Jump to content
[[Template core/front/profile/profileHeader is throwing an error. This theme may be out of date. Run the support tool in the AdminCP to restore the default theme.]]

Community Reputation

1 Neutral

About curly

  • Rank

Personal Information

  • Name
    Guy Smeyers
  • Location
  1. Hi, I've always been a fan of Newtek Lightwave. The software combines well with Houdini. Lightwave is know for his fast workflow. You can make fast things with it. It doesn't have the flexibility of a procedural system like Houdini. But it's very fast. That's why it's the perfect combination for Houdini. Now let's get to the point. There was a suggestion of sidefx that the would build a facial autorig. Rebell Hill is developing this: Can we have that also in Houdini? Perhaps you should also look more at the Lightwave way of getting things done. ( I just like their fast solutions for getting things done.)
  2. GPU rendering -> which one?

    So, because Redshift3d uses proxies and Octane doesn't you can't render large character scenes with Octane Render not even if you are using Alembic or OBJ sequences etc... ? Thanks for the explanation.
  3. GPU rendering -> which one?

    More procedural stuff in Octane Render. And Octane support out of core textures. I don't see a lot of limitations for the Vram problem anymore. Their are ways to work around it as mentioned in my post above. https://render.otoy.com/forum/viewtopic.php?f=73&t=38240
  4. GPU rendering -> which one?

    But with Octane Render you can use Layer ID's and object visibility to render only the shadows of the objects. You can render out in different layers your objects and they plan to include the Altus Innobright denoising system. On the forum they say you need a seperate dedicated gpu for the denoising with at least 6GB of VRAM. Their eventual goal is to get the Monte Carlo algorithm as noise free as possible without using a biased denoiser like Altus Innobright. And Octane supports the pixar Opensubdiv system -> see the youtube link below. That makes it possible to model low poly meshes with opensubdiv applied and you can add an extra displacement map if you want. You can even use a dirt map. I see not a lot of limitations anymore. http://www.cgchannel.com/2016/04/otoy-reveals-the-future-of-octane/ In addition, Otoy showed a work-in-progress ORBX-native implementation of Altus, innoBright’s promising standalone tool for denoising Monte Carlo raytraced renders.
  5. GPU rendering -> which one?

    I don't care about the fact that you should pay for every plugin. It seems normal to me that programmer who develop the plugin also have to make a living. Octane render can load obj sequences and alembic. That should eliminate the VRAM problem. I would love to see a side by side comparison rendering the same scenes with different engines.
  6. Hi, I just upgraded to Octane Render 3 for Houdini, I still need to buy a GTX 1070 for my GPU expansion board and I'm up and running with GPU rendering in Houdini. My question. I really like Octane Render for it's image quality and simplicity. Why aren't their so much character animations done with Octane Render and why seem the Redshift3D guys pushing their software much more then Otoy is doing? Can someone with experience in both render engines tell me the pros and cons of Redshift3D and Octane Render? My humble assessement is that Octane Render is much simpeler to setup and it gives better -> Maxwell Render style -> image quality -> while Redshift3D seems more like a VRAY on GPU type solution with a lot of tweaking required to get decent results. I haven't bought Redshift3D yet, but planning to perhaps in the future. They seem to push their marketing better. It would be nice to have some preset material of Octane in Houdini. -> Although from Rohan Dalvi's video's the shader system doesn't seem to be to difficult. Comments please about the comparison between Redshift3D and Octane. The do and don't about using high res textures in VRAM. Why is Otoy not pushing their software more towards character animation? Thanks in advance.
  7. Hi, What is besides L systems the best tree solution? I find Speedtree a bit expensive for what it is. Has anyone experience with Xfrog? From what I read you must export an obj sequence of files. Xfrog has also a decimator tool for polygon reduction. Can someone explain how to remap texture files and what file format is the best for importing? Comments, please
  8. I don't have Redshift3D, but looked at it. Perhaps someone can post more videos with rendertimes etc.... An overview video on the plugin in Houdini. There aren't a lot of videos over the Houdini plugin. Thanks for the replies. I had Octane Render for Lightwave and thought it was fast. More videos on the gpu stuff please.
  9. Hi, I've been watching forum posts and looking everything up over GPU rendering. My humble conclusions are: 1. Yes, GPU rendering is fast for simple scenes. 2. GPU's don't scale well. Your speed benefit drops when adding more then 3 GPU's. 3. You can't add up the gpu memory of all your cards, so your limited to 4, 6, 8, 12 GB memory. With inexpensive SSD's in mind, cpu rendering gives you more flexibility. 4. A GPU rendering solution isn't exactly cheaper then a cpu solution. If you count the price of the graphics cards, a special motherboard, the powe rsupply, it's as expensive as buying a decent Intel Xeon workstation. And you get more flexibility with cpu rendering and the cpu has a lower power consumption. My conclusion for the moment, I don't believe in the GPU hype, it's good for simpeler scenes, but once you take it further cpu rendering gives you more flexibility. Comments please,
  10. Hi, I've noticed the Redshift3D plugin for Houdini is out. Could someone post some video 's explaining the plugins functionality and of course with some work with rendertimes hardware specs etc posted? I'm interested in buying a license, but I want to see more work, an overview of the plugin, rendertimes and hardware specs.