Yeah it has been a while.
October, November and December have been super busy. Holidays have taken quite a bit of my time and focus. I had a pretty severe plumbing issue in my basement that required a bunch of it to be torn up with a jackhammer and have a pipe replaced. For that I had to cram a ton of stuff into the studio to make room, that slowed me down for a few weeks.
I also had been spending some time working on other projects, just some general home improvement stuff, which has been taking my time and energy away from filming plants, I only have a couple cameras running at the moment. I have not been completely sitting on my laurels however.
Ill start out with Plants, then Robots, then Studio/shoots.
This has been a bit of a hit or miss for me. I have a couple dozen carnivorous plants that I have been trying to grow. Some have been doing awesome, some have been limping along, and my filming was not producing the results I wanted, so I picked up a 75 gallon aquarium and turned it into a terrarium to see if I could get better results. So far that move has paid off. I have a Nepenthes bicalcarata that has been struggling, once I moved it into the Terrarium it finally started throwing a pitcher. Cant wait to see the fangs!
I have been somewhat busy on the Robot front. I had encountered some problems with Finn. This may get a bit hairy to explain, but I will try my best.
I had one copy of Dragonframe. With this, I was able to control multiple robots with no problem. I think I can connect up to 4 motion control devices to a single license of Dragonframe. Only one camera though, so Otto’s camera would record direct to the laptop, and the others would record to memory cards. I would build all of the axis controls in Dragonframe, and one by one set each robot and motion routine up. I started running into some problems though. When using Otto and Lil Wayne, everything worked smoothly, but once Finn was connected the Dragonframe software would freeze after a few days, and cost me several weeks worth of footage and effort.
In order to figure out what the heck was wrong I picked up another copy of Dragonframe and another laptop and started running into the same problems. I don’t want to get too nitty gritty here, but I did eventually solve the issue. The power supply I was using was just not up to the task. The “brains” of Finn is all powered by the laptop via USB, the power supply was just there to provide the power to the drivers/motors. I’m not exactly sure how that caused the issue, but once I upgraded the power supply the issue cleared up and Finn is running nice and stable.
Just to be on the safe side, I have decided to try to keep all my motion rigs on their own copy of Dragonframe, so if one of them runs into an issue it wont jack up all the other robots as well. I picked up a pair of the DDMX-512’s from Dragonframe so I can tie them in with the Biolapse Studio Controller. I am keeping Lil Wayne piggybacking off Otto for the time being, and I have one more laptop with Dragonframe that I will be using to test out a new crane I am building.
The Crane. Project Hector.
Yes. I am building a Crane. I am super excited about this. Honestly, after using Otto the 4 axis Finn and 5 axis Lil Wayne setups feel awfully limiting. I feel that a mini crane can achieve most of the same shots that Otto is capable of. The big bonus is the XYZ motion instead of just linear.
I am using a lot of the same materials from Finn. 1x2inch aluminum extrusion, 1/2 inch ID bearings, Actobotics hubs, mounts, and clamps. The pan/tilt is going to be the same used on Finn, but added on a crane instead of a slider.
Right now the entire structure is rickety as hell. Lots of sloppy slopp. I don’t have a CNC mill. So all the holes for the bearings were cut using a standard hole saw of the right size. The bearings are slightly smaller than the holes, giving them a little bit of a jiggle in their spot. I have an idea on how to reduce this slop and make everything nice and tight using some beer cans. I am hoping to try it out this weekend. if I can get the slip removed and make sure the bearings stay put with a night tight friction fit, this should result in a very nice and accurate motion system.
Dragonframe is awesome. It has what is called Virtual Axis. To unpack this a bit, the crane setup has an arm that swings left and right, up and down. This covers the X and Y directions, however the arm swings in an arc, so it is unable to move the camera perfectly up and down, or left and right without that arc. This is where the track underneath comes into play, it will allow the entire rig to move forwards and backwards. Dragonframe uses the Virtual Axis to compensate for this arc by moving the entire rig in and out while it swings the arm left and right, or up and down. This will allow it to be controlled in a very similar was as Otto, but smaller, easier to move around, and far cheaper.
Studio and Shooting
I am currently shooting again. I have 3 cameras up and filming. I am really please with the footage I am starting to get. These systems require a LOT of babysitting, a lot of monitoring.
I will check everything every morning, and every evening. I also use a 3rd party remote desktop application that allows me to log in via my mobile device and check each computer whenever I want and I will check several times a day while I am away.
At one point I had the foolish idea that I could just get everything up and running and leave it, come back in a month and presto, new footage. The fact is it just does not work that way, there are so many factors that have to be taken into consideration.
- First, build out the sets, and figure out my shots
- Check for updates on the laptop and let it reboot and update.
- Shut off windows update service to prevent forced updates
- Once everything is up, start programming in the moves based off my experience with the plant growth.
- Turn all cameras off, and turn them all back on.
- Verify hard drive space on computers, and ensure memory cards are fresh.
- Shut off live view on cameras (can cause issues over time)
- Start Filming!
Over the next few days I will check on the plants and progress twice a day in person, and 3-4 times via remote access.
Plant has moved in a way that is pulling it away from the focus point. I do this every few days, so far with my current shoot which is a few weeks in, I have done this at least 7-8 times.
- Pause shooting. ( the new BSC has a switch which I can flip that will pause everything, shut off the grow lights and turn on the fill lights)
- Disengage motion control for editing
- Turn on Live View
- Lock in current position on all motors with new Keyframes on the current location.
- Move Otto about 8-10 frames ahead of current position.
- Adjust focus and set a new keyframe.
- Move Otto out about 50-60 frames and judge if another correction is needed at this point. if so, make adjustment and correct.
- Move Otto back to 6-7 frames BEFORE current position to ensure any backlash is reeled in before starting shoot.
- Move Otto to the current frame waiting to be taken.
- Disable Live view
- Re-engage motion control for shooting.
- Flip the switch on the BSC to re-engage the timing and lighting.
- Check in an hour to ensure it is working as expected.
So that’s a lot of steps. I have a checklist I follow, but occasionally I find I forgot the disable Live View. Not a huge issue, but sometimes it gets funny and fails to capture an image.
Other things I have run into with the current shoot. I was tracking the pitcher on the Nepenthes and planned to follow it until it made contact to the ground then start to develop. After about 14 days of shooting, and multiple focus adjustments, the Canon 6D shut off on me. I did not realize that, I just noticed that from about 10pm to 8am 0 images were taken, so it missed about 10 shots which was enough to cause a big enough of a jump that I made the decision to scrap the current and start over.
Along with daily checkups, and remote monitoring, I also have a very nice output from the Biolapse Studio controller so I can tell what it is doing, and how it is performing, and it even helps me debug when things go awry, which I will get into in a minute. The BSC already has some built in troubleshooting skills to make sure the grow lights shut off. In my past blogs I was whining about pink frames, these are images where the grow lights would not shut off when the picture was taken and the pink color they produce was contaminating the images. The new BSC has built in skills to verify and monitor the lights to ensure they are in the correct state, and backup relays that will kill power to grow lights if the primary relays fail. It works like a charm, and I have not had a single “Pink frame” since I replaced the old Biolapse Control Module (BCM) with the Biolapse Studio Controller (BSC).
To help further, I can connect via hyperterm into the BSC and get a feed on what it is doing, time stamps, and get indications of when relays are engaged/disengaged, humitity, temp, and see exactly what happens every time it takes an image.
I am exploring the options of giving the BSC capability to send me emails in case it runs into relay problems, and possibly having it monitor the hot shoes on the cameras to ensure it does see the shutter release. That was if a camera shuts off on me, it will see the camera never triggered and notify my via email to my mobile device.
On the new shoot, I have already made 5-6 focus adjustments, a few adjustments to the pan and tilt, and also found a bug in the Biolapse Studio Controller. If the Humidifier was OFF when it started a routine, it would not engage the fill lights. it caused 2 frames to be taken with a pitch black image.
This put me in a bit of a situation. I needed to increase the humidity of the room. I had built a small vent system that would direct the humidifier (this is listed in equipment, it is a home build humidifier that can turn my whole house into a swamp) output directly to some live sets I was growing, but it was not getting the whole room nice and muggy. So I unhooked it from that vent and have it just dumping out into the room. It would hit the high humidity level and shut off, then the images would come out black. So in order to fix this I had to modify the code on the BSC (Biolapse studio controller) and update the software on it. It was not terribly difficult, but 2 weeks worth of footage was on the line at this point, one mess up and everything would be ruined. I opted NOT to set everything back and re-shoot those images. It was only 2 ruined, and they were not both in a row. Losing 2 frames like this is very minor and you would not even be able to see that it happened in the final result, however, moving everything back and re-shooting could cause a slight jump in the plant growth which CAN be seen.
While Otto was shooting the Nepenthes, I did get Finn setup to shoot a flowering Butterwort. I should have the footage rendered sometime soon. I hit it from 3 angles, and hope it turned out ok. It was sort of a last minute rush as I saw the Butterwort was flowering and scrambled to get the 2nd set setup.
Generally I don’t like filming plants without observing them, but flowers seem so rare with carnivorous plants I did not want to risk missing the chance.
Filming the Butterwort was interesting, I had no idea how far the stalk would go up, and it ended up facing the wrong direction. I have about 700 frames on this and 3 different angles.
The new Biolapse Studio controller came in and worked like a champ. The Nepenthes pitchers are normally filmed about one frame about every hour to hour and a half. They are just slow growing. The flowering Butterwort was growing much faster, and needed quicker intervals, every 15-20 minutes. The BSC was made to handle this sort of thing. It has 8 camera trigger outputs, and each one of those camera triggers can be setup to skip frames. Trigger 1 was connected to Otto, and Trigger 2 was set to Finn. I set Trigger 1 to only trigger every 4th interval, and trigger 2 was set to trigger every interval. I set the BSC’s intervals at 20 minutes, so this resulted in Finn taking an image and moving every 20 minutes, where Otto only did it once every 80 minutes. And it worked like a charm! This means I can film multiple plants that have different growth rates at the same time 🙂
After the flower finished blooming I decided to try filming the Cephalotus Follicularis, which is also called the Australian Pitcher plant and is one of the craziest looking plants I have ever seen. It is currently filming right now, and I hope it turns out well. I am keeping it simple and just put it on a turntable.
This work just requires a lot of work. I miss the days of setting up a rail on a mountain and setting it to go, then relaxing until it is finished. Carnivorous plants are very challenging because most of these timelapses need anywhere from 2 to 6 weeks to film, and the longer the shoot the more problems I will have to work around. There are guaranteed to be modifications, troubleshooting, and reworking that is done on the fly while shooting, and the trick is to overcome all these obstacles without ruining the footage. I am starting to get pretty good at it, but with my current small studio I am limited on how much output I can do with these plants. I don’t want to keep shooting the same plants over and over, and need to give some of the other plants some time in the terrarium in peak health before I film them. So I might be moving to some other plants for a while.
I will try to give more updates on this blog. I am not a very good writer, so hopefully this makes sense and is not too boring or painful to read. I do have some developments happening that I am super excited about. Hope to be able to talk about these soon!