Retro Tv (Topology Help please )

Narukessa

Hello so this model I was working on I was trying to get to be all one piece but I have some questions about the topology. One, I've been told that it should be pretty uniform throughout a model which is what I think I achieved, but two, I was also told that it should use as few faces as possible....I don't know if the topology from these images works....I think I would fix the corners so that they come in at an angle if I kept this version but also its a lot of faces.....



or if I should try to reduce it  as I'm doing below....my concern with the reduction I am working on is that I also remember my one teacher saying that you should avoid creating stars ( places where more then two lines meet on a vertical ) and also this is not really uniform at all....so I think this is resulting in an actually much worse mesh......( I know the one I am reducing is not done but I decided to stop myself while I'm ahead because I also think that doesn't seem right at all. ) 

I also attempted quad drawing over my original and trying to make a  map onto the lower poly version but I didn't like the edges that resulted from the quad draw or the map.....

.....I'm wondering what would be the appropriate way to solve this. I am most interested in making props for games ( because I know sometimes things depend on what it is for ). 

jStins

What's the purpose of the model? Is it going to be subdivided and smoothed for a high poly and baked down? Or is it going straight into a game engine? If it's the latter you should try to accurately capture the shape with as few polys as possible. Static game meshes can have ugly topology as long as it's still reasonable to work with and doesn't cause any shading errors. Your low poly version looks OK, but try to avoid long skinny tris as those can cause frame rate issues. Poles (stars as you call them) are fine as long as it's not an insane number of edges meeting at the vert.

Also, why model this as one mesh? It's generally easier to mimic the real-world construction and use separate meshes for knobs, switches, etc...

Narukessa

I modeled it as one mesh because when I built it originally, I was just using it as a practice build and timing how fast I could build it because I was trying to do it in under an hour. Also I am a little weird where I get obsessive about the math of things.....so I like figure how many lines things will be  for each edge and then just extruding them in order from the part that is the most in to the part that's the most out....like some sort of weird rubrics cube...or solving sudoku....( that part is hard to explain without showing  it, but modeling it as one thing made the puzzle a little harder....). ....But I was looking through my models because I wanted to practice painting in substance today...and then...I thought...well I have to fix the uv's first...then....I started thinking well maybe the topology is no good if there is so many faces....so I started reducing it....and then I thought wait is what I am doing even right....so I figured I'd ask, I also picked through these...and decided I couldn't use those either...




I think I've been doing 3d for a few months and now I'm getting anxiety that everything I'm doing is bad/wrong lol..but I haven't been doing things long enough to actually know what is wrong with them when I look at them. 

sacboi

I think I've been doing 3d for a few months and now I'm getting anxiety that everything I'm doing is bad/wrong lol..but I haven't been doing things long enough to actually know what is wrong with them when I look at them.

If you're interested in creating real time assets, either mod or make a game.

Conversely pick a subject you like, gather resource material too eventually finalize end to end, definitely a learning curve but nonetheless in my opinion worthwhile investing the time and effort.  

Narukessa

sacboi said:

Mod or make a game.

I would like to make models that are good for games. 

Ashervisalis

Narukessa said:

 I'm getting anxiety that everything I'm doing is bad/wrong

Welcome to art hahahahahah. First thing I'll mention is to not rush things. Don't try to do the TV in under an hour, try to do it right. If you're trying to make a model of a TV, it should probably go through a sub-d phase first, whether its for animation or for a game. Here's what great sub-d topology looks like;

https://www.artstation.com/artwork/ynde5

I'll also mention to not make your assets out of one continuous mesh. That'll just overly complicate everything.

sacboi

Narukessa said:

sacboi said:

Mod or make a game.

I would like to make models that are good for games. 

Ok, well pick a subject you like probably something small and achievable, gather resource material too eventually finalize end to end.

Definitely will see positive results, just my past experience when first learning this stuff which can be a hard and grindy process, but fun too.

rexo12

my 2 cents would be not to stress about the technical stuff too much at this stage. It's a really easy thing to fixate on when you're doing something for the first time because it has quantifiable metrics (a "good" model is such and such polygons, with such and such texture resolution), whereas artistic quality is a lot less easy to define. However, prematurely trying to fuss about optimisation (root of all evil and all that) is going to significantly hamper the quality of your work, and it's no good having a really well optimised piece that doesn't look good.

Not to say that it isn't important, because it is important to be doing things right and not taking shortcuts from the onset so you build good habits, but don't let it control your work, particularly at this early stage.

Applying this to your situation - if your learning goal was just to get a first look at Substance Painter but you've instead gotten bogged down in trying to make a really tight and low-poly mesh and you haven't been able to get into Substance at all yet, then maybe it's better to take a step back and just use your initial mesh for painting? It looks like it will be easier to do a good unwrap with, and you will see marginal (if not zero) difference in Substance performance between the two. It will only become a problem when or if you try to incorporate the model into a larger scene in a game engine.

Veraliot

Hello i'am having the same problem i'am pretty good at making topology for humans and simpliers stuff, but when it come to an Retro tv or something with a mix between rectancular corners boxy shape and rounded one i just don't know how to deal with the topology especially when everything is stuck on one object, i know this topic is old but did you found a way to fix it ?

Can someone tell me how can i make a good topology on a tv ?

sarasumm

@Veraliot

This is a model I made a while ago that I think might help give you some ideas with how to handle topology on a similar object. I try to keep my models as simple and boxy as possible for most of the modeling process. Then, when I bevel edges where necessary to round the edges. Also, don't be afraid to use separate meshes within the same model. For things like the volume knobs and keyboard keys, I used unconnected meshes.

Topology can be overwhelming at first - I spent my first year modeling obsessing over making my models 'perfect'. While that can be good practice, it's best to spend time thinking about "What can I get away with?" rather than "How can I make this perfect?". Many AAA game models have horrible topology, but still look great because the artist knew how to work within their pipeline.