Which way is "up"? And why?

[Slade]

Whenever I see pictures like the ones on this site, I wonder what choices have been made in how they're presented. How much scope is there for the people working on these photographs to impose their own aesthetic preferences?

For example, who decides which way up they go? Is this a purely arbitrary choice? If not, what does the concept of "right way up" mean in a context like this? The right way up relative to what?

I'd also like to know to what extent the colours shown are "real" By that, I mean, are they the same colours we'd see if viewing the real objects with our own eyes? Does any colour "correction" or sharpening go on before we see the pictures?

Forgive if these questions seem crude, but I am genuinely curious about this. Can anyone enlighten me?

[Orca]

1. Whenever I see pictures like the ones on this site, I wonder what choices have been made in how they're presented. How much scope is there for the people working on these photographs to impose their own aesthetic preferences?

2. For example, who decides which way up they go? Is this a purely arbitrary choice? If not, what does the concept of "right way up" mean in a context like this? The right way up relative to what?

3. I'd also like to know to what extent the colours shown are "real" By that, I mean, are they the same colours we'd see if viewing the real objects with our own eyes? Does any colour "correction" or sharpening go on before we see the pictures?

1. Often the most picturesque images in the visible spectrum, while beautiful and awe-inspiring, aren't the bits scientists really glean the most information from. At the same time, the sense of exploration demands something "tangible," especially for the public (who are after all footing the bill).

I think APOD has always done a great job of including a spectrum of images from the purely idyllic (such as a starscape from Earth) to more informational diagrams and images.

2. Who decided the orientation of the Earth that we are familiar with? Ultimately, no, in space orientation is always relative to some other frame of reference. We just get used to seeing objects presented in certain orientations. The Horsehead Nebula is a good example. Wouldn't "make sense" if that horse's head was upside down, would it!

3. This is a point I find interesting. Human beings tend to assume that what's "real" corresponds to the way they perceive the world around them. Sure, that nebula image is pretty, but what would it look like if I was there? For one thing the human eye can only detect a tiny portion of the electromagnetic spectrum. X-rays and radio waves are every bit as "real" as EM radiation between 400 and 700 nanometers...what is known as "visible light." However, we can't see them on our own and must use arbitrary color schemes to illustrate them. Also, even when we deal with the visible spectrum we often find it useful to filter parts of it out; thus we can view structures that would otherwise be blotted out (for example, cloud formations in Jupiter's atmosphere, or lesser gasses in nebulae such as oxygen). These structures are also "real," we just needed to narrow the band to be able to observe them.

[apodman]

Orientation

If the horizon appears in a sky picture, it is traditional to make the horizon horizontal with the sky above it. Even if the horizon is not in the picture, it is still usually shot this way if it is a picture of a phenomenon that involves the horizon such as sunrise, sunset, crepuscular rays, comet near sun, or zodiacal light.

If the horizon does not appear in a sky picture, orientation is photographer's choice. "North up" to match the sky atlas is a popular choice. "Top up" for a familiar object in the name of a constellation or nebula is another popular choice (the North America Nebula upside down or sideways doesn't look the same, though the Horsehead is vivid enough that the brain can identify it even if it is not right-side-up). Whatever direction puts all the objects you want in the frame is another popular choice.

If you take a film photo, you can rotate it any way you want afterward with the same quality results. If you take a digital photo which is a square grid of pixels, you can only rotate it in whole multiples of a right angle if you want to preserve the integrity of the pixels, in which case you should plan ahead for the photograph's intended display orientation when you shoot.

Color

Some photos are shot in natural color, some are shot with one or more narrow-band filters to capture detail and then composed artificially with near-natural colors, and some are shot with one or more narrow-band filters to capture detail and/or non-visible wavelengths and then composed artificially with false color.

In the case of photos shot in x-ray, infrared, ultraviolet, or any wavelength outside the visible spectrum, false color is obviously the only choice for viewing.

Time exposures always bring out more detail and more vivid colors than the eye will see or a snapshot will capture.

The accuracy of color conversion from wavelength to photographic representation to brain's perception is a vast subject. A search of this forum will locate some introductory commentary for you, while a web search will keep you busy for a long time.

Enhancement

Color correction, sharpening, and other effects are obviously an available option for anyone with a digital picture and a computer. Either the photographer or the camera chooses the brightness and the white balance even before the computer gets involved - is that striving for realistic accuracy or meddling with the facts? Some photographers or artists may tend toward the pure photograph as shot while others may tend toward obtaining the desired effect. Enhancement can be for artistic purposes or to make detail visible.

[Chris Peterson]

For example, who decides which way up they go? Is this a purely arbitrary choice? If not, what does the concept of "right way up" mean in a context like this? The right way up relative to what?

Scientific deep sky images usually follow the convention of north up, east left. Amateur images are often defined by the aesthetic sense of the imager, or by equipment limitations. Amateurs often work with narrow enough fields-of-view that they must rotate the camera carefully just to fit certain objects. Also, many amateurs use cameras with built in guide sensors, and the cameras need to be rotated so as to get a suitable guide star on that sensor. It is unusual for imagers to significantly rotate and crop their images, so we end up with an orthogonal, usually landscape display.

I'd also like to know to what extent the colours shown are "real" By that, I mean, are they the same colours we'd see if viewing the real objects with our own eyes? Does any colour "correction" or sharpening go on before we see the pictures?

There's really no such thing as "real" color for astronomical objects. Nearly all are too dim to show any color to the eye under any conditions. "Color" has a physiological meaning, not a physical one. Images called "real color" are typically made through filters similar to those that you would use for terrestrial photography, so the final image resembles what the eye would see if the objects in it were simply brighter. Even so, there is a large variation simply due to processing techniques and the preferences of the imager.

[Slade]

Thanks very much for that everyone. Now, let me see if I've understood this correctly.

1) The Star Atlas provides a single, objective standard allowing us to know which direction is North, and that direction will point towards the top of the picture.

2) The vast majority of professionally-produced astronomical photographs - no matter how distant the object they show - will adhere to the Star Atlas standard.

3) The colours I see in such professionally-produced photographs are "real" in the sense that they represent wavelengths within humans' normal visual spectrum.

4) Variations in photo-processing technology aside, I can generally assume that the colours have been made brighter than they otherwise would be, but that no other significant tampering has taken place.

Yes?

[Chris Peterson]

1) The Star Atlas provides a single, objective standard allowing us to know which direction is North, and that direction will point towards the top of the picture.

Specifically, north up, east left. But that's just convention, and not even every atlas (to the extent there even is such a thing as an "atlas" anymore) will necessarily adhere to it.

3) The colours I see in such professionally-produced photographs are "real" in the sense that they represent wavelengths within humans' normal visual spectrum.

Approximately. And only if the documentation is clear that "real" color is the intent. Many, maybe most, astronomical images are made with different color maps entirely- such as the Hubble palette- which are still made up only of visible wavelengths, but do not represent anything close to natural color (mapping red, green, and blue to specific narrow bands).

4) Variations in photo-processing technology aside, I can generally assume that the colours have been made brighter than they otherwise would be, but that no other significant tampering has taken place.

No, because there is no color ordinarily. "Brightening" that would just give you lighter shades of gray. The colors are similar to what you would see if the object were brighter, which is very different from saying the colors were made brighter.

[bystander]

Thanks very much for that everyone. Now, let me see if I've understood this correctly.

1) The Star Atlas provides a single, objective standard allowing us to know which direction is North, and that direction will point towards the top of the picture.

2) The vast majority of professionally-produced astronomical photographs - no matter how distant the object they show - will adhere to the Star Atlas standard.

3) The colours I see in such professionally-produced photographs are "real" in the sense that they represent wavelengths within humans' normal visual spectrum.

4) Variations in photo-processing technology aside, I can generally assume that the colours have been made brighter than they otherwise would be, but that no other significant tampering has taken place.

Yes?

I can't agree with #3. Many professional astroimages are false color. Chandra and Spitzer, for example, produce images of spectrums we can't see. Many Hubble images are produced using special narrow band filters and color coding the output from each. One such combination has even become known as the Hubble Palette. It uses red, green, and blue to represent ionized sulfur (S-II), hydrogen (H-α), and oxygen (O-III), respectively. Most false color images will be labeled as such, usually even explaining what the colors represent.