|
Heintje posted:Yeah approximation is fine. Sorry I didn't mean to come off sounding like approximation isn't a valid concept to use when modeling, I'm sure it has its uses. When it boils down to it, there were only two incidents that really made me scratch my head in the tutorial in question: One was when he modeled handlebars for the "jet-bike". Instead of capping the extruded surface he just scaled in the end verts. I guess that's OK but it just seems like a sloppy way of doing it if you can cap your object. The other was when he eyeballed the placement of something that should have been symmetrical. No huge deal either way. I've been working in 3D for a month and this guy has been teaching for years so I guess he has his reasons, and again it was a beginner's tutorial so I guess perfection wasn't the ultimate goal 
|
#
?
Jul 21, 2008 23:24
|
|
|
|
| # ? Apr 23, 2024 20:39 |
|
|
Seems to me architecture and engineering (AutoCAD, Max) are very precise whereas character modeling and animation (Maya, Max, XSI etc) are a little more organic / informal. My brother uses AutoCAD and Rhino almost exclusively and he could never rationalize how I "eyeball" stuff in Maya. The truth is you want some asymmetry in your character models and some imperfections here and there to encourage an organic feel. That is to say, even though we are essentially symmetrical as human beings, we have random imperfections which make us "human". These imperfections feel more organic and allow us to relate to the character a little bit more. I think Maya users tend to eschew exact measurements sometimes for a "look", whereas packages like Max and AutoCAD are more firmly rooted in engineering, architecture etc. The control one gets in Houdini is so close to the control one gets in programming, it is hard for me (as a fine artist) to wrap my head around. Sometimes I don't want to punch in numbers to get the right result and I go running for a paintbrush or pencil because 3d software can be so drat picky/complex. .... Sorry for the rant. Houdini has been making me think about this type of thing a lot lately. (ie: control vs ease of use and how there will always be a trade off)
|
|
#
?
Jul 22, 2008 01:01
|
|
|
sigma 6 posted:Anytime you want to rig a robot or something inflexible, you just need to group your geometry under your bones. The individual pieces of the geometry under each corresponding bone that is. This is what I needed, thanks!
|
|
#
?
Jul 22, 2008 02:12
|
|
|
Martytoof posted:Sorry I didn't mean to come off sounding like approximation isn't a valid concept to use when modeling, I'm sure it has its uses. When it boils down to it, there were only two incidents that really made me scratch my head in the tutorial in question: maybe he scaled and welded the end-verts? I haven't seen the video, but what I'll do for some shapes when I'm doing edge-extrusion modeling when I'm capping stuff is rather than mash the cap button I'll pull out another very short loop and collapse the verts (which is like scaling down and then welding). When you smooth this it winds up giving you a slightly rounded cap instead of a flatter cap, which is similar to some shapes (like rubberized grips on a bike). Eyeballing something that should have been symmetrical seems weird, though. Symmetry is dirt simple to do. Was it obviously eyeballed and imperfect or maybe it he was taking a shortcut or two that wasn't explained that made it look eyeballed? Like he mirrored a piece, and then dragged it over into place, but what you didn't see was him turning on grid snap or something?
|
|
#
?
Jul 22, 2008 16:58
|
|
|
It's entirely possible he did turn on some grid snap feature, I know I did when I worked my way through the tutorial. I don't think he mentioned it but I could have missed something because I was concentrating on following the workflow myself. In any event, I capped and snapped so I came out ahead  I also went ahead and bought "Urban Environment Creation" because that's what I'm really more interested in. It sounds like it's got a different instructor so it should be interesting to see how the teaching styles differ.
|
|
#
?
Jul 22, 2008 17:21
|
|
|
comic geek alert http://www.apple.com/trailers/wb/watchmen/hd/  OK - not entirely related to 3d but hey . . . nice VFX!! Do yourself a favor and read the graphic novel before the movie comes out. Martytoof: Check out http://www.opacity.us/ if you need reference for decayed urban environments.
|
|
#
?
Jul 22, 2008 18:57
|
|
|
sigma 6 posted:comic geek alert I don't know anything about Watchmen, but that trailer made the movie look like utter dross covered in a thin veneer of over-the-top effects. Didn't sell me at all, I'm afraid. Not that I think you care, of course.
|
|
#
?
Jul 23, 2008 05:44
|
|
|
Read a book. http://en.wikipedia.org/wiki/Watchmen http://en.wikipedia.org/wiki/Watchmen_film quote:Watchmen was published in single-issue form over the course of 1986 and 1987. The miniseries was a commercial success, and its sales helped DC Comics briefly overtake its competitor Marvel Comics in the comic book direct market.[38] Watchmen has also received several awards spanning different categories and genres including: Kirby Awards for Best Finite Series, Best New Series, Best Writer, and Best Writer/Artist, Eisner Awards for Best Finite Series, Best Graphic Album, Best Writer, and Best Writer/Artist, and a Hugo Award for Other Forms.[39] Definitely read it before the movie ruins it for you. EDIT: Moore (the writer) "refuses to have his name attached to any...films" V for vendetta League of Extraordinary Gentlemen From Hell Swamp Thing Sorry to derail the thread slightly in the name of comics. sigma 6 fucked around with this message at 08:04 on Jul 23, 2008 |
|
#
?
Jul 23, 2008 07:46
|
|
|
Gromit posted:I don't know anything about Watchmen, but that trailer made the movie look like utter dross covered in a thin veneer of over-the-top effects. Didn't sell me at all, I'm afraid. I think the trailer was mainly to amp fans. Hence no mention of a plot or really anything else. Alan Moore is a dickbag, and V for Vendetta was better as a movie than a comic.
|
|
#
?
Jul 23, 2008 08:50
|
|
|
Been going through the Urban Environment CDs and my emotions went something like this: - Wow this is awesome  - Dang I can do this! - I can't wait to try this on material not covered in the tutorial!  - Huh? What are UVs? Oh. That's right, I remember those  - Well I suppose laying out UVs on this entire scene can't be all that diffic 
|
|
#
?
Jul 23, 2008 15:33
|
|
|
I suppose this would go here. Not 100% CG, but this was an experiment in motion tracking, and fake environment reflections using a big panoramic picture I created from the front of my house. http://www.vimeo.com/1395386 Here's some images to play with if someone wants to give a go at multi pass compositing. I use After Effects, but it should be straight forward on any compositing software. Background:  Click here for the full 1440x810 image. Shadow (I don't know whats up with the white specs around where the wheels should be):  Click here for the full 1440x810 image. Diffuse:  Click here for the full 1440x810 image. Ambient Occlusion:  Click here for the full 1440x810 image. Reflection:  Click here for the full 1440x810 image. Reflection layer is a bit tricky. Make the reflection layer it's own composition with this layer and a copy of the diffuse layer below it. Set the blending mode on the reflection layer to "difference". This will cancel out anything that's the same from the diffuse pass, leaving only the reflection. Drop this in the main composition over top the background, shadow, diffuse, and ambient occlusion layer. I am new to 3D Studio Max, and have only really focused on rendering because my primary interest is adding CG to live action. I downloaded the car models from a site I found. If anybody is interested I'll look for the link.
|
|
#
?
Jul 23, 2008 19:34
|
|
|
sigma 6 posted:Definitely read it before the movie ruins it for you. I've read Watchmen and while I thought it was pretty good, I still don't understand half of it. Absolutely loved Halo Jones though. But judging by this trailer and 300 I really don't know if this movie will live up to the comic. Anyway I think I'm about finished with this Treasure Hunter character, but I'm very much open to crits.  Click here for the full 680x684 image.  Click here for the full 520x782 image.
|
|
#
?
Jul 23, 2008 21:59
|
|
|
stuckeys: Never read Halo Jones but I was so pissed off at The League of Extraordinary Gentlemen adaptation I wanted to demand my money back. He takes his name off of his movies because they are generally horrible adaptations. V for Vendetta being the arguable exception. As for your treasure hunter: Why did you normal map the eyes? The clothes look like they could have a few more wrinkles sculpted in. Overall pretty decent low poly character. Are you the stuckeys which is always posting speed paintings in the draw everyday threads? If so -- I love your work. ACanofPepsi: I tried out compositing your stills in photoshop but was not able to make the reflection pass to look good with "difference" blending over the diffuse. Using "screen" seemed to work much better. I always thought "add" should be used for reflection or spec but since there is no add in photoshop, screen must be used instead. sigma 6 fucked around with this message at 03:33 on Jul 24, 2008 |
|
#
?
Jul 24, 2008 03:16
|
|
|
sigma 6 posted:Definitely read it before the movie ruins it for you. I'm not going to derail this any furhter, but I don't think I'm into this genre enough to start. The trailer made me *not* want to see it, and I'm unlikely to want to have to read a book in order to "get" the movie. Perhaps this is fan material only. Getting back onto 3D, does anyone know why the Burly Brawl scene in the Matrix sequel looks so poor compared to what 3D could do at the time. Did they have to rush that scene or something? All the guys in it look like plasticine. I thought Wallace and Gromit were going to show up.
|
|
#
?
Jul 24, 2008 03:27
|
|
|
Gromit posted:I'm not going to derail this any furhter, but I don't think I'm into this genre enough to start. The trailer made me *not* want to see it, and I'm unlikely to want to have to read a book in order to "get" the movie. Perhaps this is fan material only. On another topic, am I the only one exploring Cinema 4D right now? The GI in Advanced Renderer looks so incredibly good for stylized motion graphics related animations. I can't get used to the sub-d modelling though. :/
|
|
#
?
Jul 24, 2008 05:21
|
|
|
sigma 6 posted:As for your treasure hunter: Thanks! Yeah, that's me in the drawing threads, though I haven't been posting in them for a little while now. Appreciate the crit too. For the eye, I saw a couple of 3d examples (I forget where exactly) that showed the iris area modeled as a bulge, so I basically just mimicked that. But I guess that doesn't really apply to game normal maps?
|
|
#
?
Jul 24, 2008 06:24
|
|
|
A few textures, and a bit better lighting. I might do something about the lanterns, but that can wait.   Again, comments and crits welcome.
|
|
#
?
Jul 24, 2008 07:09
|
|
|
sigma 6 posted:there is no add in photoshop This breaks my brain. 
|
|
#
?
Jul 24, 2008 13:29
|
|
|
Handiklap posted:This breaks my brain. This is what Jeremy Birn said. That there was no "add" in photoshop due to some weird copywrite issue. There is an "add" in After Effects and Shake however. I am only repeating what I remember him saying. I imagine he knows what he is talking about but please enlighten me if I am wrong. stuckeys: I have never heard of normal mapping the bulge in an eyeball before. Certainly if it is a subtle effect, I am sure it doesn't matter so much. Looks a little extreme in the normal map to me and I doubt it is really necessary. sigma 6 fucked around with this message at 17:13 on Jul 24, 2008 |
|
#
?
Jul 24, 2008 15:37
|
|
|
Isn't Overlay pretty much Add? I can't imagine normal mapping an eye being of any use. Oh well. ceebee fucked around with this message at 18:44 on Jul 24, 2008 |
|
|
#
?
Jul 24, 2008 18:41
|
|
|
sigma 6 posted:stuckeys: I would actually recommend this. Its currently whats being done on a project I'm with and it looks fantastic. It can be overdone, but if its done subtly it really makes things stand out.
|
|
#
?
Jul 24, 2008 21:03
|
|
|
The cornea definitely bulges out more on an eyeball and I can see where it would be done but it just looked a bit extreme in the normal map. Also wouldn't this depend on how close the camera gets to the eye? Can someone clear up these compositing questions? I only took a few VFX courses in college and I consider myself a newb in the area. What would photoshop's equivalent of "Add" be? Or am I to assume by Handiklap's comment that putting one layer over another in "normal" blending mode would be the equivalent of "add"?
|
|
#
?
Jul 24, 2008 23:44
|
|
|
Portable Staplefrog posted:This is what I needed, thanks! That way works but with large setups you can lose track of model bits embedded in your skeleton. What I like to do is similar to large production companies is to keep the deforming and non deforming geometry out of the actual character skeleton. I guess in maya terms that would be doing bind via sets and groups, so each geometry would have one influence at 100% for those elements. Keeps scene organization real easy and neat, and you only have one or a few objects in your scene instead of 1000's of small bits of geometry scattered all over the place on your skeleton.
|
|
#
?
Jul 25, 2008 01:39
|
|
|
re Add: I'm pretty sure it's Linear Dodge.
|
|
#
?
Jul 25, 2008 01:39
|
|
|
Heintje posted:re Add: I'm pretty sure it's Linear Dodge. edit: Actually, Photoshop now has "(add)" written beside Linear Dodge. IHeartBoobs fucked around with this message at 05:00 on Jul 25, 2008 |
|
#
?
Jul 25, 2008 04:58
|
|
|
^^^ Hah! Didn't see that! Thanks. BigKOfJustice: Nice tip!
|
|
#
?
Jul 25, 2008 05:12
|
|
|
Yeah screen works well too, but just remember that it's not a mathematically linear add, it's doing other poo poo too.
|
|
#
?
Jul 25, 2008 05:47
|
|
|
Screen is a bit different. From the Shake help button: The Screen function mimics the effect of exposing two film negatives together. Technically, it inverts both layers, multiplies the two layers together, and inverts the result. It is particularly handy for reflections and luminescent elements, as it preserves the highlights. I use it for adding rotoshapes together. Also, wooo I think I'm gonna have a steady job on the sarah conner chronicles. Yay no more freelance commercials!
|
|
#
?
Jul 25, 2008 08:11
|
|
|
PowerLlama posted:
This is bad rear end. What kind of work? Compositing, CG??? I'm really excited to see the new season start. Also, with my car post up there, I assumed people would be combining the images in After Effects because that is what I did the video in, obviously since I posted still images people would jump to photoshop. I tried combining the images I posted in photoshop and I just don't get the same results After Effects gave me. Lesson learned, I guess Coke truly is better...
|
|
#
?
Jul 25, 2008 20:58
|
|
|
^^^ I would like to see screen shots of the differences. I just tried using Linear Dodge (add) in photoshop with a reflection pass generated from max->mental ray and it looked almost identical to "screen".
|
|
#
?
Jul 26, 2008 02:54
|
|
|
1. This is the reflection precomp. (Just a fancy word for a composition that will be nested in another composition) Click here for the full image. 2. Change the reflection layer blending mode.  Click here for the full image. 3. Afterwards put the precomp in your main composition, and change it's blending mode to Add. Afterwards you can tweak the reflection layer's opacity and change the intensity of the reflection.  Click here for the full image.
|
|
#
?
Jul 26, 2008 05:38
|
|
|
ACanofPepsi posted:1. This is the reflection precomp. (Just a fancy word for a composition that will be nested in another composition) Just a quick note...the thick black lines on your texture absolutely destroy the work. They completely ruin it. Also there is no proper shadows....just a faint shadow with no thick shadows near where the tires actually touch. It could be pretty good (you have some nice camera matching and the modeling is sufficient), but those two things just take me right out.
|
|
#
?
Jul 26, 2008 07:45
|
|
|
BonoMan posted:Just a quick note...the thick black lines on your texture absolutely destroy the work. They completely ruin it. Do you mean the thick black lines on the hood/headlights of the car, etc? or the outlines from the matting/motion blue or the CG car? The lighter outline of the CG itself can be taken away with settings, but if you were commenting on the model itself, yeah it's pretty basic. You're absolutely right as well, the shadow SHOULD be darker where the wheels touch the ground. Time to play around some more, once again thanks, I didn't even notice that before and now I can't stop noticing. Fake edit: I actually didn't model this car FYI, I found it online. I just match moved it, positioned it, and added a fake environment reflection that I made out of a big 360 degree picture I took out front of my house. bring back old gbs fucked around with this message at 16:14 on Jul 26, 2008 |
|
#
?
Jul 26, 2008 16:12
|
|
|
ACanofPepsi posted:Do you mean the thick black lines on the hood/headlights of the car, etc? or the outlines from the matting/motion blue or the CG car? The lighter outline of the CG itself can be taken away with settings, but if you were commenting on the model itself, yeah it's pretty basic. Oh ok well that's more reassuring! Yeah I'm referring to the actual black lines on the texture. They make it seem like a toy car with the lines "painted on" to hint at the gaps in real cars where the panels come together. Are you rendering out of Maya? Mental Ray?
|
|
#
?
Jul 26, 2008 18:52
|
|
|
^^ 3D Studio Max using VRay to render.I find it easier to render in passes with Mental Ray, but it just takes way too long to render image sequences with it.
|
|
#
?
Jul 27, 2008 06:16
|
|
|
A little side project I've started to tinker with when there aren't any retirement homes needing elevations rendered. A cellular tower turned drum machine. The different antennas open up, small metal arms come out and start drumming on the "covers". Just making the different parts before I populate it more. I picture the larger one as a base drum, the smaller coned one as hi hats and the vertical ones as some akin to a xylophone or keyboard. There will also be some that turn into speakers. Something for the strobe lights too, haven't thought that far yet.  Click here for the full 619x1080 image.
|
|
#
?
Jul 27, 2008 23:13
|
|
|
I should've looked around for this thread a long, long time ago. I'm in my second year of a Digital Animation & Game Design program, trying to get into the character modeling side of things. Like many others, unwrapping was a nightmare for me. I say was because someone in one of my classes recently dug up a handy little tool called Roadkill. I can't recommend this enough if you haven't tried it yet. It integrates into Max, Maya, and XSI, and makes the unwrapping process on organic objects a breeze. Hell, even using it for hard surfaces isn't half bad. On a side note, this is my current work in project for the character class I'm taking over the summer. Workflow was base mesh through Max, sculpting and texturing through Zbrush.  Click here for the full 960x720 image (low poly).  Click here for the full 960x720 image (high poly). My current struggles are in mapping and lighting. My rendering and lighting class in fall last year was a nightmare for me to say the least. Does anyone have some good links for setting up nice lighting in Max? Certainly wouldn't mind any information provided on shaders as well - Zbrush matcaps look nice, but they count for nothing outside the program.
|
|
#
?
Jul 28, 2008 02:10
|
|
|
Well, it's probably easiest to show us what you think's wrong rather than just asking. Show us a few images to see where your texturing and lighting skills are at, and why're they not where you want them?
|
|
#
?
Jul 28, 2008 03:18
|
|
|
Been adding some textures. Now I need to start thinking how to animate it. I've never really done any animation in MAX previous. I really only use it to render architectural stuff. Does anyone have an idea of the proper MAX technique to easily replicate a series of keyframe snippets to match a beat?  Click here for the full 714x1080 image.
|
|
#
?
Jul 28, 2008 04:25
|
|
|
|
| # ? Apr 23, 2024 20:39 |
|
|
Here's some reading on the mathematical side of the blend modes. I used this, along with the Adobe Blend Modes Addendum to translate all of them over to maxscript, for native comping. I don't use it a lot lately, but it really was useful for overnight batch rendering with our old multipass pipeline. multipass 
Handiklap fucked around with this message at 14:49 on Jul 28, 2008 |
|
#
?
Jul 28, 2008 14:45
|
|