Hello from another rank newbie.So, take the following for what that's worth...

I have been able to bring in 3 Skype callers using NDI on the same machine (not the screen/window capture method
vmix recommends, but directly from NDI output of Skype).

The 3 callers were in group Skype call. You have to use the Content Creator version of Skype to get the NDI out.
Once that is working, get the group Skype call with ten people connected.
If you create an NDI input, you should see each caller in its own window to choose for that input. There will also be 2 extra; one for 'local' which is you and one for 'active speaker' which is what it says. Usually you should see each account name in the input window top border.
You have to create nine more inputs to have an input for each caller. (Need sufficient version of vMix)
If you want to combine callers in one screen, make a base input and then do multiview overlays choosing from the multiview drop-down menus, and resize/reposition them with the POSITION menu.

This much is working for me, but then comes the challenge of routing the sound correctly. There is a tutorial about connecting Zoom from same computer which I believe should work with Skype, too. I'm struggling through this part now.

If anyone can explain why doing as above is a bad idea compared to the window/screen capture method, I would appreciate hearing. As long as the PC is high enough spec, it seems to work.

Please share any successes.

Yes i need to send different sound (4 languages) for each people.
So skype non good if i cant send a special BUS for each

So the best solution is VMIXCALL
but i have 10 people (visio)
only 8 call on vmixcall
Can i use another machine with another license and have 2 more vmixcall and take it via NDI in my Principal machine?

Thanks a Lot
David

You might also want to look at Zoom's interpreter function...

@mjgraves - thank you for that input. I've only checked out Z on this and for our use, it has a huge limitation in that once the interpretation feature is engaged, no one can hear just the original meeting - i.e. people who understand all the languages are forced to listen to interpreter for half the time (in the case of two languages).

I'll surely check out ZipDX.
If you would like to point out anything else, would be glad to hear.
though I guess we are getting of topic...

ZipDX is an audio conference service. No video. We have a patented system for simultaneous interpretation over-the-phone (or computer.) It requires professional conference interpreters.

We do not support integration with third party platforms like Zoom, Webex, GotoMeeting, vMix, Wirecast. All that means is that if you want to do that, you do it on your own.

You can use "Floor" in Zoom but you need to define it.

Find a language nobody uses and define it as floor. all ettendees who want to listen to the oroginal signal need to select your defined floor signal.

Is video something you need? If not switch to ZipDX if you need video it gets realy ugly combining ZipDX with Zoom (It will work but the track active speaker will not work as expected. only the person where you plug in the audio from ZipDX will show up automaticaly.)

You can use the raise your hand feature to identify who speaks. And then force Spotlight the person for all.