Photo correlation in latest - errors

[urbanrambler1]

For a couple of weeks the track-related correlated imported photos have been displayed in incorrect positions (in latest). I thought it might be down to my equipment but on loading 'tested' I find that there are no problems and all correlated photos are shown in the correct locations.

[Don-vip]

Could you please add more info? Attaching your status report, a screenshot and a sample file would be useful.

[anonymous]

Latest vs Tested using same GPX data and photograph directory

[anonymous]

Hi,
Thanks for prompt response. I seem to have exceeded attachments but will add GPX file if needed. How do I generate a status report? The top screen shows Latest, for which the correlate screen omits the ':second' field from the photo of the GPS time although it allows the digits to be entered. Tested has all three time fields shown when it starts the 'correlate from photo' sequence. I hope that you can see where the correlated photos occupy different positions, using the same GPX file and the same photo directory and the same 'correlation time' from a photo.

[skyper]

Replying to anonymous:

Hi,
Thanks for prompt response. I seem to have exceeded attachments but will add GPX file if needed. How do I generate a status report? The top screen shows Latest, for which the correlate screen omits the ':second' field from the photo of the GPS time although it allows the digits to be entered. Tested has all three time fields shown when it starts the 'correlate from photo' sequence. I hope that you can see where the correlated photos occupy different positions, using the same GPX file and the same photo directory and the same 'correlation time' from a photo.

You probably were not relaxed enough, as it seems to me that you unintentionally did upload the same picture twice. Trac is sometimes not that fast, please stay calm.

If you want to by pass "recap" you need to log in. The GPX file or at least an small example makes it much easier to spot the problem.

Please, have a look at the wiki page to find the status report.

[urbanrambler1]

GPX file

[urbanrambler1]

In the first photo (of GPS device) it was showing 10:11:09, which was used to correlate. Josm-latest updates daily, so version in report is a couple of days later.

[holgermappt]

Hi,

I changed the component to Core image mapping as this is not related to the PhotoAdjust plugin. I don't know exactly that "image mapping" refers to, but it sounds reasonable, at least to me.

I can confirm that there is a regression in the image to GPX correlation dialog (Geotagged images -> Correlate to GPX -> Use photo of an accurate clock -> Gps time (read from the above photo)). Only hours and minutes can be specified. Seconds can be entered (in the German version, the English version complains about a wrong time format), but they are ignored. I tested with JOSM 7319.

Holger

[urbanrambler1]

Many thanks for your efforts. I only do photo correlation about weekly, so any more feedback will emerge slowly. I would note that this error 'came to me' via JOSM-latest/plugins updates, because I have been happily using this without problems for a year now, and have not altered my setup.
With regard to your second paragraph, this 'missing seconds' feature is very recent, so is presumably related to whatever causes the malfunction.

[Don-vip]

ok thanks for the info it might be a regression of r7299.

[Don-vip]

In 7320/josm:

fix #10283 - wrong time format in GPX/images layer leading to wrong correlation (regression from r7299)

[anonymous]

Replying to Don-vip:

In 7320/josm:

fix #10283 - wrong time format in GPX/images layer leading to wrong correlation (regression from r7299)

This problem is back again. JOSM 'latest' is showing incorrect positions whilst 'tested' works OK. I'm fairly sure that the problem was sorted, but it seems to have been reintroduced (probably with the latest update today).

[skyper]

Please, reopen a ticket otherwise your comment/problem might get lost.

Could you try with the built of latest of the last days to find the problematic change ? You'll find them under download.
Thanks.

[Don-vip]

I cannot reproduce. The reintroduction of the bug is unlikely as we didn't touch this part of code again.
Can you please try with recent versions like skyper said, plus attaching your status report and give us a complete step sby step reproduction method, with sample gpx and image files, thanks?

[anonymous]

Hi, thanks for response. Unfortunately I can't reproduce it either now, so I'll withdraw my bug re-notification.