U3M 1.1 Open Discussion

[vizoomartin]

We would like to collect features and changes that should go into the next version of the Unified 3D Material, with the release planned end of Q3 19, beginning of Q4 19.

Physics files in the common section Right now vendors can place their physical measurements already in their respective Custom section. As the Custom section is per definition undefined, this makes it difficult or impossible for 3rd parties to read or write physics file into the u3m format. E.g., in our Vizoo specific case, to add existing physical measurements of a vendor to a u3m scan. Hence it could make sense to have a list in the common section, consisting of the path to the physics file as well as the name of the vendor.

"material": {
  "physics": [
    { "name" : "vendor_1", "path" : "relative path to physics file1" },   
    { "name" : "vendor_2", "path" : "relative path to physics file2" },   
    ...
],
....}

Preview image as part of the common section Preview images right now are part of each vendors custom section. That makes it difficult for 3rd parties to utilize them, e.g. for a database preview. So one suggestion is to have one preview image referenced in the common section, that is generated in the same way and recreated/overwritten by every application eiditing a u3m file. For this, a formula and some rules need to be defined on how the image is created from the texture maps.

Additional meta data Let's take the information provided in https://github.com/vizoogmbh/u3m/issues/5#issuecomment-431591927 as a basis and discuss what is missing or might cause problems. We should schedule a (online) workshop in the coming months with all parties interested in saving their material information in u3m.

[RenateOnVizoo3d]

Further topics for Version 1.1:

• dpi in x and y direction
• add supported formats to the schema ( e.g. image formats like jpg, supported compression)
• add supported bit depths ( most probably 8 bit, 16 bit)
• add color profiles in images
• spec U3MA

[ReinhardOnVizoo3d]

do we need dpi at all? It can be calculated from image size and width/height.

[malkosh]

Is there a need to split it up on x and y? Just curious as I’ve never seen it. Are there any examples?

[ReinhardOnVizoo3d]

have also never seen it in real life :D but it's technically possible maybe it's not very relevant to u3m and we could also just make a rule that dpi must be consistent for x and y axis. if we want dpi at all.

[malkosh]

The only issue I have seen where dpi matters is if someone is looking for it. I’ve seen many people manipulate scans post scan and ruin the true size of the images losing scale and defaulting dpi to 72 but that’s a different issue which I’m not sure the dpi here will help. But that’s just my personal opinion.

[vizoomartin]

Outdated, please see newsletter from this week