I started this thread because I found a new modeling tool which I found refreshingly simple and intuitive. I wanted to get some feedback from people thinking they would actually try out the software.
Instead I got unsolicited advice from someone who, as far as I know, hasn’t even tried using Sculptris. Granted I am no modeler, and both you and Tiles probably have greater knowledge than I in this area. However it’s plainly obvious (even to me) that many of the criticisms aren’t actually valid with regard to Sculptris. They seem like blanket criticisms against sculpting based modeling for anything used in game development.
This is highlighted by the fact Tiles surrendered on all arguments except the one regarding edge loops. This one I have been researching myself and finding to be a somewhat valid criticism. I am not trying to win this argument, merely wade through all the bias and uncover some actual information. I don’t find the antagonistic tone of your comments to be at all helpful. I do not want to put anyone off-side and if I have offended either you or Tiles, let me offer my sincerest apologies.
The model I was exporting had 5500 tris, that is well within the budget of most games.
Your phrasing suggests that you yourself are not a modeler. The implication you are making is that manual vertex manipulation is faster/easier than sculpting. My personal belief is that user experience which mirrors real life is generally more fluid. I wont speak on behalf of people other than myself, but with my limited modeling experience I found that within 10 minutes of using Sculptris I was significantly more productive than I was after a few weeks of using Blender.
Even though I mentioned it several times, I will say it again: The vertex subdivision algo, that makes the sculpting so easy and effective, do increase the poly count significantly. That is why Sculptris comes with powerful poly reduction tools. In a couple of minutes I reduced the poly count of my model to 5.5k keeping all the detail I required so this is a moot point as far as I am concerned.
I wont respond directly to the rest of your post because it seems much of it is based on misconceptions about Sculptris. I think this discussion will progress faster if I just write up a summary of my findings so far, and we can go from there.
Findings so far
The biggest benefit of Sculptris is in the simplicity and ease of use. You can create a low poly model then automatically generate UVs and then paint textures and normals directly onto the model. All this can be done without leaving the Sculptris tool. You can then export the model and textures directly into Unity ready to go.
There are some major downsides worth being mindful of, and these are as follows:
Messy UVs
The automatically generated UVs look like a mess of triangles. This makes it incredibly difficult to modify the textures in any image manipulation software. Hand crafted UVs would also give the user great control allowing for greater detail on the face for example. This is not possible with the auto generated UVs in sculptris.
Mesh Topology
There will be no edge loops and this can be detrimental to the quality of animations, particularly facial animations.
There are some people who are quite dogmatic about the importance of clean topology, but I would like to see a reasoned argument or demonstration of a model without clean topology animating poorly.
I can understand that in the days of very low poly counts that anything but a clean topology would be completely out of the question. I will also concede that to get an equivalently good animation with poor topology might require a greater poly count; how much greater I don’t know. I don’t know if anyone has actually tested this to find out. I am currently working on animating the 5.5k model I have made as a proof of concept to see what, if any, issues arise.
Being a programmer, this seems to me quite similar to debates had a long time ago about assembly vs. C. People said you can never write a game in C it’s far too slow, everything must be hand coded in pure assembly. This was a long held view and some people stuck to their guns longer than others. Over time hardware improved and development speed was valued over performance and the industry moved to writing code in C and reserving use of assembly code for performance critical code. Nowadays most companies write most of their engine code in C++ with game logic in scripting languages. Unity is leading the way for rapid-development and in doing so willingly sacrifices some performance.
Taking that into consideration, It’s not overly surprising that this has met such staunch resistance. Game development is an ever-changing world, with new technology, new techniques and new tools. The most important thing is to question everything and accept new ideas. I’m not convinced that the dogma surrounding clean topology is still justified.
Since Sculptris offers faster development and a shallower learning curve I think it fits quite well with the philosophy of Unity3D. Even if this comes at the expense of clean topology that needn’t entirely rule it out of game development. I haven’t heard any reasons why it couldn’t be used for static meshes. Even for animated meshes it could at the very least have a place in rapid prototyping.
-Pete