How should I make time lapse observations, for molds growing over days' time?

[booksandcookies]

Say I have an old piece of food (bread, fruit, etc) and I set it in an undisturbed place, with a camera taking a picture of it once or twice a day for 7-10 days to see what molds develop. I haven’t done this yet, but I think it would be interesting and I was wondering how to upload the results to iNaturalist.

I think it makes more sense to make this 1 observation rather than a string of many observations because 1) it is easier to identify if seeing the different life/growth stages, and 2) nothing may be visible the first few days. The only problem is what time/date should I use? When I first begin? Of course, all individual photographs would be timestamped.

[fffffffff]

You should post them separtely, to connect them use observation field "similar observation set’ or a similar one.

[bouteloua]

That’s an interesting one and I think the technical response would be that since we don’t have the option to choose a date range, you should leave the date blank. But I’d probably be liable to fudge the rules and use the “end date” of the timelapse (stopped recording on July 16th - observation date = July 16th), and note in the description that the recording started on, e.g. July 10th. You could alternatively use a still photo from the last date in one observation and a still photo from the first date in a separate observation, then use the description section or comments to link to a gif or video of the recording. oh, and welcome to the forum!

[arl0045]

I have the same question when I revisit some plants. Though I tend to add the latest picture to the initial observation and edit the note or add a comment to mention the new image(s).

[murphyslab]

I agree that the date issue might be the confusing point. Really iNaturalist should have an accuracy or range associated with the temporal data in the same way as it’s associated with the spatial data.

This is a bit tangential, but since @bouteloua suggested a GIF, I may be able to offer a how-to suggestion that might help in making your observations.

I’ve found GIFs to be rather informative and since they can be uploaded on iNat, why not? Here’s an example that I uploaded of a giant water bug that I found on the sidewalk late one night. Individual screenshots were a little blurry, but collectively it’s very clear when animated:

Note: this was derived from a short video clip, but image series are essentially the same.

To make a good GIF, you ideally would not move the camera in any way, but this might not be feasible. One option that might work is to create a holder for the camera (perhaps out of cardboard, taped down or anchored to a particular spot) to ensure that it gets aimed the same way each time.

It can also be helpful to include some reference points that will be in the frame that might allow you to align the images in the future. And if you are going for high quality, some kind of colour reference may be in order, especially if the lighting, or even the coverage of a particular colour, changes substantially in the course of the experiment.

If you’re really feeling keen, a RaspberryPi with the associated small camera accessory can be set up to take photos at whatever interval you want - I’ve done this on a few research projects and it can work excellently. Plenty of guides are available online and it only takes a few lines of code to collect periodic images.

When you do have a series of images, it’s possible to import that series in FIJI (a version of ImageJ) as a “Stack”. A stack can be used to create a GIF or a montage. There are limits to the upload size on iNaturalist, so it’s often helpful to crop the GIF to only show the relevant area. It’s also possible to label the individual slices (images) within the GIF to show the date & time.

In any case, I’m curious to hear how your mold-observing experiment goes!

[bouteloua]

If they are separate images on different days, you should make separate observations. If you’re making a timelapse gif, which was how I interpreted the topic, that’s a bit different.

[upupa-epops]

@bouteloua

we don’t have the option to choose a date range

(feature request here: https://forum.inaturalist.org/t/select-a-date-range-when-creating-observations/11395)

[prokhozhyj]

“You should leave the date blank” – without a date it will be a “gray” observation. It is not the best variant.

[zookanthos]

I would say make the observation date be what is shown in the first observation photo. It is a good idea to upload a still photo as the first image and not just upload a single animated gif timelapse because that can make loading pages slow for people on bad internet connections. Also, animations are not supported on at least the android app so people will just see your first frame, which in this example would be a piece of decidedly non-moldy bread. So take your best still photo (probably the last one you took) and then make that first and use the date of that photo as your observation date.

Making separate observation for each photo is officially what you are supposed to do, but I agree with your reasons for combining them together. I wouldn’t want to bother opening up 14 different tabs to see all the photos for what in my mind is just a single observation. It would just seem spammy, and I don’t see any actual benefit for doing so here aside from wanting to follow the rules for the sake of following the rules.

[kitty12]

Why not do a reverse timelapse gif, from newest to oldest, so the first frame would be the most developed? Then it would at least show the end result as the first frame.

[robotpie]

As otehrs have mentioned, I can see two ways to do this:

1. Make different observations of the growth vs. time and link them with a “similar observation set” tag.
2. Turn the pictures into a gif and then just upload one observation, maybe making a note in the description that says, eg. “5 days growth”