gets triangulated anyways...you may get issues if the mirrored triangulation creates smoothing artefacts different from the other side (that the normal won't correct)
looks good, a bit too brushed for my taste, but good nonetheless, reminds me of wannabe_'s (I think that was his name) Diffuse only CS weapons work, back in the old PC days.
I like it! Mind if I ask for another view of the HP from a front angle?
Hey of course not man, already had this render posted on another forum:
If there are any other shots any of you want just gimme a shout, have somebody helping me to rig this for Crysis as we speak but I feel animating is going to force me into a straight jacket :poly142:
Looks like the spec is just a copy of the diffuse, which = MAJOR FAIL. Do a proper spec map dude, work on the materials a bit more too, the metal and plastic both seem to read as the same dull(not dull as in you suck, but literally dull as in no reflection) material. Pop out a bit more shine, and some specific details in the spec(contrast, scratches, different texture, some subtle color variation, etc) and the texture will be up to par with the model.
Also are the textured lowpoly shots realtime screenshots or renders? From the look of the geometry on your lowpoly i would expect some smoothing errors in the normal, but that doesnt seem to be the case. Remember that max scanline render does NOT = what it will look like in real time, as the normal interpolation is more precise, and everything will look great in your render, but may be fucked in game if you arent previewing with an actual realtime shader.
[edit] Looking closely at your normals, looks like you've got everything set up with different smoothing groups. Care to write up a little bit about how you're getting seemingly-seamless results with that method?
Am I the only one who's noticed that he's baked the lighting into his diffuse or am I confused here?
He used a ground plane, skylight, and light tracer to bake a little diffuse lighting. The effect is minimal but will help keep the weapon looking good in all lighting conditions.
agrreed^ will look confussed in lighting NOT from above
personnally id just use AO no ground plane and then more for dust dirt and saturation values than for over darkening
Thanks guys, I agree that it could use more work so I will go over it. I always get slightly bored when it comes to the specular map. On the subject of the Spec Earthquake I'm going to be quite tentative with it. I've seen way too many people just overlay what looks to be pictures of the side of an oil rig and then just contrast it by 80. An exaggeration I know but most all assets I see now tend to look as though they have been worked upon with a chisel. I'll add more unique texture to the specular (ie not just what is already there on the diffuse intensified) and possibly more dark patches to break the lighting a bit more but I'm afraid I don't plan to make every edge white, to me cleaner textures are much more pleasing to look at but I seem to be in the "escaped from an institute minority" :poly121:
About the baked lighting looking again it is too strong, I'll have to mess around with that
SHEPEIRO: I think I will keep the baked lighting, especially in Unreal 3 I've noticed without an extremely heavy specular assets can look pretty bland when not in optimal lighting conditions.
E: Earthquake do you happen to use MSN? if not which engine do you recommend I preview it in? I don't own Unreal (been told it's extremely easy to export to).
Making a quality texture has nothing to do with making every edge white, but understanding that what makes a good, dynamic material is how your diffuse, spec and normals all work together. The spec should never be a second thought, and in fact with materials like metal, is much more important for believability than the diffuse.
Also, answer my questions about normals or face BANHAMMER.
You bake with xNormal right? xNormal hates smoothing groups and renders them with seams that alias like mofos. I use a mixture of a max bake (with cage) and an xnormal bake. If you have the uv maps split where the smoothing groups oppose it should bake perfectly in 3dsm.
max2009 gives a good bake on split edges, but xnormal does better away from the edges. You still get aliased edges with smoothing groups, but the projection is so good, it's no big deal when stuff is in motion or when antialiased like you have here, if I'm correct Kimono?
Someone please make an animated gif of EQs avatar.
Pretty much, I've rendered it with just the bake that xnormal did, as you can see some of the seams are quite noticeable. It is great dealing with single smoothing groups probably because I'm guessing it was geared towards organic work (inbuilt z-brush features and the ability to have a good bake without a cage)
oh so moderators edit posts to conform with their personal tastes here
awesome
anyway seeing as my opinion wasnt good enough first time around, i'll try for round two.
First person weapons seldom are seen upside down... And (of course, depending on the engine) it's generally a sick idea to give art some extra pop by baking a LITTLE light (because in games there's MAYBE like 1% of the actual light sources and conditions youd find in reality)
Having said that, full dynamic lighting is also cool.
What's not cool is pretending theres a right and wrong way to present your artwork. I think people are getting some tall poppy syndrome because kimono's artwork kicks the ass out of pretty much any commercial weapon art, particularly in the area of realism.
which probably would be best summed up as "oh for fuck's sake" @ the amount of swelling, pulsating rancid ego in this thread.
I think people are getting some tall poppy syndrome because kimono's artwork kicks the ass out of pretty much any commercial weapon art, particularly in the area of realism.
Leave the ego at the door kiddo, or preferably, just go back to cdg. That attitude will get you nowhere here...
Perna, I'm not the one editing the posts of people who don't blindly agree with me.
And tall poppy refers to a syndrome where humans feel a need to drag down anyone who excels. I wasn't calling anyone "a tall poppy."
And the ego? Ego refers to the self, im talking about Kimono's work.
The incoherance of all this really doesnt help me feel sympathetic to your "camp."
dude, relax ok?
random bullshit posts like the one you made aren't well suited for this forum.
like i said, this isn't CDG. each forum has it's rules, and here, posts like that dont contribute to the quaility of the forum.
why didnt you post earlier what you said in this last post of yours?
Looks like the spec is just a copy of the diffuse, which = MAJOR FAIL. Do a proper spec map dude, work on the materials a bit more too, the metal and plastic both seem to read as the same dull(not dull as in you suck, but literally dull as in no reflection) material. Pop out a bit more shine, and some specific details in the spec(contrast, scratches, different texture, some subtle color variation, etc) and the texture will be up to par with the model.
Also are the textured lowpoly shots realtime screenshots or renders? From the look of the geometry on your lowpoly i would expect some smoothing errors in the normal, but that doesnt seem to be the case. Remember that max scanline render does NOT = what it will look like in real time, as the normal interpolation is more precise, and everything will look great in your render, but may be fucked in game if you arent previewing with an actual realtime shader.
[edit] Looking closely at your normals, looks like you've got everything set up with different smoothing groups. Care to write up a little bit about how you're getting seemingly-seamless results with that method?
you talking about this?
i dont know what you are reading. but it must be diferent from what im reading. all i see is advise. maybe you are reading a diferent stuff that im not aware off.
Hey Kimono it's good to see you doing next gen stuff, I've seen your stuff progress a lot in the last few years.
I agree with Earthquake though, you need to really spend some time popping your spec and focusing on bringing out the materials. The lighting in your diffuse could cause some issues and is probably worth looking into once put in engine to make sure it's not conflicting too dramatically with your lights. The texture as a whole seems very flat gray across the board right now too, and with the spec fixes it will pop more but I also suggest darkening your diffuse grays with value changes to separate materials, I think that will help pop it as well. I think your shell should be hue shifted a little and given some more saturation as well.
What do people think of using an AO as a base for a spec map? Then adding scratches, and things to it? Seems like it works good as a base, since it will darken where things go down or where two planes meet.
I really dig the shotgun, dude. Great detail work!
oh so moderators edit posts to conform with their personal tastes here
awesome
anyway seeing as my opinion wasnt good enough first time around, i'll try for round two.
First person weapons seldom are seen upside down... And (of course, depending on the engine) it's generally a sick idea to give art some extra pop by baking a LITTLE light (because in games there's MAYBE like 1% of the actual light sources and conditions youd find in reality)
Having said that, full dynamic lighting is also cool.
What's not cool is pretending theres a right and wrong way to present your artwork. I think people are getting some tall poppy syndrome because kimono's artwork kicks the ass out of pretty much any commercial weapon art, particularly in the area of realism.
which probably would be best summed up as "oh for fuck's sake" @ the amount of swelling, pulsating rancid ego in this thread.
hah talk about swelling heads here.
Yeah weapons are not upside down usually but quite often there on their side laying on the ground or perhaps a table. Baking the shadows into the diffuse like he has it would look silly for a weapon laying on its side.
The spec map is crap. For starters why is the bottom less shiny where the AO was baked in? Nobody is saying make it a streaked oil rig.
OK guys, please, I'm working on it today. Ironhawk, I've seen that before (Earthquake posted it on some other forums) it's an amazing looking piece but the actual texture of the metal is pretty strange, looks kinda like it was laying in water for a while.
What I'm trying to achieve seems to be quite harder in a dynamic environment and (as I've been told) I'm apparently failing. I don't want 100% realism but I want something close. I've already said the textures could use some work so I'm going to see what I can do.
Not to be another fresh guy lecturing you on the basics of texturing, but maybe we should see how it looks in the game before he goes and makes a bunch of changes based on advisement rather than empirical evidence.
oh another thing that's worth noting..
When lighting is dynamic, it of course means that it very rarely will come from any of the "cool angles" that you want. When you pose a real gun for a photo you're careful to create the best lighting conditions for that particular photo. When baking light in your diffuse you choose that best lighting condition, meaning it will "always look cool". Yes, but it will also look dull and incorrect in movement.
So to balance out the fact that you get poor results most of the time with dynamic light (exactly like in real life), you exaggerate things, you push the art side.. and by doing this you are moving away from "realism". It has to be done however.
Generally for games, you're not making 3d assets that will only be viewed from one angle or under one light setup. I would say your approach is more geared towards making 1 good-looking render more than to make something that works well ingame.
I understand exactly what you're trying to do and why, and I certainly appreciate it, you are doing a very good job. But it may be relevant for you to ask why established professionals are not doing the same.
Well I'm not looking to change the world. As Kaskad said I want to view it ingame, not as an offence to anyone here, I respect you all and this is a great community. Just that I can be quite stubborn and to learn on my own (mistakes?) will be the best course of action
Yeah, except it actually appears to be of a high standard in practise.
What ever happened to going by whether something looks right rather than make adjustments based purely on what's the norm? Its a first person model, it's not a world model, it will be "right-way-up" 100% of the time. Clearly the world model would be different, and at the very least not use the same diffuse map (baked light could be removed)
I totally understand where youre coming from perna, but the way I see it it's either, bake a little lighting on and have it look correct in 90% of lighting situations, or have the diffuse even intensity across the map and have it look correct in like 10% of lighting situations. The problem is, in my opinion, that dynamic lighting just isn't at a stage yet where things can be lit in a truly natural way.
Of course though, I'm open to the idea that he might put it ingame and it turn out looking like complete shit.... Because I'm not the kind of guy to just assume I know the deal without actually observing evidence. He might as well at least try with what he has now, and then if it looks like the back of his nuts then it's time to change it, and perhaps apply the lessons to his work in future.
Oh and, if overlaying photos of pubic hair is what you consider "material definition," the standards and expectations of this place are way below what I was hoping for.
Kaskad: After you tie your shoe laces, will you walk with uncertainty for some time before you affirm that you have done it correctly and will not trip up and fall?
That little metaphor is as far as I'll go in participating in a debate over these very conventional and established art theories.
Your metaphor doesn't hold up m8, being that this is the first time he's put something in the engine and not the 5,000th. The very first time I tied my shoe laces, yes, I was a bit uncertain that I had done it right even though my parents, who had years of experience with shoe tying, told me that I had. Telling him not to check it in the engine and trust what you're saying blindly is dumb. I doubt you learned that way.
It sounds like he's just tucking his shoelaces into his shoes rather than tying them
or maybe he has velcro straps
or maybe he's wearing flip-flops without any laces at all
That little metaphor is as far as I'll go in participating in etc etc...
This thread has been horribly de-railed, please keep it back on track with more appropriate critiques (such as what has been mentioned with the specular map), and less of the person-bashing.
Anyone who posts personal attacks after this will be facing infractions, leading to a ban if they persist.
I'm sorry if I misunderstood you but I got the impression that you were suggesting that checking it in-game would have been a waste of time because you had already told him a better way to do things. If that is not what you were suggesting then I apologize. If that is what you were saying then I do not. Either way, I agree that a calm and objective discussion is what is preferable. The only lighting he's got in the diffuse is top-down and non-specular. I don't think that this is particularly unrealistic in a game such as Crysis which takes place largely outdoors; in fact I think it'll probably do a good deal to keep the weapon looking interesting when not in a particularly interesting lighting situation. Is this realistic? Nope. But you yourself said that sometimes you have to break realism and push the art side. I don't see how what he's done is anything else.
And by the way, I agree with Earthquake about the specular. I'm only talking about the lighting on the diffuse.
Harry:"The problem is, in my opinion, that dynamic lighting just isn't at a stage yet where things can be lit in a truly natural way."
Since the gun is in the game world, and not the real world, shouldn't it match the game lighting? The point is to make the gun STICK into the world. Not look perfect on its own terms.
Harry:"I totally understand where youre coming from perna, but the way I see it it's either, bake a little lighting on and have it look correct in 90% of lighting situations, or have the diffuse even intensity across the map and have it look correct in like 10% of lighting situations."
it will look wrong, technically, in 100% of lighting situations as is, unless its this static shot. That is what they are getting at.
Again, great gun! I'm very jealous that you have it.
So now on that spec? Is starting from a AO map a good way to do it? Thats what I do.
man, lets all whip out or cocks and measure! seriously though, most of everyones the points are valid re: some baked light does look coolio etc, but I totally agree on the whole no 100% black and beyond. some light will almost always bounce into cracks and crevices so it should never be totally dark. thanks for the example per, good illustration of what you mean.
Nice highpoly model man, looks great. I really cant wait to see your new spec, the original images grabbed my eye so im eager to see it once you've givn'er a bit more to refine the material type. nice work so far man now sift through this curfuffel of a thread and apply the crits and comments for a smokin' result.
What do people think of using an AO as a base for a spec map? Then adding scratches, and things to it? Seems like it works good as a base, since it will darken where things go down or where two planes meet.
I really dig the shotgun, dude. Great detail work!
For both my diffuse and spec i use the same AO/LM. What i do here is just create a layer group above both my diffuse and spec layers in my psd. The reasoning for this has always seemed pretty sound to me, the less diffuse light something is getting, the less specular reflection you're likely to get as well.
So would i used an AO bake as the base for my spec? No, but i would use the same AO bake i have on my diffuse, multiplied on my spec. Generally its good to use your diffuse as a base for your specular, and identify which areas will need more/less reflectivity depending on material type, conditions, wear, etc and make adjustments accordingly.
Mathematically the gradient continues beyond black, but once you bake those values into a texture, you lose that gradient, thereby losing all shape and definition in that area if it's lit.
You never let baked light in your diffuse go down to black, let alone beyond black. I'll probably receive a bunch of hatemail now for using the word "never" where I suppose I should have said something like "maybe less frequently, perhaps, in my opinion, and I don't wish to offend anyone and I preach love and understanding across the universe".
Per dropping some science here. Another very important thing to consider when dealing with very dark values is compression, not that this necessarily has a lot to do with Kimono's asset, but dealing with darker values in the low color rangers can really hurt you especially with gradients, when you have compressed textures(most games). You're already dealing with a very low amount of precision, being at the bottom of the spectrum of color range, and compression tends to really make this obvious.
Replies
well...
the normalmapped model looks good, so who cares.
looks good, a bit too brushed for my taste, but good nonetheless, reminds me of wannabe_'s (I think that was his name) Diffuse only CS weapons work, back in the old PC days.
Hey of course not man, already had this render posted on another forum:
If there are any other shots any of you want just gimme a shout, have somebody helping me to rig this for Crysis as we speak but I feel animating is going to force me into a straight jacket :poly142:
Also are the textured lowpoly shots realtime screenshots or renders? From the look of the geometry on your lowpoly i would expect some smoothing errors in the normal, but that doesnt seem to be the case. Remember that max scanline render does NOT = what it will look like in real time, as the normal interpolation is more precise, and everything will look great in your render, but may be fucked in game if you arent previewing with an actual realtime shader.
[edit] Looking closely at your normals, looks like you've got everything set up with different smoothing groups. Care to write up a little bit about how you're getting seemingly-seamless results with that method?
He used a ground plane, skylight, and light tracer to bake a little diffuse lighting. The effect is minimal but will help keep the weapon looking good in all lighting conditions.
personnally id just use AO no ground plane and then more for dust dirt and saturation values than for over darkening
About the baked lighting looking again it is too strong, I'll have to mess around with that
SHEPEIRO: I think I will keep the baked lighting, especially in Unreal 3 I've noticed without an extremely heavy specular assets can look pretty bland when not in optimal lighting conditions.
E: Earthquake do you happen to use MSN? if not which engine do you recommend I preview it in? I don't own Unreal (been told it's extremely easy to export to).
Also, answer my questions about normals or face BANHAMMER.
Pretty much, I've rendered it with just the bake that xnormal did, as you can see some of the seams are quite noticeable. It is great dealing with single smoothing groups probably because I'm guessing it was geared towards organic work (inbuilt z-brush features and the ability to have a good bake without a cage)
this isn't CDG dude.
mkay?
awesome
anyway seeing as my opinion wasnt good enough first time around, i'll try for round two.
First person weapons seldom are seen upside down... And (of course, depending on the engine) it's generally a sick idea to give art some extra pop by baking a LITTLE light (because in games there's MAYBE like 1% of the actual light sources and conditions youd find in reality)
Having said that, full dynamic lighting is also cool.
What's not cool is pretending theres a right and wrong way to present your artwork. I think people are getting some tall poppy syndrome because kimono's artwork kicks the ass out of pretty much any commercial weapon art, particularly in the area of realism.
which probably would be best summed up as "oh for fuck's sake" @ the amount of swelling, pulsating rancid ego in this thread.
Leave the ego at the door kiddo, or preferably, just go back to cdg. That attitude will get you nowhere here...
And tall poppy refers to a syndrome where humans feel a need to drag down anyone who excels. I wasn't calling anyone "a tall poppy."
And the ego? Ego refers to the self, im talking about Kimono's work.
The incoherance of all this really doesnt help me feel sympathetic to your "camp."
(at kimono's work and at the behavior of forum regulars in this thread)
makes me wants to jump on to the cs replacement saddle again... haha keep at it dude.
random bullshit posts like the one you made aren't well suited for this forum.
like i said, this isn't CDG. each forum has it's rules, and here, posts like that dont contribute to the quaility of the forum.
why didnt you post earlier what you said in this last post of yours?
it wouold have been more productive.
wanna talk about quality feedback
you talking about this?
i dont know what you are reading. but it must be diferent from what im reading. all i see is advise. maybe you are reading a diferent stuff that im not aware off.
I agree with Earthquake though, you need to really spend some time popping your spec and focusing on bringing out the materials. The lighting in your diffuse could cause some issues and is probably worth looking into once put in engine to make sure it's not conflicting too dramatically with your lights. The texture as a whole seems very flat gray across the board right now too, and with the spec fixes it will pop more but I also suggest darkening your diffuse grays with value changes to separate materials, I think that will help pop it as well. I think your shell should be hue shifted a little and given some more saturation as well.
Keep at it, model looks good!
What do people think of using an AO as a base for a spec map? Then adding scratches, and things to it? Seems like it works good as a base, since it will darken where things go down or where two planes meet.
I really dig the shotgun, dude. Great detail work!
hah talk about swelling heads here.
Yeah weapons are not upside down usually but quite often there on their side laying on the ground or perhaps a table. Baking the shadows into the diffuse like he has it would look silly for a weapon laying on its side.
The spec map is crap. For starters why is the bottom less shiny where the AO was baked in? Nobody is saying make it a streaked oil rig.
http://boards.polycount.net/showthread.php?t=53731&page=5 Would be a good source for some pointers.
It's a decent weapon for CS source but for dropping it in an engine like Crysis a few of us are saying it needs more work.
What I'm trying to achieve seems to be quite harder in a dynamic environment and (as I've been told) I'm apparently failing. I don't want 100% realism but I want something close. I've already said the textures could use some work so I'm going to see what I can do.
spherical harmonics is what you need- or for the purpose of renders a very blurred env.map use the ati cube map generator to produce a good one.
What ever happened to going by whether something looks right rather than make adjustments based purely on what's the norm? Its a first person model, it's not a world model, it will be "right-way-up" 100% of the time. Clearly the world model would be different, and at the very least not use the same diffuse map (baked light could be removed)
I totally understand where youre coming from perna, but the way I see it it's either, bake a little lighting on and have it look correct in 90% of lighting situations, or have the diffuse even intensity across the map and have it look correct in like 10% of lighting situations. The problem is, in my opinion, that dynamic lighting just isn't at a stage yet where things can be lit in a truly natural way.
Of course though, I'm open to the idea that he might put it ingame and it turn out looking like complete shit.... Because I'm not the kind of guy to just assume I know the deal without actually observing evidence. He might as well at least try with what he has now, and then if it looks like the back of his nuts then it's time to change it, and perhaps apply the lessons to his work in future.
Oh and, if overlaying photos of pubic hair is what you consider "material definition," the standards and expectations of this place are way below what I was hoping for.
or maybe he has velcro straps
or maybe he's wearing flip-flops without any laces at all
That little metaphor is as far as I'll go in participating in etc etc...
Anyone who posts personal attacks after this will be facing infractions, leading to a ban if they persist.
Cheers!
And by the way, I agree with Earthquake about the specular. I'm only talking about the lighting on the diffuse.
Since the gun is in the game world, and not the real world, shouldn't it match the game lighting? The point is to make the gun STICK into the world. Not look perfect on its own terms.
Harry:"I totally understand where youre coming from perna, but the way I see it it's either, bake a little lighting on and have it look correct in 90% of lighting situations, or have the diffuse even intensity across the map and have it look correct in like 10% of lighting situations."
it will look wrong, technically, in 100% of lighting situations as is, unless its this static shot. That is what they are getting at.
Again, great gun! I'm very jealous that you have it.
So now on that spec? Is starting from a AO map a good way to do it? Thats what I do.
Nice highpoly model man, looks great. I really cant wait to see your new spec, the original images grabbed my eye so im eager to see it once you've givn'er a bit more to refine the material type. nice work so far man now sift through this curfuffel of a thread and apply the crits and comments for a smokin' result.
cheers!
For both my diffuse and spec i use the same AO/LM. What i do here is just create a layer group above both my diffuse and spec layers in my psd. The reasoning for this has always seemed pretty sound to me, the less diffuse light something is getting, the less specular reflection you're likely to get as well.
So would i used an AO bake as the base for my spec? No, but i would use the same AO bake i have on my diffuse, multiplied on my spec. Generally its good to use your diffuse as a base for your specular, and identify which areas will need more/less reflectivity depending on material type, conditions, wear, etc and make adjustments accordingly.
Per dropping some science here. Another very important thing to consider when dealing with very dark values is compression, not that this necessarily has a lot to do with Kimono's asset, but dealing with darker values in the low color rangers can really hurt you especially with gradients, when you have compressed textures(most games). You're already dealing with a very low amount of precision, being at the bottom of the spectrum of color range, and compression tends to really make this obvious.
Where do I sign up!?