Geomitry mismatch on Gen8 in blender mirror

Marchosius

Good day all, Been playing around with custom morphs for Gen8F,

The issue I came acros was with ear morphs. What I tried was to split the mesh in half and use a mirror modifiier so what ever change I make one side is reflected on the other.

What I have noticed was when splitting the Gen8F figure in half the face count is not exactly half. So I am assuming the figure is not exactly symetrical?

Is there a way to make this workflow easier so I dont have to shape both ears by them selves?

felis

You can't use a mirror modifier in blender, because that will change vertex order, and then your morph will be a mess. In sculpt mode you can mirror along the axis.

DaremoK3

As soon as you cut it in half you destroyed vertex order.

You can not do anything that destroys vertex order  --  The obvious ones are Add/Remove topology in any Mode, but this also carries through with Joining/Separating meshes (Example; Separate eyelashes, fingernails, etc.) as this also changes vertex ordering on the coding level via BMesh.

Now, for the B28.x series, I have been doing more coding than morph target work, but I believe this will still be the same as is in the 2.7x series (Just checked -- Still works the same).

You can use sculpting with the virtual mirroring as felis above has pointed out, but if you are more comfortable creating your morphs in Edit Mode you can utilize the technique I use for asymmetrical to symmetrical (half of model to other half of model [mirrored half - virtual only -- Do Not Cut Mesh]) morph targets.

You do this via use of Shape Keys, and the SK's transfer functions.

* EDIT :  I forgot to mention  --  You can use Sub-D Modifier for your MT work, but only keep it virtual (as a modifier stack).  Do Not Apply!  Set it to what ever viewport level you want to approximate for Daz Studio Sub-D, but make sure to TURN OFF the Sub-D modifier before you export your MT or Blender will hard-set the Sub-D level in the exported vertices to your current level which WILL destroy vertex order and count.

Example:

1.  In Object Mode Set up a shape key for a morph target for your ear morph, enable the 'use' in Edit Mode toggle, and set your SK to 1.000

2.  Go to Edit Mode and create your morph.

3.  Go back to Object Mode and play with the SK slider - you should see your morph work.  If all is well, click on the drop-down arrow underneath the '+/-' (add/delete SK's) for the shape key functions and while your morph is set to 1.000

4.  Select the one to create a 'New Shape From Mix' -- This will create a duplicate SK.

5.  Set all SK's to 0.000.  Select the new SK, set it to 1.000, and go back to the function menu.

6.  This time select 'Mirror Shape Key' (there should be 2 -- Default and mirror topology -- I always use default).  There you go - a mirrored morph.

If you want to combine them to a single morph, set them both to 1.000 and repeat the 'New Shape From Mix' function.  Rename them as you see fit (if needed/wanted).

For export to use a MT back in Daz Studio, only set the SK you want as a morph to 1.000 -- All others should be set to 0.000.

Hope this helps...

Example Images:

marble

@DaremoK3 ... Thanks for your explanation here. I hope I'm not side-tracking too much but I have to ask about your references to "virtual" and Sub-D in Blender. I use Blender to make the occasional morph (usually for clothing but less so since the advent of dForce). I prefer the Sculpt tools to Edit in Blender but I'm always wary of anything that might alter the vertex count. So my question is: what is the point of sculpting in Sub-D or virtual mirroring if you have to turn it off again before sending the object back to DAZ Studio?

DaremoK3

Okay, marble, fair-enough -- good questions...

"Virtual" just means as stored in (RAM) memory for the Sub-D (subdivision) modifier.  That's the Subdivision Modifier for Sub-D modelling that contains algorithms for Catmull-Clark (B2.7x and below only, with Daz Studio equivalent - Catmull-Clark[Legacy]), Simple (no rounding - squares stay square), and Pixar OpenSubDiv (for B2.8x series -- Replacing legacy Catmull-Clark).

Not to be confused with the sculpting only Multiresolution Modifier for sculpting multi-level sculpts inside of Blender.

As far as preferring the Sculpt Tools to Edit Mode vertex morphing (including Proportional Editing), the good news is you can use the non-destructive Shape Keys MT work in Sculpt Mode as well.

I use both in tandem.  As long as you have a SK set up per morphing task, you can have an advanced array of limitless choices/changes/reworks, and compounding non-destructive morphing to get what you envision.  I create in the 100's range per average session  --  Think along the lines of the Daz3D Store work of PA Zev0 and his/her enormous morph target products catalog.  I have been creating at that level for decades  --  I am a content creator, but not a seller/PA/what-have-you in any store  --  Some of us create for our art, and some have to purchase from those who create for the masses who choose not to create for whatever reason.  But, I digest...

To answer your question regarding the point of utilizing Sub-D when you just have to turn it off for the morph target upon export;  If you want your morph to be better represented in DS at whatever DS Sub-D level you are shooting for, you should be creating your MT's at the same level for fidelity.  There will never be a true 1-1 conversion, but you will be close enough for it to work better.

If you only create at Base level in Blender -- It's fine if you only want to utilize Base level in DS, but you might encounter not seen anomalies if you use Sub-D on your figures/props in DS at different levels.  It might not happen with simple MT's, but some advanced ones can be quite different from there base counterparts where it is smooth transitions at level 3 or 4, but have crossed topology (overlapping edges/vertices) at base level  --  at base the morph might look horrendous, but look great at the level you were shooting for.

Virtual mirroring  --  I was only referring to the mirroring algorithms in Sculpt Mode, which are great.  They are permanent modifications on the mesh for all enabled mirroring unless you utilize the Shape Key non-destructive morphing workflow I outlined above.  Same as with Edit Mode, without it, all changes are permanent - with it, you can make any changes, including all the way back to original.

* EDIT:  marble, thank you, for your questions  --  You gave me something to think about to add to my TODO list for the DS <--> Blender Bridge Addon for morph target work.  I should add auto-setup of Sub-D Modifier at user defined level toggle button to my other user enable-able import settings.  But, I will hard-code auto-OFF for the Sub-D for the Export button to force 1-1 vertex count.  Thanks again...