The color of the MFMP dogbone and the Lugano hotCat are more complicated than what you would expect from blackbody radiation. Both devices had internal heaters at higher temperature than the surface of the ribbed ceramic convection tube. At visible wavelengths alumina transmits a good percentage of the photons. Thus, the color the eye sees in these devices consists of the sum of reflected light, surface blackbody radiation, and transmitted light from the heater coil filtered by the intervening alumina.
Because of the porosity of the alumina, the ceramic is a reddening (low pass) filter, so the light from the much hotter internal heater coils will be reddened as it is transmitted. I have experimented with up to 5mm of alumina ceramic with a red diode laser, and while there is a lot of scattering, there is also a lot of bulk transmission. So if the heater is at 1200C, and the surface is at 800C, in a dark room, the surface will look a color somewhere in between the surface blackbody color and the blackbody color of the higher temperature heater coil. Then, as Ryan mentions, almost all digital cameras have an infrared cutoff filter, but they do not all have the same cutoff wavelength. Some cameras aim to boost their perceived low-light performance by admitting more of the infrared into the red channel. To make things worse, some cameras have light leak in the blue channel from the infrared that is transmitted (causing purple fringe effects). The cameras that pass more infrared will take a picture that looks more red in this case. Further, the exposure is very important because the red, green, and blue channels individually saturate. If overexposed, the color will change because first the red channel will saturate first from this type of light. At much higher exposure, all channels saturate and the result is white. Saturation actually begins at 10% below full well in the sensor and the recorded signal becomes nonlinear. These effects conspire to make a simply taken picture taken nearly worthless in determination of temperature for this style of device. On Mon, Dec 22, 2014 at 1:37 PM, Jed Rothwell <[email protected]> wrote: > There has been a lot of discussion about these photos published by the > MFMP on Facebook: > > > https://www.facebook.com/MartinFleischmannMemorialProject/posts/881440191886681 > > Look at the lower two photos. They look bright white. However, someone > from the MFPM commented at Facebook that they do not appear bright white to > the human eye. So I asked Ryan Hunt the following: > > > "I have a question about this [attached Facebook] photo from the MFPM > experiment. Someone on Facebook . . . said the camera shows this as white > but to the human eye it is more of a red color. They were emphatic: 'Not. > White. Hot.' > > . . . In other words: Is the color an artifact of the camera, or does it > look this way to the human eye as well?" > > > His response: > > > "The bright appearance was mostly an artifact of the camera detecting IR > in addition to visible. It was a red/orange color, but not nearly as > bright as any of the photos appeared to show it. All the modern cameras > auto adjust so much I didn't know how to make them show what I was seeing." > > > Consumer cameras usually capture colors pretty much the way the human eye > sees them, but not in this case. I guess there is a lot more IR coming out > of this gadget than the objects in a typical family photo at the beach or > campfire. > > I asked him if the orange color resembles the color in Fig. 12 of the > ELFORSK report. His response: > > > "Yeah, I would say it is very similar. I would judge that to be roughly > 800 to 900C on the surface. I found that the amount of ambient lighting > made a big difference to how the camera viewed it. That's why I suggested > we try to make a visible reference at a controlled temperature to compare > to in the photo." > > > My comment: > > I also judge that shade of red to be around 800°C based on this color > scale: > > > http://en.wikipedia.org/wiki/Incandescence#mediaviewer/File:Incandescence_Color.jpg > > - Jed > >
