General > General Technical Chat
Looking for RGB or RGBW values of a sunrise
<< < (7/7)
tooki:

--- Quote from: Twoflower on September 17, 2017, 07:53:12 pm ---For a small project I need a table of RGB or RGBW values over time of the sky-color from the morning just before the sun rises until the sun is nicely over the horizon**. In the end I would like to play back this using a LED strip. Actually two strips: One RGB and one white. Thus the RGBW. But I think I can convert an RGB table to RGBW.

As I don't have a single pixel RGB sensor I need to find another way. My idea so far was to setup a cam looking into a white, opaque globe to get the sky colour as whole. Do a picture every minute or so. And reduce the rescale the picture to 1x1 pixel to further obscure the light and get the RGB value from the sky from that remaining pixel.

Any other ideas, or is there somewhere an existing table I can use? An existing table would be nice as here the winter fall is coming and the chances to capture this is not that high any more.

**I'm speaking of a rising sun as seen from Earth. And a nice one is preferred ;-)

--- End quote ---
What makes you think that every sunrise is the same?

Honestly, I don’t see the point in attempting to scientifically capture something that is so wildly variable. Just play around with values until you get the desired effect.
Twoflower:
I never claimed that all sunrises are the same. I was looking for a base-point to start of the tweaking you suggested. Especially what color do one would take: The color of the sun, the color of the sky (close to the sun, opposite, ...) a mixture of all. And also was looking for suggestions if either something like a dataset is available, a simulation tool or how to record the color composition. And I got very valuable feedback from the forumites (big thanks about that).

I get every time surprised if there are implied suggestions that I want to do that scientifically correct**. I just want to re-build a kind of Phillips sunrise alarm clock. Nothing more, nothing less :-)

Unfortunately the project got pushed out even further :-(


** In the meantime I was visiting the PTB here in Germany. They have a scientific setup for testing solar-cells. I forgot how many different wavelengths they use. I have only a picture taken of the big setup: they had also a smaller but more accurate testbench.
Twoflower:
To provide some additional input: The latest version of the render software Blender (free&open source) includes a physically based sky renderer called Nishita (a bit down is a short demo reel including the source files for that setup). The code behind is documented here.

The result looks impressive and might be a reasonable good source of daylight color transitions. Especially you can replay and modify it as you need not like using a real world measurement.
macboy:
Colour is funny business. By funny I mean extremely complicated.
You ask for RGB values. You can't just take RGB values and drive the R, G, and B portions of the LED strip to those brightnesses and expect to get the same colour. No way. Because.. what is Red or Green or Blue? The actual specific color coordinates of those primary colours produced by your strip will not match any standard. Then there is the brightness of those primary colours produced by the strip.

So even if you can convert from same arbitrary standard to something you can directly program into the strip, you have another significant problem. The R, G, and B LEDs each produce a very narrow spectrum of colors. For a display device (something meant to be viewed directly, like a TV or monitor) this can be highly desirable as it allows a wide gamut of colour to be produced. A property of the human eye called metamerism allows this spiky, uneven spectrum to appear exactly the same as any other different spectrum which happens to excite the eye's three types of cones in the same proportions. However, for a illumination source (something meant to illuminate other objects like a light bulb) this is highly undesirable, as not all wavelengths are represented in the output. Brightly coloured objects are especially affected. So while you can calibrate and tweak the LED strip to mimic any colour exactly, for direct viewing, as soon as you bounce that light off some other non-white object, it will look very different. Not even close.  This includes the paint on your walls and your furniture, your skin, and everything else in the room. It will look strange.
Twoflower:
Yes, there's lot of things to do to get a reasonable output. Just for the LOLs I fed the 'RAW' data taken from a set of frames to the LEDs made me wonder if I mixed up the the channels ;-). The white channels doesn't simplify things either but will hopefully improve the result as it does not have this extreme RGB-peaking. I hope to use the VEML6040A3OG color sensor to find the right relationship between the individual channels, tweak the whitepoint to get close enough to my goal. At least I think driving the LEDs with PWM provides reasonable good linearity to not run a full calibration process. And of course the setting is valid for the same type of LED combination.

The light-strip will facing a wall*. No direct light. The problem with direct light to colored objects are not given. At least as close as it doesn't hurt the eye. And the plan is to use the RGB channels only to 'tweak' the whitepoint (not correct therms, but I think you understand what I mean) to the right direction. That reduces the peaking of the RGB channels to the minimum.

*) Which is unfortunately not white but a light yellow/orangish color. So probably a final tuning is needed in the end.
Navigation
Message Index
Previous page
There was an error while thanking
Thanking...

Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod