That's weird. I use the same API. The initial version won't have a tutorial on the API, if I delay this any longer people are going to start posting hate mail ;-|
Maybe we can write a tutorial together or blog post or something...
Simple to use, effective, and live preview ... couldnt get any better.
I just submitted to the Asset Store Fingers crossed it makes it through first time.
I'd like to thank all the beta testers for your help! And I'd also like to thank everyone for their patience. I didn't think it would take so long to get this thing to this stage.
with your plugin, is it possible to bake high poly details as bump map like blender does, and use it with low poly?
Unity 4 Pro (iOS Pro, Android Basic, Flash Basic), Modo 701 (ACS,NPR,SLIK)
iMac i7, MacBook Pro i7 HD 6750M, (Mavericks, Windows 8.1)
iPhone 4, iPad 2/3, (iOS 7), Nokia Lumia 925 ( Phone 8 )
The plugin is now live in the asset store!
I have started a new thread now that the preview is over
If an admin wants to close this thread, please do so.
@atmuc Sorry, not possible. This plugin just reduces geometry.
Can your reduction process merge triangles based on the angle between them and their size?
I ask because it always amazes me that there isn't a SINGLE decimator available that can automatically take a cube made of for example 50 or 2000 polygons, and reduce its size down to just 12 polygons.-What that also implies is that for all other models also, it doesn't take into account the importance of the polygons properly!
What a really good polygon reduction programme might do is the following-
1-take away all polygons that are on the same plane as a polygon next to them.
2-slowly reduce polygons according to their adjoining angle, and preserve polygons with the highest angles for example teeth ends.
3- measure the areas wit the model with the most angle changes per volume, and areas with angle changes that add together to make protruding structures and try and preserve them as much as possible
4-don't apply 3 large polygons to represent the entire back, flank, or side of model, when the model has 2000 polygons! The program should definitely check the length of a potential new polygon and only making a new polygon if the length is something like 1/10 of the total model length.
As far as I know, no other programs including yours have done the optimisations number 1 and 4. Please tell me why your algorithm makes 3 polygons for the flanks and back of a 2000 polygon model? and his arms are flat!
and would your program be able to decimate planar surfaces properly? for example a procedurally generated tower block with lots of coplanar polygons?
Last edited by ZoomDomain; 11-20-2012 at 02:57 AM.
Basically Cruncher does exactly what you said. At each step of decimation it determines which vert could be removed that would have the least impact on the mesh in terms of shape while maintaining boundaries such as material boundaries so that textures have something to pin the edges to. So edge angles, etc. are taken into consideration.
For my internal testing I in fact created a cube with more than 65K verts. It didn't reduce to 12 polygons, but it came pretty close...
@bbubbinator That's all right, cool, seeing as you have a brilliant decimation algorithm that works automatically within unity, do you think it would be very useful for users to have a fully automated LOD management tool, that can automatically get all the prefabs in a project, make multiple copies of each of various levels of detail using your algorithms, and automatically manage the loading, the assignment of the different models into different distance steppings. It would avoid Indy game creators having to individually load and adjusts hundreds of prefabs into various versions, and have the advantage of them appearing in glorious full detail when you approach close to a particular mesh without having to spend long periods on managing the LOD, and also having fast rendering, all by automatically managing the prefabs within unity.
I think it would be great because you could simply load your automated algorithms into the project, set the level of detail to a number of levels and a certain gradient, press enter, and it would manage the entire project with a better detail than they could possibly hope to do manually.
Last edited by ZoomDomain; 12-02-2012 at 02:59 PM.
I have explored this idea a bit via a "dashboard" feature. I might add this in a future release. For now my focus is on making sure it works perfectly for all kinds of meshes, and then probably next I'll add support for user hinting/weighting to influence the reduction process so you can specify areas to preserve more detail and areas where you don't need it.
Thanks for the feedback.