It is based on our SDK and the frontend is open-source. Everyone interested is invited to join and contribute. We currently focused on an easy-to-use tool for basic scanning support. It should ease the handling of ReconstructMe especially for everyday and new users. If you are power-user you can of course still use the console application.
Hi Christoph, nice news :-) so let´s get it going!
although ReMe was doing its job here fine in its ConsoleVersion At firs tReMeQt installation I did not touch the old driver installation.
ReMeQt failed initializing the device (XboxKinect: Main GUI turns pale and freezes) :-(
As said, I removed all previous stuff and tried to install KinectXbox drivers via ReMeQt installer - no luck. The only device I can find in the DeviceManager is a XboxNUI Motor signed as being not installed.
If I set the installer to install XboxKinect drivers it says this would require 0KB???
Am Mittwoch, 10. Oktober 2012 18:20:55 UTC+2 schrieb MagWeb:
> Hi Christoph, nice news :-) so let´s get it going!
> although ReMe was doing its job here fine in its ConsoleVersion > At firs tReMeQt installation I did not touch the old driver installation.
> ReMeQt failed initializing the device (XboxKinect: Main GUI turns pale and > freezes) :-(
> As said, I removed all previous stuff and tried to install KinectXbox > drivers via ReMeQt installer - no luck. The only device I can find in the > DeviceManager is a XboxNUI Motor signed as being not installed.
Hi Mag, thanks for testing! We didn't encounter this issue, but I will dig into it tomorrow
> If I set the installer to install XboxKinect drivers it says this would > require 0KB???
That's ok, since in the installer you cannot estimate sizes of other installers.
One more thing, ReconstructMeQt uses the same openni drivers as the console #405 does. So you should be able to install those and run ReconstructMeQt anyway, if you don't trust the installer ;)
another thing: by default a system restore point should be created before the installer starts. maybe you can revert to this and check if everything works again.
Restore points are created here with every installation of ReMeQt, but System Restore fails - This issue here seems not to be related to ReMe, for also restore points created by routine do not work. Had this also some time ago. Have to look into that...
So I installed the latest driver package from your download: Now at the same time: ReMeConsole works as it did before (even after installing ReMeQt - until I removed that stuff) - no problem here, but same issue with ReMeQT: It crashes after some moments trying to initialize the device - Unfortunately there are no logs.
Best Gunter
BTW: My posts need quite some time to be shown here - does that happen to all users?
I'm an unregistered user. It really doesn't work well enough to be usable in this configuration. At least not for me. Camera loses track on a much too frequent basis. I didn't generate a model to look at, and though I don't mind the watermark in the active display, it'd make the model useless if it appeared in the geometry.
Beggars can't be choosers, I can hardly complain given the capabilities of the software and the price I paid ($0), but I'm sad it appears I'll be sticking with the command line version...
Am Mittwoch, 10. Oktober 2012 22:39:55 UTC+2 schrieb MagWeb:
> Thanks Christoph,
> Restore points are created here with every installation of ReMeQt, but > System Restore fails - This issue here seems not to be related to ReMe, for > also restore points created by routine do not work. Had this also some time > ago. Have to look into that...
> So I installed the latest driver package from your download: > Now at the same time: > ReMeConsole works as it did before (even after installing ReMeQt - until I > removed that stuff) - no problem here, > but same issue with ReMeQT: It crashes after some moments trying to > initialize the device - Unfortunately there are no logs.
Did you try to use some other configuration files? Could you provide the content of the installed directory as screenshot? Could you try to start it using administrator rights? What's the content of
With those settings ReMeQT crashes initializing the sensor.
As soon as I set some invalid config_path (f.e.: "config_path=C") I´m able to start ReMeQt with an error. Ignoring that, I´m able to access File/Settings and choose config path (and sensor path).
The very first time I did that I was able to initialize the sensor and could do a scan.
But restarting ReMeQt I had the same crash causing problem initializing the sensor (no logs- sorry)
It seems to me that somehow the sensor isn´t killed completely and can´t be accessed again.
My previous message isn´t published yet, but an update:
My problem is related to the used device:
Auto select device (-1) seems to choose my GPU (Barts) >>> does not work with ReMeQt
Disabling Auto select and running via GPU (Barts) (0)>>> does not work with ReMeQt
Disabling Auto select and running via CPU (Intel i7-2600) (1) >>> works with ReMeQt
Any idea why my GPU does a nice job using it with ReMeConsole while it fails using ReMeQt?
Am Donnerstag, 11. Oktober 2012 12:28:11 UTC+2 schrieb MagWeb:
> My previous message isn´t published yet, but an update:
> My problem is related to the used device:
> Auto select device (-1) seems to choose my GPU (Barts) >>> does not work > with ReMeQt
> Disabling Auto select and running via GPU (Barts) (0)>>> does not work > with ReMeQt
> Disabling Auto select and running via CPU (Intel i7-2600) (1) >>> works > with ReMeQt
> Any idea why my GPU does a nice job using it with ReMeConsole while it > fails using ReMeQt?
You might want to try some beta drivers. I'm currently running catalyst 12.9. It might be that the code basis has advanced and Console lacks behind a bit.
I have the same problem : RM-QT starts until the point : "Initializing Sensor" There it stops or waits or whatever. It is possible to go on with ESC, but the sensor doesn't work.
The same problem I have with the RM-SDK. It works fine until : reme_sensor_open() and it will never come back from this function.
64-Bit Win7 NVIDIA GTS 450 Kinect for Windows
---------- Another question : (maybe an errorsource for above problem)
Is it correct to bind - sensor_kinect_windows.txt <-> reme_sensor_bind_camera_options() - volume_1m_highres.txt <-> reme_context_bind_compile_options()
This info is missing in your 1-minute-example. Please add another 1/2 minute to show, where and how to give parameter-files to different functions.
Am Montag, 15. Oktober 2012 21:59:34 UTC+2 schrieb didi...@googlemail.com:
> Dear Christoph,
> I have the same problem : > RM-QT starts until the point : "Initializing Sensor" > There it stops or waits or whatever. > It is possible to go on with ESC, but the sensor doesn't work.
Could you show me the log?
> The same problem I have with the RM-SDK. > It works fine until : reme_sensor_open() > and it will never come back from this function.
> 64-Bit Win7 > NVIDIA GTS 450 > Kinect for Windows
What Kinect runtime do you have installed? 1.5 or 1.6?
> ---------- > Another question : (maybe an errorsource for above problem)
> Is it correct to bind > - sensor_kinect_windows.txt <-> reme_sensor_bind_camera_options() > - volume_1m_highres.txt <-> reme_context_bind_compile_options()
Yes that's correct. you would need to load the parameters though before the calls to sensor_open or context_compile. This info will be present in the upcoming release.
I checked before the Kinect driver with device manager of Win7, clicked search new driver, the result was : "Your driver is the newest one ! Be glad to use this OS !!!"
After your post I searched directly for kinect runtime 1.6., installed it and RM-QT works, also now does the SDK.
Thanks for your idea to think in the right direction. DidiWett
Am Dienstag, 16. Oktober 2012 15:48:58 UTC+2 schrieb didi...@googlemail.com:
> Dear Christoph,
> I checked before the Kinect driver with device manager of Win7, clicked > search new driver, the result was : "Your driver is the newest one ! Be > glad to use this OS !!!"
Be glad to use this OS !!!"
Was this really the message you got? :) By the way the next version of the SDK is built against 1.6 and supports calibration.
Am Samstag, 20. Oktober 2012 14:46:46 UTC+2 schrieb Gecuba:
> > I couldn't launch the old console version now - getting 'Failed to load > OpenNI driver' message. :((
> Update. After several restarting of my computer I can launch both versions > of ReMe - console and GUI.
Gecuba, this is great! Please update to ReconstructMeQt 1.0.35 as it brings longer scanning times. When tracking is lost due to licensing, just hold still for a moment and wait for the tracking to recover.
> update to ReconstructMeQt 1.0.35 as it brings longer scanning times.
I used 1.0.35... I admit, that new GUI ReMe isn't convinient for one more reason - buttons. When you are holding sensor in one arm you need to take mouse in the other hand and "travel" with it to the gui-button... and in console version you need to push keys - they are always in the exact, same, position.
Am Samstag, 20. Oktober 2012 20:04:39 UTC+2 schrieb Gecuba:
> > update to ReconstructMeQt 1.0.35 as it brings longer scanning times.
> I used 1.0.35... > I admit, that new GUI ReMe isn't convinient for one more reason - buttons. > When you are holding sensor in one arm you need to take mouse in the other > hand and "travel" with it to the gui-button... and in console version you > need to push keys - they are always in the exact, same, position.
Wait for our next blog entry, and you'd be suprised :)
On Sunday, October 21, 2012 2:04:39 AM UTC+8, Gecuba wrote: > > update to ReconstructMeQt 1.0.35 as it brings longer scanning times.
> I used 1.0.35... > I admit, that new GUI ReMe isn't convinient for one more reason - buttons. > When you are holding sensor in one arm you need to take mouse in the other > hand and "travel" with it to the gui-button... and in console version you > need to push keys - they are always in the exact, same, position.
Do you have a smartphone??? Install hippo remote (iOS) or similar for android and control your mouse via your phone. Works a treat for me.
