Please, Critique My 50s Radio Model

anthonybp

Hi everyone,

I'm currently modeling a radio for a school project and I simply wanna know if my topology is good.

If you have any comments on anything else like the design, do not hesitate.

Thank you !

CybranM

What do you mean by "wrong with my topology"?
Do you want feedback on the design or is this the lowpoly model and you want feedback on the polycount of the model?

Its a lot easier to answer a specific question than a very vague or unclear one.

anthonybp

CybranM said:

What do you mean by "wrong with my topology"?
Do you want feedback on the design or is this the lowpoly model and you want feedback on the polycount of the model?

Its a lot easier to answer a specific question than a very vague or unclear one.

Oh, sorry.

I just want to know if my topology is good, because I'm not sure. But, if you have feedback about the design as well, I open to it

Brian "Panda" Choi

Your topology is fine.  It's a static object; it just needs to be efficient and look rounded where it needs to.

You could throw more polys at this if you want this to stand on its own.  No need to be niggardly with more tris.

Rossua

From your second and third picture, it looks as if some edge loops stop randomly in the model. Which makes me think you have ngons in your mesh, which is not good. correct me if im wrong, but apart from that it looks fine

anthonybp

Brian "Panda" Choi said:

Your topology is fine.  It's a static object; it just needs to be efficient and look rounded where it needs to.

You could throw more polys at this if you want this to stand on its own.  No need to be niggardly with more tris.

Thank you for your feedback. I thought that my quads should be the same size as much as possible. I though it was a problem since I got huge quads and tiny ones.

And, what do you mean by "stand on its own" ?

Brian "Panda" Choi

Stand on its own as in you're going to take close up shots of this.

anthonybp

Rossua said:

From your second and third picture, it looks as if some edge loops stop randomly in the model. Which makes me think you have ngons in your mesh, which is not good. correct me if im wrong, but apart from that it looks fine

It's actually two separated mesh

anthonybp

Brian "Panda" Choi said:

Stand on its own as in you're going to take close up shots of this.

Oh okay. It's an important asset, so I'll do a high poly version after the low poly is done. Thank you !

Rossua

anthonybp said:

Rossua said:

From your second and third picture, it looks as if some edge loops stop randomly in the model. Which makes me think you have ngons in your mesh, which is not good. correct me if im wrong, but apart from that it looks fine

It's actually two separated mesh

That's good then, i forgot to ask if there were multiple meshes.
Solid start, looking forward to seeing this develop

anthonybp

I did quite a lot today. I want to know if the green topology is okay. It doesn't affect the render or anything, I just want to know if it's good practice ?

Do not hesitate to ask for more details.



Thanks.

Carabiner

Hey there,

It's okay ... ish? One big thing this is NOT good practice for is avoiding quad overdraw. Basically having lots of converging long skinny triangles will cause your game engine to have to work harder (I don't remember the exact details here, our tech artist explained it a while ago). If you're wanting to avoid that in an actual game, you'd probably want to modify the topology to avoid the long skinny converging triangles (terminate the edges earlier). But if you're just working for your portfolio, what you've done is usually the most efficient method, so it's not terrible if you just are aware of the issues it can cause.

ETA: This all applies only when it's being seen by the player/camera, so if it's on a back face and won't be seen much, it would be okay.

Also, if there isn't a change in angle, you probably don't need that holding loop around the very edge.

Alex_J

About overdraw :

from thread:

https://polycount.com/discussion/comment/2674671#Comment_2674671

TL:DR

A person still learning to pull verts could probably stick with the basic rule of, "don't have long thin triangles like that."

anthonybp

Carabiner said:

Hey there,

It's okay ... ish? One big thing this is NOT good practice for is avoiding quad overdraw. Basically having lots of converging long skinny triangles will cause your game engine to have to work harder (I don't remember the exact details here, our tech artist explained it a while ago). If you're wanting to avoid that in an actual game, you'd probably want to modify the topology to avoid the long skinny converging triangles (terminate the edges earlier). But if you're just working for your portfolio, what you've done is usually the most efficient method, so it's not terrible if you just are aware of the issues it can cause.

ETA: This all applies only when it's being seen by the player/camera, so if it's on a back face and won't be seen much, it would be okay.

Also, if there isn't a change in angle, you probably don't need that holding loop around the very edge.

Oh, I didn't know that. Thank you for your feedback

P.S. : It's for an animated short, not a video game !

Carabiner

anthonybp said:

P.S. : It's for an animated short, not a video game !

Oh, that makes a huge different with regards to how people are going to give feedback on your model, actually! In that case, real-time considerations like overdraw aren't as important as far as I know (might increase render time? Really no idea though). 

Ashervisalis

Hey, as I said in the other thread, if this is for animation, it should be sub-d modeling! Use this topology;



For something that will subdivide into this;



It'll look hella clean and proper

Remember to use a bevel with a division of 2 in order to get edges that hold properly while sub-dividing. You do not need to be as weary of poly count when using it for animation.

anthonybp

Ashervisalis

Oh my god, thanks again my friend. That look WAY better