rjshearman any news?, ryan can you send me what i asked you?, and graphic designers are you working on something? bye see ya!

__________________
Please we need help in developing an application to measure (HP, Lateral G's, Braking distance) in our cars...

http://www.hackint0sh.org/forum/show...ght=likes+cars PLEASE CODERS AND DEVELOPERS....HELP NEEDED READ POST #80 FOR MORE INFO!
THIS IS NOT AN OVERSIZED LETTER SIZE!

REP UP IF MY POST HELPED YOU IN ANY WAY!

Tomas.. sorted for you.. cheers Ryan

Ryan already posted a pm thanks replying your pm rjshearman thanks for the update on what you're on hope the problem of the idling gets solved because is annoying and less accurate solving that we can move on bye and thanks for you effort see ya!

__________________
Please we need help in developing an application to measure (HP, Lateral G's, Braking distance) in our cars...

http://www.hackint0sh.org/forum/show...ght=likes+cars PLEASE CODERS AND DEVELOPERS....HELP NEEDED READ POST #80 FOR MORE INFO!
THIS IS NOT AN OVERSIZED LETTER SIZE!

REP UP IF MY POST HELPED YOU IN ANY WAY!

Lots of user support for a 6-axis calibration. As lovemygti observed, it needs to be done only once per iPhone. Hopefully the results can be saved to a .plist file or something. In case people don't want to do the procedure, they can work with the standard values (which can be preset in the .plist shipped with the application) and accept the <10% error.

Now this will require six measurement runs, each of a sufficiently large number of data points (maybe measuring for a few seconds). We'll need the averages of x, y and z for each of the six runs. The phone must be in flat, flat upside down, straight up, straight upside down, on right side, and on left side orientations. Since I don't expect that users will be able to do this exactly, I will correct for deviations from the exact orientations -- also because we can't be sure that the accelerometer is mounted exactly straight inside the phone.

The measurement results will be six sets of averages (x,y,z): x_up is the average x value when the phone was stright up, y_up ...., x_left ... (you can figure it out). Eightteen numbers in total.

These results will then be converted to six calibration results for the zero-level and sensitivity of each of the axes: xzero, xsens, yzero, ysens, zzero and zsens. These are the characteristics of the phone's accelerometer, independent of its orientation in the car. They should be stored somewhere on the phone (maybe in the /var/mobile/Library/Preferences/com.rjshearman.cartrack.plist file or some other official settings location that would not be affected by an update of the program -- I believe the iPhone API has a way to tell the operating system to save an application's preferences and it will pick the right location automatically).

Now to find its orientation relative to the car, we'll still need further measurements. We need a forward acceleration run, to determine what "forward" means in terms of the accelerometer axes. In the old system we did this by recording the direction of gravity during the first calibration, so we could subtract it (filter it out) from the forward acceleration run (the second calibration). I'm thinking of a way to eliminate the gravity measurement and go straight to forward acceleration, maybe spanning a longer period so there will be enough precision in the data. However it's done, the user will need to accelerate the car forward (on a straight flat road) so the orientation can be figured out.

I'm leaving temperature dependance completely aside, since we don't have a thermometer in the phone.

__________________
iPhone 3GS / 3.1.2 JB (PwnageTool) / 04.26.08 carrier-locked, Carrier Logo Fixer / Cydia / 1000 posts on Hackint0sh

Installing Cydia programs on a phone that has no internet connection: read this.

Editing binary .plist, .strings, .nib and .xib files:
* on your computer: Windows tool / conversion website.
* on your iPhone: convert those files in a terminal with plutil (installed with Erica Utilities) or edit them with iFile (Cydia links).

Hey Jim,

Nice post. There is no problem storing the values as we are already doing this for the current Beta. These values are resistant to updates/reinstalls and could be over-written at any time by re-calibrating. We can also input default values at start up if it's the first time they've started.

I'm thinking we could present the user with an image on screen showing which way to hold the device and then display a button to press when ready. This would then start a count to 3 seconds where we build an array of accelerometer values.

Once we have this i'd need the following:
- how you plan on correcting for deviations
- how you would calculate xZero, xSen etc

I also understand your point about the forward calibration but i'm guessing you'll leave that until after the 6 point calibration is worked out?

My main concern is that the phone be completely still during these calibration runs. It can't be in someone's hand or we'll be measuring their heartbeats and nervous tics... They have to balance the phone on a desk somehow.
Taking the averages should be our correction for deviations. The positive and negative deviations are expected to cancel out on average. So as with the earlier versions of the calibrations, you can just add up the totals and divide them by the counter; no need to keep all the individual results. Eightteen averages is all we need.
I'm creating a monstrous set of calculations that will do this. (At the end of my creative session there will be six convoluted formulas which the iPhone will handle with ease.)
Yes, there are a number of options and I will try them out on your earlier data dumps before making a choice.

__________________
iPhone 3GS / 3.1.2 JB (PwnageTool) / 04.26.08 carrier-locked, Carrier Logo Fixer / Cydia / 1000 posts on Hackint0sh

Installing Cydia programs on a phone that has no internet connection: read this.

Editing binary .plist, .strings, .nib and .xib files:
* on your computer: Windows tool / conversion website.
* on your iPhone: convert those files in a terminal with plutil (installed with Erica Utilities) or edit them with iFile (Cydia links).

Hi guys, I am very interested in your project here.

I would love to be involved. I am a fulltime C developer (Work for digital TV company) but I have virtually zero Obj-C. I have been meaning to get into Obj-C for sometime and this seems the perfect opportunity. I am also very into cars.

Unfortnatly I also do freelance so my time for the next few weeks is limited but I would still like very much to be involved and offer suggestions and maybe a bug fix or two (if i can get my head around obj-c).

As a bonus i have Dynolicious on my iPhone

PM if I can be of any help

Shouldn't people be able to hold the iphone against an edge (eg: side of door frame, corner of wall etc)?
Sure, no problem.
Looking forward to it.
OK, let me know how you go with that data. I didn't realise until afterwards that the values were truncated to 6 decimals (this is the default in the Log statement) - hopefully there's enough detail to work with.

Dynolicious has one setup phase where you hold the phone in different directions (in hand except for two) and then each time you do a 'run' it does a small calibration phase.

I assume this is to normalise the standard vibrations from the car when it is idle and also to get the readings for its orientation when mounted in car.

I have pulled some stats out of the database file for the dynolicous app if anyone is interested.

I agree, the app appears to let you calibrate by hand. When in each position it seems to wait until you've held it steady (or still) for about 2 seconds. Then it will buzz and move on. If you hold flat to a wall then it goes a lot quicker.

You're probaby right on the extra calibration for the idling - i'm guessing we'd only do this once if required - not each time.

I too checked out the database but the results table, from what I can see, only contains the time at the various intervals and a record date for a specific ride.