The Triumph expression looks crazy after importing with the new morph system.

[东方夕惕]

I’ve read the newest blog and tried the new morph system. I import the Triumph expression, but the result looks weird.

‌

[Alessandro Padovani]

I don’t get triumph but the other expressions work fine here. Have you tried the latest commit 93c07e4 ?

[东方夕惕]

I’ve tried the latest commit, but the issue still exist. I don’t know why the blog’s example looks right, but my test looks so weird.

[Alessandro Padovani]

Is it triumph only ? I mean, do all the other expressions work fine for you ? For example afraid and angry that come with the base pack work fine for me.

[东方夕惕]

Yes, I only find triumph don’t work so far. Afraid and Angry work fine for me too.

[Thomas Larsson]

Should work now. The problem manifested itself differently in your file than in mine, but I had it too, it was just too small to be visible. Don’t understand the reason for this difference, but anyway it seems to be gone now.

[Alessandro Padovani]

Thomas, it seems there’s another issue. I can’t import the surprise expression in the base pack. That is, there’s the div2 driver but I don’t get the expression slider.

steps

1. import G8F
2. import expressions

‌

edit. oops .. too fast, going to check the latest commit

edit. As for commit ddcdf08 I can confirm that surprise doesn’t work.

[Thomas Larsson]

Surprise/Surprised is the only expression where the ejcm comes before ectrl in lexical order. The slider is there if you turn on Load Hidden Morphs. Apparently the ejcm file hides the ectrl slider otherwise.

[东方夕惕]

My file’s scene unit scale is 0.001 and the unit scale in global settings is 10. Maybe those settings cause the difference, I guess. But anyway the Triumph expression works fine now, thank you!

[Alessandro Padovani]

Thomas, so do we have to always load hidden morphs to get the sliders ? If so the option should be checked by default. Or do you plan a fix for this ? And what is hidden morphs intended for anyway ? I see it’s in the debug section so I guess it’s not intended for the user.

Below it is what I get with hidden morphs.

‌

[Alessandro Padovani]

东方夕惕, is there any reason why you keep a global unit of 10 meters ?

[Thomas Larsson]

Alessandro, you should definitely not need to load hidden morphs once all bugs have been ironed out. That’s why it is listed as a debug option, something that the user would normally not need to change. However, if you are loading hidden morphs, they are enclosed in brackets to show that they are hidden.

The name of that option has to change. Hidden morphs are always loaded, the option controls whether you create sliders for them.

Anyway, I think the problem is gone in the latest commit.

[Alessandro Padovani]

Thomas, thank you for the explanation. Commit caf6d0d works fine here with all the face expressions, including the div2 HD base morphs. To me we can mark as resolved.

[东方夕惕]

@Alessandro Padovani Because my scene unit scale is 0.001, which means 1 unit is 1 mm. I have to set add-on’s global unit scale to 10 to import Daz character as real world size. The reason why I set the blender scene unit scale to 0.001 is that I need to simulate small rigid body objects, which is more stable than default unit scale.

‌

[Alessandro Padovani]

@东方夕惕 Thank you for your reply. Didn’t know that the unit scale affects the simulation precision. Not that I play with simulation a lot, but when I did the default unit scale was just fine. There were a couple bugs some time ago but they are fixed afaik.

Do you have any reference to the issue with unit scale ? I’d be interested myself.

https://developer.blender.org/T61313

https://developer.blender.org/T62631

[东方夕惕]

  <div class="preview-container wiki-content"><!-- loaded via ajax --></div>
  <div class="mask"></div>
</div>

</div> </form>

[东方夕惕]

@Alessandro Padovani

There are two blend files you can play with. One’s scene unit scale is 1, another is 0.001. I’ve set up rigid body properties. On my computer, 0.001 works fine, 1 even can not simulate, once you click bake, blender will not responding.

‌

[东方夕惕]

As far as I know, the chain size in real word is too small for rigid body simulation when scene unit scale is 1. But if we set the scene unit scale to 0.001, the chain becomes large relatively.Then the rigid body simulation can calculate properly.

https://docs.blender.org/manual/en/2.92/physics/rigid_body/tips.html

‌

[Alessandro Padovani]

@东方夕惕 I see, this is a limit specific to rigid bodies that honestly I never used. Thank you for the unit scale trick, I’ll keep it in my toolbox. Though I fear you’re actually taking advantage of a flaw may be.