As you can see, there aren't even bright enough stars in this field to make natural RC star spikes, that's why the artificial ones looked so out of place.
As you can see, there aren't even bright enough stars in this field to make natural RC star spikes, that's why the artificial ones looked so out of place.
Mike
Yeah - point taken Mike. I think the artificial spikes have their place but perhaps not with this type of image.
Steve, that's absolutely stunning! One of those images that make you go 'oh wow' out loud as they appear on (or should I say 'pop out of') your screen.
Quote:
Originally Posted by strongmanmike
As you can see, there aren't even bright enough stars in this field to make natural RC star spikes, that's why the artificial ones looked so out of place.
Creating "real" fake diffraction spikes is a bit of an art and requires the right tool. StarSpikes Pro in my - admittedly biased (because I got something better ) - opinion isn't it.
The most obvious errors people make is 1. keeping the spikes too sharp (they need a slight blurring consistent with the rest of the image), 2. not treating every source of light (e.g. all other stars) the same and 3. choosing an incorrect compositing mode for blending in the starlight (an art in itself)
More subtle giveaways are incorrect (or non existent - rays shooting out of the stars do not a diffraction pattern make) diffraction patterns for the combination of angular size of the object, focal length, aperture and general expected obstruction profile for the type of scope that's being emulated.
I did a quick modeling of a 12" RC (guessed the focal length : Edit: Doh! just saw F/9) and subsequent re-synthesis of the stars using the resulting point spread function (hope you don't mind Steve!).
The one error I can still see though (as in the original), is that the subtle color shift due to dispersion is wrong; it's diffracting in RGB palette but should have also been in the HST palette like the rest of the image...
Steve, that's absolutely stunning! One of those images that make you go 'oh wow' out loud as they appear on (or should I say 'pop out of') your screen.
Creating "real" fake diffraction spikes is a bit of an art and requires the right tool. StarSpikes Pro in my - admittedly biased (because I got something better ) - opinion isn't it.
The most obvious errors people make is 1. keeping the spikes too sharp (they need a slight blurring consistent with the rest of the image), 2. not treating every source of light (e.g. all other stars) the same and 3. choosing an incorrect compositing mode for blending in the starlight (an art in itself)
More subtle giveaways are incorrect (or non existent - rays shooting out of the stars do not a diffraction pattern make) diffraction patterns for the combination of angular size of the object, focal length, aperture and general expected obstruction profile for the type of scope that's being emulated.
I did a quick modeling of a 12" RC (guessed the focal length : Edit: Doh! just saw F/9) and subsequent re-synthesis of the stars using the resulting point spread function (hope you don't mind Steve!).
The one error I can still see though (as in the original), is that the subtle color shift due to dispersion is wrong; it's diffracting in RGB palette but should have also been in the HST palette like the rest of the image...
Thanks for your comments Ivo. I rarely use Star spikes pro - I think I pressed "buy" too quickly on my computer screen <g>. Your rework is interesting but I'm wondering now whether fake spikes are really worth the trouble.
Very nice Steve but I am the outsider here. I like the version with the spikes. Like most images/ imagers we all have difering likes and dislikes. My opinion of the first image was the spikes highlighted the the center of the image and looked nice, sharp and clean.
Do whatever rocks your own boat, too many experts to please these days so do it for yourself.
Thanks for your comments Ivo. I rarely use Star spikes pro - I think I pressed "buy" too quickly on my computer screen <g>. Your rework is interesting but I'm wondering now whether fake spikes are really worth the trouble.
Yeah, it's a tough one. Personally, I do really appreciate diffraction spikes from my humble Newt as they give an extra visual cue as to the brightness, color and just general prominence/depth of a star. It's like a form of instant (but coarse) photometry. They're also helpful in discerning the properties of close together stars that would otherwise bleed into one in a single elongated star in a refractor.
Plus I grew up with the Hubble images being the absolute holy grail of astronomical imaging.
Most poorly executed fake ones do the opposite for me - they confuse - as they give me false cues; one has a prominent spike, but the other, nearly as bright, doesn't because the imager 'forgot' to put them on.
They are really distracting when they're not exactly as blurry as the rest of your real world image, because your eyes are drawn to their sharpness. You can't help but interpret that as a (false) depth cue - sort of a false depth of field effect with the sharp star being in the foreground ("more in focus").
Quote:
Originally Posted by Stevec35
So what tool do you use?
<spruik>I use my own processing suite (www.startools.org).</spruik>It's got a module called 'Synth' that allows you build a virtual telescope to the desired specs. Then it uses those specs to calculate how the light would refract around the different components in your scope and applies that to every star in the image (after it has performed some basic photometry to detect all stars, their magnitude and their color).
This procedure can be abused to obtain absolutely perfect real looking diffraction spikes, but its main purpose was really to 'augment' the stars in an image so they would look tighter, brighter and more '3D'. The technique is incredibly effective on star clusters.
Hi Steve,
I much prefer this non 'spiked' version.
Definitely!
Never been a fan of artificial spikes.... at all.
Refractors trying to be reflectors
When you first posted this, I though.... nah, what were you thinking.
But then I read you had this software, so, mmm, ok...
But you got a reflector man.... and a lovely RC at that...
Be happy for those rare times when you don't see a spike and revel in the majority of images that do...
Didn't mind the star spikes too much in the original. Starspike pro works well. You just have to blend them in manually on a different layer and apply a bit of gaussian blur if needed so they don't stick out. Like what Ivo did. Then it looks natural. It's a great shot with or without.
Thanks for the comments guys. I guess that star spikes are a somewhat controversial issue. Marcus, I await your Tarantula with great interest. I'm sure it will be a beauty.
Fabulous image. Such incredible central detail and those bright central stars are all individually differentiated. Terrific. I like the star spikes myself but agree they could do with subtle gaussian blur to match the seeing in the image.
Some of your stars got cut off by something during processing. A minor thing though. Alignment error?
Ooooooh ...don't let us average 6"ers down now Marcus, Steve's is pretty high res ...bah! you can do it
MIke
ps Steve...he hasn't gotta chance (err?..do you think Marcus heard that )
I think he'll do great but the RCOS maybe up for sale if a 6" produces equivalent resolution
Quote:
Originally Posted by gregbradley
Hi Steve,
Fabulous image. Such incredible central detail and those bright central stars are all individually differentiated. Terrific. I like the star spikes myself but agree they could do with subtle gaussian blur to match the seeing in the image.
Some of your stars got cut off by something during processing. A minor thing though. Alignment error?
Overall a terrific image and very high impact.
Greg.
Thanks Greg. Yeah I think there's a bit of alignment error in there which requires investigation. It doesn't show up all the time though and what with clear nights being so few and far between I'm reluctant to spend the time tracking it down.
) - opinion isn't it.
: Edit: Doh! just saw F/9) and subsequent re-synthesis of the stars using the resulting point spread function (hope you don't mind Steve!).
Too true Doug....
)
