Is this acceptable topology for a model pack?

So I am new to subsurf modeling, and I have spent the two weeks working hard on making these models (all are quads only) and looking back, I’m not sure if my topology is acceptable for a model pack that is meant to be subdivided. I read somewhere that quads should be square and your model should subdivide evenly, which mine doesn’t. However, since most of my surfaces are flat, reflections and materials work perfectly even with this messy topology. All 22 of my models have a similar topology. Is this a problem? Will potential users be turned off by seeing this wireframe?



Let’s clarify a couple of things, when you’re using subsurf modifier it will take a face and will subdivide into small faces,but he’s not that smart, so he’s gonna subdivide every faces by the same number. However, even faces with less the space is supposed to be divided into the same number resulting in small geometry that could cause underisable shading artifacts. That’s why is recommended to have evenly spaced quads. Quads are desirable because de subsurf modifier can produce bad shading artifacts with ngons or tris* and for a convention when exporting to other softwares or gaming engines.But keep in mind, at the end the engine will convert every face to tris anyway.

For your question, a “acceptable” topology is hard to say, there’s no rule, but isn’t a rocket science either. First ask yourself where your models are suposed to be, for game, it won’t bottleneck that much when alone, but in a realtime with lot’s of artifacts in scene, it could be overkill(specially when the other models is having high number of topology) and demand more hardware resource that it really neads. In a single render, it’ll maybe give you a little delay at the end of the render, but depending on your rig, won’t be that bad.

In both cases i would recommend you to make a low-poly version with baked normals. I believe you already know about them. But Just in case :

NORMALS Normal maps is a technique used for faking the lighting of bumps. Are commonly used to represent a HighPoly mesh into a low poly, or add details visible beacuse it says to render where the light has bumps or not.

Here’s an example, it’s a single plane with small quads, but with the normal map it looks like way more gorgeous.


I’d use a proper socket model only if I need a closeup of it. Otherwise I use a single flat plane with some textures on it (normal, id, AO etc).
Obviously it will fall flat when viewed at high angles, but since it’s not the main focus of my renders I’m not too bothered.
So a socket is never something that I will require a high poly model for, or a model at all, but a simple plane will suffice.


how many fricking levels of subsurf has that green pic got?! surely no need on Earth for that amount, for a wall socket!

I’m not trying to be harsh, but no…I do not think anyone would pay for that…you need to capture the shape with as little geometry as possible…while avoiding Ngons…and then use mapping as shown above to bring out the smaller details. You should look at some wireframes for reference maybe google them…go to poly count or pinterest.

I’m turned off by the asymmetry, personally. There’s no reason the left side shouldn’t match the right, other than the polarized plug, but that shouldn’t effect topology.

Some plugs have one small hole and one large hole so that’s not really the issue…

To improve the model I would have had the sockets separate from the wall cover. There isn’t a need to have them attached to each other since that isn’t how they’re manufactured. Since your model is a flat surface you can try and incorporate more diamonds outside of the edge loops to minimize the amount of polygons. It’s a bit difficult to tell what’s happening in your first screenshot because of the large dots on the vertices but I’d imagine you could have made it lower poly.

Some people just use a normal map for this kind of stuff if it’s really in the background and to never really be rendered front and centre.

Yeah, that’s the polarized plug I mentioned. The asymmetry in the edgeloops doesn’t instill confidence in the quality of the modeling.

You have to remember that the market for pre-made assets is mostly professionals. If some pro needs an outlet plate for their scene, they can spend 15minutes making it, or they could purchase an asset. If you are selling an outlet for less than a quarter of their hourly rate, it might be worth it to them to purchase the asset, rather than model it. Since professionals are your market, you have to have a professional grade product.

The proportions on this outlet are decent, but the edgeflow is noticeably unrefined. A professional will see that and recognize it as an indicator of your experience, and you won’t make many (probably wont make any) sales.

Look at this comparable product: https://www.turbosquid.com/3d-models/max-light-switch-wall-plug/723799

The edge flows look intentional and systematic. It looks like it was made by someone who knows what they are doing.

I mean there isn’t a huge difference between what you linked and what OP posted. They just need to clean theirs up a little -e.g. just clean the socket areas a little. The major issue with his model, isn’t the small asymmetry, since it’s very minor and easy to fix it’s just the over use of polygons.

For most use cases, unless it is a close up shot what you are going to end up using is a 512k normal map image on a lightly modeled image plane for something like this.


The topology I used for the model to generate this is somewhat comparable to the one in turbosquid and unless it is an ultra close up shot many game devs would be sticking to the plug plate to put as needed to flesh out their level.