When does processing go too far?

[Ags]

At what point does processing of an image cease to be a fair representation of the object and become creative illustration? That is, you are no longer imaging something in the real world, but just making a digital drawing loosely based on some real world data?

For example, I have very bad chromatic aberation around one or two of the bright stars in my images. would it be evil to floodfill the blue halos with sky background? On the one hand it is painting over a portion of the image, on the other hand I am manually correcting for an obvious flaw in my equipment...

[Psychobilly]

Oh goodie we haven't had this debate for a while...

What is the purpose of the image ..? is it Art or Science...?

I take the view that the only person who has to be happy with an image is the person who produces it...

And as long as its their own data they are using anything goes...

Next...

[Ags]

Yes, but the debate has not yet specifically addressed the issue of the removal with extreme prejudice of pernicious chromatic aberation... :-)

[Psychobilly]

Kill it, get rid of it it is the devils spawn...

[ncjunk]

Get rid of it as it is correcting a failure in the equipment to produce a corrected picturr of the object. Add two photos of m42 together so the central area isn't burnt out and i will tar and feather you and then call you a heathen...thats just creating lies...misrepresenting the brightness...nay, tis slander and thou shalt burnnnnn.

I hope that adds to the debate ;-)

[RikM]

Get rid of it. Could you colour select it then drop the levels to match the background? Same result without the dirty feeling...

[brantuk]

Easy - it's prcessed enough when it looks nice to me and I like it. Please note this criteria may be different for other people lol

[Ags]

I've also tried using a green luminance layer on the blue channel to cut away the fringing - but this doesn't always work on all images.

Another wierd thing I've seen is a shadow of the CA in the green channel - noise seems suppressed in the green channel in the area where the blue fringe would be. I would have expected the three color channels to be independent of each other, but that is not what comes out of the camera.

[JimStan]

The whole debate seems rather silly to me when professional imagers take weeks if not months to create images from the Hubble, Spitzer, and Chandra telescopes that are strictly visible representations of the different wavelengths of very restrictive filters, translated into something we can see with our eyes.

You have seen the "great observatory" photos, which are stacked outputs from these very different telescopes into photographic results that we all love and enjoy !

Since our eyes can only see dim whisps of gray , I guess we should not attempt to show anything but monochrome images. Oh, maybe some of you would not appreciate that so much ! Thought so!

And how about the images from the big radio telescopes?

Jim S.

[Earl]

id call it more of an artistic interpretation of the data gathered processed to your personal taste.

When i crack it ill let you know, but I warn you it might be a very long wait

[astromerlin]

Get rid of it. Could you colour select it then drop the levels to match the background? Same result without the dirty feeling...

I used to copy/paste background over unpleasant artifacts but now I have learnt a bit more at using processing software I use the method above to reduce CA. I image for personal pleasure and as long as I'm happy with the result that's all that matters. My idea of a pleasing image is if I have processed out as much of the equipment/operator defects but not introduced too many processing artifacts, and also the enjoyment of looking back over earlier images to see how much I have improved.

[ollypenrice]

I mostly use levels and curves to reduce star bloat. I don't really get CA from Takahashi or TEC but the bright stars do flare. If they're on a nebulous backround I will use clone stamp in Lighten to kill haloes as well.

I'm trying to make beautiful pictures of the night sky. I want them to be informative as well but they are to science what a Constable is to botany. Nothing!

Olly

[rfdesigner]

My view:

Each image has a purpose. If the image fulfulls that purpose then it's good.

So a narrow band image of a nebula will casue 'distortion' by dimming the stars.. but that is desirable to show the nebula. Personally I like to use a log function to compress high dynamic range images, doing this it's possible to show nebula like M42 without any burnt out stars or missing nebula, it's even possible to work out true brightness of each object/area (within the optical band) directly from the image.

What I dislike though is heavy handed unsharp masking as that can leave bright areas of an image darker than other supposedly dim areas.

But if the purpose of the image is to be 'pretty' then frankly anything goes... it's all down to the purpose.

Derek

[Sam]

There's a fine balance between too much and just enough processing. Like others have said everyone's taste is different and it depends if it's for art or science. Personnally I don't care much for the processing that cuts and pastes, dodges and burns or adds artifacts like diffraction spikes - some people like them added, I only like the real ones.

[antimorris]

I vote for doing whatever makes you happy! Personally, I do this hobby as I enjoy it so take the same approach to processing. If your processing/processed version makes you happy then so be it. Get rid of the chromatic aberration!!

[Sparrow]

I'm trying to make beautiful pictures of the night sky. I want them to be informative as well but they are to science what a Constable is to botany. Nothing!

Olly

I couldn't have put it better myself Olly. I think what I, and many are doing with the imaging is close to art but not quite. I will happily use curves, levels, cotrast, colour balance and various tools like that. Where I personally draw the line for my own images would be painting in extra stars for effect or something like that. BUT - if someone else wants to do that for an artistic effect then as long as they are honest about it, I don't see why they shouldn't.

At the end of the day, it's a hobby for personal enjoyment so doing whatever you want with your own data is fine by me as long as when the images are shown on here, there is some honesty about it.

Regards

John

[ollypenrice]

I don't hesitate to consider myself a 'deep sky artist.' I hesitate more than a little to call myself a good one, however!! I'm talking about my intention, not my acheivement.

I spent this afternoon in a superb art exhiition in Aix en Provence with my wife, a professional painter. It has made me think about what I do and what I'd like to do in representing the night sky. Where this thinking will lead I don't know. But...

I was recently asked if I did serious astronomy or did I just take 'pretty pictures.' Now I give fair warning. I intend to tolerate no more of this insulting blather from old school spectroscopists or position angle measurers. I don't insult what they do and they can damned well learn to respect what I do or go somewhere else. Am I narked? Yes, I'm narked.

I work very hard indeed at trying to do justice to the quite incredible beauty of the night sky and the amazing skyscapes it presents. Pretty pictures? Jump in the lake. The next person to use that expression will get a sound exposition of my theory of aesthetics - which might just resemble an ear-bashing!

Sorry, a rant but a rant based on why I spent 55 hours outside at -15C to image Orion.

An unrepentant Olly

[Psychobilly]

Olly, I can understand your frustration... I for one appreciate the fine images that you and your guests produce and the service and facilities you offer...

I hope you wouldn't get the same "treatment" from the members on here...

Peter...

[ncjunk]

Olly,

I suppose it depends how the question was asked. Was it asked out of ignorance (not understanding that it could be deemed insulting) or did the person mean to imply that the pictures contained no merit or were not real astronomy?

I have my views on pictures and processing but as far as i'm concerned each to his own.

In processing each picture to provide a representation, all be it pretty, of the object you gain an insight into its makeup, i.e. HA, the faintness, structure and formation. You have to spend time tweaking each part in order to produce your "Art" and the amount of work involved in taking and processing these subs to produce the end result is nothing short of amazing.

I would suggest they didn't understand what they were saying or the truly don't realise the wealth of astronomical detail and information that can be gained from this "Art".

But then that shows another level of ignorance on the part of the person asking the question.

I'll still never really like an m42 which is a merging of two pics to show the core and fainter detail....i'll still scream it's a lie BUT i'll appreciate the time and effort that has gone in to the picture and applaud the detail. I don't like Manchester united but it doesnt mean they dont play good football or are a bad team.

[barkis]

Well, we have incrediblely beautiful scenic pictures taken of our planet, the mountains, seas, rivers, forests, snow scenes, desert and polar wilderness tracts. They are beautiful because we have the wonderful illumination of these scenes by the star close by.

No one is criticising those images, although no doubt many are tampered with to improve them.

The fact that images of the universe and it's contents need some help, due in most part to very low light levels, in the form of filters, and clever, and intuitive processing, should not diminish the credibility of those images.

Are we to deny ourselves the beauty and magnificence of the night skies, simply on the grounds of it's not cricket. I think not.

You capture the data, then the artist gets to work.

No different to Picasso, Rembrandt, et al.

I have never understood what anyone saw in a Picasso painting, but who wouldn't like to own one.

Science isn't being neglected here, it is being enhanced.

Surely we can't knock it all, simply because our eyes are too inefficient, and the vast distances have reduced the energy so much, that we have to employ tools extra tools to reveal it.

Sorry, bit of a rant really.

Ron.

[ollypenrice]

Oh, perhaps I didn't express my mood too well. Sorry. I don't mean to come across too strongly. Narked is only narked, after all, not incandescent! (For incandescent get me started on Windows Vista...) But let's be clear, the word 'pretty,' to me, is intentionally demeaning. It is to damn with faint praise. There is also a habit, in astronomy, of using 'pretty pictures' to mean scientifically useless imaging. Well, I want that to stop. I've said this several times before but I know dozens of DS imagers and precisely NONE of them goes out at night to take 'pretty pictures.' I want to declare war on that particular bit of astronomical vocabulary and I will never ever let it pass unchallenged again. But let's not get too irate! Maybe I was looking that way.

HDR renditions of M42 are a lie? In what sense? If you take the human eye as the arbiter of reality then yes. But don't forget that the human eye performs a log stretch on all that you see. Is that a lie?

For me, M42 is M42. It is not M42 as defined by my eyes. It has faint outer reaches and it has some OB superiants in the Trapezium. One thing that it certainly does NOT have is a huge white starless blob where the OB supergiants lie. To me, this image (combining three exposure lengths) contains more truth than any one of the three data sets that went into making it. There is a decision to be made. I made the Trapezium as bright as I could without losing resolution. I brightened it a bit following a critique on the French forum making your precise point. So I know what you mean, yes.

Olly

Edit, Ron, good post. There were some Picassos in today's exhibition. They are challenging, but Picasso said he didn't paint what he saw, he painted what he thought. I was looking at one of his distorted portraits of a girl with the eyes 'in the wrong place.' When you looked at one of those eyes, suddenly the whole girl was looking straight at you! Eerie.

[MartinB]

Is removing red eye art or science? I don't suppose the question gets asked very often because pupils and red eye photos are all in our every day experience. The only difference with asto photos is we are presenting something that we just can't see normally and so we want to be looking at a fair representation of the object. That isn't the same as science but it does perhaps require a little more faithful process than would be the case with pure aesthetics. However it all gets very messy, I sometimes add star spikes artificially because a still photo can't really capture the scintillation and sheer dynamic range of stars. So it's not just a matter of what's out there but how we perceive things.

I think if you want to paint in detail which wasn't in the original data then it would be good practice to explain that you have done that. Correcting CA around stars is just the equivalent of red eye correction though.

[ollypenrice]

Good point about spikes, Martin. I'm not a fan but the scinillation, or an impression of it, does come across in spikes especially when they open up the spectrum, as they often do.

Olly

[Ags]

Well, now I feel validated in my CA-zapping. If you're using an ST80 there's a lot of CA to zap :-o

[dph1nm]

Hmm - isn't the 'correct' way to account for CA to degrade the point-spread function in the two best channels to that of the worst ... !

NigelM