GPU mode is in the very early stages, If performance has not fully proven across platforms. The majority of users won’t find a benefit of using the GPU mode. We kept it in to see on a large scale how many it worked for. So far, it doesn’t work well for very many… we will continue to work with the developers and enhance for future releases.
I’m not sure what you mean it doesn’t see anything. Turning off smooth filter, will make it not show anything in the smooth filter window. Then you adjust highpass and threshold. Sounds like you might have the threshold too high, or might need to also turn on amplify.
I have tried everything, I have turned on Highpass and Amplify, and Threshold is around 4-5 and even if i take it up to the max, it picks up nothing. It only then starts registering blobs as soon as i put the smooth filter on again. But the blobs are very feint in size compared to those of Touchlib. It is indeed strange but i can post screenshots of the 2.. with my settings.
Here i am posting 3 images, 2 with tbeta with and without amplify filter and the other with touchlib.
As you can see tbeta is not registering my blobs :~(
Please excuse the large images, I am new to posting images and im not sure how.
I found the problem.. The problem seemed to be with the GPU mode, I have a 8800GT and when i deselect GPU mode it works without the smooth filter, but the main image under pressure map just shows as black. Whereas if it is in GPU mode then the picture or the webcam actually shows. Very weird but it works awesome now. Thanks
I see the problem. You’re using GPU mode. You do need the smooth filter on for GPU mode I believe (I’ll have to fix that in next release).
Unfortunately, GPU mode is not fully finished yet. To tell if it’s actual worth using GPU mode, you should check the DSP value in GPU mode to see if it’s less than without GPU mode. If it’s not, then you shouldn’t use GPU mode. Also, Amplify doesn’t currently work on GPU mode.
None of the c++ demos work with tbeta right now. They will have to be recompiled and integrated.
As for your (0,0) issues, that most likley do to noise and IR leakage. When a 0,0 point is found in touchlib, it is a point outside of the calibration grid, meaning it’s seeing a touch outside of where your screen is. So, make sure to cover IR leakage and other things beforehand.
@ M. van Veen: Yeah, you can’t use the CMU driver with tbeta as it’s not supported. But, the good thing is, the directshow driver for tbeta is just as fast as the CMU driver. So go back and use the normal driver that came with your fire-i and it’ll work.
I tried it with the newest fire i driver and it works perfectly. Also it is far more accurate after calibration than our old software.
Thnx! keep up the good work.
Mijo: You have to be more specific. For example, what resolution are you using and have you changed the fps in the config.xml file? Also, I recommend running amcap or other video capture software to set everything up and make sure it’s running at the desired rate, then opening tbeta since tbeta will use those last used settings.
Hi i have what is probably a very simple problem to fix. When i try and launch Launch_FLOSC_Gateway.sh in terminal once i press enter this happens /Users/my name/Documents/demos/Launch_FLOSC_Gateway.sh: Permission denied Why is this? Probably really simple to most but i am stumped.