BACKGROUND SKY LEVEL

14 replies596 views
Background sky level
Multiple choice poll 23 votes
35% (8 votes)
39% (9 votes)
26% (6 votes)
You must be logged in to vote in this poll.
Riccardo Civati avatar
Good morning,
I'm curious to know how far you stretch your images. What ADU values do you achieve for the sky background? (normalized from 0 to 1).

In my latest images, I stopped at about 0.07 on a "real" sky background point, which I used as a reference.

Please note that I don't clip dark pixels.
Well Written Respectful
Norman Hey avatar
Like a lot of a-p:  it depends. I guess for my tastes and most of my data, somewhere closer to 0,1 than 0,2.
andrea tasselli avatar
It depends on the sky background and how faint the subject is (or the faintest portion of the subject you want to show).
Well Written
John Hayes avatar
It depends on the object and I covered this in my “Galaxy Studio” presentation on TAIC a couple of weeks ago.  For galaxies, 0.06 to 0.07, which is 15-17 for an 8-bit color channel, is about right.  Regions with a lot of nebula will be higher but again that depends on the object.

John
Riccardo Civati avatar
John Hayes:
It depends on the object and I covered this in my “Galaxy Studio” presentation on TAIC a couple of weeks ago.  For galaxies, 0.06 to 0.07, which is 15-17 for an 8-bit color channel, is about right.  Regions with a lot of nebula will be higher but again that depends on the object.

John

Thanks, John!
It would be interesting to read your article! Can I find it anywhere?

Riccardo
Well Written Respectful
Scott Badger avatar
I usually go a bit higher, 0.08-0.09, but just taste. Though, maybe monitor quality/type/calibration is a factor?

Cheers,
Scott
John Hayes avatar
Riccardo Civati:
John Hayes:
It depends on the object and I covered this in my “Galaxy Studio” presentation on TAIC a couple of weeks ago.  For galaxies, 0.06 to 0.07, which is 15-17 for an 8-bit color channel, is about right.  Regions with a lot of nebula will be higher but again that depends on the object.

John

Thanks, John!
It would be interesting to read your article! Can I find it anywhere?

Riccardo

https://www.youtube.com/watch?v=1XBon7x6kio
Sam Badcock avatar
This has got me interested now!! Never knew people took this factor into account when processing!! 

Always just assumed it was a case of “taste” 

May have to pay attention more to Adam Blocks video when he talks about the sky background level!! May have to start considering this factor in my processing
Riccardo Civati avatar
John Hayes:
Riccardo Civati:
John Hayes:
It depends on the object and I covered this in my “Galaxy Studio” presentation on TAIC a couple of weeks ago.  For galaxies, 0.06 to 0.07, which is 15-17 for an 8-bit color channel, is about right.  Regions with a lot of nebula will be higher but again that depends on the object.

John

Thanks, John!
It would be interesting to read your article! Can I find it anywhere?

Riccardo

https://www.youtube.com/watch?v=1XBon7x6kio

thanks John!
your video is very much interesting. From now on I will also follow the TAIC videos

Riccardo
Brian Diaz avatar

hi

around 16/255=0.06

CS

Brian

Alan Brunelle avatar
I guess I cannot dissagree with anyone wanting a metric to decide what "sky" background is "appropriate".  However, consider that this is an art and eventually one should find and do what they wish.  I remember when I started out doing this and posting here, I would get some comments that my sky background was too dark.  I took that to lesson I suppose.  I present no image that contains pixels valued at zero.  But back when I started, I saw much too many images with clearly milky backgrounds.  Much too high in my opinion.  I think more and more, the artists here are presenting deep space galactic images with darker backgrounds and that to me shows a better image.  Still some stick to their guns and present milky backgrounds.  For what ethic, who knows?

I personally take an attitude that I am not photographing object millions of lightyears away as sky images.  As art, I am trying to present these objects as objects in deep space.  Unless I am unaware of a human's vision is capable of seeing the Cosmic Microwave Background, space above our atmosphere is essentially perfectly black, except where out visual line intersects a light emitting object.  Our suffering through photographing through Earth's atmosphere need not be applied to my photos!  If metrics were important or follow some photographer's honor system, would I need to present my Bortle 4 skies with a higher "sky" background for my image of the Virgo Cluster at some higher level as for my M31 taken in Bortle 1?  Nonsense!  So then what is the artistic reason for setting an objective metric?  That said, I do think that a "Space" background of not too dark a background dimishes what can be a harsh appearance and tough on the eyes.  But that certainly depends on the amount of stretch and how bright/dynamic an image is or how dynamic an image the artist wishes to present.  But if one is practicing your art for purely personal reasons, then make it what you want.  If practicing your art for the broader public and commercial success, you will need to learn your customers' wants.  

In my personal images that I present as A ExtraGalactic Traveler's View of whatever galaxy I am imaging, I go one step further.  Not only do I try not to mimic anything like a sky glow in my images, I actually remove any Milky Way stars, planetary nebulae, other features, as well.  Reasoning for the purpose of the presentations are that Milky Way stars are only relevant to how a distant galaxy view appears to those stuck firmly on Terra Firma.  Secondarily, removing Milky Way stars allows the viewer a better understanding of the structure of the target galaxy(s), the myriad of distant galaxies, galaxy clusters, quasars, and even stars for galaxies such as M31, M33, etc, which are close enough to see stars within those galaxies.  See example here.  My backgrounds for these images I want dark, but not lose background objects, because I want to be sure that the very faintest of the distant galaxies can be clearly seen, even if the viewer needs to zoom in to a larger degree.  Another example.  Having a noticable background gives the feeling to the observer that they are not missing anything.  (Fact is, since how deeply can be seen is limited by the conditions and optics.)
David Néel avatar
Ppl with hdr screen will go less than 0.1 (even that on regular screen is strong dark)
Ppl with standard screen will go around 0.15/0.20. (faint dark to Grey)
Scott Badger avatar
Alan Brunelle:
I guess I cannot dissagree with anyone wanting a metric to decide what "sky" background is "appropriate".  However, consider that this is an art and eventually one should find and do what they wish.  I remember when I started out doing this and posting here, I would get some comments that my sky background was too dark.  I took that to lesson I suppose.  I present no image that contains pixels valued at zero.  But back when I started, I saw much too many images with clearly milky backgrounds.

I don’t think anyone is necessarily *proscribing* background levels, just reporting where their images are typically at. I agree, though, that the ‘space isn’t black’ mantra is often pushed too far. Is the darkness of our own sky a factor? Do darker images look more ‘natural’ too me because I live in a bortle 3 area?

Also agree that the monitor is a factor. I don’t have an HDR monitor, but still go a bit below 0.1. At .06, as John Hayes and a couple others have mentioned, then the background is going to start looking clipped on my monitor.

Cheers,
 Scott
Alan Brunelle avatar
Scott Badger:
I don’t think anyone is necessarily *proscribing* background levels, just reporting where their images are typically at. I agree, though, that the ‘space isn’t black’ mantra is often pushed too far. Is the darkness our own sky a factor? D


I think the question was a very good one and did not mean to imply that any answers in this thread are incorrect.  I was referring to what your second sentence points too, and that if asked, it is not uncommon to get basically a rote answer that it should never look black, or something similar.  As I said, my experience is that was more common in the past and I would get those comments unsolicited.  The question in my mind is what is the reasoning behind any answers?  I believe that what the sky looks like in the area it is shot is irrelevant.  (My answer has exceptions in certain cases, such as landscape astro work!)

Your other point about how one should present the image is, in my opinion, the biggest objective constraint on this issue.  You've made it twice.  It was late for me so I have not watched John Hayes' video and he may have addressed this.  I intend to watch his presentation.  But if the image is intended to be viewed directly from monitor, is one thing, but what if printed?  Or to be viewed by any reflected light?  I have returned many a print because of this issue. But I think that is a different question. If we restrict the issue to how an image looks on a monitor, there is a huge range of how any chosen background can look depending on local ambient lighting.  I'll bet John covered that in spades in his talk.  But a lot of the onus has to be on the viewer in all situations. 

When I go back and critique my own images, I find my tastes change. But if I have one complaint, it is that my primary galaxy targets are almost too bright.  Less a background issue.  Most evidence is that galaxies are typically ethereal in nature. One sees evidence of this when finding background galaxies coming through the disk of a large spiral, even through a spiral arm. Sometimes my galaxies look like a plastic Frisbee floating in the air!  The same issue with Mikky Way nebula data and I only do OSC!  Now I'm back to the problem of how many of the different presentations do I want to publish from the same data, all of which seem to have merit?  Choices choices...
Tony Gondola avatar
I would also point out that targeting a background level as a limit to stretching is a bit different from John's recommendation in adjusting the sky background of final 8 bit render. Both are aesthetic choices depending on a myriad of factors.