Jump to content

danio1011

Members
  • Posts

    20
  • Joined

  • Last visited

Everything posted by danio1011

  1. Right, well depending on the tree model the resource usage is either high or low. But by using VRayProxies to optimize things you could cram in more detail where it counts :-)
  2. Yeah it's a great read. Definitely worth it. One thing he brings up is the method of using just a few unique VRayProxies for the last branches in the tree system (and the leaves of course.) Since this is where most of the polys are it makes things a lot more efficient. It would be cool to see a feature where you could choose within a GrowFX Path or Mesh Builder to 'repeat' only about 4 iterations of a given last branch and leaf system and then proxy them straight out of GrowFX. So then your final output from GFX would be 1 mesh of the trunk and early branches + however many unique Proxies you chose for the final branch systems. It would mean major poly savings and very lightweight trees, I think?
  3. I was just looking through my new copy of the Third and the Seventh 'making of' book by Alex Roman (arrived today.) Reading through the nature section they mentioned that one software package can generate natural transitions from trunk to bough. The book had a brief, pretty positive review of GrowFX and commented on the 'exceptional internal wind module.' No other tree design packages were mentioned. Anyway - thought it was cool to see that! *Goes back to reading the book* Dan
  4. Hi Eduard - Amazing upgrade! You guys give the best support. The displace + meta mesh is going to be great for foreground trees. It might be me but some of the example files seem to be missing...? Like the Oak, the copy examples, etc. Cheers, Dan
  5. Cool! You guys have great support. Thanks. Dan
  6. Hi Eduard - I've noticed in 1.8.6 and Max 2013 that if I save the scene and reopen it later my distributor names are in a funny font. Sometimes chinese, sometimes just squares (a font I don't have?) If I reopen the GrowFX file from scratch this goes away. Do you think this is something on my end that would be fixed by a reinstall or is it a known issue? Thanks! Dan
  7. Eduard - Thanks for the reply. I have turned these into proxies. They are still fairly heavy if I do the poly count I really want. I'm planning on moving from 16gb to 32gb soon anyway so not a big deal, but it'd be cool to be able to do really high poly conifers using a proxy as an instance or something. Thanks! EDIT: Thanks for the complement :-) Mostly thanks to GrowFX being an awesome tool to work with!
  8. BTW, Here are a couple of images showing the tree I'm working with plus some other plants I've done in GrowFX lately:
  9. I tried some VRay proxies. Didn't seem to work well...GrowFX chugged for a while and then it almost seemed like memory footprint was bigger. I will try again soon. Tree with high needle count is 2.5 mil polys, tree with 1/2 as many needles is 1.5 mil. I think I can get by with fewer needles but it would be nice to know if I'm 'missing' something with my worfklow that could make confiers more accessible. Proxied needles would bring me down to VERY low loads!
  10. Hi - I am doing some nice fir trees in GrowFX right now and am noticing that I'm getting a really 'heavy' tree as far as memory. My 'needles' are just flat, skinny planar leaves basically. I'm really happy with the results, just kinda amazed at how HEAVY the tree is (5-6gb at render time). If I go lighter on the needles I don't get the realism. I only have one size of needle. Would it work to model it from a plane outside of growfx, convert it to a VRay Proxy, and then use it as an 'instanced geometry' in GrowFX? Does the memory savings from a VRay Proxy carry through after you convert your GrowFX tree into a mesh\editable poly and proxy that? I had tried doing long leaves that are parallel to branches and using a cutout map with skinny needles in the opacity map. Maybe 20 needles to a map. This never seems to look as good tho. Thanks for any tips anyone has, Dan
  11. Wow! That works like a charm. I was only using Standard, you are right. Is 'Vector Direction' the best way to cup the leaf along its length to curl it? 'Angle Across' cups it the other way, but I'd be curious if there is a way to do that in the other axis. Thanks for the great support! Dan
  12. Hi, I've been using your software for a long time and love it! I just had a quick question about a potential feature. I do almost all of my leaves with Opacity Maps and flat plates. Sometimes, though, the flat plate isn't that realistic for foreground trees. I was just wondering if you'd ever consider adding a feature where you could do leaf plates that have a variable for resolution (number of polygons per plate) and perhaps bending or cupping amount? I used to use Onyx and that is the only thing I miss about it. I know I could probably do this with instanced geometry, but it would be so sweet to be able to have Opacity Mapped textures assigned to plates that were made up of 4 polygons rather than just 1 flat one. It's amazing how much realism a slight bend adds. Anyway, you guys do amazing work. Cheers! Dan
  13. Thanks Eduard! Great support, really appreciate it.
  14. Wow, that's great. I was hoping it wasn't just me. Thank you!
  15. Hi - I just installed the new 1.8.5 beta. It works great in Max 2012, but in 2013 when I try to open a gfx file it says 'File xxxx is incorrect!' and doesn't open. Again, I can open any files in the new beta in '12, but not 13. Is that a known issue or is it maybe something up with my install? Thanks! Dan
  16. Ok, perfect! That makes up my mind about upgrading to '13. Thanks for the quick response, Dan
  17. Hi - I use growFX in all my projects and was thinking of upgrading to Max 2013...do you have any kind of an ETA for when you might release a compatible version? Thanks! Dan
  18. Thanks for the update, I would love to see that feature implemented! Thank you for making such a great product - I will be introducing my architecture students to your software this spring. Dan
  19. Hello - I'm considering buying GrowFX shortly. It's an amazing product! However, I'm using the demo now and I'm noticing that when I convert a complicated path to meta mesh, it takes a while. I checked my CPU usage and it's only operating on one core. Are you folks considering coding in the ability for it to use multiple cores\threads at once? Sometimes for complex bough systems it takes quite a few minutes to convert and for me, for example, it would be 8x faster since I have 8 cores (2600k @ 4.5ghz). Best regards and thank you for any information, Dan
×
×
  • Create New...