|
Pretty sure that Maya has layered shaders that you can use for that purpose.
|
#
?
Apr 2, 2014 01:30
|
|
|
|
| # ? May 29, 2024 18:44 |
|
|
tuna posted:Ah, I meant UE4. Yea it's FBX 2013 for UE4, it throws up a ton of errors when you use 2014.
|
|
#
?
Apr 2, 2014 03:27
|
|
|
If I wanted to animate something like cellular division, what area of animation should I be looking at? Would this be done using particles?
|
|
#
?
Apr 3, 2014 00:48
|
|
|
Possibly particles with some type of mesher like Frost from Thinkbox Maybe those meshers work on stuff that isn't particles too, like you could just have two spheres scaling and moving apart and it could mesh them together
|
|
#
?
Apr 3, 2014 01:42
|
|
|
Metaballs in Cinema4d can do this, but you'd have to keyframe a bunch of things to keep the scaling correct and I don't know how it would react to transparency or splitting the material
|
|
#
?
Apr 3, 2014 01:57
|
|
|
SynthOrange posted:Pretty sure that Maya has layered shaders that you can use for that purpose. Oh cool, thanks! This is a good reminder to never tell a client something can't be done (which I just nearly avoided doing). Because it probably can, I just don't know how.
|
|
#
?
Apr 3, 2014 02:39
|
|
|
Had a day or two off from projects so I've been trying to pump out some quick things and learn some stuff! https://www.youtube.com/watch?v=Os3S9wujIxc None of it's super refined or anything, but it's nice to get some stuff done without it having to look too presentable. If there's anything that looks really obviously crappy let me know though, can always stand to add some tips to my arsenal.
|
|
#
?
Apr 3, 2014 20:58
|
|
|
This is cool. I'd get rid of that blue screen at the end though - you should do a little fake writeup (maybe a mockup of the 'implant') and post it around the internet as a leaked prototype. I can totally see gangs of people sharing this on facebook.
|
|
#
?
Apr 3, 2014 21:30
|
|
|
I'm trying to do some compositing between c4d and ae - having trouble tracking the camera. I'm using foundry camera tracker which seems to work well for shots with a bit of distance but it keeps reading the points on this shot as more or less sitting on one plane in z-depth - the shot rotates slowly to the left to show more of the room and I need to put an object behind the bed. I havn't tracked any shots as close up as this before and I have a few to do - is there another way round it? Could upload a sample if needed... 
|
|
#
?
Apr 3, 2014 22:12
|
|
|
 I can finally post this after like 6 months, since our project has gone public (I did not do the logo or the character art, just the models and textures.)
|
|
#
?
Apr 3, 2014 22:47
|
|
|
Dad Dick Embrace. posted:I'm trying to do some compositing between c4d and ae - having trouble tracking the camera. I'm using foundry camera tracker which seems to work well for shots with a bit of distance but it keeps reading the points on this shot as more or less sitting on one plane in z-depth - the shot rotates slowly to the left to show more of the room and I need to put an object behind the bed. I havn't tracked any shots as close up as this before and I have a few to do - is there another way round it? Could upload a sample if needed... Can't tell from your still, but if the camera is just turning on a tripod(nodal pan) a 3d solve won't really be possible since there is no paralax to pull depth info from. A nodal pan looks like a bunch of points projected onto a sphere when viewed in 3D. These points can still be useful for making masks.
|
|
#
?
Apr 3, 2014 23:43
|
|
|
homo punching bag posted:Can't tell from your still, but if the camera is just turning on a tripod(nodal pan) a 3d solve won't really be possible since there is no paralax to pull depth info from. A nodal pan looks like a bunch of points projected onto a sphere when viewed in 3D. These points can still be useful for making masks. thanks, thats what I figured in the end - sadly most of the shots i have to work with are like this. might try ae cc and cineware and set the depth manually
|
|
#
?
Apr 4, 2014 01:29
|
|
|
Dad Dick Embrace. posted:thanks, thats what I figured in the end - sadly most of the shots i have to work with are like this. might try ae cc and cineware and set the depth manually If you have access to AE:CC just do it with Mocha. You can track that back wall no problem. Mocha is the best. You can either use the corner pin data for tracking and masking, or track position/scale/rotation and just roto the bed/wall with a simple mask. Personally I'd corner pin since the area is on the edge of the frame and will probably have weird distortion that the position/scale/rotation method won't take into account. I'd also do the headboard masking by hand since it's just straight lines and I'm assuming the camera is steady.
|
|
#
?
Apr 4, 2014 01:43
|
|
|
homo punching bag posted:Do you need to extend the room past the bed, or just add a poster/TV screen to the wall? I can help with the latter.  i have to place an animated object standing between the bed and the walls to the left of the frame which, judging by the lighting, will be casting shadows on the wall behind it and maybe onto the bed - so my hope was to get the camera motion tracked with a few accurate null points and build some simple shadow catchers in c4d to go with the object render. i've used mocha for 2d stuff - can you use it to track camera movement and get it into c4d? edit: I've tried a boujou nodal pan solver and a mocha pro camera solve with no luck - as far as I can work out the only way forward is to create some tracked planes with mocha, take into ae and make a 3D scene with them, then export that to c4d and somehow transfer the movement of that scene to a camera, or just have the entire scene move past the still camera... unless i'm missing something really obvious and nodal pans don't really need all that due to their 2 dimensional nature - so I can just track camera movement and estimate in the object and shadow catchers? having a bit of difficulty visualising how to put this together. sweet to chat fucked around with this message at 13:21 on Apr 4, 2014 |
|
#
?
Apr 4, 2014 09:18
|
|
|
Redshift 1.0 is out for Maya and Softimage. It is amazing. https://www.redshift3d.com/
|
|
#
?
Apr 4, 2014 13:54
|
|
|
Dad Dick Embrace. posted:unless i'm missing something really obvious and nodal pans don't really need all that due to their 2 dimensional nature - so I can just track camera movement and estimate in the object and shadow catchers? having a bit of difficulty visualising how to put this together. This is probably what I'd do, it should be fairly simple. There won't be any parallax movement so it just needs a 2d track, and it'll probably be simplest to eye what looks best with shadows etc. and just put them in in AE.
|
|
#
?
Apr 5, 2014 14:41
|
|
|
OtherCubed posted:This is probably what I'd do, it should be fairly simple. There won't be any parallax movement so it just needs a 2d track, and it'll probably be simplest to eye what looks best with shadows etc. and just put them in in AE. cool, thanks, I guess i just got swept up in the idea of making a really cohesive 3D scene!
|
|
#
?
Apr 5, 2014 16:15
|
|
|
This is really cool.
|
|
#
?
Apr 5, 2014 16:45
|
|
|
Dad Dick Embrace. posted:cool, thanks, I guess i just got swept up in the idea of making a really cohesive 3D scene! One of the most useful things I ever read on here was someone saying '3d is an illusion'. It's all about doing the simplest thing you can to create the most elaborate result! itiot posted:This is really cool. cubicle gangster posted:This is cool. I'd get rid of that blue screen at the end though - you should do a little fake writeup (maybe a mockup of the 'implant') and post it around the internet as a leaked prototype. I can totally see gangs of people sharing this on facebook. Cheers guys! Yeah, I might do a more in-depth one at some point, it was pretty fun but I was kind of doing it as a testing ground for some ideas. Now I know I can do it I'll probably try and make something a bit more advanced at some point.
|
|
#
?
Apr 5, 2014 17:09
|
|
|
Dad Dick Embrace. posted:
Take a cube roughly matching the dimensions of the the room and eyeball it!. It'll take 30 mins tops. For sanity check, take another box, roughly matching the bed for cross reference.
|
|
#
?
Apr 5, 2014 20:19
|
|
|
Downloaded the demo of Headus Uv Layout after seeing it at work and holy poo poo why didn't I do this before. gently caress UV'ing in anything else after using this.  Also Mari 2.6 is out and it is amazing as usual. Too bad it costs 2.5k because I can only use it from 15 day trial to trial. Hope it goes on sale later this year.  http://www.thefoundry.co.uk/products/mari/
|
|
#
?
Apr 6, 2014 05:13
|
|
|
keyframe posted:Downloaded the demo of Headus Uv Layout after seeing it at work and holy poo poo why didn't I do this before. gently caress UV'ing in anything else after using this. UV Layout might look like it belongs on a mid-90's Unix workstation (I find it kind of charming really), but yes, it is indeed probably the most painless UV solution out there.
|
|
#
?
Apr 6, 2014 06:27
|
|
|
On that note, I've got a texturing question for you guys: I've modeled this clock and I'm trying to determine if there's a better way to texture the face than manually creating 60 different text layers in Photoshop. Is there some kind of Photoshop action or plugin that would allow me to create a given number of layers rotated around the center of the canvas (like a 'Duplicate Special' in Maya, for instance)? If it could increment a numerical value with each copy as well, that would of course be amazing, but I'd settle for anything that would make this a bit easier. I'm not stuck on Photoshop either - if you have a suggestion for another program that could do the job I'd love to check it out.
|
|
#
?
Apr 6, 2014 19:07
|
|
|
That sounds like a complete bugger. Off the top of my head, I think I'd go about that by: 1. Creating separate images for every number by keyframing values in after effects and rendering as a png sequence. 2. Laying them out in maya as separate textured planes, would probably have to use a little python/mel, and then render with an orthro camera. That's kind of insane, there's probably a much easier way to go about it.
|
|
#
?
Apr 7, 2014 00:09
|
|
|
I've been tooling around with a couple ideas: - Illustrator has a Type Along Path tool, which is almost perfect except that I don't actually want my text to follow a curve, but to create discrete text layers along points on that curve. - Using good old Duplicate Special in Maya, I can easily get text splines arranged the way I want them, and with some MEL scripting I'm pretty sure I could even get the right numbers on them. If I convert the exported DWG from Maya to an earlier version of the format I can get it into Illustrator. This seems like the most promising avenue so far. The After Effects idea is interesting - not sure I would have thought of that one myself.
|
|
#
?
Apr 7, 2014 01:07
|
|
|
Illustrator can do it accurately, in the sense that you can make a 60 pointed star and accurately align all 60 bits of text to each point, but yeah, I dont know of any tools that will do it right away. It's going to be messy one way or another.
|
|
#
?
Apr 7, 2014 01:11
|
|
|
I managed to get something kinda close in Adobe Illustrator using the Area Type function, but getting the positioning and spacing right that way may not be possible. I'd still suggest using Illustrator, just an alternate method in it.
|
|
#
?
Apr 7, 2014 01:27
|
|
|
Edit: Nevermind.
mutata fucked around with this message at 03:41 on Apr 7, 2014 |
|
#
?
Apr 7, 2014 03:39
|
|
|
Well, looks like high time to bone up on my Illustrator skills. Thanks for the ideas guys - I'll be sure to post a final render when I'm done with everything.
|
|
#
?
Apr 7, 2014 04:20
|
|
|
3d text inside max that's pivot aligned to the verteces of a sixty sided cylinder?
|
|
#
?
Apr 7, 2014 05:08
|
|
|
yeah i'd do that entire face in max and render it out to use that as a texture. you can also create a text object, lock the inherit rotation and link it to a dummy - you'll be able to rotate text around a larger circle while keeping it vertical. you could write a script that increases the number by 1, rotates it 6 degrees, selects the number, copies it, breaks the link to drop it in place and loops 60 times.
|
|
#
?
Apr 7, 2014 16:22
|
|
|
I just had a catastrophic After Effects crash where it failed during an auto-save so my original file and the version it incremented as it crashed were both screwed up. The file wasn't corrupted though, luckily. It would load, but instantly crash on the frame it was trying to render. Very stressful since both versions were like this, but opening the file with CAPS LOCK on prevented the frame from loading and AE from crashing. Usually I don't get so lucky.
|
|
#
?
Apr 8, 2014 19:21
|
|
|
 Something I'm working on for university. We needed to make a very heavily illustrated concept document based on an 80's cartoon IP. I chose a relatively obscure cartoon that's very popular in Scandinavia and Japan. Rather than draw the character a billion times I thought I'd do a model and pose it. Cute dog! He's been fun to model and I've gotten a lot out of some tutorials I've been looking at lately. 
uglynoodles fucked around with this message at 00:22 on Apr 10, 2014 |
|
#
?
Apr 9, 2014 21:47
|
|
|
I dunno if this is the proper place to ask it, but: can anyone recommend a good, free resource for some kind of poseable 3D human model? I've been trying to practice some less common angles with traditional drawing (such as a more top-down perspective), and I can't find a great deal of stock that's in the right sort of angles for it.
|
|
#
?
Apr 10, 2014 04:57
|
|
|
http://www.posemaniacs.com/ ? It's not interactive apart from being able to rotate the models, but there's a hojillion poses on there.
Synthbuttrange fucked around with this message at 05:03 on Apr 10, 2014 |
|
#
?
Apr 10, 2014 05:01
|
|
|
Thank you, and sorry if the question was a bit out of place - I figured the 3D thread would know best for this sort of thing. v  v
|
|
#
?
Apr 10, 2014 05:03
|
|
|
Daz3d is basically a free version of poser.
|
|
#
?
Apr 10, 2014 06:13
|
|
|
Hey, have a workflow query - I'm working on a project with another animator who is using 3DS max - I'm using C4D. He's doing the rigging and animation and I am compositing the scenes, doing environmental effects etc. I'm aware that there's a number of problems with trying to export rigged animation objects from one software to another, but I understand that you can bake the animation in max into the object and use that in c4d? Secondly, one of the animated objects needs some strands of dynamic hair added - is it possible to add this onto the baked object in c4d or will it need to be made in max and turned into a polygon object for rendering in c4d? e: we tried out a sample and it works suspiciously smoothly - you can just drop hairs on anything. sweet to chat fucked around with this message at 18:46 on Apr 10, 2014 |
|
#
?
Apr 10, 2014 09:03
|
|
|
I've been doing more water r&d for potential future boat jobs. This is the first test done in my spare time to convince my ceo to give me a few days to give it an environment, a few cameras & turns etc. https://vimeo.com/91627172 There's an emitter at the back of the boat to create the wake caused by the propeller but it needs turning up a bit. The power was right but it was too big, so I scaled it down and forgot that the lower surface area would drop the power too. Easy to do, but I couldn't be arsed waiting 1 night to sim and another to render for such a small change. Simulation time was 30s a frame at the start, up to 1min at the end (1,000 frames in total, took 300 to establish a good foam pattern) Render time was 8 minutes a frame at 720p. all in, 700 frames were simmed and rendered on 4 machines here over 2 nights (no days, just 8pm to 10am). I think it's nearly production ready.
|
|
#
?
Apr 10, 2014 18:08
|
|
|
|
| # ? May 29, 2024 18:44 |
|
|
What simulator did you use for that work?
|
|
#
?
Apr 10, 2014 18:56
|
|