Announcement
Collapse
No announcement yet.
[Plugin] Leap Motion - Event Driven
Collapse
X
-
Olivierus repliedThe current version of this plugin is pretty slow, if there is another version where people are working on, where could we find this?
-
getnamo repliedOriginally posted by Nicolas4 View PostHello everyone,
I am very new at this, i am an archviz artist and trying to step into VR and Unreal, i am using the Oculus rift along with the leap motion and trying to add the Leapfloatinghands to the HMDlocomotion pawn from the unreal VRtemplate, and i am not being able to see my hands, actually i tried but i don't know how to do it, could someone help me with a step by step guide to set it up please.
Thank you very much [MENTION=548]getnamo[/MENTION]
Leave a comment:
-
Nicolas4 repliedHello everyone,
I am very new at this, i am an archviz artist and trying to step into VR and Unreal, i am using the Oculus rift along with the leap motion and trying to add the Leapfloatinghands to the HMDlocomotion pawn from the unreal VRtemplate, and i am not being able to see my hands, actually i tried but i don't know how to do it, could someone help me with a step by step guide to set it up please.
Thank you very much [MENTION=548]getnamo[/MENTION]
Leave a comment:
-
Nicolas4 repliedHello everyone,
I am very new at this, i am an archviz artist and trying to step into VR and Unreal, i am using the Oculus rift along with the leap motion and trying to add the Leapfloatinghands to the HMDlocomotion pawn from the unreal VRtemplate, and i am not being able to see my hands, actually i tried but i don't know how to do it, could someone help me with a step by step guide to set it up please.
Thank you very much
Leave a comment:
-
Syed repliedOriginally posted by ParagonVRAdmin View PostI think you can lower that Data of the Leap hands if you eliminate the knuckles and Just track the Fingertips locations and palm location. That would allow for less bandwidth to be broadcast. right? Or, am I missing something?
Leave a comment:
-
ParagonVRAdmin repliedOriginally posted by getnamo View PostFinally what is the best way to have this working from a developer point of view? Not all leap experiences will be multiplayer, so there should be a toggle, then you need a way to determine ownership of tracked hands, perhaps check if a rigged actor's parent is a pawn or player controller and determine ownership that way? This isn't fully settled yet so if you have good ideas about how to expose the functionality in an intuitive way, please contribute thoughts, ideas, maybe even code
Leave a comment:
-
ParagonVRAdmin repliedI think you can lower that Data of the Leap hands if you eliminate the knuckles and Just track the Fingertips locations and palm location. That would allow for less bandwidth to be broadcast. right? Or, am I missing something?
Leave a comment:
-
getnamo repliedOriginally posted by MariuszWaclawek View PostHi [MENTION=548]getnamo[/MENTION], first of thanks for your work on this plug-in it's a great start.
We are working on a project that requires Leaps to be replicated properly over network, out of the box we get our own hands mimicked to all players and players can't see tracked hands of others just their own copies everywhere.
From what I understand the new rewrite of the plug-in will have improved latency which we are excited about. Does the rewrite also integrate well with Unreal replication? Will it be supported out of the box or do we have to replicate all transforms by hand like [MENTION=14253]maffew[/MENTION] did, it does not sound like a good solution though?
Do you have any blog on the progress of the rewrite or even a short list of upcoming improvements, features?
PS. If you are willing to share the github repo I would really appreciate the access.
Thanks,
MariuszOriginally posted by Syed View PostAFAIK, there is no replication done in getnamo github. And I am now also trying to get replication done here.
Edit: No replication in the existing github. I have no idea about rewrite version.
For VR this is especially true, since 2 players at 90fps of just 3 transforms (e.g. HMD, left and right) will saturate the 10kb/s default bandwidth, so we will likely need to raise the cap e.g. https://answers.unrealengine.com/que...-of-setti.html. Leap motion fingers have ~ 22 per hand or nearly 15x the data of head + motion controllers.
The other consideration is how to have a good experience with a ranging number of people streaming their hand positions at these higher data rates. I'm thinking some kind of interpolation will be needed with the more aggressive kind being used when the other person is e.g farther away from you, while more accurate streaming happens when they're near (or maybe leave that as an option)
Finally what is the best way to have this working from a developer point of view? Not all leap experiences will be multiplayer, so there should be a toggle, then you need a way to determine ownership of tracked hands, perhaps check if a rigged actor's parent is a pawn or player controller and determine ownership that way? This isn't fully settled yet so if you have good ideas about how to expose the functionality in an intuitive way, please contribute thoughts, ideas, maybe even code
Originally posted by Juangea View PostHi getnamo
With the latest beta I can´t manage to enable hands collisions (I´m using the low poly hands) do I have to do something special about it?
Cheers.
Leave a comment:
-
Syed repliedOriginally posted by MariuszWaclawek View PostHi [MENTION=548]getnamo[/MENTION], first of thanks for your work on this plug-in it's a great start.
We are working on a project that requires Leaps to be replicated properly over network, out of the box we get our own hands mimicked to all players and players can't see tracked hands of others just their own copies everywhere.
From what I understand the new rewrite of the plug-in will have improved latency which we are excited about. Does the rewrite also integrate well with Unreal replication? Will it be supported out of the box or do we have to replicate all transforms by hand like [MENTION=14253]maffew[/MENTION] did, it does not sound like a good solution though?
Edit: No replication in the existing github. I have no idea about rewrite version.Last edited by Syed; 03-27-2017, 03:14 AM.
Leave a comment:
-
Juangea repliedHi getnamo
With the latest beta I can´t manage to enable hands collisions (I´m using the low poly hands) do I have to do something special about it?
Cheers.
Leave a comment:
-
MariuszWaclawek repliedHi [MENTION=548]getnamo[/MENTION], first of thanks for your work on this plug-in it's a great start.
We are working on a project that requires Leaps to be replicated properly over network, out of the box we get our own hands mimicked to all players and players can't see tracked hands of others just their own copies everywhere.
From what I understand the new rewrite of the plug-in will have improved latency which we are excited about. Does the rewrite also integrate well with Unreal replication? Will it be supported out of the box or do we have to replicate all transforms by hand like [MENTION=14253]maffew[/MENTION] did, it does not sound like a good solution though?
Do you have any blog on the progress of the rewrite or even a short list of upcoming improvements, features?
PS. If you are willing to share the github repo I would really appreciate the access.
Thanks,
Mariusz
Leave a comment:
-
getnamo repliedOriginally posted by maffew View PostHi Getnamo,
Is there any update on the rewrite? I've managed to get the hands to replicate over the network but it involves sending each transform every frame and means it looks very slow and laggy!
Thanks,
Matt
Originally posted by Juangea View PostSo I have an update.
I implemented it and it does exactly the same as the included plugin in 4.15, so no improvement, eagerly waiting to the new plugin, please hurry up!
Cheers and thanks!
Leave a comment:
-
maffew repliedHi Getnamo,
Is there any update on the rewrite? I've managed to get the hands to replicate over the network but it involves sending each transform every frame and means it looks very slow and laggy!
Thanks,
Matt
Leave a comment:
Leave a comment: