What is Astrophotography?
Before starting to read the book 'Photographing the Invisible' by Vicent Peris, as an exercise I began to describe in my own words what astrophotography is for me, firstly because it was something that I had never stopped to do, and secondly, to establish a posteriori what influence the book had on my vision. What follows is nothing more than the result of that exercise.
We are limited by our perception: our eyes are sensitive to a small window of the electromagnetic spectrum, and only if the light they receive has enough intensity. Technology allows us to record light beyond the limitations of our senses and thanks to this we can capture the universe in images.
The purpose of astrophotography is to transform the data generated from the capture of light into an image adapted to our visual ability. It can be understood as an information visualization process, in which the astrophotographer determines the processing to apply on the raw data resulting from the capture, to obtain an image that efficiently displays the information using the visual elements that we perceive best: contrasts of luminosity (textures and shapes) and color. It is in this process where we have room for subjectivity and creativity.
Since this visualization process has a subjective component, astrophotography cannot be considered scientific imaging, although it uses scientific resources and techniques throughout the creation process. For this reason, it is a fundamental and ideal tool for science communication.
Astrophotography must obey ethical criteria, otherwise, the final representation could give a false perception of the object represented. Although the ethical limits can be diffuse and subjective, the basic criteria is summarized in not inventing, distorting or destroying information, which would separate the image from representing reality. These ethical criteria must be kept in mind at each step of the processing. In any case, the best option to avoid confusion is to accompany the image with an explanatory text. All results have their place as artistic work (even unrealistic compositions, nebulae with stars removed, etc.), but for an image of an astronomical object to be considered astrophotography, from my point of view, it must be true to the nature of the represented object.
The process starts from the capture of light from the object with instrumental resources (cameras, optics, filters, etc.). With this we obtain the raw data that, within the physical limitations and our instrumental limitations, contain information about the object. In this context, in the raw data we have signal (object information) and noise (defects).
These raw data must be distilled through a calibration process, whose objective is to correct as far as possible the instrumental defects (sensor irregularities, non-uniform illumination, etc.). There are defects that cannot be corrected and generate missing values (e.g., permanently saturated or dead pixels).
During the capture, noise is inevitably added to the data, from natural or artificial origin but unrelated to our instruments (background brightness, photonic noise, cosmic rays, satellite traces, passage of high clouds, etc.) that must be treated through processes of reduction (alignment / registering, stacking / coaddition / integration, etc.) that statistically combine the captured data, generally with rejection of extreme values or outliers. These processes also try to correct any missing values that may exist after calibration. Sometimes, it is not possible to avoid them completely and it is necessary to impute them, that is, to extrapolate them in a coherent way with the rest of the data that we have: it is a purely cosmetic correction, to prevent these defects from becoming protagonists in the final image.
These pre-processes, although adjustable by the astrophotographer, require experience, but in practice leave little room for creativity. Their objective is to recover all the latent information in the raw data that has been captured. Sometimes, it is not even the astrophotographer who is in charge of the capture, calibration or even reduction; nowadays we have public archives of data captured by scientific teams that operate professional telescopes. This emphasizes the high degree of collaboration that can take place to produce an astrophotography, even involving people who do not know each other or have structured as a team: from the person requesting the observation in a scientific project to the astrophotographer who produces the final image, somehow involver many people: those who operate the instruments, those who built them, those who keep them operational, those who developed the algorithms and the software for each step, who manage and curate the data archives, etc. In this regard, even the astrophotographer who individually performs the entire process from capture to production needs a whole arsenal of optics, hardware and software, all of huge technical complexity, which have generally been developed by other people, so it seems difficult to justify individualisms in astrophotography.
This representation of the object as latent information has to go through a pipeline or computational graph of mathematical transformations built by the astrophotographer to finally obtain a new representation of the information of the object that we can interpret with our senses and efficiently uses our perception abilities (contrasts, shapes, colors), complying with ethical criteria (do not invent, distort or destroy latent information). The representation we work with is purely numerical, abstract: we constantly need to render data previews to guide us in the processing.
In this sense, the image processing applied in the pipeline is intended to visualize the information, and this process can be interpreted as a projection of the latent information of the object in the space corresponding to our perceptual ability, always trying to preserve the maximum information that is of interest about the represented object. We usually work with data from different bands of the spectrum, so this projection often involves a reduction in dimensionality. The efficient use of our visual perception ability implies that this projection must try to use the entire range of luminosities and colors, within what is possible or ethically admissible, so almost invariably the processing will include an expansion or amplification of the captured data.
In the case of planetary astrophotography, the objects are mostly luminous and we detect them with the naked eye, and also normally we capture approximately the same spectral range that we perceive with the eye; This type of astrophotography is characterized above all by trying to fight against atmospheric blurring or seeing, however, fundamentally the same ethical and aesthetic criteria are applied as in deep space astrophotography.
Vicent Peris's book adds another dimension that has become a revelation for me: astrophotography is an artistic expression of its own merit. In retrospect, being absorbed with the technical details of this complex discipline and not having artistic training, it was easy for me to ignore that when I’m looking for a particular framing of an object, enhancing textures or adjusting the color palette looking for a specific rendering, I am not only applying subjective or creative criteria, as I wrote in my notes; I am making art. Astrophotography is art. Astrophotographs are works of art. To be fully aware of this, Vicent had to arrive with his vision as an artist, and give astrophotography its philosophical, ethical and aesthetic basis.