Jump to content

iline

Members
  • Content Count

    749
  • Joined

  • Last visited

Everything posted by iline

  1. in terms of computation, it is far far better to use either Connectors in Collider mode than real full-fat dynamics. Making chains of connectors is arduous, look up Chris Smith's latest NAB video for a tip where he clones a grouped object and connectors with a piece of 'predictive xpresso' = it only works once you make the cloner editable, but the xpresso tag sets the relative paths of Object A and Object B to look in the same place in hierarchy, so when you have twenty of them they are automagically hooked up instead of setting 40 connections by hand tediously. Because the pivot points will have been set carefully, and the dynamics Connectors do the swinging motion for you (make sure to use a restriction tag so they only swing on the Z-axis or whatever) the chain will interact with other RBD/Collider objects in dyn calcs Sorry it's not easier than this but you'll just get exploding chains otherwise. Forget Hair dynamics, even with Iterations set to 100 they'll still melt and sag out of Initial Pose and clones will not align/distribute along them. I promise.
  2. Hey, Vozzz is in Budapest. Matt Frodsham did an internship at Zeitguised in Berlin but generally the mograph scene there is pretty horrible, basically all the studios make explainer videos for startups. It is a bit depressing. If you look at the heavy hitters like Fuchs & Vogel, AixSponza, Sehsucht - they're not in the capital but rather spread out mostly in the south-west. The corollary is that a lot of German mographers work on car ads, and the motor companies are in Bavaria. Simon Fiedler, Cornelius Dämmrich, Holger Biebrach are all freelancing in the south of Germany where there are bigger studios with bigger clients. If you do end up in Berlin, try to attend the Faux Images events (organised by Mate Steinforth) as you'd be aamzed how many mographers there are in a city with such a dearth of decent motion work (and how attractive their girlfriends are). HTH
  3. I have the same beef with Arnold. As far as I could tell the Arnold tag with subdivisions is necessary to force sub-d computation. Also did you use a Normal Displacement node for the disp?
  4. ARRRGHH WHERE'S FAST BLUR??!! ...actually finally, Gaussian Blur now has 'repeat edge pixels' and is GPU accelerated, so Fast Blur has been deprecated. Just so you know.
  5. TeamViewer got haXX0red http://www.theregister.co.uk/2016/06/01/teamviewer_mass_breach_report/
  6. and this is why i want a little 512gb M.2 just for cache
  7. iline

    GTX 1080

    From the discussion on the Octane Facebook page, OTOY is going to probably take till August to make Octane 3 Pascal-friendly, with or without optimisations. By that time, the non-Founder's Edition cards will be out (bout £100/$150 cheaper). The 1080 is an 8Gb vRAM card - I'm curious how many millions of polys you can get in one scene before you hit that threshold, but hey, forcing people to model more economically and use instancing is no bad thing. But yeah, I am foaming at the mouth. I couldn't afford a Titan when they came out, and now I can just about afford two 1080s, which will soon enough inherit the throne of Octania. Until that point I'm thinking of piecing together a dual-xeon build so that Vray speeds the fuck up. When's Stefan gonna bring Vray RT for C4D? Anyone?
  8. https://aws.amazon.com/marketplace/pp/B00FGB3528/ref=sp_mpg_product_title/177-1529843-8539910?ie=UTF8&sr=0-2 I thought the fact that OTOY were selling AWS cloud licences implied you could cloud render. Also Octane does rather create the situation of relatively cheap hardware scalability - creating our own render farms - because it uses all GPUs on all networked render nodes (yes, not all the vRAM, I know) so two of those 4-titan watercooled badboys would pay for themselves pretty fast, I think.
  9. Substance Designer offers a subscription that confers ownership after 18 months of subscription payments. Finance, in other words. This is the model that small businesses and freelancers need.
  10. Yep. £385 for 3 months = £128.33 per month. At least Adobe had the sense to wean pirates onto payment schemes with lowered prices.
  11. Pretty sure the GPMU (Graphic and Printmakers' Union) would welcome you in with open arms, but they're myopic and Trotskyist and fossilised. Taking a legal fight to Adobe is likely pretty far from their list of stated aims. I'm really looking forward to having a subscription for C4D on my home computer. I use Adobe apps once a week so thank f'k I'm still on the promo rate of £26 p/m for CC.... Maya is doing a sub for £123 per month, I'm guessing C4D is similar as the two are almost the same price for box software?
  12. tumbleweed... though C4D Cafe is full of null objects who'd answer your question for you. I'm guessing your problem is related to baked objects not being live for collision calc, and there's a few ways to crack your particular nut, but C4D is so easy you probably worked this out about a minute after posting.
  13. Here's a modelling trick you might have missed: If you're modelling for subdivs, then you know creasing is a pain in the balzac. To make lovely even creases, double click an edge to get the whole edge (up to an angle), select the bevel tool and set it to Solid (there's other options, i'll save that for a video) and hey presto, lovely even creasings.
  14. I have to say - I loved CC 2014, but RAM Preview is somehow borked on CC 2015. I'm gonna use this dinky M.2 gumstick from my laptop as a cache disk for AE and see if that helps. I'm gonna down-save to CC 2014 tomorrow, download and install CC2014 and open this music video that is becoming an invidious challenge due to RAM Preview being a dick. Rendering: You know that multi-machine rendering trick with opening as many instances of BG Render as you have CPU threads? I can't be bothered to scroll up, preferring the sound of my own invective, but didn't we talk about AE being a multi-threaded application? Rather than opening AErendercore.exe instances which is, quite honestly, silly. Dual 8-core Xeons when I was working at Sky = 36 AE instances to load with 3GB of data each. That render won't start till Wednesday, but when it does, boy, you should see 'er go! (Todd, you know I love you man, but I hadn't started on this video when I did the survey)
  15. I wonder if I can convert wordpress portfolio posts into Beyonce posts automagically, because it does seem tedious to try to stand out in a windy cornfield filled with people who can do what I do.
  16. I don't take the site seriously, but I went to Faux Images in Berlin (200 people there with their cute partners, what? There's sod-all mograph work in Berlin) and the presenters said they were approached by multiple prospective clients after putting shiny HDRI nonsense on their Behance page. Can anyone attest to this?
  17. Layer Groups Ability to own a copy of AE forever after a year of rental, even if it's a deprecated version without Character Animator (shock!) Expressions for dummies: like a pickwhip tool on the top toolbar that can link two values of anything toComp/toWorld expressions to be just a context menu item on transform props twirly Node-based expressions More helpful error messages on expressions error dialogues ("needs to be of value 3, not 1" isn't gonna help anyone learn) with a clickable link to help page of probable causes
  18. I think cloners could work, if the bricks were really small. Then you could extrude a duplicate of each letter to make a cage around it, set this cage as the falloff shape on a Plain Effector (set to Visibility, Max=0 Min=0 to make the clones disappear) and then hit Invert on this falloff to make any bricks outside the edges of the cage invisible. If you went the UV map>displacement route, try to imagine the UV map you'd need to make it work - yeah Cubic might work, if you don't have curvature... Third way I can think of is doing the UVs, then using a tiling brick texture heightmap as a brush alpha in Sculpt mode, or in a third-party sculpting app (Sculptris is free, 3D-Coat inexpensive) All depends on the design concept as Mylenium says...
  19. iline

    Drive advice

    Binky - is a cache SSD that good/useful/noticeable? I always have a pang of guilt setting the media cache folder in AE to the OS drive.. On the note of SSDs - Apple and others do PCI-e x16 SSDs now, and they're likely raidable stripe 0 to give you scary fast xfer speeds. For laptops, you can get laptops with 3 M2 SSD slots which raided are something like 1gfps or something ridiculous. I'm weighing up laptops at the moment - narrowed it to the MSI WS60, the Dell XPS 15 and the ubiquitous MBP. Despite the macbook using onboard/AMD graphics its read/write speeds are the fastest of any laptop. Something to do with having welded their proprietary components to the mainboard. wankers. i digress.
  20. iline

    Cinema4D R17

    +1 on BP... They could just buy out 3D-Coat and implement their retopo/UV/Sculpt tools. As I keep saying, one thing that stops C4D being an innovatoris their stubborn refusal to address the 90s legacy components of BodyPaint, xpresso, Content Browser (and those fucking assets therein...) and may i suggest that they look to Autodesk for ideas about the viewport because fuck me it's 2015
  21. iline

    Cinema4D R17

    ufff...Not sure it's really fair or relevant to compare them, but as they are the same price, it becomes necessary. NO, is the answer. Viewport 2.0 is not coming to C4D, maxon haven't intimated that they see anything worth addressing there, and what with both vRay and Octane having live render viewports, they're likely disclined to. With regard to heavy scenes: cloner objects, xRefs, instances are pretty much fine but with dense polys - again, no viewport 2.0 means no vRAM means wireframe boxes and 1fps playback. I know Autodesk have absolutely smashed this with timeline caching in Maya 2016 Bake to texture in vray - okay, you can't compare vray in Maya with Cinema 4D vray. If you want to build amazing shader networks, you wouldn't touch Cinema. How is retopo? Well, they made the Polygon Paint brush... no, to be fair, Maya and C4D are almost evenly matched for box modelling. Although little things like parametric knife loops and hitting G to repeat last command are missing, actually, Cinema is nicely alternative, even with the extra tools that Maya brought in in 2015. To that end I'm looking at Modo>>> http://www.digitaltutors.com/tutorial/2190-MODO-for-CINEMA-4D-Artists
  22. Er, so I tried to play around with Beeple's Zero Day project files. They're pretty huge and stuff and don't play, because he's a mad genius who doesn't optimise anything. Anyhoo, I only have a demo of Octane but I do know that Vray and Octane are more or less identical in implementation for c4d - same tags, very similar materials. It's child's play to pull up the call commands for all params as well. I wanted to write a script that could convert all Beeple's lovely Octane materials automatically into Vray. But then, it would be easy enough to convert the camera, env, light tags as well. Anyone got an idea? I suck at python, I'd love to see what Lennart could pull out, and I am disinclined to use this to teach myself python if there is already snippets floating around I can mangle into shape. I'd like to see a similar material conversion script to see how to put a new vray material into memory and populate its params from (at least) the active Octane material and then replace it.
×
×
  • Create New...