Skip to content

Astrofotografia II Mystery Cache

Hidden : 7/30/2024
Difficulty:
3 out of 5
Terrain:
2.5 out of 5

Size: Size:   micro (micro)

Join now to view geocache location details. It's free!

Watch

How Geocaching Works

Please note Use of geocaching.com services is subject to the terms and conditions in our disclaimer.

Geocache Description:


PL
Astrofotografia, jako podzbiór fotografii, niesie za sobą wiele wyzwań które zna każdy fotograf. Jest jednak pod wieloma względami unikalna. Pewne szczególne wyzwania wiążą się z fotografowaniem obiektów układu słonecznego, jednak w tym wpisie skupię się na DSO (deep space object) czyli "obiektach głębokiego nieba" takich jak gromady gwiazd, galaktyki oraz mgławice. W obrazowaniu tych obiektów przeszkadzają nam 2 dość oczywiste fakty: są bardzo ciemne i są bardzo daleko. 
Chociaż, jak być może niektórzy wiedzą, istnieją DSO które można zobaczyć gołym okiem lub przez zwykłą lornetkę (zakładając że jesteśmy pod bardzo ciemnym niebem daleko od sztucznego światła i wiemy gdzie patrzeć), to można je policzyć na palcach jednej ręki. W dodatku, w najlepszym razie zobaczymy je jako strzęp mgiełki wyróżniający się od otaczających je gwiazd wyłącznie pewną nieostrością. 
Jak więc osiągnąć efekt taki jak na poniższym zdjęciu? 

EN
Astrophotography, as a subset of photography, carries many of the challenges that every photographer knows. However, it is unique in many ways. Some particular challenges are associated with photographing solar system objects, but in this post I will focus on DSO (deep space objects) such as star clusters, galaxies and nebulae. We are hampered in imaging these objects by 2 rather obvious facts: they are very dark and they are very far away. 
Although, as some may know, there are DSOs that can be seen with the naked eye or through ordinary binoculars (assuming we are under very dark skies far from artificial light and know where to look), they can be counted on one hand. In addition, at best we will see them as a shred of haze distinguished from the surrounding stars only by a certain fuzziness. 
So how do you achieve an effect like the one in the photo below? 

PL
Nie da się oczywiście wszystkich tajników przedstawić w tak krótkim tekście, jednak w miarę oczywiste jest, że wymaga to przede wszystkim bardzo długiego czasu naświetlania. Samo w sobie wiąże się to z pewnymi trudnościami, bo złośliwa Ziemia swoim obrotem sprawia, że jeżeli po prostu wycelujemy aparat w niebo i zrobimy długie zdjęcie, to gwiazdy zobaczymy jako kreski a nie kropki. Zapomnijmy jednak chwilowo o tym problemie. 
Skupmy się za to na innym aspekcie, tj SNR. Z angielskiego "signal to noise ratio" czyli stosunek sygnału do szumu.
Okazuje się, że nawet jeżeli zrobimy (nieporuszone) bardzo długie zdjęcie, to i tak na naszym zdjęciu pojawi się ogromna ilość owego "szumu". Wynika to z wielu rzeczy. Zanieczyszczenia światłem, zniekształceń obrazu przez drgania atmosfery, przelatujących w kadrze satelit, zabrudzeń optyki, a nawet szumów samej matrycy z "bad pixelami" włącznie. W efekcie tylko część fotonów które uda nam się złapać pochodzi z obiektu który próbujemy obrazować, a nawet spośród nich nie wszystkie trafiają dokładnie w to miejsce na matrycy gdzie byśmy chcieli. Spójrzmy na przykładową 5-minutową klatkę przedstawiającą ten sam obiekt co powyżej.

EN
Of course, it is impossible to present all the secrets in such a short text, but it is fairly obvious that, above all, it requires a very long exposure time. This in itself involves some difficulties, because the malignant Earth, with its rotation, makes it so that if you simply point the camera at the sky and take a long picture, you will see the stars as dashes rather than dots. However, let's not think about this problem for now. 
Let's focus on another aspect, i.e. SNR - signal to noise ratio.
It turns out, that even if we take a (unmoved) very long photo, there will still be a huge amount of this "noise" in our picture. This is due to many aspects. Light pollution, image distortion by atmospheric vibrations, satellites flying in the frame, dirt on the optics, and even noise of the sensor itself including "bad pixels". As a result, only a fraction of the photons we manage to catch come from the object we are trying to image, and even among them, not all end up exactly where we want them to be on the sensor. Let's look at an example of a 5-minute frame showing the same object as above.

PL
Część różnic, przede wszystkim bardziej blade kolory, wynika z różnicy w "końcowej obróbce". Jednak widać ewidentnie, że zdjęcie to zawiera przede wszystkim znacznie mniej detali. Żeby zobaczyć to dobrze spójrzmy na centralną część zdjęcia czyli "głowę konia".

EN
Some of the differences, especially the paler colors, are due to the difference in "final post-processing." However, it is evident that this photo contains, above all, much less detail. To see this well let's look at the central part of the photo i.e. the "horsehead".

PL
Nie muszę chyba nikogo przekonywać, że na końcowym zdjęciu widać "nieco" więcej. Zatem wracamy do pytania - jak to osiągnąć? 
Odpowiedź jest stosunkowo prosta: zebrać więcej materiału. Jeżeli czas naświetlania wydłużymy z 5 minut do 5 godzin, to "szum" statystycznie musi się do pewnego stopnia uśrednić. Proste, prawda? Niestety nie do końca. Chociaż teoretycznie nie jest to niemożliwe, to z wielu powodów jest niepraktyczne i bardzo trudne. Nie wyeliminujemy w ten sposób efektów zabrudzonej optyki, szumy matrycy wraz z czasem naświetlania również rosną, w ciągu tak długiego czasu pewnie przeleci nam na niebie samolot lub w kadr wejdzie chmura, a noc nie jest nieskończona. 
Astrofotografowie obchodzą więc ten problem w taki sposób, że wykonują takich (na przykład) 5-minutowych naświetlań wiele, a potem "stackują" je czyli nakładają na siebie, każdy piksel po kolei komponując z odpowiedniego miejsca na wielu zdjęciach poprzez zastosowanie średniej arytmetycznej, mediany lub (najczęściej) bardziej skomplikowanych algorytmów których nie czuję się na siłach wyjaśnić. Jeżeli dodatkowo w międzyczasie poruszymy lekko teleskopem to wyeliminujemy błędy związane z uszkodzeniami matrycy, a obracając aparat o 180 stopni mozemy pozbyć się ze zdjęcia brzydkiego paprocha który akurat wylądował na szkle. Ów proces "stackowania" jest sam w sobie dość skomplikowany. Mamy w końcu dziesiątki lub setki zdjęć, każde może być nieco obrócone, każde może mieć środek w minimalnie innym miejscu, a co gorsza przesunięcie może wynosić na przykład 2.5 piksela a nie równe 2 czy 3. W rzeczywistości wykorzystuje się do tego zatem odpowiednie oprogramowanie które potrafi poradzić sobie z tymi trudnościami, chociaż można sobie z tym poradzić do pewnego stopnia "na piechotę". Ciekawym zagadnieniem jest to jak robiono to przed erą fotografii cyfrowej (a robiono!) ale to temat na zupełnie osobną rozmowę. 
Poniżej załączam serię zdjęć tego samego obiektu o ekstremalnie niskim SNR. Polecam zmierzenie się samemu z problemem na codzień męczącym astrofotografów.

EN
I don't think I need to convince anyone that the final photo shows "a little" more. So, we return to the question - how to achieve this? 
The answer is relatively simple: collect more material. If we increase the exposure time from 5 minutes to 5 hours, the "noise" statistically must average out to some extent. Simple, right? Unfortunately, not quite. Although theoretically not impossible, it is impractical and very difficult for many reasons. We won't eliminate the effects of dirty optics this way, sensor noise also increases with exposure time, during such a long time an airplane will probably fly in the sky or a cloud will enter the frame, and the night is not infinite. 
Thus, astrophotographers circumvent this problem by making such (for example) 5-minute exposures many times, and then "stacking" them. Composing each pixel from the appropriate place of many photos by using the arithmetic mean, median or (most often) more complicated algorithms that I do not feel able to explain. If, in addition,, we move the telescope slightly in the meantime, we eliminate errors associated with damage to the matrix, and by rotating the camera by 180 degrees we can get rid of the ugly fuzz that just landed on the glass. This process of "stacking" is in itself quite complicated. After all, we have dozens or hundreds of photos, of which each may be slightly rotated, each may have the center in a slightly different place and, what's worse, the offset may be, for example, 2.5 pixels and not equal to 2 or 3. In reality, therefore, appropriate software is used, that can cope with these difficulties, although it can be done manually with to some extent. An interesting issue is how this was done before the era of digital photography (and it was done!) but that's a topic for a completely separate conversation. 
Below I attach a series of photos of the same subject with extremely low SNR. I recommend facing the problem that torments astrophotographers on a daily basis yourself.

 

 

PL
Fotografia użyta jako ilustracja jest mojego autorstwa. Przedstawia "mgławicę Koński Łeb" (występującą pod numerem IC434) oraz jej okolice. Fotografia powstała z 75 8-minutowych klatek czyli okrągłych 10 godzin naświetlania w ciągu kilku nocy.

EN
The photograph used as illustration is my own. It shows the "Horsehead nebula" (catalogued as IC434) and its surroundings. The photograph has been created from 75 8-minute frames - exactly 10 hours of exposure over multiple nights.

Additional Hints (Decrypt)

zntarglx

Decryption Key

A|B|C|D|E|F|G|H|I|J|K|L|M
-------------------------
N|O|P|Q|R|S|T|U|V|W|X|Y|Z

(letter above equals below, and vice versa)