APS Excellence in Physics Education Award
November 2019

Education Prize Logo
Science SPORE Prize
November 2011

NSF Logo
The Open Source Physics Project is supported by NSF DUE-0442581.

Data export issue V 6.1.2 post and replies

Return to the Data export issue V 6.1.2 thread
Login to post to this thread

Data Export with start frame offset
Michael Peschk
8 Posts

I noticed a problem with the data export.

how to reproduce:

- The first frame is skipped because it only contains reference objects for calibration -> set startframe to frame 001 (or higher)
- track point mass
- set zero time at a specific frame step
- set number format (wish I could save my own format rules)
- set axis to traced point at time 0 (I wish there was a button that sets the origin to a point of my choice in the active frame)

Then I exported the point mass via File>Export>Data as formatted. However, the timestamp is shifted by one frame. I suspect that the skipped frames are still used for the timestamp calculation during export.

If you export the data via the Data Tool then the time stamp is correct, but unfortunately you cannot format the output here.

Attached File: tracker_6.1.2_issue.png


Replies to Data Export with start frame offset

Re: Data Export with start frame offset -
Douglas Brown Avatar
Douglas Brown
455 Posts

Hi MIchael, unfortunately I can't reproduce the problem. I followed your instructions but the data looks right when exported from both Tracker and Data Tool.  Can you save a TRK file (with data but no video) that exhibits the bug and attach it to your post?  Thanks. Doug


> I noticed a problem with the data export.
>
> how to
> reproduce:
>
> - The first frame is skipped because it
> only contains reference objects for calibration ->
> set startframe to frame 001 (or higher)
> - track point mas
>
> - set zero time at a specific frame step
> - set number
> format (wish I could save my own format rules)
> - set
> axis to traced point at time 0 (I wish there was a
> button that sets the origin to a point of my choice
> in the active frame)
>
> Then I exported the point mass
> via File>Export>Data as formatted. However, the timestamp
> is shifted by one frame. I suspect that the skipped
> frames are still used for the timestamp calculation
> during export.
>
> If you export the data via the Data
> Tool then the time stamp is correct, but unfortunately
> you cannot format the output here.



Re: Data Export with start frame offset -
Michael Peschk
8 Posts

I use Windows version (if this matters)

Attached File: TRACKER.ZIP



Re: Data Export with start frame offset -
Douglas Brown Avatar
Douglas Brown
455 Posts

Hi Michael, I'm sorry but the text data in the test.dat file you sent looks exactly the same as the data in Tracker's datatable when I open your test.trk. And when I export the datatable using File|Export|Data I again see the correct text data. I agree the data in the pair of screenshots you included in your first post look different but I can't do anything if I can't reproduce it.  Doug



Re: Data Export with start frame offset -
Michael Peschk
8 Posts

OK,


I was amazed why it sometimes works and sometimes not.

What I've found out so far is that this appears to only occur with image sequences in my case PNG images.

I converted the pendulum drag video into single images.

ffmepg -i pendulum_drag.mov ./PNG/pendulum_drag_png_%03d.png

Then tracked a section and right-click on the slider "Set time to zero" somewere in the middle of the sequence.

The time is now set to zero in the table, but the start time in the clip settings is not calculated and updated.

When exporting the data probably the start time, which is in the clip settings, is used for start counting. If I set start time to -1.400s then my exported data looks like in Tracker table.

Save to *.trk with -1.400 s showing but -4.700s set in clip settings will save the first frame as -4.700s, thats why my exported project and data looks the same for you.

Attached File: tracker_6.1.2_issue_2.png



Re: Data Export with start frame offset -
Douglas Brown Avatar
Douglas Brown
455 Posts

Well done, Michael, this time I was able to reproduce it and fix it :-)  It was indeed a problem only for image videos. Thank you for persevering and nailing down the issue!

The fix will be in the next Tracker release which I hope will be soon--2-3 weeks?? Doug

> I was amazed why it sometimes works and sometimes not.
>
> What I've found out so far is that this appears
> to only occur with image sequences in my case PNG images



OSP Projects:
Open Source Physics - EJS Modeling
Tracker
Physlet Physics
Physlet Quantum Physics
STP Book