by Roger N. Clark
The true color of the Trapezium region in the Great Nebula in Orion, M42, is shown to be blue-green, and best described as teal. Digital cameras can record excellent color for the Trapezium. The natural true colors of nebulae can be computed using recently available spectral image cubes and the standard spectral response of the human eye.
The Night Photography Series:
The Color of the Trapezium Region
Human Vision and Color Response with Age
Narrow Band Imaging
References and Further Reading
As we saw in Part 2c of this series, "The Color of Nebulae and Interstellar Dust in the Night Sky," emission nebulae show a beautiful range of natural human visual colors. In this article I show the natural true color of the Trapezium region of the Great Orion nebula, M42. The true color presented here is a direct calculation using spectra of the Trapezium with the standard spectral response of the human eye (presented in Figure 8 of Part 2c, and the technical section below). Then I will compare the color response of digital cameras.
The human visual color ranges from about 4000 Angstroms in the blue (400 nm, 0.4 microns) to about 7200 Angstroms in the deep red (720 nm, 0.72 microns). See "What Wavelength Goes With a Color?" for more information. It turns out that the emission lines from nebulae like those in the Trapezium, have emissions dominated by hydrogen and oxygen, are in the edges of the human eye spectral response. That means the digital camera filter response that tries to mimic the spectral response of the human eye also has the nebulae emission lines at the edges of the filter response.. Thus, a small variation in the filter response can result in a large error in recorded color. In fact, these emission lines provide a more stringent test of a camera's color accuracy in many ways than ordinary color test charts.
The data set used for the trapezium is from the MUSE instrument and data and the M42 scientific results are described in Weilbacher et al. 2015 A&A 528, A114 (arXiv:1507.00006). Specifically, if you want to download the data download the Full Cube, get the improved wavelength sampled data from here, using any of the 3 locations for the 110 GiB FITS file. Warning: you need special tools to work with this data and many of the downloadable FITS programs are compiled for 32-bit systems, so work only with image files less than 2 gigabytes, and will not work with this huge data file.
While a digital camera provides 3 broad spectral bands covering the visible spectrum for each pixel, the MUSE instrument produces a high resolution spectrum covering the visible to infrared spectral range in 5614 wavelengths. Of course, so many bands makes for huge data files. The data file used for this study is 117,183,222,720 bytes (117 GBytes) to make a 1766 x 1476 pixel image, 32 bit floating point per pixel. The true color image derived from this data set is shown in Figure 1. The natural true color is shown to be blue-green, best described as teal.
Next I compare results from digital cameras (more cameras will be added). Figure 1 shows the colors from a Canon 7D Mark II digital camera. The production of the color image was exactly the same as for other recent color astrophotos on this site: convert the raw file from the camera in a raw converter using a standard tone curve, daylight white balance, and Adobe RGB output. The web jpeg image is sRGB color space made from the 16-bit tif Adobe RGB image. The work flow is a standard color-managed work flow. A daylight white balance out of camera jpeg produces almost the same colors. The Canon 7D mark II results show that the camera is producing excellent true color images of nebulae. Compare the 7D II color accuracy to the colors from a Canon 10D in Figure 3a, and a 4-camera comparison in Figure 3b.
The above comparisons show that older model cameras, like the Canon 10D have less accurate color than newer model cameras. The Canon 7D Mark II stands out as the most color accurate camera model of the models tested. If you would like to contribute images from other model and manufacturer cameras and give me permission to publish examples on this web site and magazine articles (e.g. Sky and Telescope, Astronomy, others on color in astrophotography), make a single image at ISO 1600, f/5.6, 4 seconds, with white balance set to daylight in the camera, and focal lengths of at least 300 mm, and send me the raw file with a statement allowing publication.
The MUSE data set is an image cube, 1766 x 1476 pixels, with 5614 wavelengths from 4595.0 Angstroms to 9366.1 Angstroms (0.45950 to 0.93661 microns) with 0.85 Angstroms per channel, 32-bit floating point values. The calibrated data are in units of 10-20 ergs/second/cm2/Angstrom, which enables precise responses to be calculated. For easier handling of such a large data set, the image size was reduced by binning 2x2 to 883 x 738 pixels (the image shown in Figure 1 is the full binned resolution). Further, the data were limited to a maximum of 7200 Angstroms for the human eye response. The MUSE data do not cover all of the blue wavelengths of the human eye response, so the low level dust was extrapolated from 0.4595 to 0.4 microns, and using published spectra of the Trapezium region and published line intensities, e.g. Flather and Osterbrock, 1960, The emission-line spectrum of the Orion Nebula, Astrophysical J. v132, 18F, the significant spectral lines were added into the extrapolated region. Table 1 lists the major emission lines that are responsible for the dominant coloring in the Trapezium. Emission lines in Table 1 at wavelengths less than 4595 Angstroms used the intensity and shape of an emission line in the MUSE data set to synthesize lines in the extrapolated region. For example, the 4363 O III line was synthesized from the 5007 Angstrom line at 10% the intensity. This was done on a per-pixel basis so that variations in emission intensity are reasonably tracked throughout the nebula. An average spectrum of the bright area below right of the 4 Trapezium stars is shown in Figure 4, and with the eye spectral response in Figure 5a and 5b.
Table 1: Emission lines in the Trapezium Wavelength Angstroms microns Origin 4101.0 0.41010 H delta 40% of the H-beta line, 80% of H gamma 4340.5 0.43405 H gamma half of the H-beta line 4363.2 0.43632 O III 10% of 5007 line 4387.9 0.43879 He I 30% of 4472 line 4472.4 0.44724 He I half of the 5876 line 4861.9 0.48619 H beta 4958.8 0.49588 O III 5007.2 0.50072 O III 5755.2 0.57552 N II 5876.0 0.58760 He I 6562.8 0.65628 H alpha
The M42 Trapezium spectrum from Figure 4 is plotted with the relative spectral response of the human eye in Figure 5. Several interesting properties are illustrated in this graph. The hydrogen-alpha and 0.5-micron oxygen lines are near the edges of the eye spectral response functions. Note too that green light subtracts from the red response, so presence of oxygen emission reduces the eye's response to hydrogen-alpha! The position of the emission lines in nebulae mean that a camera filter response must be very precise in order to come close to true color, and digital camera filter response has only recently been refined enough to produce accurate color in emission nebulae post circa 2005. Also apparent is that there are multiple hydrogen and oxygen emission lines contributing to both the eye blue and green response while mostly only hydrogen-alpha contributing to the red response. That alone illustrates why the color must be blue-green.
The spectral response standard for the human eye was based on young people with normal vision. As we age, the lens in our eyes become yellower, meaning blue is absorbed. So the perceived color of the Trapezium through a good-sized telescope will appear greener as we age, losing the blue component. If you have cataract surgery that replaces the yellow lens, the blue-green teal color will likely be seen again.
To illustrate the power of a data set like that analyzed here, examine Figure 6a. The color scheme shows velocity of the hydrogen gas toward (blue) or away from us (yellow to red), with green neutral. This is a way to show the Doppler shift in a color image, and is called a false-color, narrow-band image. The image in Figure 6b is another narrow band image showing different compositions. Images like that in Figure 6b tells us what is causing the colors in natural color images, and what is causing the colors we might see through a telescope. This also illustrates that color is used for other purposes than simply true visual color images, and that both true color and false color images can be beautiful.
Some amateur astronomers have objected to my analyses and hold up as the "gold standard" the amazing color works of professional astronomer David Malin, and of Robert Gendler. To be clear, the following is not meant to take anything away from the amazing images these men produce. Their images are really impressive and stand for themselves. The only question is are they true natural color to be used as standards to compare to other images that are supposedly natural color?
David Malin's web site is here. David Malin describes the filters and film he used for his color astrophotos while at what is now called the Australian Astronomical Observatory, AAO, at this web site. Here are the details regarding the filters used and spectral response of the film and filters. Quoted information is from this web site.
Red:"The red emulsion was 098-04, a code with no technical meaning. It was a late addition to the range and is not shown above, but it had a sensitivity range similar to the 'F' sensitising above. It was thus panchromatic, but a somewhat coarser-grained, more contrasty material than the IIa emulsion, with a sensitivity that extended to almost 700 nm. Used with an RG (red glass) 610 filter, which absorbed UV, blue and green, it gave the R passband, 610 to ~700 nm." Note: a passband turning on at 610 nm is already a longer wavelength than the peak red response of the human eye (see Figure 5b above), thus this passband is dominated by longer wavelengths than the human eye response, and blocks the more than half of the short-wavelength response of the human eye, so is not close to true color.
Green:"The green-light plate was a IIa-D, which was the IIa-O brew to which had been added small amounts of special dyes during manufacture. The dyes extended the spectral sensitivity to about 625 nm, and the plate was used behind a Schott GG 495 filter. This absorbed both blue and UV, leaving the V (green) passband, 495 - 625 nm."
Blue: "The IIa-O plate was exposed behind a 2 mm thick Schott glass filter, type GG 395, which absorbs any light shortward of 395 nm, i.e. the UV. The longward cut-off is the upper end of the emulsion sensitivity itself, about 500 nm, so the B-band recorded in this way was 395 - 500nm."
Summary, it should be obvious that the red channel alone is beyond the human eye red response. The filter starts turning on after the peak in red sensitivity of the human eye. The other filters and film response only approximately cover the blue and green human response.
Robert Gendler put together this amazing image of the Trapezium region using Hubble Space Telescope data, and this image has been cited by amateur astronomers as showing the true colors of the Trapezium. I emailed Robert asking about the colors in the image. Here is what he had to say:
Thank you for the kind feedback about my images. I have no pretense about there being perfectly true color in my images. People should not hold up my images as highly precise representations of true color. Its such a controversial topic especially among amateur astrophotographers. Interesting that even visible light Hubble images are not precise true color images. For many of those images the red channel is 814nm which is actually near IR. That's why many of the galaxies imaged by HST are devoid of HII regions unless they specifically acquire 656nm data."
There are many beautiful color images of the Trapezium region on the internet, whether true color, false color infrared, ultraviolet, or narrow band. But the natural true color response is blue-green or teal. This should not be construed to mean that I am saying all images of the Trapezium should look teal. But if you want to produce natural color images, here is a good reference.
The Trapezium is an easy bright target, requiring only a few seconds of exposure time and is a good test of the accuracy of a digital camera color response.
The advantage of a stock digital camera in astrophotography is that the color balance is close to that of the human eye, and shows compositional differences very well.
References and Further Reading
Clarkvision.com Astrophoto Gallery.
Clarkvision.com Nightscapes Gallery.
The Night Photography Series:
First Published February 18, 2016 Last updated February 28, 2018