I hope this does not become a word salad. I have so many things to write about that are all inter-related, but it is such a “thought-mess” that I don’t even know where to begin, so I am just going to start tugging at strings.
This has been a bit of a marathon shoot. This is about the longest I have filmed a single subject by about 2 weeks. I am done shooting this particular plant, and really looking forward to moving onto something that grows a bit faster.
I am going to start with the eye candy, then move into the BTS videos and wrap up with discussion about equipement.
The Nepenthes. 43 day timelapse
When most people think of carnivorous plants the Venus Fly Trap comes to mind. However Nepenthes are the true kings of the carnivorous plant world and produce the greatest variety of species. They can grow to enormous sizes and are incredibly varied in color, shape, and size. This one is a “Nepenthes x ‘Louise”, and it has been in my care for a number of years. It actually made its first film debut as a juvenile back in 2015 on my Carnivora Gardinium short film at the 50 second mark. It has grown quite a bit in the past 3 years and is developing pitchers that are about 6 to 7 inches in height.
Nepenthes are pretty slow growers, but when you look at what they do its sort of makes sense. They start by producing a leaf with a little nub of a tip, the leaf stretches way and eventually lowers down. Once that tip makes landfall, it starts to grow and develop into a pitcher shape. The lid will eventually open up and secretes a sticky fluid that attracts the prey. Bugs and small mammals will crawl up the pitcher to get to the nectar and fall into the trap which is filled with a digestive fluid and down. The plant will absorb nutrients as the prey is broken down. The fluid itself is pretty much safe, and does not digest quite the way a human stomach does, at best it seems to encourage the prey to break down a little bit faster.
I have been using this plant as a test for learning how to control Otto for interactive filming. The first 14 days (it would have been 57 days total) were trashed due to the camera shutting off while filming and I did not catch it for about a day, and at that phase the growth is pretty quick and the resulting skip was too noticeable. Here is the original beginning of the footage.
I scrapped that shoot, and changed the opening to start out looking at a fully developed pitcher then down to the growing nub. The intervals started at 10 seconds and increased as the camera moved closer to the tip where I extended them to 1 min, 5 min, 10 min, 30 min, then out to an hour. Generally i fins with these plants the intervals between frames works best when set between 45 -90 minutes. Starting at one hour, and I can adjust the intervals throughout the shoot to help match the growth of the plant.
With this sort of work the depth of field (area that is in focus) is so shallow that keeping it in focus requires daily checks to verify it is in focus and growing as expected. When it does not, I shut the system down, lock in the current position, move ahead about 20 frames and verify the focus is still going to be in the right spot, and maybe another 30-40 frames out for another check, then I will move it back to its current position and resume. The results are pretty seamless, as I had to make 8-10 adjustments to the program in this clip to accomplish the shoot.
There was one part around 34 seconds where for some reason dragonframe stopped responding to the signal from the BSC, and I lost about 8 hours of capture. This is seen as a small jump in growth on the plant, and I have a plan to prevent that from happening again, more on that later.
Butterwort Flowering, 14 day capture
Two days after I started on the nepenthes I saw that one of my Butterworts was flowering. it was a Sunday night at 10pm. This plant was a new one that I have had for less than 2 months and I did not want to lose the chance to film it. I scrambled to put together a set, it was very hasty and not one of my best, but for all I know this plant only blooms once a year so I had to film it!
The set was super basic, using a foam set piece I made years ago, a few fake plants, and a poster for the background. Due to the rush and lack of time to get proper motion setup, I opted to go with a static camera and a Ken Burns approach to motion by panning and tilting via moving the 1080p frame within the 6K 24 megapixel frame capture.
Flowering plants grow much faster than regular plant growth, and I was already filming the Nepenthes at 45 minute intervals which is WAY too long between images for a flowering plant. Luckily I had anticipated this with the new Biolapse Studio Controller (BSC) which has 8 camera trigger ports and the ability to skip frames. So with Otto on port 1, and Finn on port 2, I set the interval to 15 minutes and set Port 1 to only trigger the camera every 3 intervals. This allowed me to capture 2 plants at different intervals at the same time. Worked like a champ! The end result was not perfect, but I am pretty happy with the result. This one filmed for 2 weeks.
After I felt I had finished enough of the Butterwort I moved on again to a Cephalotus, Australian Pitcher Plant. I love this plant, the pitchers are so crazy looking with big ribbed mouths and short fat hairy bodies. I have had this one for about 10 months or so and for the first 6-7 months It was the size of a quarter and hardly grew. Then one sat it just started taking off, It just took a while for me to figure out the right conditions.
I skipped the backdrop and went with a minimalist approach with this one, and put it on a rotary table for basic movement. This one was filmed for 33 days at intervals between 2 and 3 hours. I used the same trick with the BSC, but this time set Port 1 (otto) to trigger every interval and port 2(finn) to trigger every 3rd or 4th.
I nearly got a full pitcher growth and opening, but before the pitcher opened it was drifting out of focus.
I am looking forward to getting back to this one again in the future, but hopefully in a proper set.
I always try to film the process of filming. I have always wanted to use some go-pro cameras for this stuff, but honestly hacking the bluetooth wifi triggers that gopro uses just beyond my ability. So in the past I had been using my old Fuji XE-1 and XM-1 (rest in piece)
I recently found some cheap knockoff gopro cameras on amazon that run about 40 bucks, and the best part is they have a remote trigger that runs off radio and it has a button! Super easy to hack, i just open the remote and solder a couple wires on the button and connect them to a 3.5mm headphone jack. I can plug that jack directly into the studio controller, and every time it takes an image it will short out the connection under the button on the remote, and all the Gopro knockoffs take an image.
The image quality is nothing to scream about, the video has flicker as there are no manual settings, but the output is good enough that if something strange happens during the shoot I have footage of it happening to hopefully debug the problem and fix it. The cool BTS videos are just an extra bonus.
The video above was filmed by my XE1, and the output is a lot better. The exposure is pretty even, but there is some flickering/pulsing of the light levels off the LED panel in the back. I don’t think I am having light output consistency problems though, i think the flicker is extra glare coming from the atomized water coming from my humidifier. This would be harder to confirm with the action cams, but the Fuji XE-1 generally gets good output. Also there does not seem to be any flickering on the subject itself, and the humidifier is located a little bit behind the LED panel and aimed towards the set. The movement seen is from me bumping into the Fuji’s tripod i believe, as none of that movement was seen from Otto. You can see as i add new set elements as Otto continues to move around near the end.
This one is from above over the rear backdrop on ActionCam1, note the quality is pretty crappy, but the footage would still be useful for learning what went wrong if encountering a problem.
Actioncam 2 had rear sourced lighting, as before you can see as i expand the set. I prefer not to do this, but the pitcher developed growing the wrong direction and some effort was needed to continue the camera on the desired path. It worked out pretty well!
The BSC (Biolapse Studio Controller)
The new BSC has worked out wonderfully! I packed a lot of capability into it, and so far it has performed flawlessly. This recent nepenthes ended up being a total of about 1,200 images, and not a single pink frame caused by a grow light failing to shut off.
As mentioned earlier, the BSC has 8 camera triggers, which provided the flexibility to shoot multiple plants at independent intervals, and also allowed me to adjust any interval/parameter on the fly. There are some limitations to this. I built the system knowing it would be triggering multiple cameras and gave each trigger the ability to skip. So whatever the set interval is, it can trigger other systems on multiples of that number. So if it is set for 15 minutes, I can trigger other systems at 30, 45, or 60 minutes.
Taddlebox (getting nerdy here)
This is something I am excited to get started on. I picked up a YUN shield for arduino. Along with temboo this promises to give me the ability to sent emails from the Arduino.
Why is this important?
There has been a few times where Dragonframe does not seem to see the shutter signal from the studio controller. It is sort of an all-or-none scenario where the only way to fix is to reset dragonframe. this has not happened more than a handful of times. It would do the same with the older retired studio controller as well. This may happen once every few months, but it does happen.
Another scenario is if the camera shuts off. I do run AC power for the cameras using OEM battery adapters. 3rd party brands always seem to cause problems, while the OEM ones are fairly stable, however they are not 100% reliable either. My Canon 6D has shut off quite a few times over the last year and a half, and this is an easy way to kill perfectly good footage.
The taddlebox, as I envision it will have 8 camera trigger inputs, 8 camera trigger outputs, and 8 hot shoe monitors. When it receives a shutter signal from the studio controller it will trigger the camera, and it will also monitor the camera’s hot shoe to make sure the camera did in fact take a picture. If the camera fails to take an image, it will send me an email alerting me that my attention is needed. Often when this does happen I will be home, but I don’t sit in my studio all day watching this stuff. So if it fails to trigger the camera it will let me know. Generally speaking if i lose 2-3 frames its not a big deal. So depending on what I am filming, i have anywhere from half an hour to several hours to get home to my studio and check to make sure the cameras are all working and figure out which one is not working, and hopefully resolve it before it becomes noticeable on the footage.
hope you enjoyed, thats all for now