everfresh

keep in mind that c4d rigging people are comparably rare, that's a good thing if you are freelancing. competition is much higher in the maya world. i'm only aware of a handful of c4d rigging people as good as you are, and given the fact that you're only into this for 2 years now i'd say your future is bright. so maybe not such a bad idea to stick with c4d, just sayin. ;) just like dan i also suck at math, and also coding, and i'm also surprised that i'm able to understand most parts of expresso setups i look at, but i still find it hard to come up with more complex solutions in xpresso by myself. my main enthusiasm for rigging is purely the fact that i want to animate comfortably, but i don't really enjoy dealing with priorities (i guess no one does though) and the general process of debugging. and although i'm much less capable in rigging than you are i've managed to get rigging gigs every now and then. people just need to know you even exist, so don't forget to work on that as well.

like bezo said, you have extra points floating around in your floor object... if you run the optimize command to get rid of those extra points the sim is working as expected. if you have set the collision shape of an object to automatic, c4d draws an invisible bounding box around the object for simulation, because it's faster to calculate. and since your floor object has those extra points further up, that invisible bounding box is colliding with the sphere and makes it shoot up.

thanks for sharing, much appreciated. i've also tried this in the past but couldn't get it to work properly.

can't tell, i haven't really worked in r20 yet, at least not with xrefs... but my guess would be that it's still present. it's been around since r16, so i'd be surprised if they fixed it in r20 :D just remember to untick that box each time you create an xref, and you'll never get the greyed out buttons again. a little annoying to do those extra steps every time, but you'll get used to it. ;)

sounds like you already know the pros and cons... i think it all comes down to what you want to do, for instance if you are going to have scenes with dozens of characters all interacting with each other it will get tough in c4d. if the video contains a lot of mograph elements and stuff then that's another thing to consider, maybe c4d would be the better choice then. i'd say if you feel more comfortable in c4d, just go with c4d. there's not much maya can do that c4d can't do in character animation and rigging, it's just much harder to find tutorials, scripts and plugins related to character animation in c4d since it's not it's core area. but if you really need help with the project it will be much tougher to find. regarding your questions what keeps me animating with c4d? i just feel comfortable with the application and i also don't just do character animation. for the work i do i feel that it's the best choice overall, and although i do a lot of character animation, as a hobby and also commercially every now and then i didn't really feel the need yet to switch to maya. no doubt it would be the better choice if i would just do CA exclusively or if i was looking to get employed, but i don't. i'm a freelancer, i do graphics design, motion graphics and character animation for a living, and since c4d character animation people are rare it would be somewhat foolish to switch to maya, at least from an economical standpoint, i feel like this is a pretty good niche nowadays.

if you need it just for a still, i'd just comp in an image. if it needs to be animated you could also do it in post with trapcode particular for instance. i've managed to get pretty realistic cigarette smoke out of it in the past. if you need to do it directly in c4d you could either do it with xparticles by emitting from only the shell, not the volume of a cylindrical object and using curl noise. be aware that you will need millions of particles to get a dense uniform smoke. also you can try to make it with explosia fx in xparticles, in that case i'd try to emit from a plane that is animated, rotating and moving around at the same time. if you don't have xp you could try it with a texture only solution. make a few planes, add a white material with an animated turbulent gradient in the alpha channel to them and animate the material offset in y. additionally add a wind effector to those planes to get an extra layer of wavyness. i've done this in the past for cartoon projects, looked more realistic than i wanted it to look.

there's a bug in xref objects that messes up the file path after a while, resulting in the greyed out reload button. but luckily there's a fix: every time you add an xref object and loaded in your file, you need to change the xref file path from relative to absolute. in the xrefs options there's a tab "file" with a "relative" tickbox. untick that and you're good to go.

here's a version also using booles, if you don't have too many letters it should run ok in the vp... topotypo.c4d

ah, i see.... sorry i missed that in the beginning. that's because the realflow mesh isn't actually moving. it's an independent mesh built for each frame. i'm afraid you will have to render sub frame motion blur directly instead of doing it in post then... motion vectors won't work here unfortunately.

for best results render your motion vector 32bit. after import right click on your motion vector footage in AE and select "interpret footage" >> "main", then go to colour management tab and check "preserve rgb"... in a perfect world that would be it, but we're not done yet. you also need to render out an object buffer for your to be blurred objects and use that as a mask for your motion vector. precomp that and use that comp as your RSMB input. wait, no, still not done. for optimal results precomp that precomp again to be able to apply a simple matte choker to the first precomp and play with the values until you get correct looking results. so you would tell RSMB to look at the second precomp which contains the first precomp which has the choker applied and contains the masked motion vector. sorry for the convoluted workflow, but that's the only way i know of to get motion vectors to work correctly.

i'm already rolling, although i might not be dead yet. on exhausting days like today i kinda feel like it though :D @Igor should be an easy fix, if you still have the vector file... if not just say it and i'll send it to you again. ;)

last night i gave prorender in r20 a shot, and well, still unimpressed. :/ first i opened one of the example scenes from the content browser, the one with the raspberries in a glas and the pudding. the 100 iterations took 24 minutes and it was still noisy as hell. then i made a simple cube on a plane slapped a pbr material on it and rendered it. it took about 9 min (500 iterations) to get the noise to an acceptable level. i ran those tests on a 2016 mbp with a radeon pro 460, haven't done any tests yet on my mac pro trashcan with the 2 fire pro 600 cards, but i don't think it will be significantly faster. so what sort of card(s) would one need to run to get a somewhat decent render speed out of prorender? i was really hoping prorender made a big jump in render speed compared to r19, but i can't see anything of that.

@Fastbee i'm a redshift user, so honestly it doesn't even matter to me that much. also hoping redshift will come up soon with a build for r20. i'm just stunned how slow prorender still is.i know it isn't developed by MAXON, but i don't really follow the development of prorender, so i imagined it might have become faster over the course of a whole year. apparently not. at least not as far as i can tell. really havin a hard time wrapping my head around what's the point of a gpu render engine that's slower than most cpu renderers. when a few examples were posted at the time r20 was announced it didn't seem too bad regarding render times. the one with the kid in the garage took what, 22 minutes or so? was a quite complex indoor scene with a lot of stuff going on, even redshift would have taken around 5-7 minutes on that on a 1080ti. somehow doesn't add up with my experiences with prorender.

that's actually great stuff! a little jitter is part of the hand drawn look, in fact i tend to add fake jitter to my 3D stuff to get a more hand drawn feel.

yes, directly in after effects. i usually go for solids with masks. a lot of people use photoshop, especially if there's shading involved since it's a bit more convenient to draw in photoshop. but for simple shapes like i mostly use after effects is the more efficient choice i think. this video and breakdown might also be of interest for you:

proxies are placeholders, like models that just roughly have the shape the final object should have. they are there to establish the motion and animate the camera in c4D and will be replaced in post with paint-overs. usually for this sort of animations my characters are 3D and most of the rest is 2D, here and there some 3D proxy elements as a guide to paint over to create those melting effects. and yes, you could also make those animations with shading, it just means (a lot) more work. i'm very much looking forward to r20 with fields and volumes at this point for this sort of stuff, since we will be able to do a lot more in this regard with those new tools.

morphs like these are usually just mostly tricks to the eye... and if there is an actual morph going on a lot of times it's just animated frame by frame. i've done a couple of smooth transition morph animations with a mixture of 3D and 2D techniques. the way i go about it is to animate the elements or proxies of the elements in 3D and then refine those in after effects with frame by frame animation.

your sketch looks fine, the problem with your 3D version is that you're clinging on to too much detail that fights with the cartoony proportions. notice that in your sketch the whole body - from chest to feet - form a clean line, while your 3D model still has detailed and somewhat realistic muscle definition, defined hips and so on. i mean it's all about which style you wanna go for, maybe you just have something different in mind when you say cartoony than i do. but most people would characterize cartoony with much simpler and cleaner shapes.

not really that much difference and still a bit too realistic if you strive for cartoony... other than his head being bigger in relation nothing essential has changed. try really messing with the proportions, i'd suggest making the torso much longer and legs much shorter and thinner. also i'd make the arms thinner. btw, is that a self-portrait? if so, look at the mirror and try to find characteristics of your body that you could exaggerate.

half man half cat. ;)

they always do ;)

thanks... characters and cat tree and title are c4d, the rest is mostly frame by frame animation in after effects.

ok, that doesn't work anymore... the problem seems to be that you have a dynamic tag on your target nulls. just make another null, make it a child of your target null, reset psr, drag the dynamic tag onto that new null and then in your connector link to that instead of the target null... then it works.

seems to work just fine here as it is... just opened your file, didn't touch anything... i'm on the latest r19 version.... https://gyazo.com/b59656b1f4a296be454855b7fa7cb94a

chris schmidt once did something like that in an ask gsg episode... of course with soft body dynamics. ;) worked surprisingly well. but i agree with maliohammad here, simply rigging it up with joints and animating it by hand gives you the best possible control over the movement.