tpyo kingg
Posts: 809
Joined: Mon Apr 09, 2018 5:26 pm
Location: N. Finland

Direct sunlight versus camera board

Sun Feb 03, 2019 12:48 pm

I've got a Raspberry Pi Camera board v 1.3 NoIR which I'd like to point out a window and do some time lapse photography. However, it would be catching direct sunlight for up to a few hours per day by midsummer if I try that. At first the sun will be filtered a bit through some trees but later be crossing over the tree tops and directly in for some months. Can the camera handle that direct light or will it be ruined? I'm guessing the latter.

User avatar
HermannSW
Posts: 2657
Joined: Fri Jul 22, 2016 9:09 pm
Location: Eberbach, Germany
Contact: Website Twitter YouTube

Re: Direct sunlight versus camera board

Sun Feb 03, 2019 2:58 pm

I don't know why the sensor should break.

I once did record directly into bright lamp without issues. Because of 600fps framerate the shutter time was not high though:
https://www.raspberrypi.org/forums/view ... 0#p1243594
Image
https://stamm-wilbrandt.de/en/Raspberry_camera.html
https://stamm-wilbrandt.de/en#raspcatbot
https://github.com/Hermann-SW/raspiraw
https://github.com/Hermann-SW/Raspberry_v1_camera_global_external_shutter
https://stamm-wilbrandt.de/github_repo_i420toh264

pfletch101
Posts: 623
Joined: Sat Feb 24, 2018 4:09 am
Location: Buffalo, NY, USA

Re: Direct sunlight versus camera board

Sun Feb 03, 2019 3:19 pm

HermannSW wrote:
Sun Feb 03, 2019 2:58 pm
I don't know why the sensor should break.

I once did record directly into bright lamp without issues. Because of 600fps framerate the shutter time was not high though:
https://www.raspberrypi.org/forums/view ... 0#p1243594
There is a lot of difference (orders of magnitude) between 'a bright lamp' and the sun! What happens if you use a lens to focus the sun on a piece of paper? I would expect the same thing to happen to a sensor element with the sun's image focused on it.

User avatar
HermannSW
Posts: 2657
Joined: Fri Jul 22, 2016 9:09 pm
Location: Eberbach, Germany
Contact: Website Twitter YouTube

Re: Direct sunlight versus camera board

Mon Feb 04, 2019 12:01 am

pfletch101 wrote:
Sun Feb 03, 2019 3:19 pm
What happens if you use a lens to focus the sun on a piece of paper? I would expect the same thing to happen to a sensor element with the sun's image focused on it.
https://www.raspberrypi.org/documentati ... /README.md
Sensor image area is 3.76 × 2.74 mm / 3.68 x 2.76 mm for v1 / v2 camera.
As understand there is no single point focus.
I would risk one of my 7 v1 cameras for an record directly into sun experiment, but weather in winter does not allow for that right now.
There is a lot of difference (orders of magnitude) between 'a bright lamp' and the sun!
I used mich brighter Ikea 1000lm lamp than the lamp for above mentioned video.
And to avoid dark recordings for very high framerate videos with Raspberry cameras (>500fps) I used 5000lm light.

Perhaps time to record directly into 5000lm light ...

I have no cooling for 5000lm light, so can run it only for up to 20s because it gets very hot by 50W LED driver:
https://twitter.com/HermannSW/status/980729703899246592
https://stamm-wilbrandt.de/en/Raspberry_camera.html
https://stamm-wilbrandt.de/en#raspcatbot
https://github.com/Hermann-SW/raspiraw
https://github.com/Hermann-SW/Raspberry_v1_camera_global_external_shutter
https://stamm-wilbrandt.de/github_repo_i420toh264

pfletch101
Posts: 623
Joined: Sat Feb 24, 2018 4:09 am
Location: Buffalo, NY, USA

Re: Direct sunlight versus camera board

Mon Feb 04, 2019 1:26 am

HermannSW wrote:
Mon Feb 04, 2019 12:01 am
pfletch101 wrote:
Sun Feb 03, 2019 3:19 pm
What happens if you use a lens to focus the sun on a piece of paper? I would expect the same thing to happen to a sensor element with the sun's image focused on it.
https://www.raspberrypi.org/documentati ... /README.md
Sensor image area is 3.76 × 2.74 mm / 3.68 x 2.76 mm for v1 / v2 camera.
As understand there is no single point focus.
I would risk one of my 7 v1 cameras for an record directly into sun experiment, but weather in winter does not allow for that right now.
If the camera is focused on the far distance - and I got the impression that the OP's camera probably would be - the image of the sun would be in focus as it tracked across the sensor, and (since it is of finite width) each element in its path would be exposed to the image for a significant length of time. Don't forget that the concept of 'shutter speed' for the Pi's camera is just that - a concept. The camera has no shutter (or iris diaphragm) in the normal sense, so that light from what it is 'looking at' is continuously impinging on the sensor.
There is a lot of difference (orders of magnitude) between 'a bright lamp' and the sun!
I used mich brighter Ikea 1000lm lamp than the lamp for above mentioned video.
And to avoid dark recordings for very high framerate videos with Raspberry cameras (>500fps) I used 5000lm light.
1000 lumens is about the light output of an old-fashioned 150W incandescent bulb - a long way from sunlight!
Perhaps time to record directly into 5000lm light ...

I have no cooling for 5000lm light, so can run it only for up to 20s because it gets very hot by 50W LED driver:
https://twitter.com/HermannSW/status/980729703899246592
You would need to expose for (at least) some minutes to start to model what the OP is envisaging, and even then I think that the spot intensity would be much lower with the artificial light..

User avatar
HermannSW
Posts: 2657
Joined: Fri Jul 22, 2016 9:09 pm
Location: Eberbach, Germany
Contact: Website Twitter YouTube

Re: Direct sunlight versus camera board

Mon Feb 04, 2019 7:58 am

OK, it makes no sense to test with 5000lm light.

I read that "whatever can damage your eyes can damage a rollling shutter camera directed into sunlight as well".
Global shutter cameras with reasonable shutter times are not affected, but since smartphone/Raspberry cameras are rolling shutter cameras, the image sensor will be exposed the whole time to sunlight.

While I read that total sun has c×10²⁸lm, I did not find a calculation of how much lumen end up on earth/m²/camera sensor by inverse-square law.


P.S:
Found this (unrelated) nice animation of rolling shutter effect wrt lightning:
https://www.metabunk.org/will-an-slr-cr ... ing.t6718/
Image
https://stamm-wilbrandt.de/en/Raspberry_camera.html
https://stamm-wilbrandt.de/en#raspcatbot
https://github.com/Hermann-SW/raspiraw
https://github.com/Hermann-SW/Raspberry_v1_camera_global_external_shutter
https://stamm-wilbrandt.de/github_repo_i420toh264

muth
Posts: 27
Joined: Sun Aug 05, 2012 3:32 pm

Re: Direct sunlight versus camera board

Mon Feb 04, 2019 8:25 am

I experienced CCD permanent marks on camera V1.3 with direct sunlight. I had it with a time-lapse as well (https://hackaday.io/project/6402-photog ... me-journey)
On the path of the sun, the colors are a bit washed out, it seems it is not the silicon that is damage but more the RGB bayer filters.
If you look at the video bellow, you can spot a lighter trace on the top-right corner after the first day, at 0:08. (there is two traces because the orientation was a bit different in a previous test)

https://youtu.be/VR9EkZVwe64

Image

Return to “Camera board”