How to insist on camera meta data ? or exclude observations that don't share camera meta data in a project?

Hi

So was wondering if there was a way of insisting on observers posting their camera meta data. Could this be a parameter of a project ?

The reasons I was exploring this I Have seen photos which are stripped of all camera meta data - making it a bit unsure as to the accuracy of the time and date of observation

I understand that some apps and tools strip the meta data off of pictures for privacy (or other reasons)

But on the other hand I suspect “some users” have been stripping all data from their images - which in some cases looked suspiciously like they wanted to boost their numbers within a project period so were using images from other time periods.

Thanks

I don’t have an answer to your question but I am sure there are lots and lots of common reasons for metadata inadvertently getting stripped off. For instance, the tool I use for batch cropping and adding watermarks to images also strips off the metadata (I have to remember to use a different less user-friendly cropping option for images I want to post on iNat).

It is also worth noting that retained metadata often isn’t accurate. Generally all pics I post on iNat are taken with my phone or DSLR, so they have the correct time and date embedded. But a while ago, I was posting a pic from a little pocket camera I have (which I don’t usually use for wildlife) and noticed that the embedded date was something like 10yrs ago! It had never ever occurred to me to set the date on that camera. Any camera that doesn’t have GPS and can’t connect to WiFi will need its date setting manually and I’d be willing to bet a very large proportion of people just don’t bother.

2 Likes

iNaturalist automatically strips all metadata from uploaded photos, saving that information separately. This is for privacy reasons.

4 Likes

Am a bit confused

Inat does show a tonne of meta data (or have i got the terminology wrong). When one clicks on the i icon on a picture is shows pretty much everything - date, time, location, camera models , and a lot more. Or is this that not part of the meta data.

Then what exactly is stripped away ?

3 Likes

Thanks, yes –

Some tools when I export to .png it strips away everything . In some tools there are options to retain or remove meta data or even add new fields.

What you say about not connected devices is true too.

I think what @trh_blue meant was that iNat removes the metadata from being literally embedded in the image, ie if you save an image from the iNat website it won’t have the metadata in it anymore. Indeed you are right that iNat displays the extracted metadata alongside the image when you click the info logo.

5 Likes

oh ok i understand, thanks -

means when we download the image(or the related observation) the data is gone. That I think is a good point.

But am still not closer to any answer on it is possible to ignore observations that don’t have meta data.

The short answer to your question: no it is not possible. When you create a project you can choose from an handful of basic parameters, and that’s not one of them.

2 Likes

Thanks,

Will have to find a solution someway

I suppose the essence is that I was looking for a technical solution for a human issue - people will always find a way to try and beat the system.

yes and if they were really committed, they could easily falsify the metadata!

6 Likes

My camera provides a date (accurate I think) but no location data with the pictures. So I guess you’ll have to exclude all my observations from your project. (Though the locations are accurate.)

That is what I am facing in a project :-(

Actually no - its the date that is being falsified

many cameras don’t have location data – mine doesn’t

what I was asking is that “any orginal” data from the camera be uploaded to provide info that it is not a manipulated file,

Provenance

1 Like

Does it? I know GPS coordinates are stripped, but I think camera settings are preserved by default (though for some cameras I’ve noticed it’s screwed up to the point of being unreadable)

I don’t know details. This isn’t my area of expertise

iNat strips basically all metadata from photos (there might be basic stuff like size, etc) when they’re uploaded to iNat. We display metadata on the photos page (eg https://www.inaturalist.org/photos/150862379) but the metadata are not in the photo file on iNat.

If the photo is used for an observation that has obscured or private geoprivacy, we hide location, date, etc from public view (the observer can still see those metadata). eg I can see GPS and date/time info on this page, but no one else should be able to)

Correct, it’s not possible.

Yeah. Does the project provide incentives - prizes, notoriety, etc?

I’ll also note quite a few photographers strip metadata from photos because they don’t want anyone to copy their techniques and settings. And it’s possible some members of the project might be removing metadata in order to remove location, not knowing iNat does that already.

3 Likes

Agreed. It is really easy to edit metadata. I have been doing it frequently because my trail cam has this weird thing where any photo it takes is the correct time I set, but any video it takes is marked as 6 hours earlier. I noticed the problem when I had it set to take one photo then take a video if there was still movement.

It’s useful to check if “computer screen photos” have the same date captured as observation to know if date is incorrect, so hiding all metadata would have some minuses, but if it’s a user’s choice we should respect that and force anyone to show it if there’s no clear reason to.

Do you know any posts where screenshots were uploaded from files? I’d like to check that, pretty sure forum doesn’t save that info, but I post screenshots without saving them beforehand.
You were clear, that’s why I meantioned this case where seeing it by ider can be crucial, ofc we can just mark all such observations as having incorrect date and ask to change files, but it’d be a strong action to do and not marking anything will be wrong too.