top of page

Splash 3, 4 and 5



Summary: hard work, a lot of experiments, and great models provided some good preliminary results, see above and see my instagram and the group instagram. I expect even better results in the upcoming shoots. The text below is about the boring technical details, so you can stop reading now, unless you are a nerd like me :)

Text for nerd, full of boring technical details

If a 360 camera does not give you the result you were hoping for, and you are kind of stubborn, what do you do? Indeed: research other options and keep experimenting to get better results. And so came underwater shoot “Splash 3”, a swimming pool group photo shoot with several models. And Splash 4, and Splash 5. Each shoot was a test, and each time problems were solved and new problems were defined to keep improving.


Gear: some camera’s give great results, but are unlikely to be used underwater, some camera+housing are proven but are a bit pricey. My idea is to combine images and video of multiple action camera’s and stich them into one 360 image and video. I fell in love with the Insta360 Ace Pro action camera, because of its novelty and its amazing capability to shoot 8K footage and use of AI in lowlight conditions. And its water resistance.


I encountered several problems

Interference with led lights while shooting video resulted in black bands in the resulting images. Solved by advice from Martin to set the framerate to 25 fps, which is half of the 50 Hertz of the European electricity network. By synchronising with the electricity network frequency, the flickering of light, which is invisible for the human eye, can be mitigated.


Images were blurred, even with slow moving models. Solution was up the framerate to 50 fps and to use even more light. Next to 2000 watt of led lights above water, I added an underwater led fish light of 400 watt.


The models were only partly visible during their swimming. Solution, set the camera tripod higher and set the camera view to 4:3.


The resulting images could not be stitched due to lack of distinguishing features in the swimming pool. We added black and white markers. The markers floated away. We fixed them to heavy boards.


When light travels from water to camera lens it breaks different than when traveling from air to lens. This feature diminishes the FOV (field of view) of the cameras, resulting in images that can not be stitched to other images due to lack of overlap. It seems that instead of the specified 155 degrees FOV of the ace pro, underwater the cameras only sees 100 degrees horizontally, and 75 degrees vertically. My approach to mitigate this is to simply use more camera’s.


The cameras need to be positioned correctly, so far I made wooden frames, but I will work on a metal frame for the next iteration.



6 camera rig made out of cut out parquet, a mouse mat and a large bolt and nut


There are many camera setting that have an effect on the resulting images. So far I came to these settings:

  • Framerate 50 fps

  • Video resulting 4K

  • Form 4:3

  • White balance: manual 8000K

  • Stabilisation: off

  • ISO: manual 100

  • Video sharpening: medium

  • Shutter speed: automatic

  • Aperture: automatic

These settings provide satisfying results, but I assume that these settings can be improved, by using rigorous testing.


The cameras have limited battery capacity, resulting in maximum of about 1.5 hour of continuous shooting. We charge all cameras parallel after a first shooting session, using high capacity (120 Watt) chargers. During the charging, we have lunch and make group bodyscape photos with the models. Which is fun.




Freak accident, 80 percent of the footage of Splash 4 was lost due to corruption of an SD card. I bought a better reader and quality Samsung SD cards.


Happy accident: one of the new models is not only a great dancer, but a professional choreographer as well.


Lessons learned: combining images of insta360 ace pro action cameras to create 360 degrees images is an original idea, but so far no-one has done that, I am afraid for a good reason. But being a stubborn person I must continue.

To do: build a rig for 8-10 cameras, prepare a test protocol to test out camera settings, ask additional models to participate in this adventure

Ideas for future shoots: would a 360 bodyscape photo be possible? Keep rigorously testing camera settings, keep shooting with groups of kind models.


Uitgelichte berichten
Recente berichten
Archief
Zoeken op tags
Er zijn nog geen tags.
Volg ons
  • Facebook Basic Square
  • Twitter Basic Square
  • Google+ Basic Square
bottom of page