Hi, new sketch and new library is still early days. The in-game reset interval does seem to vary between people. I would suggest using the exponential setting in the sketch as this reduces the effect of small levels of drift.
Just uploaded with that commented in, will take another look, may not be right away as wife just home from work, but thanks for helping.
I had similar issues with the v1 as well.
While that looks like a lot of Yaw drift its over a very short space of time the units are very sensitive and keeping it stable on your head is like trying to balance a football on a pencil.
I've not tested the V2 software in the current Alpha but was getting better than 15 minutes with the V1 software.
You can try monitoring the DMP Yaw value over an hour with the ED tracker still.
It shouldn't change by more than 0.5 plus or minus in 30 minutes.
Remember that the software is till in the early days of testing and V2 might have bugs that the V1 software didn't.
Keep posting your findings here I'm sure Pockmoon will take note and look in to any issues we pick up.
The ability to reset the tracker with a keypress within the UI makes me wonder if it could be used along with something like Voice Attack, to avoid having to take your hands off the controls in order to reach for the button.
(VA can send commands to arbitrary applications, iirc).
Shouldn't even need voiceattack. Haven't exactly used the processing IDE or written anything complex in Java but you should be able to just listen for a specific keypress or bind even when the app doesn't have focus.
What's the state of EDtracker 2 at the moment. I would like to try it out with the second unit I built but am a bit unsure of the state at the moment as this seems to be the last entry in the blog.
16/05/2014 – There is a bug is the Calibration Sketch which prevent it doing what it’s supposed to to. So until I get a fix out (tonight) it’s eye candy only. The yaw drift calibration can still be performed in the main sketch but the in-game experience isn’t as drift free as it should be.
Is this the best place to pick up the latest information. I think there was talk somewhere of a separate website to save trying to follow progress in this thread.
What's the state of EDtracker 2 at the moment. I would like to try it out with the second unit I built but am a bit unsure of the state at the moment as this seems to be the last entry in the blog.
16/05/2014 – There is a bug is the Calibration Sketch which prevent it doing what it’s supposed to to. So until I get a fix out (tonight) it’s eye candy only. The yaw drift calibration can still be performed in the main sketch but the in-game experience isn’t as drift free as it should be.
Is this the best place to pick up the latest information. I think there was talk somewhere of a separate website to save trying to follow progress in this thread.
Just think og github as place for the code to live. If you follow the link below there should be a button near the bottom right that says 'Download Zip'
That gives you eveything in the EDTracker2 repository.
Just think og github as place for the code to live. If you follow the link below there should be a button near the bottom right that says 'Download Zip'
That gives you eveything in the EDTracker2 repository.
The ability to reset the tracker with a keypress within the UI makes me wonder if it could be used along with something like Voice Attack, to avoid having to take your hands off the controls in order to reach for the button.
(VA can send commands to arbitrary applications, iirc).
Been looking a bit a that. It might be possible to have the UI sit in the background and let you do the reset. Maybe.
Java alone can't monitor your keyboard input unless it's the app currently in focus so gets a bit tricky.
There is a keybind for headlook reset (at least in Alpha 4, not sure if it's new), if it does what it sounds like the EDtracker UI doesn't new to handle that, not for ED anyway.
Is this the best place to pick up the latest information. I think there was talk somewhere of a separate website to save trying to follow progress in this thread.
Currently this and the GitHub would be the place to keep an eye on. There will be a site but it's still being worked on. Will be at edtracker.org.uk if I recall correctly.
A programming running in the background which sends a 'R' character to the ED Trackers serial port which it detects a defined key press will trigger a recenter (nasty spelling!)
BTW, FaceTrackNoIR with Hatire allows setting of a recentre key which can then be mapped in your game or even on your joystick/throttle
You can set up hot keys directly in windows so the program only needs to send the 'R' when ran. Not sure if 'CTRL + ALT + C' can be mapped in E: D though.
You can set up hot keys directly in windows so the program only needs to send the 'R' when ran. Not sure if 'CTRL + ALT + C' can be mapped in E: D though.
#v1.1 - Simplified, figured there is no point trying to identify the device as we would have already performed a 'destructive' test to do so. Simply iterate through active com ports and throw an 'R' at each of em...
import serial
def get_id(i):
print("Querying port COM%i"% (i + 1))
try:
ser = serial.Serial(i, 9600, timeout=1)
except:
print("Failed")
ser.write("R")
for i in range(256):
try:
s = serial.Serial(i)
s.close()
print("Active port found: COM%i" % (i + 1))
get_id(i)
except serial.SerialException:
pass
It requires pySerial (which in turn is happiest under python 2.7).
I have then just invoked the above via a batch file, which in turn I have added to Roccat Powergrid. Net result is that I can reset the tracker within having to reach for the button on the tracker itself
import serial
import string
import re
def get_id(port):
print("Querying port COM%i"% (i + 1))
try:
ser = serial.Serial(i, 9600, timeout=1)
except:
print("Failed")
ser.write("I")
cycle=0
p=re.compile('I.*EDTracker.*')
while(cycle<5):
cycle+=1
line = ser.readline().rstrip()
#print("Received: %s" % line)
if(p.match(line)):
print("ED Tracker Found")
reset_tracker(port)
def reset_tracker(port):
print("Resetting Tracker on COM%i" % (i + 1))
ser = serial.Serial(i, 9600, timeout=1)
ser.write("R")
ser.close()
for i in range(256):
try:
s = serial.Serial(i)
s.close()
print("Active port found: COM%i" % (i + 1))
get_id(i)
except serial.SerialException:
pass
It requires pySerial (which in turn is happiest under python 2.7).
I have then just invoked the above via a batch file, which in turn I have added to Roccat Powergrid. Net result is that I can reset the tracker within having to reach for the button on the tracker itself
I've been experimenting with making the drift compensation adjust itself (small increment or decrement depending on which way the axis has drifted) when the reset button is pressed. It does seem to help ... or maybe the sense that pressing the button does more than just re-centering has a placebo effect on my perception of how long it remains stable
I have just used the latest code from and including the change an hour ago and these are my findings in the short term.
The calibration seems to be working spot on as expected and the values are saving.
The edtrackerII uploaded and monitored and values stored again as expected.
Now in Elite i seem to have better results than before so far but need more testing, stays centered well and no drift i could see.
But i do have the issue still where if i look to far to the sides i get chucked across to the opposite side? very odd behavior.
Also when i look at the info panels the left one comes up if i look at the left of my screen and down a little, but the right one i have to look past the screen and down a fair bit and is not really user friendly, is this part of the work in progress or elite issue?
I have the define exponential commented in ie // removed.
Thanks again for all the work looking very good and heading the right way.