Learning and practicing in 3DS...
-
christianclavet
- Posts: 1638
- Joined: Mon Apr 30, 2007 3:24 am
- Location: Montreal, CANADA
- Contact:
Humm.. I think it goes around 40-50k (triangles), or 15-20K (quad) per chesspiece and yes they have modifiers (Mostly Edit Poly, Symmetry, Turbosmooth) on them since they are still in WIP (Work in progress)
I've heard that there are tool in max to create a low level poly version and bake the maps... Teacher will probably show them in the next week.
The current objective was to model the chesspieces, next exercise will be to UVMap them. We should be taugh how to create the maps for the characters.
I've heard that there are tool in max to create a low level poly version and bake the maps... Teacher will probably show them in the next week.
The current objective was to model the chesspieces, next exercise will be to UVMap them. We should be taugh how to create the maps for the characters.
subsurf for games? well, maybe your teacher looks for the long shot plan, and doing multirse stuff for normal maps or sth.
You could have a 3k tris model per piece, easily, and with more detail.(40k tris is way too much)
Sorry, over loaded of personal work... if not, I'd have helped a bit as I said in posting a fast sketch over those shots...
the woman is the one having more probs...
I guess anyway your milestone date is ended
You could have a 3k tris model per piece, easily, and with more detail.(40k tris is way too much)
Sorry, over loaded of personal work... if not, I'd have helped a bit as I said in posting a fast sketch over those shots...
the woman is the one having more probs...
I guess anyway your milestone date is ended
Finally making games again!
http://www.konekogames.com
http://www.konekogames.com
Really I don't think these are for games. They are just practicing their modelling, and that is generally how all courses start on modelling. I feel it is harder to express yourself, as a beginner/intermediate, when you have limits imposed on yourself.
He will probably, by the end of the course, have them modelling in low poly instead. Although I don't know, there has been a wave of strict hi-res modellers getting hired these days...
He will probably, by the end of the course, have them modelling in low poly instead. Although I don't know, there has been a wave of strict hi-res modellers getting hired these days...
TheQuestion = 2B || !2B
-
christianclavet
- Posts: 1638
- Joined: Mon Apr 30, 2007 3:24 am
- Location: Montreal, CANADA
- Contact:
Hi. The work had to be submitted yesterday.
It's as you said Halifax. We'll continue working on thoses meshes now in texturing and we'll surely do a lowes version because I've seen in the program we'll learn how to "bake" some textures and produce normal maps.
I think the way to go theses days are to create a hires-mesh then create a lower rez based on that one.
I practiced myself over the weekend to export the mesh in B3B using the new B3DPipeline (patch made for 3DS 9 that I'm using). I also praticed a little with Biped and checked how it worked (a real jewel compared to Lightwave). Once We'll be demonstrated how to reduce them (None of my modifiers are collapsed), I'll do a lower rez of them and make the mesh availables if someone want to create a chess set from theses characters (Would perhaps have to check with Warner's Studio before releasing a game based on theses characters). When creating the lower version 3K-4K seem a good limit as you mention vermeer.
My plan is to use thoses characters to learn 3DS within the program at Campus Ubisoft. I also want (when we'll learn how to bake and produce normal maps) to create a mega hires version in ZBrush (600 000+ poly), so I can have every tiny details on the warrior(damage, scars, etc.) and would create an animation with normal mapped characters for my portfolio.
I plan to start modeling on the Egypt Scene over the weeked for the First King since now I feel confortable to model in 3DS. (Will model in Hires first) (The idea is to model in hi detail in 3DS, later do a "damage" pass in ZBrush then create a lower poly model again in 3DS with baked and normal/specular maps applied to that low poly model. Modeling will start on the exterior scene first (desert, Gizea Plateau, all piramids & Sphinx, then the interior scenes (underground tunnels) after.
It's as you said Halifax. We'll continue working on thoses meshes now in texturing and we'll surely do a lowes version because I've seen in the program we'll learn how to "bake" some textures and produce normal maps.
I think the way to go theses days are to create a hires-mesh then create a lower rez based on that one.
I practiced myself over the weekend to export the mesh in B3B using the new B3DPipeline (patch made for 3DS 9 that I'm using). I also praticed a little with Biped and checked how it worked (a real jewel compared to Lightwave). Once We'll be demonstrated how to reduce them (None of my modifiers are collapsed), I'll do a lower rez of them and make the mesh availables if someone want to create a chess set from theses characters (Would perhaps have to check with Warner's Studio before releasing a game based on theses characters). When creating the lower version 3K-4K seem a good limit as you mention vermeer.
My plan is to use thoses characters to learn 3DS within the program at Campus Ubisoft. I also want (when we'll learn how to bake and produce normal maps) to create a mega hires version in ZBrush (600 000+ poly), so I can have every tiny details on the warrior(damage, scars, etc.) and would create an animation with normal mapped characters for my portfolio.
I plan to start modeling on the Egypt Scene over the weeked for the First King since now I feel confortable to model in 3DS. (Will model in Hires first) (The idea is to model in hi detail in 3DS, later do a "damage" pass in ZBrush then create a lower poly model again in 3DS with baked and normal/specular maps applied to that low poly model. Modeling will start on the exterior scene first (desert, Gizea Plateau, all piramids & Sphinx, then the interior scenes (underground tunnels) after.
yeh.Halifax wrote: He will probably, by the end of the course, have them modelling in low poly instead. Although I don't know, there has been a wave of strict hi-res modellers getting hired these days...
well. am too old skool. Or maybe just old, lol. I do model with subdivs always, so low cage and hi res things go togehter(do even hi res small details in subdiv, if feel in the mood). Learning the low cage thing well allows both things, as like in traditional drawing, there's no good detail without underlaying good , solid concept and good structure. Low cage subdiv allows this, and with some clever things, you can easily drive later your low cage to low model. Anyway, retopo painting is a lot of fun , too. So am not against it. Is just I emphasize how important is knowing well the base structure/volumes/shapes. And then after an step is solid, go to the next detailing step.
Anyway, that above is quite off topic.
Maybe.... 3k / 4k ...is a bit to high depending on the engine, I said too fast.
32 * 3k .... (unless if were instances, but dunno which engine allows that.... I know of one I used inside certain house...custom one) 96,000 tris. Well...could be depending on the card, machine, engine, and rest of cpu (etc) eaters around... 3k is a bit bellow UT 2003 (I modded for that one) , and by experience is a count that allows quite detail .Think of those characters, usually were 3.500 even 4k, a pair of 1024 textures (one for the head!) , if my memory is working ok. I did one of a bit aboe 2.5k. And had detail and smooth surfaces.With extra elements in account.
Your teacher for sure will tell you a wise pipeline.
Finally making games again!
http://www.konekogames.com
http://www.konekogames.com
Well I don't know. I am just going to quote Epic Games here, but in the year 2006 they said "Renderable Mesh: We build renderable meshes with 3,000-12,000 triangles, based on the expectation of 5-20 visible characters in a game scene. ... Detail Mesh: We build 1-8 million triangle detail meshes for typical characters. This is quite sufficient for generating 1-2 normal maps of resolution 2048x2048 per character. ... Bones: The highest LOD version of our characters typically have 100-200 bones, and include articulated faces, hands, and fingers. ... We are authoring most character and world normal maps and texture maps at 2048x2048 resolution. We feel this is a good target for games running on mid-range PC's in the 2006 timeframe. Next-generation consoles may require reducing texture resolution by 2X, and low-end PC's up to 4X, depending on texture count and scene complexity. "
So based on what they have laid out as a projection for 2006, I would say that 3k-12k with normal maps is a good range for the PCs of today. But eh, that's just Epic Games.
So based on what they have laid out as a projection for 2006, I would say that 3k-12k with normal maps is a good range for the PCs of today. But eh, that's just Epic Games.
TheQuestion = 2B || !2B
heh, that's with Unreal engines. 
As well as Unreal games.
All factors count.
And... I spoke about ut 2003. 3 years before... usually is a good measure for indy games...as the main prob is not the engine (i repeat, am told irrlicht is fast) but also the techniques for optimizing inside a game stduio are crazily good. I know as have seen make wonders some mates of mine. Like a few models of 7k tris , in aheavy level, fly with machines much more underpowered than those at 2003....but....*but* those were absolute master in optimizing and dirty tricks. Specially certain man... Now, take the fact indy ppl have very varied experience, some very little or none. And some engines are not thought to work for every game, many are dedicated from start to what certain game is gonna do...lots of things like this. I do prefer achieve wonder with little polies...why? because I freaking know pretty well to throw a million polies detail thoughtless and get my uber hi quality mega monster. But with optimized meshes, i can get also a result that will make coders have les pains. I mean, the low model, if low, runs smoother . Not at the cost of detail, of course, but if one is smart there, then waaay better. I know am bit biased as have worked 3d for phones and other low machines, not only pcs, but imo, the tris you throw carelessly is cpu, card, or whatever later...better use that for ai...
As well as Unreal games.
All factors count.
And... I spoke about ut 2003. 3 years before... usually is a good measure for indy games...as the main prob is not the engine (i repeat, am told irrlicht is fast) but also the techniques for optimizing inside a game stduio are crazily good. I know as have seen make wonders some mates of mine. Like a few models of 7k tris , in aheavy level, fly with machines much more underpowered than those at 2003....but....*but* those were absolute master in optimizing and dirty tricks. Specially certain man... Now, take the fact indy ppl have very varied experience, some very little or none. And some engines are not thought to work for every game, many are dedicated from start to what certain game is gonna do...lots of things like this. I do prefer achieve wonder with little polies...why? because I freaking know pretty well to throw a million polies detail thoughtless and get my uber hi quality mega monster. But with optimized meshes, i can get also a result that will make coders have les pains. I mean, the low model, if low, runs smoother . Not at the cost of detail, of course, but if one is smart there, then waaay better. I know am bit biased as have worked 3d for phones and other low machines, not only pcs, but imo, the tris you throw carelessly is cpu, card, or whatever later...better use that for ai...
Finally making games again!
http://www.konekogames.com
http://www.konekogames.com
I didn't quite feel like reading that large paragraph. Sorry I must admit, but I like it better when things are structured. 
But I wouldn't necessarily say that these polygons are placed irrationally. If they were, then I would imagine that the models look exactly like previous generation models, just with normal maps.
In fact, epic games meshes are quite optimized. And there are some rendering engines out there currently that are open-source, and free, that are quite on par with Unreal Engine in the rendering part. Which is all that really matters when it comes to rendering.
But I wouldn't necessarily say that these polygons are placed irrationally. If they were, then I would imagine that the models look exactly like previous generation models, just with normal maps.
In fact, epic games meshes are quite optimized. And there are some rendering engines out there currently that are open-source, and free, that are quite on par with Unreal Engine in the rendering part. Which is all that really matters when it comes to rendering.
TheQuestion = 2B || !2B
You can't read a paragraph? There are lots of them in books. Anyway, doesn't mind at all.Halifax wrote:I didn't quite feel like reading that large paragraph. Sorry I must admit, but I like it better when things are structured.
And is that above well structured or written? Do you refer to his models (I have not spoken about them, only about possible polygon count to use in the final thing). Of course with today's machines you can pull a lot more. And some of us are able to do that level of detail and more. But optimizing keeps having its value...But I wouldn't necessarily say that these polygons are placed irrationally. If they were, then I would imagine that the models look exactly like previous generation models, just with normal maps.
In fact, epic games meshes are quite optimized.
Who would be as crazy as to say the opposite. I meant Unreal engine is quite too much an excellent/efficient engine to bring numbers from it...
I think there aren't so many. And seems you really didn't read my post , but still you are replying it. I said in there that it also counts how the coders do the game code, how they optimize every part,(ai, physics, if they count on middleware or custom solutions, 3d matters, a large etc.) how the experience in that during years, and done by professionals, counts as well, at the moment to lower a bit all requirements in the houir of planning of making an indy game.And there are some rendering engines out there currently that are open-source, and free, that are quite on par with Unreal Engine in the rendering part. Which is all that really matters when it comes to rendering.
Well, whatever.
Finally making games again!
http://www.konekogames.com
http://www.konekogames.com
Wow no need to bring in a harsh tone, although I figured that line would be a bit controversial; futhermore, I was responding to bits and pieces that I was reading at the time, that is true. And to answer your statement, I don't consider paragraphs as being sentences that are connected with ellipses, but that is besides the point.
Yes, optimization does contain a stronghold in every industry, and it is virtually impossible that it would ever leave.
I was clearly bringing numbers from the rendering part of the Unreal engine which makes all other aspects simply irrelevant. And I know quite a few engines that match up to the Unreal engine in rendering, excluding visible sight determination.
But anyways, as you said, "Well, whatever", let's get back on topic as this is going much further than I anticipated, or even wanted.
Yes, optimization does contain a stronghold in every industry, and it is virtually impossible that it would ever leave.
I was clearly bringing numbers from the rendering part of the Unreal engine which makes all other aspects simply irrelevant. And I know quite a few engines that match up to the Unreal engine in rendering, excluding visible sight determination.
But anyways, as you said, "Well, whatever", let's get back on topic as this is going much further than I anticipated, or even wanted.
TheQuestion = 2B || !2B
Christian Love your work, make sure you place the queen on her colour... A common mistake when setting up chess pieces, but I suppose it doesn't really matter if you only have one team!
Programming Blog: http://www.uberwolf.com