Currently leapmotion.pd does not let you know if the hand is a right or left one. This abstraction tries to work around this limitation, but does not fully eradicate it. It ultimately is a quirk, because the SDK should be able to know if it is a right or left hand because of the thumb position. However this information isn not available from within leapmotion.pd
The abstraction has two modes which switch depending on how many hands are in the field of view:
1 one hand becomes present: will check if the hand appeared on the right or the left side from the zenith above the device. Assuming that it is the right hand if it appeared on the right side. You can then use the full space with that hand and it will keep its designation.
Catch: make sure to initialize correctly and always approach the device with the right hand from the right and the left hand from the left.
2 two hands present: will check which of the two hands is rightmost and defines that as the right one.
Catch: never cross hands.
Hi,
I'm planning on updating chikashi's code to version 2 of leapmotions api. This will add many more data entries..
Should be straight forward and finished in about 2 weeks. I will announce the release on the list.
Cheers, Jan
On 09/08/2014 08:17 PM, Max wrote:
Currently leapmotion.pd does not let you know if the hand is a right or left one. This abstraction tries to work around this limitation, but does not fully eradicate it. It ultimately is a quirk, because the SDK should be able to know if it is a right or left hand because of the thumb position. However this information isn not available from within leapmotion.pd
The abstraction has two modes which switch depending on how many hands are in the field of view:
1 one hand becomes present: will check if the hand appeared on the right or the left side from the zenith above the device. Assuming that it is the right hand if it appeared on the right side. You can then use the full space with that hand and it will keep its designation.
Catch: make sure to initialize correctly and always approach the device with the right hand from the right and the left hand from the left.
2 two hands present: will check which of the two hands is rightmost and defines that as the right one.
Catch: never cross hands.
Pd-list@lists.iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
That is good news, good luck. IMHO it might be better to write a middleware to make the Leap data available via OSC in Pd. For example with NodeJS https://github.com/leapmotion/leapjs That would be more flexible and add less dependencies then the flext external.
On 09/09/2014 03:56 AM, Jan Baumgart wrote:
Hi,
I'm planning on updating chikashi's code to version 2 of leapmotions api. This will add many more data entries..
Should be straight forward and finished in about 2 weeks. I will announce the release on the list.
Cheers, Jan
On 09/08/2014 08:17 PM, Max wrote:
Currently leapmotion.pd does not let you know if the hand is a right or left one. This abstraction tries to work around this limitation, but does not fully eradicate it. It ultimately is a quirk, because the SDK should be able to know if it is a right or left hand because of the thumb position. However this information isn not available from within leapmotion.pd
The abstraction has two modes which switch depending on how many hands are in the field of view:
1 one hand becomes present: will check if the hand appeared on the right or the left side from the zenith above the device. Assuming that it is the right hand if it appeared on the right side. You can then use the full space with that hand and it will keep its designation.
Catch: make sure to initialize correctly and always approach the device with the right hand from the right and the left hand from the left.
2 two hands present: will check which of the two hands is rightmost and defines that as the right one.
Catch: never cross hands.
Pd-list@lists.iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
IMHO it might be better to write a middleware to make the Leap data available via OSC in Pd. For example with NodeJS https://github.com/leapmotion/leapjs That would be more flexible and add less dependencies then the flext external.
that's true, but when testing with other apps (from leapstore or others) I
noticed that the osc traffic was too high for the cpu of my x61s (not the
newest laptop). if this bridge would allow for sorting of the data, it
could be practical - or in the end other problems could happen.
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
Are you planning to make the HD video pair available from within Pd too? then an external would do make sense.
On 09/09/2014 03:56 AM, Jan Baumgart wrote:
Hi,
I'm planning on updating chikashi's code to version 2 of leapmotions api. This will add many more data entries..
Should be straight forward and finished in about 2 weeks. I will announce the release on the list.
Cheers, Jan
That's very good news. If you need testing, let me know.
I was already starting to do my abstractions also to circumvent the 1st
sdk's shortcomings.
Best,
Joao
Hi,
I'm planning on updating chikashi's code to version 2 of leapmotions
api. This will add many more data entries..Should be straight forward and finished in about 2 weeks. I will announce the release on the list.
-
Jan Baumgart -
João Pais -
Max