I just opened ipi recorder and it updated to version 1.5.077 and I see that it's saying it uses libusb-win32, which I'm pretty sure is a kinect driver. It's not finding it though. There is no build information posted about version 1.5.077. I have been messing around with brekel's mocap and so I have the NUI drivers installed for that. Is there conflict with these drivers that ipi record can't see the kinect or is there no kinect support in this version? Any clarification would save me some time from messing with drivers if this version does not work with Kinect. Thanks.

_________________

Latest iPiRecorder (1.5.0.77) features beta support for Kinect through libusb drivers. To enable Kinect support, please uninstall previous version of iPiRecorder and install new version from iPiRecorderSetup.exe (from Downloads page on our website). In the future, we may support Kinect through PrimeSense's OpenNI drivers but we will need additional time to implement it.

iPiRecorder with Kinect support cannot co-exist with PrimeSense's OpenNI drivers and Brekel plugin. If you want to restore PrimeSense's OpenNI drivers and Brekel plugin, you will have to uninstall "iPi LibUSB-win32 Driver Registration for MS Kinect" (see "Uninstall Program" in your Control Panel).

iPiRecorder PlayStation Eye drivers can co-exist with PrimeSense's OpenNI drivers and Brekel plugin. If you have problems in this configuration please send us more details about the problem so that we could implement a fix.

We did more testing and we actually cannot reproduce the problem. We see some random tracking errors in both old (build 112) and new (build 113) versions, but no statistically significant difference between versions. This does not mean there is no problem. We just need more help to track it down.

I would appreciate if you could send us your "Deform Test" raw video so that we could do more thorough debugging on that particular video. We promise to use it only for debugging and to not disclose it to third parties. To reduce the size, you can compress the video with Xvid or x264 or MP4 or similar codec. There will be some loss of quality but I think it should be very minor. We will convert the video to uncompressed AVI on our end. I hope it's not too much trouble for you to upload the video to some file sharing site and send me a link by email. Please keep in mind that there is no need to upload related calibration video, it would be better to send corresponding *.scene.xml file instead.

I think I know what you mean. I can see some small errors in 112 but they're different errors. For example, in 113, when I squat and touch the floor, the hands 'touch' the ground okay, but in 112 they pop and jitter a bit as if colliding with the floor. It's not a terrible error because it's easily fixable, but it's a difference between 113. I think I'm seeing more errors in 113, but this is just my perception. I'll switch back to 113 again tonight and see if there is a an actual difference I can document.


I will be happy to send the video and calibration .xml file to you. This data has been useful to me in a number of tests these past few months, and I was thinking of sharing this or a similar video with other iPi users eventually. I'll send you an link for the data tonight.

Thank you very much for looking into this for me. Hopefully the issue, if it exists, is minor and easily fixed. :)

G.

_________________
Greenlaw
Artist/Partner - Little Green Dog | My Demo Reels (2013,) (2015,) (2017,) and (2019)

Watch a one minute excerpt on Vimeo now!

Michael,

Email sent. Let me know if you have any problems with the files.

G.

_________________
Greenlaw
Artist/Partner - Little Green Dog | My Demo Reels (2013,) (2015,) (2017,) and (2019)

Watch a one minute excerpt on Vimeo now!

Thanks for the "Deform Test" video.

Latest status is the following. We analyzed it and I see only slight difference in tracking results for older and newer versions of iPiStudio.

In future versions, we will try to update algorithm settings to make tracking results more consistent with older versions.

One of the problems with your video was the fact that we could not uncompress it without serious loss of quality. The video was encoded using h264 codec in MP4 format. Neither DirectShow nor VirtualDub understands MP4. Seems like the only way to uncompress it on our end was to use VLC player to re-encode it into MPEG2 / AVI and then uncompress it using VirtualDub. This resulted in horribly pixelated uncompressed video. Another problem was framerate: original framerate was 60 fps but compressed MP4 video is at 29fps.

Surprisingly, tracking results with iPiStudio for this pixelated 29fps video were rather good - animation looks accurate and sharp. However, I suspect the results still can be seriously affected by the loss of video quality, and this may prevent us from seeing the difference between older and newer versions of iPiStudio.

I would appreciate if you could resend us the video in different format. I suggest compressing it into an AVI using Xvid and VirtualDub. ( http://www.xvid.org/Downloads.15.0.html , http://www.virtualdub.org/ ). It should be easy to uncompress such video on our end without much loss of quality. Your "Deform Test" video is a very interesting test case and we would love to have a higher-quality version of it in our collection. It will help us to tune tracking algorithms.

Hi Michael,

Thanks for the feedback. Sorry, I should have realized that when I compressed the video but it was very late in the evening and I'm sure I wasn't fully alert. :)

I'll try to send you a new file tonight.

G.

_________________
Greenlaw
Artist/Partner - Little Green Dog | My Demo Reels (2013,) (2015,) (2017,) and (2019)

Watch a one minute excerpt on Vimeo now!

Ah! The Kinect Sensor arrived today. Hope I can find time to do some tests soon.

Sorry about the delay with the video. I'll definitely send the new file tonight. I was feeling a little sick last night and wound up going to bed early.

G.

_________________
Greenlaw
Artist/Partner - Little Green Dog | My Demo Reels (2013,) (2015,) (2017,) and (2019)

Watch a one minute excerpt on Vimeo now!

I didn't know when I'd have time at home to try it out so I brought my gear into work today. (Thankfully mocap tech is actually relevant to my job!)

I'm using my old Dell e1705 Dual Core Laptop with an aging Nvidia card, running XP Professional x32. The laptop has an external eSATA drive for video capture. (It also has an external eSATA RAID but I forgot to bring that in.) FYI, this is the computer I did my very first capture tests with a year or two ago.

I installed the latest version of Recorder, plugged in the Kinect, waited a few minutes for Windows to install all the necessary drivers. Then I launched Recorder and selected the Kinect as an input device. And that was it...I immediately see a depth video in realtime.

The video res is 640 x 480, but the frame rate seems to be limited to 30fps. Is this the actual limit or is this just my system?

I'm amazed by how little space is required to do a full body capture. This will be perfect in our living room at home.

Anyway, I haven't done any actual capture yet but maybe after lunch and after I get a chance to go through the docs. I will try to track the results at home tonight.

G.

_________________
Greenlaw
Artist/Partner - Little Green Dog | My Demo Reels (2013,) (2015,) (2017,) and (2019)

Watch a one minute excerpt on Vimeo now!

Got back from lunch and breezed through the docs. Not much else do to but click record. Unfortunately, I forgot my eSATA cable so I can't capture. (USB is too slow.) Setup was very easy so I will continue this at home tonight.

If the laptop isn't fast enough with eSATA or the eSATA RAID, I will switch to my regular computer which has the SSD.

Looking foward to the results. BTW, I noticed that there is a new release of Recorder already with new drivers. :)

_________________
Greenlaw
Artist/Partner - Little Green Dog | My Demo Reels (2013,) (2015,) (2017,) and (2019)

Watch a one minute excerpt on Vimeo now!