-1
\$\begingroup\$

I've been doing a little electrical engineering (outside of my lane a bit). I'm trying to get the RGB values for Edison light bulbs. Can anyone provide me with those?

\$\endgroup\$
12
  • 2
    \$\begingroup\$ it is unclear what you are asking ..... are you talking about color temperature? \$\endgroup\$
    – jsotola
    Oct 17, 2018 at 2:22
  • 1
    \$\begingroup\$ I'm asking for the closest RGB match to the golden glow that an Edison bulb emits so I can use it with my LED light strips and my smart bulbs. \$\endgroup\$
    – Ravaal
    Oct 17, 2018 at 2:28
  • 1
    \$\begingroup\$ Photoshop isn't as precise as getting an exact match \$\endgroup\$
    – Ravaal
    Oct 17, 2018 at 2:29
  • 1
    \$\begingroup\$ There is no "exact match". Incandescent bulbs produce a continous spectrum governed by en.wikipedia.org/wiki/Planck%27s_law. Those "RGB" values in contrary are completely arbitrary and depend on the pigments used for the color mask of your TFT screen. Or, on the non-continous emission spectrum of the LEDs you use. Each LED type has another one. There's dozens of red types, orange types, yellow types, etc. \$\endgroup\$
    – Janka
    Oct 17, 2018 at 2:38
  • 2
    \$\begingroup\$ If you want an accurate answer buy a colorometer. They cost as much as a high end DVM. Some offer both RGB and CMYK values. \$\endgroup\$
    – user105652
    Oct 17, 2018 at 2:40

4 Answers 4

Reset to default
10
\$\begingroup\$

I used Firefox Web Developer Eyedropper to grab the RGB from a photo of an Edison Light Bulb.

enter image description here

#FACA08
(250,202,8)

enter image description here

A lighter pixel
This is very close to the color when a 2700K 97 CRI LED was reflected off bright white paper: #F4D4AB (244,212,171) just a little less blue.

#F5D483
(245, 212, 131)

enter image description here

I do not believe the LED actually illuminates using the color you see when viewing the filament. The filament LEDs are no different from other lighting LEDs. Many of these "Edison Lights" have Color Correlated Temperature (CCT, e.g. 3000K) and Color Rendering Index (CRI, e.g 80). With that you can get the CIE x,y chromaticity coordinates based on CCT and CRI. Both CCT and CRI are needed although CCT only will suffice. Once you have the x,y chromaticity coordinates they can be translated to RGB.

LED datasheets often have the CIE x,y chromaticity coordinates as shown here:

enter image description here

Using CIE x,y chromaticity coordinates from the datasheet of a 2700K CRI 90 LED (red x,y).

I wrote an SVG app to plot the x,y on a chromaticity diagram to compare two LEDs.
The red x,y are from a Citi CLU036-1205C1-273H5G3 2700K 90 CRI
The blue x,y are from a Citi CLU036-1205C1-30AL7G4 3000K 70 CRI

I used the eyedropper to grab the RGB off the chromaticity diagram for the 2700K 90 CRI.

#F7AF60
(244,175,90)

enter image description here

enter image description here

Using CIE x,y chromaticity coordinates from the datasheet of a 2700K CRI and using a CIE Color Calculator I get:

#FDAC5A
(253, 172, 90)

enter image description here


Some various CCT and CRI reflected off high grade bright white paper.
enter image description here

#F06F5C 1750K CRI 98
#E2CBAC 3000K CRI 80
#F4D4AB 2700K CRI 97 (244,212,171)
#E96C66 Luxeon Fresh Focus Red Meat

2700K is a very warm light. 97 CRI is very close to (natural) sunlight (CRI 100 = sunlight)

I'm asking for the closest RGB match to the golden glow that an Edison bulb emits

There is no difference between a "Edison Bulb" and any other LED light bulb.
In my opinion a nice warm light is 2700K with a high CRI.

Grabbing the RGB with the eyedropper from the reflection of a 2700K 97 CRI LED off a bright white paper we get this:

Not a very "golden glow"
enter image description here


2700K CRI 97
Measured with a StellarNet BLUE-Wave Spectrometer

The PPFD is a measurement of the number of visible photons in µmols.
So this PPFD measurement shows exactly what wavelengths are being emitted (as a plant "sees") before they are adjusted for photopic luminous efficacy (human perception).

2700K CRI 97 Quantum µmol/m²/s (number of photons)
enter image description here

Slightly lower CRI, more green and less red.

2700K CRI 90 Quantum µmol/m²/s enter image description here


When the number of photons are adjusted for photopic luminous efficacy it is not what most people would expect. This is exactly the same LED being measured (seconds later) with the same spectrometer as the first image with the measurement units flipped from PPFD to Lux.

2700K CRI 97 Photometric Lux
enter image description here

Examples of Low and High CRI

Examples of Low and High CRI apples

Examples of Low and High CRI strawberries

\$\endgroup\$
13
  • \$\begingroup\$ Put the RGB values please just for future sake \$\endgroup\$
    – Ravaal
    Oct 17, 2018 at 2:42
  • \$\begingroup\$ This result will be specific to the color balance of the camera that was used to take that photo. It's not some sort of absolute. \$\endgroup\$
    – user39382
    Oct 17, 2018 at 3:01
  • \$\begingroup\$ That's a very orange light, probably <1800K. Also you might want to scale it so one of the numbers (the red in this case) is FF(255). \$\endgroup\$
    – Spehro Pefhany
    Oct 17, 2018 at 3:02
  • 1
    \$\begingroup\$ @Misunderstood, that is not the color of light that the lightbulb produces ..... a closer match would be if you got the color of a white sheet of paper that has been placed a short distance from the bulb \$\endgroup\$
    – jsotola
    Oct 17, 2018 at 3:13
  • 1
    \$\begingroup\$ Answering myself: yes, it is. hypertextbook.com/facts/1999/AlexanderEng.shtml about 3000k for a tungsten filament. Also (very) helpful: planetpixelemporium.com/tutorialpages/light.html . 40W tungsten is about 255, 197, 143 according to them. \$\endgroup\$
    – RJR
    Oct 17, 2018 at 5:14
2
\$\begingroup\$

do a visual comparison using something like this

enter image description here

\$\endgroup\$
2
\$\begingroup\$

This website has some empirical matching of RGB values to apparent filament temperatures. For a 40W incandescent bulb with a color temperature of 2800K, the suggestion is (255,197,143)

Here is what that looks like (there's quite a bit more blue in it that some other answers):

enter image description here

Here is the table (credit to the above-linked website, above graphic is mine).

enter image description here

\$\endgroup\$
1
  • \$\begingroup\$ This is works well enough but it has issues. The CCT to RGB algorithm often requires red to be 255. The algorithm floating all over the Internet does not take CRI into consideration. With red being the major color factor in CRI and red is fixed in this CCT algorithm it ignores the most important R9 color palette. See "Enter R9": medium.com/@yeutterg/r9-more-important-than-cri-468f0efc0d58 \$\endgroup\$
    – Misunderstood
    Oct 17, 2018 at 18:08
1
\$\begingroup\$

RGB value doesn't have units. It depends on the specific monitor and GPU combination and calibration.

To measure color, you can measure its temperature, use colorimeter or compare it to a known calibrated color such as a Pantone swatch.

\$\endgroup\$

Not the answer you're looking for? Browse other questions tagged or ask your own question.