Screenshot Photos

I’ve been running into observations where the user has posted a photo of monitor instead of posting the original image. I feel this leaves too much open to where the photo came from, date, location, etc and have been marking these as “human” for an ID. If the user took the photos to begin with I don’t understand why they don’t post them.

I can’t find any official documentation on this practice so thought I would post here for more input.

1 Like

If you think photos are “too good” or something else is suspicious – flag them. There’re many users doing this, not everyone knows there’s a website and not only an app, mostly it’s easeir to leave them as they are and add comment with user tag, but you can check if exif of file has date saved, if it’s the same as date of observation you can mark it as wrong date, but usually they don’t have it. Some observers just refuse to do differently while multiple people asked them to and I can’t say why.
Marking them as humans isn’t the right way, it can be a totally ok observation.

8 Likes

Like @marina_gorbunova says, a decent number of people genuinely don’t know about the existence of the iNat website and uploader. iNat too often just gets describe as an “app” so people think they must upload photos via their phone. Other folks are usually either testing out CV, are too lazy to use the web uploader (I’ve heard that explanation before), or are genuinely posting photos that don’t belong to them. In most of these situations my standard recommendation would be to ask them about the photo unless it’s obviously a copyright violation.

4 Likes

When I encounter images like this, I usually inquire if the photo is theirs and if date and location are correct. I mark the date and location as incorrect in the Data Quality Assessment until they respond to clarify. I don’t think it makes as much sense to identify them as human.

11 Likes

Agreed, it’s not an image of a human, it’s an image of an image.

2 Likes

I think the idea is that basically anything can be IDed as human, including cameras or computer monitors, because they are evidence of human workmanship. Identifying as human provides an easy way to get an observation to casual without violating any obvious rules. (I’m not agreeing with this approach, just explaining it.)

2 Likes

I am not a fan of the image of an image style posting, but I don’t think marking the photos as “human” is the correct approach. Better to contact the person and/or use one of the tags at the bottom of the observation page that deals with issues like incorrect date. etc.

Some folks have said that they want to post things from the field to get accurate location data, so they use their phone to take a photo of the image on the camera screen and post that.

To my this is beyond ridiculous and often results in terrible quality images, but that’s one of the reasons I’ve seen mentioned here.

6 Likes

I’ll hop on to advocate as one of the people that does this. When I’m in the field, sometimes I’ll take temporary photos of the LCD right after I take the photo so I have the GPS coords and the data is available (I don’t always upload these, but sometimes I do). Going through photos and processing them for upload takes a lot of time when you compare it to the large magnitude of observations and if it’s a common species I’d rather upload something like that than nothing at all (or a phone photo of the specimen in the vial, where you can barely tell what it is). So I can empathise when I see images like that (plus they can often be much better than photos taken of an organism with a phone). I always try to eventually upload the original image itself (since it’s winter I don’t think you’ll see any images like this on my account by this point), but when the season ramps up I expect this will be the case.

Photos of computer monitors though, I don’t get (though as others have said, some think of iNat only as an app where photos can only be uploaded from a phone).

Oh and just a point about image quality. The vast majority of images uploaded to iNat generally are not very sightly and I think that’s fine based on the way the site operates. An observation is identified to what is evident in the observation and the more that is shown, the more specific the identification can be. Unlike BugGuide, where curators can (currently) frass images that don’t need to be added to the overall guide, iNat holds a record of every observation. Most images of hoppers that I go through are pretty bad—and I love to see that! For so many reasons.

4 Likes

I get why people take that approach, but for me, at least, I prefer to spend the extra time with the processing of the images taken with my cameras and then to accurately place them using the map locator.

Image quality is often very important for correct observations (depends on the species, of course), and often an image will need post-processing to even be recognizable to species, especially of high-flying birds against a bright background, so taking a photo of the camera LCD can wind up with a low quality image that lacks the detail necessary to properly id it.

At this point I rarely post any images taken with my phone, although there are sometimes ones as that’s what’s always in my pocket.

Again, this is just for me, but a big part of the image/ID recording aspect is artistic as well, not just a simple recording of existence. This has been the case in natural history since the beginning of the field. Look at the care taken in old natural history illustrations to arrange everything in a pleasingly, and often creative, artistic arrangement. Photos, to me, are the same. Not only are they a record of what was seen and the context, they’re also a mini story, capturing a moment in the life of the organism seen, maybe an element of its “personality” (as conceived of by the observer), as well as something about what caught the observer’s attention and their own mindset.

4 Likes

My thoughts are that it’s better to have a low-resolution observation than no observation at all when it comes to iNat (and you can always leave a comment if you want a bit more from the observer). Users have asked me before (usually the ones taking alright photos) “are my pictures bad?” or “am I uploading too many photos?” and I always try to encourage them with the concept that every image counts here.
but really I just get giddy at the concept of more hopper photos

I share that sentiment—I usually spend a painstaking amount of time with post-processing (lack of efficiency, mostly) and I occasionally do observational illustrations/paintings as well. And I think it’s awesome that others don’t—you can really be any person of any talent or skill and participate in naturalism! Leafhoppers are tiny and while most photos on here are taken with phones, I can still give a pretty good identification most of the time. Some users really are not tech savvy and others log into this forum and talk about the source code. I have come to see iNat as a highly versatile tool and it’s been a trip sacrificing that perfectionism in the interest of sharing data that would otherwise go unseen. I consider it as much a part of my workflow as Lightroom or my camera itself. Additionally, a photo of a DSLR image is usually going to be better than a phone photo on its own (especially when it comes to 3mm bugs). Having images that are of lower resolution are still useful if they can be identified to anything at all imo.

I would certainly never generalise naturalism as being a mere recording of existence. My favourite part is how interdisciplinary it is in the way I can bring all my interests together (and how fundamental it is to being human!). I don’t think uploading these types of images for a technical purpose discounts that.

When it comes to the artistic side, I have many other platforms to utilise and I guess I just don’t think of iNat the way I think of a personal website or Flickr. iNat is more than an image sharing site, which is something that personally took me a while to adjust to since most photos are not ones I wish to share or have associated with my name :P. I guess I can’t really say that I’ve overcome this though—my ratio of photos uploaded to photos taken is… heavily skewed. I tell myself it’s because of my temporal limitations and not because I’m a terrible photographer.

So I guess for me, my perspective has changed over the years. I know you mention this is about personal preference, but I think it’s also important to stress to other users that most photos are welcome here and there isn’t really a set threshold for image quality. We can each choose how we “curate” our own accounts!

5 Likes

Well put. I couldn’t agree more.
We don’t have to be sloppy just because we’re doing citizen science.

2 Likes

Consider this observation: https://www.inaturalist.org/observations/71750108. The user has made a screenshot of a photo and used that for the observation. This seems to be dangerous as the photo metadata gets lost, so the user will have to manually edit it. It would not be surprising if most users don’t bother. Any recommendations on if these should be flagged somehow? Or are identifiers just expected to waste a lot of time questioning the data accuracy of such observations?

https://forum.inaturalist.org/t/screenshot-photos/20812
As ider you don’t have to do anything you don’t want to.

2 Likes

There are 3 main problems with screenshots as evidence:

  • it may not be their photo
  • it likely has the wrong date
  • it likely has the wrong location

All of these are flaggable/addressable in the Data Quality Assessment. When I come across screen shots, I usually leave a comment like

It looks like this is a screen shot rather than the original photo. Can you confirm that the date and location correspond to the place and time the original photo was taken? It’s important to confirm these things so that the data associated with the photo is correct. Can you post the original photo instead of the screen shot? Thanks!

And then I usually mark the location and date as incorrect in the Data Quality Assessment. If they leave a clarifying or correcting comment then I update my DQA votes.

4 Likes

when it comes to obvious screenshots, I usually reverse image search and leave a comment asking about it. I will likely upload screenshots of my raw photos in the future if I don’t have time to process them and have been asked to share a photo of a specimen that has been ID’d through a specimen.

1 Like

Agreed. If the location looks correct (and isn’t obviously wrong like someone’s living room), I tend to trust it.

I think Carrie’s approach is spot on. Consider someone showing a new person about iNat, and then that new person goes scrolling back through their camera at old photos, and makes an observation of those photos. This is usually the scenario I encounter with this. Location and Date are more likely to be wrong than right, and marking the DQA only makes it casual so the datapoint is still available to those that want it. And it’s an easy “fix” to review the DQA setting once a confirmation/correction occurs. I would tend to default to this approach, unless there is clear evidence to suggest it is accurate (eg comments to support that)

2 Likes

Well, I used screenshots whenever my observation is a video. Since I cannot upload even a 3-second video, the only practical choice I know is to use a screenshot from the video and manually enter date and location. On occasion (not often), I also upload the video to YouTube and link to that in the Note.

If you know of a better way I can deal with video, that would be good news.

3 Likes

You can use gif.

I looked for a while; but I did not find a video to gif converter that isn’t a hassle.

2 Likes