I noticed that a lot of models for games with mid poly models seem to have their vertices' weight value have very specific values, for example: 0.1, 0.2, 0.3, 0.4 etc. When they have in betweens of those, it's also specific numbers such as 0.125 or 0.15 Never things such as 0.48739 which I get when regularly weight painting for example. I presume the specific values are to save up on file size space in some manner.

What I've always wondered was how this effect is achievable?

Mid poly model example

Higher poly model example that uses the less specific values I mentioned earlier

  • $\begingroup$ Havent tried, but check Vertex Weight Edit modifier Fall-off > Median Step or something like that, that rounds weights ... docs.blender.org/manual/en/latest/modeling/modifiers/modify/… $\endgroup$
    – vklidu
    Commented Sep 13, 2022 at 14:13
  • $\begingroup$ I would guess that it's painted normally and the weights are quantized by the exporter as part of the export process. For previewing, there could be script which writes the quantized weights back to the model. $\endgroup$
    – scurest
    Commented Sep 13, 2022 at 16:32
  • $\begingroup$ I located median step but it doesn't really seem to do much of what I desire, I don't see an option to customize the values it can give me. $\endgroup$
    – Cerulean
    Commented Sep 14, 2022 at 0:48

2 Answers 2


I don't think this is gonna save anything noticeable on file size and calulations, but you can select some vertices in edit mode, go to the vertex properties, vertex groups tab, select the appropriate vertex group, set a weight value and click assign.

enter image description here

  • $\begingroup$ What I'm more specifically looking for is how to make it while weight painting, it only goes between values such as 0, 0.025, 0.050, 0.075, 0.1 etc $\endgroup$
    – Cerulean
    Commented Sep 13, 2022 at 13:13

If you want, you can use a "quantize" operation in weight painting. The operator panel will allow you to specify which groups to affect, as well as a number of "steps". Weights will be clamped to 0 or to the nearest step-- so if you choose 2 steps, weights will be 0, 0.5, or 1; if you choose 5 steps, weights will be 0, 0.2, 0.4, 0.6, 0.8, or 1. Note that this can break weight normalization.

I don't recommend this though. I've never had cause to use this operation.

  • $\begingroup$ Where do I find this operation? Also if it doesn't allow for auto normalize, that winds up being an issue. $\endgroup$
    – Cerulean
    Commented Sep 14, 2022 at 0:17
  • $\begingroup$ You can find "quantize" on the searchbar, which is where I'd call it (where I call other weight paint operations that I actually use, like "smooth".) Presumably, it's also on a menu somewhere, but searchbar is easier. I don't know that quantization with normalization is really a process that makes any sense, but you can normalize manually afterwards with a "normalize all" operation (with "lock active" disabled in the operator panel.) Of course, then your weights will no longer necessarily be quantized. $\endgroup$
    – Nathan
    Commented Sep 14, 2022 at 1:49

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .