Even though it was pretty cold outside (-15°C), we bravely 😉 chased northern lights yesterday evening. I just finish processing the data and update my playlist on youtube. Enjoy 🙂

For those who want to have some details about the acquisition and processing, here they are.
The pictures used for this sequence are acquired with a gopro hero 4 black on the settings night time lapse:
– exposure time: 5s
– continuous acquisition
– 12 MP and wide FOV
– spot metter: OFF
– white balance: 5500K
– color: gopro
– ISO: 800
– sharpness: low
– Exposure bias Value: 0.0
– focal length: 3 mm
– aperture: f/2.8
Then I used ffmpeg to denoise with the owdenoise filter, here is the command:
ffmpeg -i input.jpg -vf owdenoise=depth=10:ls=10.0 -y output.jpg
Finally, the pictures are concatenated to form a video, also with ffmpeg:
ffmpeg -framerate 4 -pattern_type glob -i ‘*.jpg’ -pix_fmt yuv420p -c:v libx264 output.mp4
which create a 3 frame per second (fps) video that is converted into a 30 fps:
ffmpeg -i input_fps4.mp4 -q:v 0 -pix_fmt yuv420p -r 30 -y output_fps30.mp4

A bright morning in Tromsø

Today I will upload the photos freshly taken. It was a bright, sunny day, so we decided to wake up early (not too early because daylight is coming pretty late nowadays) and walk in our neighborhood before heading to the docks.

I can’t wait to show you the pics 🙂

GoPro Hero4 Black, aperture: f/2.8, exposure time (auto): 1/5 s, ISO: 100, focal length: 15.0 (35mm film), 3.0 (lens). No post processing apart from the downsampling using the default settings of imagej run(“Size…”, “width=2000 height=1500 average when downsizing=true”)

Here it is! The picture are on the page and it’s blue.

Just a random photo

Hi folks, I’m working on the photos and films I recently did, so it will take a little more time before the next page is coming up. Here is a random photo I took today, so that you can wait 🙂

GoPro Hero4 Black: exposure time 1/3 s, iso 200, focal length 3.0 mm, aperture f/2.8. Processing: denoising based on wavelet (ffmpeg -i in.jpg -vf owdenoise=depth=10:ls=1.0 -y out.jpg), downsampling 1:2 (gaussian blur with sigma=1.0, and decimation by 2) and contrast enhancement using imagej.

Happy new year :-)

Hey everyone, happy new year! I wish you the best, and all the usual stuffs… I’m sorry, I’m not really good at this exercise. Anyways, I will soon add pictures of the period from Xmas till new year. Here is a little preview of what happened yesterday. Fireworks were everywhere in Tromsø it’s just amazing 😀

The photo: taken with a GoPro Hero 4 on the night mode time lapse with auto contrast, and I run a simple post-processing using ffmpeg (yes, it works also with pictures :-)) for denoising based on wavelets (probably Haar). Here is the command: ffmpeg -i input.jpg -vf owdenoise=depth=10:ls=10.0 -y outpout_d10_ls10.jpg