Why is it recommended to make all UV seams hard?

3D4Eva

I was having some uv problems and the person helping me told me to make sure I split my uvs up if there is a hard edge which I understand. I am confused about being told that all uv seams should be hard because if I have a human face or maybe a small animal and I cut up my uvs to fit in my map I dont want to make those uv seams hard otherwise I will get hard seams.

Is this supposed to be the way to do things? I never had problems keeping my seams soft for those types of models. I only split if I have a hard edge and make that edge a seam as well to break it apart.

poopipe

Ah, dogma....

Splitting UVs and adding hard edges allows you to deal with changes in surface direction that are too great for a normal map to cope with. 
Eg. A 90 degree outside corner wants a hard edge , UV split and padding because the change in surface direction is 270 degrees and a normal map has a total range of 180 degrees.

Anything under 60 degrees and its theoretically safe to do whatever - although compression will knacker that in practice.

The hard edge and split 'rule' largely comes from hard surface work where sharp corners abound 

Use your eyes and judgement 

3D4Eva

Thanks. I understand why when dealing with hardsurface, but I couldn't understand the reason for say a round ball looking creature which doesn't have those degree shifts and by making every seam hard you'll see hard seams when baking when you don't want it.

FrankPolygon

In general: hard edges require corresponding UV splits for padding but UV splits do not require hard edges.

Hard edge placement is about controlling low poly smoothing behavior and baked normal gradation. UV seam placement is about optimizing usable texture space while also limiting texture distortion. Bringing the two together it's about balancing visual quality with resource efficiency. Using fewer hard edges and UV splits tends to be more resource efficient but it's also important to make sure the final product looks good.

There is an individual resource cost to both UV splits and hard edges but once the mesh is split by either it's split so stacking them doesn't incur additional overhead beyond the first split. This is one of the reasons why it's generally considered best practice to strategically pair and place hard edges and UV seams along natural breaks in the shapes.

Hard surface modeling tends to naturally produce shapes that have a lot of matching UV splits and hard edges but this is less common on organic shapes. The idea that every UV seam on every model needs to have hard edges is unconventional. There may be some edge cases where this is done for artistic or technical reasons but it wouldn't be standard practice.

Bellow is some additional documentation that explains these concepts in detail.

Marmoset Toolbag baking tutorial:
https://marmoset.co/posts/toolbag-baking-tutorial/
https://marmoset.co/posts/toolbag-baking-tutorial/?page=Map Types#maptypes
https://marmoset.co/posts/toolbag-baking-tutorial/?page=Toolbag Setup#basics
https://marmoset.co/posts/toolbag-baking-tutorial/?page=Best Results#bestresults

Controlling shading behavior by Alec Moody:

https://www.youtube.com/watch?v=ciXTyOOnBZQ

3D4Eva

Thank you

FourtyNights

3D4Eva said:

Thanks. I understand why when dealing with hardsurface, but I couldn't understand the reason for say a round ball looking creature which doesn't have those degree shifts and by making every seam hard you'll see hard seams when baking when you don't want it.

No, this is the typical misconception. You definitely don't need to put hard edges on all seams, but you need seams on hard edges.

Repeating this again with a good golden rule to memorize this "All of your hard edges are always UV seams, but not all of your UV seams need to be hard edges."

3D4Eva

FourtyNights said:

3D4Eva said:

Thanks. I understand why when dealing with hardsurface, but I couldn't understand the reason for say a round ball looking creature which doesn't have those degree shifts and by making every seam hard you'll see hard seams when baking when you don't want it.

No, this is the typical misconception. You definitely don't need to put hard edges on all seams, but you need seams on hard edges.

Repeating this again with a good golden rule to memorize this "All of your hard edges are always UV seams, but not all of your UV seams need to be hard edges."

thank you for explaining this.

I'm also wondering if I should even bother doing my bakes by splitting because I could just bake using averaged normals and just forget about the need to worry about uv splitting and having smoothing groups? I've had people tell me that I can just do this because of synced workflow due to mikkt being available in so many renderers. Another person said that non averaged baking is an old method no longer used. I was surprised because I see people teaching this. He told me to just make my low fully smoothed and baked with averaged normals in Painter.

FourtyNights

3D4Eva said:

thank you for explaining this.

I'm also wondering if I should even bother doing my bakes by splitting because I could just bake using averaged normals and just forget about the need to worry about uv splitting and having smoothing groups? I've had people tell me that I can just do this because of synced workflow due to mikkt being available in so many renderers. Another person said that non averaged baking is an old method no longer used. I was surprised because I see people teaching this. He told me to just make my low fully smoothed and baked with averaged normals in Painter.

I usually avoid sharp angles and hard edges by beveling/chamfering those edges, and going with everything smooth. It means more polygons, but less drastic gradients on the low poly shading for the normal map to compensate when baked. To further improve shading on beveled low poly models, I use face weighted vertex normals.

3D4Eva

I see thank you. 

I hope someone can also comment on the different workflows for baking. I'm still confused on why some people say that using the splitting and smoothing group one is bad, and using averaged normals is now better since mikkt.

gnoop

3D4Eva said:

I see thank you. 

I hope someone can also comment on the different workflows for baking. I'm still confused on why some people say that using the splitting and smoothing group one is bad, and using averaged normals is now better since mikkt.

Using  averaged normals  still provides more seamless  look  but generally not very good for hard surface objects having flat  surfaces . Especially shiny ones because of gradients.

in certain conditions you can use face weighted normals  without edge splitting but not pure averaged.   Face weighting transfers shading  gradients to  less visually important areas.

But in any case  you should keep in mind how second lod would look like.  I would even say   do the baking always for the second lod and  the splitted edges + face weighted normals  is usually a best way  considering lod 02

3D4Eva

Do gradients not matter because of having mikkt ? I thought it wasn't a big problem as long as the baker/renderer was synced?

I didn't think about lods thank you!

gnoop

in our  in house game engine mikkt made no difference, cured nothing .     My few tests in Unreal4 last year showed it wasn't a magic fix there too.  I still saw gradients at dawn/dusk lighting.  After all normal maps are  heavily compressed  two  8 bit channels usually.    I doubt they would be ever nice with its gradients

FrankPolygon

There's a lot of good advice in the replies here. Something to consider is what's the source and broader context behind the idea that using hard edges with UV splits is bad and using a single smoothing group with averaged normals is good?

Simple explanations like this can be tempting but blanket statements made without proper context tend to just be dogma. Sometimes these ideas can be based on a little bit of truth but are repeated so often they become distorted. Context maters because what makes sense for one project or pipeline may not make sense for another. It doesn't have to be a binary decision to only ever use one method or the other.

There's some examples of when and how to use these different strategies towards the end of the video on controlling smoothing behavior and here's another discussion that may help provide additional context for when these different strategies make sense. https://polycount.com/discussion/177780/difference-between-substance-painters-average-normals-baking-and-a-traditional-cage-workflow

pior

Whenever someone gives you some oddly simplistic, dogmatic, bruteforce piece of advice ... one good thing to do is to have a look at their own work, up close. If it's clean and tidy they probably know what they're talking about and are genuinely trying to help you. But more than likely you'll spot shading artefacts, wonky surfaces and glitchy detailling - meaning they're just talking out of their ass and have no idea why things are done this or that way.

It's the "oh don't worry about it, it'll work out fine" kind of people. You're more than welcome to slap them in the face

3D4Eva

Thank you everyone for your answers.

My confusion really came because one person was saying to only used averaged normals when baking hard surface, and another had told me about UV splitting and smoothing groups. I did notice on hard surface it didn't look proper with some details and they looked skewed and off when using averaged normals.

It really  threw me off when I was told no other work flow should be used but averaged normals because of mikkt and there is no reason to split hard edges and use smoothing groups. Such information really throws me off. 

I'll read up on that link again, I did see it before. I noticed about blending at the bottom of the thread which is interesting.