Forum Archive › COLLAB2021 – HQ – Stylized Fantasy Environment
No worries if you commit last minute! We’d love to have you join of course 🙂
Lol I hope I used those words correctly 😅
@yukinoh1989 Yes, I’d like to explain the asset list during the first stream (April 6th) at the very latest. I want to post the list earlier so people can get started earlier but we’ll see if I can finish this current course soon enough to have that time.
Are we bound to also 1 piece or if there is spare time can we even go for a second or third piece?
You’re not limited to 1 piece. I’d love to see contributors challenge themselves with multiple pieces, as several did last year 👍
Glad you signed up! Also glad you’re open to work on anything. I think we’ll need open and willing people to fill in the gaps 👍
We also used Google Sheets to confirm our submissions, I’m assuming [Kent] we will have a similar process for this COLLAB?
Yes absolutely. I’m getting it started now with an official roster of contributors as registrations come in. I will share it out with everyone once it’s finalized and not so WIP.
I really appreciate your clarification @jakeblended and completely agree as well. As Orson says, limitations (and I’m adding “structure”) within an artistic team project is crucial to its success.
On that note, one of favorite aspects of Collabs is the potential to evolve and take on a life of its own. Courses and Classes need to be more strict with less deviation in order for specific things to be learned. However with Collabs this isn’t the case and I really like that. In my opinion this increases potential for the spark of CG / Blender passion to fan into flame through the course of a Collab. I’ve already seen this evidenced by the enthusiasm for the TSFHI initiative!
When I chose this art for the project I knew it would be cool and impressive. But I didn’t know what would take it to the next level. TSFHI is pushing it to the next level already. I wish I wasn’t recording a course so I could participate! 😅
It looks like I’m a bit slow to catch up with everyone… I’ll get to work taking the recommended courses right away.
I wonder when SpikeyPotter and Adrian Finnigan will show up and teach me to stupify?
@blanchsb look in the TSFHI post ( https://youtu.be/LZ0EO7gSdAM ), I think you’ll be stupified 😱
I like your new profile picture William, it looks like those chocolate candies people use to clearly visualize how UV Unwrapping works.

@theluthier Kent I have a question, maybe more than one…
Should we absolutely avoid Subdivision Surface Modifiers when modeling the interior assets? It is a low poly project right? And so can we use that technique were you enable out smooth under normals in the object data properties? and then go through all the stuff of selecting edges that should remain sharp. But then it’s a Blender only work flow? Is that ok?
@theluthier , awesome i really look forward to the collab 😀
Great questions – questions I haven’t answered myself yet. My thoughts:
Should we absolutely avoid Subdivision Surface Modifiers when modeling the interior assets? It is a low poly project right?
The main reason I lean toward low-poly (non subsurf modeling) is so we could possibly have an optimal, complete environment running in real time in an Eevee viewport. I just think that’s a cool final product to shoot for. One that could probably be uploaded to sketchfab as well.
If we allow subsurf it’s very unlikely it will be navigable in realtime.
NOTE I’m not saying subsurf can’t be used for the high-res versions of assets for the purpose of baking to the low poly. On the contrary I expect hi-poly sculpting to lo-res baking to be a common technique throughout this project. Still within that the MASTER models would be low-poly.
can we use that technique were you enable out smooth under normals in the object data properties? and then go through all the stuff of selecting edges that should remain sharp. But then it’s a Blender only work flow? Is that ok?
I know normal data is exportable to other apps. We’ve done it many time from Blender to Unity. So I’m pretty sure hardening normals will translate to any app. BUT I don’t care too much about the project’s smooth transition to other apps. Not enough to influence our project’s protocols.
What do you think?
I think that I think exactly what you think you’re thinking. So basically the rules of the game are:
– We shoot for real time so no heavy dense meshes
– Do your high res stuff but bake it to low poly ones
– Thumbs up on the harden normals- auto smooth-mark sharp- super trick
– Not like the DOG when we were aiming at universal modeling language, this can be geared towards Blender.
– (Extra space if another rules comes up)
@theluthier Speaking of low poly… For me low poly is at most 2000 vertices per person model. About 300 vertices per table for example. But that’s my criteria. Do you have any number to give on an example?
I don’t know if there’s a specific number, but I would say SubD is a big no no, triangles are good friends and I remember JL courses, like the low poly tank of the sci-fi door, where he was always sliding verts that don’t alter the shape in order to get the poly count even further. So I’d say as low as you can get given the model you’re trying to create.