back on, just in case.
-
Okay, well, here's a preliminary conversion attached. It's a tilemapped, unLOD-ed version. What I intend to do is unwrap this model onto a single UV-space and bake these textures onto it, and combine it with an AO bake.
I'm aware of two texture-distortion issues in the current model, which I'll fix before baking - see picture below. If you spot any other errors, please let me know.
[image expired]
[attachment deleted by admin]
-
These are looking really good!
-
About done unwrapping detail1, I'll create the detail0 mapping based on this one.
[image expired]
[image expired]
-
In PCS2, Header tab, there's a section called "Cross Sections". The original Knossos has 25 of them. What is that, does it actually do something?
Also, do I copy mass, centre of mass offset and MMOI from the retail model or does that not matter?
Edit: Oh, and does anyone know how to bake textures with alpha channel in Blender?
-
Copy over all the retail data that you can. Don't change things unless you have a good reason to.
In fact, when the Mass was incorrect on the high-poly Cain model, it actually broke a mission.
-
Right, if nobody knows what those cross-sections are for, I'm not gonna put them all in manually. So yeah it's done.
MediaFire download
If anyone feels like improving on the maps, that would be much appreciated; because of my inexperience with texturing, the shinemap is just the diffuse map without AO, and there's no normal map. But otherwise, it's just plug and play.
Edit: Download here (http://www.hard-light.net/forums/index.php?topic=72377.msg1443927#msg1443927)
-
Edit: Oh, and does anyone know how to bake textures with alpha channel in Blender?
I don't. But I know how you can cheat:
Extract the Alpha channel from the image in photoshop/Gimp. Use it as a diffuse map. Apply it to your model. Repeat for as many tiling textures as you have.
Bake the Black-and-white textures to a single black-and-white texture.
Use the black-and-white bake texture as an Alpha channel for the final texture.
Profit.
I suppose it's for the shinemaps?
P.S. Very good job, BTW.
-
Loving these - I just finished another run through FS2 (with BP shaders and a new hud, although antipodes is still rather unstable), and I couldn't help but notice the knossos as it is now. an upgrade has been a long time coming, thank you!
-
Enioch is awesome :)
Shinemap (http://Shinemap)
Normal map (http://Normal map)
Edit: See here (http://www.hard-light.net/forums/index.php?topic=72377.msg1443927#msg1443927)
-
In PCS2, Header tab, there's a section called "Cross Sections". The original Knossos has 25 of them. What is that, does it actually do something?
According to a comment in the code they affect explosions of big ships.
-
ingame shot:
(http://img220.imageshack.us/img220/7439/htlknossosig.png)
-
SCREENSHOTS.
-
Light settings?
It looks like there's no normal/shine map above >.>
As for maps, we don't actually need the ao one, right?
And moar screens, this time with the better normal/shines. (I did remove the AO map though, since that seemed to already be on the diffuse). There's a lot of sparkle going on though. Wonder if the normal map should be made a little less noisy? That, and I'd like to ask some kind soul to take the maps and make them .dds. I don't think I have software that does that, and the 55 fps should ideally be closer to the 100 mark.
[attachment deleted by admin]
-
No, you don't need the AO map, just the four you're used to. I included the AO bake in case people want to play around with the texture.
The sparkle on the normal map... dunno, it was just copied over from the MVP maps. It might look better when you scale down the normal map.
In PCS2, Header tab, there's a section called "Cross Sections". The original Knossos has 25 of them. What is that, does it actually do something?
According to a comment in the code they affect explosions of big ships.
Damn, so they do have a use? Will I have to copy all 25 of them manually now? :blah:
-
Importing the header from the original one in PCS2 should work.
-
You mean, like, go to the header, press "Load", select the retail model? Doesn't work with me...
Using the November 30 version of PCS2, and I don't know if this is related, but I also get an error warning each time I start it up.
file '', line 52: '=' expected.
-
Here are the textures scaled down to 1024. http://www.mediafire.com/file/789v81yhs77hg89/smallerKnos.rar (http://www.mediafire.com/file/789v81yhs77hg89/smallerKnos.rar)
(http://thumbnails2.imagebam.com/11010/f7c469110092054.jpg) (http://www.imagebam.com/image/f7c469110092054)
When I tried to convert to dds I got larger file sizes. :confused: These work reasonably well though. Great model, thanks.
-
You mean, like, go to the header, press "Load", select the retail model? Doesn't work with me...
Try a new build.
-
Try a new build.
Ah, that's better. Still getting the startup error, but meh.
So, let's try this again.
Plug & Play pack (http://www.mediafire.com/?1kz54fk8sty7dl7) (7MB) - contains game-ready POF and diffuse, glow, shine and normal map at 2048^2 resolution DDS.
Map pack (http://www.mediafire.com/?9j5vg64qvzigutv) (68.5MB) - contains all possible maps in 4096^2 resolution PNG.
Enjoy!
-
Awesome! Thanks for the .dds
(I'm curious though, the glow map is 5.33 megs in dds, but 192kb in png. Will the png or the dds get more performance?)
-
dds. Textures are always more efficient in dds.
-
DDS is larger mainly because of mipmaps, which are also one of the primary reason of their efficiency.
-
I'm going to venture a guess that the dds might also be closer to what's stored on VRAM, and that png would have to be uncompressed onto VRAM anyways?
-
I'd like to see better normal maps... I have never liked the ones that are just grooved without any additional depth.
-
I'm going to venture a guess that the dds might also be closer to what's stored on VRAM, and that png would have to be uncompressed onto VRAM anyways?
Exactly. DDS files can be used right away by the graphics card (even in their compressed form, all other images are uncompressed before use and hence take more space on the VRAM than a compressed DDS would.
(Note that there's also uncompressed DDS, which mainly has benefits in having mipmaps and faster loading, because it doesn't need to be decoded). The DDS files I provided are DXT3-compressed for diffuse and glow (no alpha channel needed, but DXT1 is very lossy) and DXT5 for shine and normal (full 8-bit alpha channel).
So, in short, DDS should be used for pretty much anything in-mission.
I'd like to see better normal maps... I have never liked the ones that are just grooved without any additional depth.
Additional depth? What depth is there beyond the grooves?
-
I mean a normal map that actually has curves and texture to it, created from a greyscale bumpmap instead of just a black-and-white one.
-
So that that the surface will be lobed plating resembling adipose tissue?
eh. I kinda like the sharp grooves that are here now.
-
So that that the surface will be lobed plating resembling adipose tissue?
eh. I kinda like the sharp grooves that are here now.
which would be sorta frickin awesome?
-
Awesome, I contemplated making one. No need to now. :D Will be great to replace the old Knossos.