New release
20 views
Skip to first unread message
Tony Gonçalves
Jan 20, 2011, 10:26:13 AM1/20/11
to osce...@googlegroups.com
Hey guys,
I pushed a new version that i hope solves the strange crashes that were happening.
As a bonus I added saving and playing back from ONI files, but i didn't have the time to test it much so your mileage may vary.
Have fun :)
I pushed a new version that i hope solves the strange crashes that were happening.
As a bonus I added saving and playing back from ONI files, but i didn't have the time to test it much so your mileage may vary.
Have fun :)
Ramo
Jan 20, 2011, 2:57:52 PM1/20/11
to OSCeleton
Thanks alot for implementing the playback feature. Perhaps a silly
question, but how do I configure OSCeleton so it will playback the
*.ONI file?
Cheers,
question, but how do I configure OSCeleton so it will playback the
*.ONI file?
Cheers,
Tony Gonçalves
Jan 20, 2011, 6:22:50 PM1/20/11
to osce...@googlegroups.com
You have to run it with the "-i" command line option. Like:
osceleton -i filename.oni
If you don't know how to do that, just tell me what your operating system is (Windows, Linux, OSX?), and i'll try to help out :)
--
Tony Gonçalves
Partner
IPN Incubadora
Rua Pedro Nunes
3030-199 Coimbra, Portugal
Tel : +351 927 474 665
Tel : +351 239 700 358
Fax : +351 239 700 301
Ramo
Jan 21, 2011, 5:31:49 AM1/21/11
to OSCeleton
That's all I needed to know,Thanks!
Cheers
On 21 jan, 00:22, Tony Gonçalves <t...@sensebloom.com> wrote:
> You have to run it with the "-i" command line option. Like:
>
> osceleton -i filename.oni
>
> If you don't know how to do that, just tell me what your operating system is
> (Windows, Linux, OSX?), and i'll try to help out :)
>
>
>
> On Thu, Jan 20, 2011 at 7:57 PM, Ramo <ohertg...@gmail.com> wrote:
> > Thanks alot for implementing the playback feature. Perhaps a silly
> > question, but how do I configure OSCeleton so it will playback the
> > *.ONI file?
> > Cheers,
>
> > On 20 jan, 16:26, Tony Gonçalves <t...@sensebloom.com> wrote:
> > > Hey guys,
>
> > > I pushed a new version that i hope solves the strange crashes that were
> > > happening.
> > > As a bonus I added saving and playing back from ONI files, but i didn't
> > have
> > > the time to test it much so your mileage may vary.
>
> > > Have fun :)
>
> --
>
> Tony Gonçalves
> Partner
>
> t...@sensebloom.com
>
> www.sensebloom.com
Cheers
On 21 jan, 00:22, Tony Gonçalves <t...@sensebloom.com> wrote:
> You have to run it with the "-i" command line option. Like:
>
> osceleton -i filename.oni
>
> If you don't know how to do that, just tell me what your operating system is
> (Windows, Linux, OSX?), and i'll try to help out :)
>
>
>
> On Thu, Jan 20, 2011 at 7:57 PM, Ramo <ohertg...@gmail.com> wrote:
> > Thanks alot for implementing the playback feature. Perhaps a silly
> > question, but how do I configure OSCeleton so it will playback the
> > *.ONI file?
> > Cheers,
>
> > On 20 jan, 16:26, Tony Gonçalves <t...@sensebloom.com> wrote:
> > > Hey guys,
>
> > > I pushed a new version that i hope solves the strange crashes that were
> > > happening.
> > > As a bonus I added saving and playing back from ONI files, but i didn't
> > have
> > > the time to test it much so your mileage may vary.
>
> > > Have fun :)
>
> --
>
> Tony Gonçalves
> Partner
>
>
> IPN Incubadora
>
> Rua Pedro Nunes
>
> 3030-199 Coimbra, Portugal
>
> Tel : +351 927 474 665
>
> Tel : +351 239 700 358
>
> Fax : +351 239 700 301
>
> i...@sensebloom.com
> IPN Incubadora
>
> Rua Pedro Nunes
>
> 3030-199 Coimbra, Portugal
>
> Tel : +351 927 474 665
>
> Tel : +351 239 700 358
>
> Fax : +351 239 700 301
>
>
> www.sensebloom.com
Eidur Arnason
Jan 21, 2011, 11:46:01 AM1/21/11
to osce...@googlegroups.com
Is it possible to invert the XYZ parameters ?
Eidur
Eidur
Message has been deleted
Tony Gonçalves
Jan 21, 2011, 12:35:20 PM1/21/11
to osce...@googlegroups.com
Why would you need to do that?
Eidur Arnason
Jan 21, 2011, 1:01:25 PM1/21/11
to osce...@googlegroups.com
I need to invert most if not all the signals in Max -- and I was just wondering if there was a simpler way to do it, rather than building all the objects in Max Msp
2011/1/21 Tony Gonçalves <to...@sensebloom.com>
paganmuzak
Jan 21, 2011, 1:22:11 PM1/21/11
to OSCeleton
thanks a lot for your work,
i was using the old version of osceleton.exe with processing exemples
on win 7 32
bit and all work fine but when i try last version processing print
only the message "got it" but dont work, am i wrong !?
paganmuzak
i was using the old version of osceleton.exe with processing exemples
on win 7 32
bit and all work fine but when i try last version processing print
only the message "got it" but dont work, am i wrong !?
paganmuzak
Tony Gonçalves
Jan 21, 2011, 4:39:02 PM1/21/11
to osce...@googlegroups.com
Ooops, seems I included the wrong executable with the last commit...
I pushed a version with the correct executable just now, does it solve your problem?
I pushed a version with the correct executable just now, does it solve your problem?
paganmuzak
Jan 22, 2011, 3:33:26 AM1/22/11
to OSCeleton
Now processing receive OSC message but it work extremly slow (near to
be blocked), i will try to make some experiment to understand why...
be blocked), i will try to make some experiment to understand why...
Sebastian Kox
Jan 25, 2011, 8:12:17 AM1/25/11
to OSCeleton
Hi Tony,
first of al, thanks so much for your time and efford on this project!
It's great that you made this available, and I'm having a lot of fun
with it!
Right now I'm trying to figure out why the 1st osx version is still
working better for me than the latest build.
Apart from the segmentation fault it seems that the joint data is
transferred differently and a bit more unstable in the latest version.
I've tried to document it in this video:
http://vimeo.com/19166878
For this setup I used one OSX machine, with the two versions of
OSCeleton (11/1 and 20/1 build) without any parameters set, a
processing app that re-routes the OSC data to a QC readable format and
one QC applet that visualises the data.
First in the video is the 20/1 version. This version segfaults more
easily (althoug less than any build after the 11/1 version) and the
leg/feet data seems more unstable. The mapping of the coordinates seem
to be different as well.
The 11/1 version remains the most stable (no segfaults) and the joint
data from the legs are more smoothly/perfect.
I've installed the v0.4 avin2-SensorKinect driver
Nite 1.3.0.18
OpenNI v1.0.0.25
Hope it helps, if you need more info/testing please contact me!
first of al, thanks so much for your time and efford on this project!
It's great that you made this available, and I'm having a lot of fun
with it!
Right now I'm trying to figure out why the 1st osx version is still
working better for me than the latest build.
Apart from the segmentation fault it seems that the joint data is
transferred differently and a bit more unstable in the latest version.
I've tried to document it in this video:
http://vimeo.com/19166878
For this setup I used one OSX machine, with the two versions of
OSCeleton (11/1 and 20/1 build) without any parameters set, a
processing app that re-routes the OSC data to a QC readable format and
one QC applet that visualises the data.
First in the video is the 20/1 version. This version segfaults more
easily (althoug less than any build after the 11/1 version) and the
leg/feet data seems more unstable. The mapping of the coordinates seem
to be different as well.
The 11/1 version remains the most stable (no segfaults) and the joint
data from the legs are more smoothly/perfect.
I've installed the v0.4 avin2-SensorKinect driver
Nite 1.3.0.18
OpenNI v1.0.0.25
Hope it helps, if you need more info/testing please contact me!
Tony Gonçalves
Jan 25, 2011, 8:36:57 AM1/25/11
to osce...@googlegroups.com
Wow, thanks for your effort!
I can't test the OSX version much since I don't have a mac, so this will be very valuable in trying to find the cause of the issue.
From what i see in the video I'm fairly certain i messed up the normalization of the coordinates (or maybe it's just OSX's gcc compiler playing games with me again)...
As for the jerky movements, it could be due to some optimizations I've done (reusing some OpenNI objects instead of recreating them every frame). I'll have to play with it a bit and see what happens.
Thanks for identifying the versions too, this will make it easier to see what has changed ;)
Have fun!
Tony
I can't test the OSX version much since I don't have a mac, so this will be very valuable in trying to find the cause of the issue.
From what i see in the video I'm fairly certain i messed up the normalization of the coordinates (or maybe it's just OSX's gcc compiler playing games with me again)...
As for the jerky movements, it could be due to some optimizations I've done (reusing some OpenNI objects instead of recreating them every frame). I'll have to play with it a bit and see what happens.
Thanks for identifying the versions too, this will make it easier to see what has changed ;)
Have fun!
Tony
George Toledo
Jan 25, 2011, 9:11:31 AM1/25/11
to osce...@googlegroups.com
Out of curiosity, did those optimizations apply to the original OS X version (not the normalization, but the reusing objects)? I have a make that was from Linux, converted to work on OS X that seems to track more stably and recognize user quicker (thanks@Zach). I had thought it was placebo.
--
--
George Toledo
The information contained in this E-mail and any attachments may be confidential.
If you have received this E-mail in error, please notify us immediately by telephone or return E-mail.
You should not use or disclose the contents of this E-mail or any of the attachments for any purpose or to any persons.
Tony Gonçalves
Jan 25, 2011, 3:01:45 PM1/25/11
to osce...@googlegroups.com
No, I was talking about very recent changes. I'm going to take look at the issues now, so I should have news pretty soon :)--
Tony Gonçalves
Partner
IPN Incubadora
Rua Pedro Nunes
3030-199 Coimbra, Portugal
Tel : +351 927 474 665
Tel : +351 239 700 358
Fax : +351 239 700 301
Tony Gonçalves
Jan 26, 2011, 7:42:37 AM1/26/11
to osce...@googlegroups.com
Hey guys,
It seems I had indeed messed up the normalization of coordinates, which caused distortion of the skeleton, and lots of jerky movement. I also found I was using some un-initialized memory (ooops!) which would cause some crashes.
Both of these issues are hopefully corrected now - the jerkyness should be gone, and at least some of the crashes too.
Somehow it seems to me that calibrating is again faster too (but this could be an illusion).
If someone could download the latest version and take it for a spin I'd appreciate it ;)
Have fun,
Tony
It seems I had indeed messed up the normalization of coordinates, which caused distortion of the skeleton, and lots of jerky movement. I also found I was using some un-initialized memory (ooops!) which would cause some crashes.
Both of these issues are hopefully corrected now - the jerkyness should be gone, and at least some of the crashes too.
Somehow it seems to me that calibrating is again faster too (but this could be an illusion).
If someone could download the latest version and take it for a spin I'd appreciate it ;)
Have fun,
Tony
2011/1/25 Tony Gonçalves <to...@sensebloom.com>
DanielG
Jan 26, 2011, 10:56:39 AM1/26/11
to OSCeleton
Hi Tony,
I just took the new version for a test drive, and yes, it does seem to
calibrate faster than the last version. Thank you!
I have another question about osceleten though, and Tohm at the
cycling74 forums told me you might be able to help. I have a question
about more rough tracking. In my testing the skeleton sometimes can
take a while to calibrate the person. For the project I am currently
working on all i need is the spatial position of each user, which the
sample programs in the openNI folder seem to register almost
immediately. Is there any way to get general user info (i.e. ID & XYZ
pos.) out via OSC so I can mess with the data in Max/MSP as soon as
the kinect recognizes them? I am using MAX 5 and OSX 10.6.6. here is
an example of the info i need: http://gogginvideo.com/daniel_goggin/blank.html
Thanks again for all your hard work, the program is the bees-kness!
DG
On Jan 26, 6:42 am, Tony Gonçalves <t...@sensebloom.com> wrote:
> Hey guys,
>
> It seems I had indeed messed up the normalization of coordinates, which
> caused distortion of the skeleton, and lots of jerky movement. I also found
> I was using some un-initialized memory (ooops!) which would cause some
> crashes.
> Both of these issues are hopefully corrected now - the jerkyness should be
> gone, and at least some of the crashes too.
> Somehow it seems to me that calibrating is again faster too (but this could
> be an illusion).
>
> If someone could download the latest version and take it for a spin I'd
> appreciate it ;)
>
> Have fun,
> Tony
>
> 2011/1/25 Tony Gonçalves <t...@sensebloom.com>
> >> gtole...@gmail.com
> t...@sensebloom.com
>
> www.sensebloom.com
I just took the new version for a test drive, and yes, it does seem to
calibrate faster than the last version. Thank you!
I have another question about osceleten though, and Tohm at the
cycling74 forums told me you might be able to help. I have a question
about more rough tracking. In my testing the skeleton sometimes can
take a while to calibrate the person. For the project I am currently
working on all i need is the spatial position of each user, which the
sample programs in the openNI folder seem to register almost
immediately. Is there any way to get general user info (i.e. ID & XYZ
pos.) out via OSC so I can mess with the data in Max/MSP as soon as
the kinect recognizes them? I am using MAX 5 and OSX 10.6.6. here is
an example of the info i need: http://gogginvideo.com/daniel_goggin/blank.html
Thanks again for all your hard work, the program is the bees-kness!
DG
On Jan 26, 6:42 am, Tony Gonçalves <t...@sensebloom.com> wrote:
> Hey guys,
>
> It seems I had indeed messed up the normalization of coordinates, which
> caused distortion of the skeleton, and lots of jerky movement. I also found
> I was using some un-initialized memory (ooops!) which would cause some
> crashes.
> Both of these issues are hopefully corrected now - the jerkyness should be
> gone, and at least some of the crashes too.
> Somehow it seems to me that calibrating is again faster too (but this could
> be an illusion).
>
> If someone could download the latest version and take it for a spin I'd
> appreciate it ;)
>
> Have fun,
> Tony
>
>
>
>
>
>
>
>
>
>
> > No, I was talking about very recent changes. I'm going to take look at the
> > issues now, so I should have news pretty soon :)
>
>
>
>
>
>
>
>
>
> > No, I was talking about very recent changes. I'm going to take look at the
> > issues now, so I should have news pretty soon :)
>
> > On Tue, Jan 25, 2011 at 2:11 PM, George Toledo <gtole...@gmail.com> wrote:
>
> >> Out of curiosity, did those optimizations apply to the original OS X
> >> version (not the normalization, but the reusing objects)? I have a make that
> >> was from Linux, converted to work on OS X that seems to track more stably
> >> and recognize user quicker (thanks@Zach). I had thought it was placebo.
>
> >> -gt
>
> >> 2011/1/25 Tony Gonçalves <t...@sensebloom.com>
>
> >> Out of curiosity, did those optimizations apply to the original OS X
> >> version (not the normalization, but the reusing objects)? I have a make that
> >> was from Linux, converted to work on OS X that seems to track more stably
> >> and recognize user quicker (thanks@Zach). I had thought it was placebo.
>
> >> -gt
>
>
> >> Wow, thanks for your effort!
> >>> I can't test the OSX version much since I don't have a mac, so this will
> >>> be very valuable in trying to find the cause of the issue.
>
> >>> From what i see in the video I'm fairly certain i messed up the
> >>> normalization of the coordinates (or maybe it's just OSX's gcc compiler
> >>> playing games with me again)...
> >>> As for the jerky movements, it could be due to some optimizations I've
> >>> done (reusing some OpenNI objects instead of recreating them every frame).
> >>> I'll have to play with it a bit and see what happens.
> >>> Thanks for identifying the versions too, this will make it easier to see
> >>> what has changed ;)
>
> >>> Have fun!
> >>> Tony
>
> >> Wow, thanks for your effort!
> >>> I can't test the OSX version much since I don't have a mac, so this will
> >>> be very valuable in trying to find the cause of the issue.
>
> >>> From what i see in the video I'm fairly certain i messed up the
> >>> normalization of the coordinates (or maybe it's just OSX's gcc compiler
> >>> playing games with me again)...
> >>> As for the jerky movements, it could be due to some optimizations I've
> >>> done (reusing some OpenNI objects instead of recreating them every frame).
> >>> I'll have to play with it a bit and see what happens.
> >>> Thanks for identifying the versions too, this will make it easier to see
> >>> what has changed ;)
>
> >>> Have fun!
> >>> Tony
>
> >>www.georgetoledo.com
>
> >> The information contained in this E-mail and any attachments may be
> >> confidential.
> >> If you have received this E-mail in error, please notify us immediately by
> >> telephone or return E-mail.
> >> You should not use or disclose the contents of this E-mail or any of the
> >> attachments for any purpose or to any persons.
>
> > --
>
> > Tony Gonçalves
> > Partner
>
> > t...@sensebloom.com
>
> >> The information contained in this E-mail and any attachments may be
> >> confidential.
> >> If you have received this E-mail in error, please notify us immediately by
> >> telephone or return E-mail.
> >> You should not use or disclose the contents of this E-mail or any of the
> >> attachments for any purpose or to any persons.
>
> > --
>
> > Tony Gonçalves
> > Partner
>
>
> > IPN Incubadora
>
> > Rua Pedro Nunes
>
> > 3030-199 Coimbra, Portugal
>
> > Tel : +351 927 474 665
>
> > Tel : +351 239 700 358
>
> > Fax : +351 239 700 301
>
> > i...@sensebloom.com
> > IPN Incubadora
>
> > Rua Pedro Nunes
>
> > 3030-199 Coimbra, Portugal
>
> > Tel : +351 927 474 665
>
> > Tel : +351 239 700 358
>
> > Fax : +351 239 700 301
>
> t...@sensebloom.com
>
> IPN Incubadora
>
> Rua Pedro Nunes
>
> 3030-199 Coimbra, Portugal
>
> Tel : +351 927 474 665
>
> Tel : +351 239 700 358
>
> Fax : +351 239 700 301
>
> i...@sensebloom.com
> IPN Incubadora
>
> Rua Pedro Nunes
>
> 3030-199 Coimbra, Portugal
>
> Tel : +351 927 474 665
>
> Tel : +351 239 700 358
>
> Fax : +351 239 700 301
>
>
> www.sensebloom.com
Ferhat Şen
Jan 26, 2011, 11:04:35 AM1/26/11
to osce...@googlegroups.com
Hi Tony,
I did a test couple of minutes ago. Latest release on OSX is way better than the previous in terms of the skeleton jerky movements and elongations :) But it seems to require a little more fine tuning as far as I see from Windows example videos. I have took videos to compare the Prev and Latest release in Processing. I'll send the links.
Thanks for the effort again :)
Ferhat
Ferhat Şen
Jan 26, 2011, 12:00:28 PM1/26/11
to osce...@googlegroups.com
Dear OSCeleton community,
Here are the links for the tests of new release on MacOSX
new release with processing http://www.vimeo.com/19216246
everything is same for both of the tests except the OSCeleton verison.
my feet sometimes goes out of the frame. i think that's why they are acting weird
best,
Tony
Tony Gonçalves
Jan 26, 2011, 2:03:01 PM1/26/11
to osce...@googlegroups.com
OSCeleton doesn't send that info at the moment, but it wouldn't hurt to add it. I'll put it on the wishlist :)--
Tony Gonçalves
Partner
IPN Incubadora
Rua Pedro Nunes
3030-199 Coimbra, Portugal
Tel : +351 927 474 665
Tel : +351 239 700 358
Fax : +351 239 700 301
Tony Gonçalves
Jan 26, 2011, 2:06:47 PM1/26/11
to osce...@googlegroups.com
Thanks for the videos :)
It seems to me that the performance is pretty close to the windows version. Any differences between the 2 at the moment are probably due to the OpenNI OSX port still having some issues.
The next version will have an option to apply some smoothing to the joint coordinates, to eliminate any remaining jerky movement on the skeleton.
Tony
It seems to me that the performance is pretty close to the windows version. Any differences between the 2 at the moment are probably due to the OpenNI OSX port still having some issues.
The next version will have an option to apply some smoothing to the joint coordinates, to eliminate any remaining jerky movement on the skeleton.
Tony
Sebastian Kox
Jan 30, 2011, 3:57:00 PM1/30/11
to OSCeleton
Hi Tony,
Just got the chance to play with the updated OSX version, and
unfortunately I still get the segfault. It seems to be less often
though. I can't really pin down the cause but it happens in most cases
when I sit down and obviously most body segment's are
tangled&untrackable.
The jerky movements in the segments seem less. The right foot/knee/hip
really lags when in QC mode (it updates once every few seconds). In
normal mode it's ok.
The coordinates still seem to be different than the 11/01 version in
standard and QC mode.
I've also tried to track 2 people at the same time. In the processing
example it worked (although slowly) in qc I only get the data of the
last calibrated user.
My setup didn't change, and the 11/01 version still works perfectly.
Hope it helps!
Sebastian.
Just got the chance to play with the updated OSX version, and
unfortunately I still get the segfault. It seems to be less often
though. I can't really pin down the cause but it happens in most cases
when I sit down and obviously most body segment's are
tangled&untrackable.
The jerky movements in the segments seem less. The right foot/knee/hip
really lags when in QC mode (it updates once every few seconds). In
normal mode it's ok.
The coordinates still seem to be different than the 11/01 version in
standard and QC mode.
I've also tried to track 2 people at the same time. In the processing
example it worked (although slowly) in qc I only get the data of the
last calibrated user.
My setup didn't change, and the 11/01 version still works perfectly.
Hope it helps!
Sebastian.
George Toledo
Jan 30, 2011, 4:17:55 PM1/30/11
to osce...@googlegroups.com
I've only had one segfault, but I've checked out an install/recent version that gets it a bit, and what seems to trigger it is when a user gets right up by the camera, or readjust the cam. I've seen people walk away and back in over long periods, and that doesn't seem to segfault (in my small sample).--
George Toledo
SteveElbows
Feb 15, 2011, 6:56:03 PM2/15/11
to OSCeleton
Tony, if the version from this date is stable on the mac, do you think
it could be the changes to the makefile for compiling it on macs, or
changes to version of oscpack that were made around this date that are
causing the segfault problems?
Cheers
Steve
it could be the changes to the makefile for compiling it on macs, or
changes to version of oscpack that were made around this date that are
causing the segfault problems?
Cheers
Steve
Tony Gonçalves
Feb 21, 2011, 5:24:35 AM2/21/11
to osce...@googlegroups.com, SteveElbows
I think it has to do with the creation and destruction of some OpenNI object in the code, but your suggestions are good places to investigate too.
I'll try with the old build flags and oscpack and let you guys know how it goes :)--
I'll try with the old build flags and oscpack and let you guys know how it goes :)
Tony Gonçalves
Partner
IPN Incubadora
Rua Pedro Nunes
3030-199 Coimbra, Portugal
Tel : +351 927 474 665
Tel : +351 239 700 358
Fax : +351 239 700 301
Reply all
Reply to author
Forward
0 new messages