 I have a new spider. I will need a camera that works at night. I believe that I now have a working infrared camera. Let's build ourselves an infrared light. Fantastic! All nine LEDs work. A quick spider update. She's continuing to do some pretty hefty earthworks. The amount of earth here is now considerably bigger and there's a mysterious hole here. Though I'm not really sure what she's trying to achieve with that. The trap itself, which is rather hard to see, particularly she's covered it with earth, is the tube here. I believe that she spends most of her time living in there. It extends underground, under here somewhere. So I've no idea what this stuff is all for. So that will be interesting to find out. So we come back to the camera. Unfortunately, there have been a few developments. I was planning to use the G-Photo program, which is a UNIX command line program for driving cameras. Unfortunately, it turns out that G-Photo doesn't really support this camera very well. It'll download images from the onboard storage, but it won't let you trigger captures, which is not very helpful. So I need to do something else. Now, while the camera will actually do video, if I can set this to video here, and once that goes away. I mean, it works. It is now recording. You don't get AV output while it's recording annoyingly. You only get 640p, which isn't very good. It's very noisy, and the 32GB SD card that's in it has only got enough space for, as you can see, 140 minutes, which isn't enough. So I'm actually going to resort to using time lapse. Luckily, if I put this back to auto, there is a feature called continuous recording, where if I push this button and go down to long period, and if I hold down the trigger, it takes continuous pictures. And I can decrease the period a bit by switching it to manual mode, which I want to use anyway, and increasing the shutter delay. So yeah, that's actually set to 2.5, well, 1, 2.5th of a second. That's not doing continuous recording, because every time you change mode, it resets. There we go. You can see the icon on the top, like so. The orange light is the autofocus. I can turn that off, simply going to focusing, manual focus, so like so. Annoyingly, it uses the same set of controls for changing the aperture and shutter interval that it does to change the focus. So when it's set to manual focus, you then can't change the f-stop or shutter, which is annoying. But this will at least let me focus. See if we can get this a bit better. That's the wrong way. Oh yeah, we also... There is something wrong with the focus. I believe I know what the problem is, which is that it relies on reflection of the... Go, that's better. It relies on reflection of the filter to make the focus work, because I took the filter out rather than replacing it with a simple piece of glass, then the autofocus now doesn't work at all. But that should now be focused. It even thinks it's focused. You can tell by the way the crosshair goes yellow when the focus changes. Come on, go yellow. I lost the focus. This will take it. There we go, yellow. So now I can take pictures and it's reasonably in focus. Of course, the problem with relying on holding down the button in order to take pictures is it requires you to hold down the button to take pictures. I did look at a few other ways to get around this, including disassembling the camera in order to solder on wires to the pads for the trigger button. That would allow me to have an actual switch somewhere, presumably hanging out the side on a wire. However, in order to get at this, you have to remove the entire insides of the camera, which involves desoldering all the flash wires and don't want to do that. So instead, I've come up with this high-tech piece of budget technology. What you do is you hook this assembly around here, like so, and then this rubber band goes on there and pulls the trigger down. Yeah, it's a little bit dubious, but here it is taking pictures, as you can see, without me touching the camera. So with luck, it will continue doing this all night. As for the other side of things, let's just push this aside a bit. Let that keep going. On the other side of things, we've got this, which is more dubiously 3D printed stuff. This is a mounting attachment for the infrared light. This hooks over the side of the spider enclosure, which allows the actual light itself to be oriented. These are a ball designed by me rather badly. These connect up to a USB power supply. The light itself just slots in rather poorly. Again, a rubber band would probably make a world of difference, just holding it together a little better. So all I need now is a power supply to set the thing up with the camera on a tripod pointing at spider enclosure. And then I need to carefully focus and calibrate it. And then I think we're ready to go, which will be exciting. So what you're looking at here is the raw footage. As you can see, it's very purple. We can easily fix the purpleness by changing the saturation and make it more visible by changing the gamma as well. The camera is taking 1600 by 1200 photos with an exposure of about half a second. You're seeing them here at 30 frames per second. Each frame takes about two seconds to take, so that means that every second of the footage here is about a minute of real time. You can see that focus is still an issue, but honestly, this isn't too bad. Off to the left is the lamp. The substrates you're looking at is a mixture of sand and cocoa fiber, which is very light and holds moisture well. It's been made bioactive by adding isopods and springtails. These will eat any waste organic matter and help keep it clean. I've added a handful of leaf debris on top, and a few dead moths collected from the window sills to give them something to eat. The moths also help provide scale. The springtails seem to come to the surface during the night. You can see lots of activity all over the place from them moving around. I wasn't expecting quite so much, honestly. On the second night, the focus didn't work terribly well, so everything's just a bit fuzzy. There was no activity all night, except just after dawn, where you can see her moving up and down the tube. She seems to be pushing earth out the top of the tube. That explains where the mound is coming from, so this suggests an extensive burrow below. Apparently she will dig quite a large chamber to live in at the bottom of the burrow. I'm surprised that there was activity quite so late. This is now broad daylight. I thought they were strictly nocturnal. The activity continues until the camera card fills up and the recording stops. She's making about one trip every few minutes. On the third night, I zoomed in on the mound and finally got the focus right, so everything is much clearer. You can see the isopods rummaging around constantly. Burrow activity starts earlier in the middle of the night. I don't have an exact timestamp. She doesn't seem to be pushing earth out this time, but something is happening inside the mound from the way it settles. I discovered reports that their burrows can be a meter deep. This enclosure isn't nearly big enough for that, but she's obviously busy. On the fourth day, I added some food from my collection of feral house crickets. A few managed to chew their way out of their enclosure a while back and I'm still attempting to round them up. The cricket is very active and with a half second exposure, it doesn't really show up clearly. It's always blurred. After a while, it finds one of the dead moths and gets very interested. It's clearly hungry and I'm glad to know that my apartment floor doesn't have very much cricket food on it. There's a certain amount of interest from the burrows the cricket comes near, but the cricket's ability to jump makes it hard to catch. From the way it moves, I think it's aware that there's something there. Finally, she emerges from her burrow and we get to see her for the very first time. First legs and then tail. By visible light, she's jet black and clearly she reflects very little infrared either, so she's still practically invisible. And then the cricket steps in the wrong place. I think the spider manages to grab it by the leg, but it's unable to bite it and kill it with its venom and eventually it gets away. Crickets are pretty strong in their jumping muscles. She then clearly spends some time during repairs and ignores the cricket completely for most of the rest of the night. It even manages to step right on top of her burrow and nothing happens. Just before the card filled up, I can see a little activity that may be interest again, but nothing happened in the footage I have. However, the cricket was gone the next morning, although there is a possibility it may have just escaped. So on the next day, I added another smaller cricket to see what would happen. I want to make sure the spider is actually eating. Absolutely nothing happened all night, even when the cricket was literally poking its head into the spider's burrow. I think this shows that the spider did actually catch the first cricket and is now underground digesting it. She'll probably start moving in a week or so when she gets hungry again. So, as projects go, I think that one's been a success. I'm getting some decent pictures, some really interesting footage of all sorts of stuff happening at night. We're not seeing much of the spider, but given that they hardly ever leave their burrows, I'm not really surprised. I would love to get some footage inside of the burrow itself, but it's just too small. While I do have a little endoscope, which I've pointed at the burrow mouse, this is what you're seeing now. It doesn't really show much in the way of details and the picture quality is terrible. You can just about make out the inside of the burrow, but that's about all. Another sign of the spider, of course, she'll be down at the bottom inside her chamber. So, as I said earlier, I'm going to wait a week or so until she starts moving and see if I can capture some more footage. I can build a better camera mount and get a closer look at the mouth of the burrow, where the action happens. One of the issues is trying to keep the cricket in as of when I set up this particular recording. The small cricket was still in the enclosure, but while I was fiddling with the camera, it vanished. So, I suspect it jumped out. It's now somewhere on my workbench. The cricket has, in fact, not escaped. It showed up later. It was just hiding. You may also notice that it has eaten most of that moth. I'm going to have to find some more for the isopod. I'm afraid this is a problem with cricket. I need to find something else that she might want to eat that's a little less prone to bouncing. Come spring, when the weather warms up, I'm going to find her somewhere safe and sheltered under a bush somewhere and let her go. These personal web spiders will apparently live from 8 to 10 years. This one's probably about four years old. So, that's quite a lot more ecological investment than the usual one-year jumping spiders. Anyway, I'm going to call that finished. I hope you all enjoyed this little series of videos and please let me know what you think in the comments.