Getting "pop" or 3D in images

[LewisM]

How do you make an image pop/stand out and not look flat?

I keep my data linear for as long as possible, and yet it still looks way too flat to me.

Is it still a matter of more data = depth, or is there a trick to give it a feeling of depth?

Here is my total reprocess of NGC 6188 - I kept it linear in PI until basically everything else was done, then STF'd, Histogram Transformed and a small saturation tweak then saved as a TIFF exported to PS for a saturation tweak (and nothing else). And yet it still looks FLAT https://www.astrobin.com/342457/B/.

Also, is it better to save as a TIFF or PNG before putting it into PS? Which retains quality better?

[xelasnave]

Something I did in the old days was to slightly blur parts of the image which seemed to give an interesting perspective to some captures.You can achieve a sence of perspective somehow.
It seems counter intuitive but I suppose it is what I would call an artist approach:D...anyways something perhaps you could experiment with when you have everything perfect..the key is to be very subtle.
alex

[RickS]

It's not a matter of more data or staying linear as long as possible. IMO, it's just a matter of getting everything completely schmicko: colour, contrast, etc. A lot of small improvements add up to an obviously better image. It also takes time, unless you're already some sort of visual artist, to develop a sense of aesthetics. Some would argue I still have a long way to go on this ;)

As Alex mentioned, there are some tricks like blurring background objects and sharpening the foreground but I've never done this with an astro image. I have subtly lightened or saturated the colour of features that I wanted to pop out, but based on a mask extracted from the data (I draw the line at "painting" masks.)



Both formats are lossless so either is good. Both support greater than 8 bits per colour channel too (16 bit in PNG and 16 or 32 bit in TIFF.) If you're going to do serious colour manipulation it is wise to go > 8 bits.

[LewisM]

Thanks Rick - I always work in 16bit TIFF in PS.

[JA]

Hello Lewis,

To me it would be more about image capture choices and capabilities of the equipment & use, rather than image processing, but perhaps you meant for data you'd already captured..... ??? If that's the case then Alex's suggestion of defocusing (selectively in software) the background or selectively sharpening /increasing contrast on the subject is on the money and used in normal photography (maybe not Astro) to create POP, where there has been little ability to control the depth of field to isolate the subject during shooting. Of course controlling the depth of field is not really do-able for infinity focus to isolate a subject, so selectively defocus. Why not? ..... Well it would probably be regarded as not-the-done-thing to alter the data in that way. Not Kosher perhaps.

Out of interest, Selective defocus is used in higher end phones (not that they're any technological goal*** to strive for, albeit that they pack alot in to a small device at increasing levels of quality) in order for a studio portrait look, typical of short to medium telephoto fast lenses used wide open. Some are even using multiple lenses for the same effect.

*** - I should be careful what I say - I've yet to read a comparison on a 40Megapixel Phone Vs Canon DSLR, Who knows?


Best
JA

[Camelopardalis]

Lewis, if you think about it rationally, almost any sense of depth is likely to be false, since with the distances involved any real foreground <> background is a small number of light years, and we don’t have that level of depth perception.

Personally, I find a STF to be similarly an unrealistic function. If you try stretching with HT (for example) you might find you eek a little more contrast out.

[LewisM]

I guess I described it wrong. Not POP so much as not making the darned images look pallid and like a coloured pencil drawing :)

I'm not talking HDR or any of that BS, I just want my images to not look "pushed" for want of a better description. I tried some things on my latest revision of Corona Australis, and it seemed to give me the look I want, but then tried it on Ara's Dragons, and still looks like an impressionist painting.

[troypiggo]

In photography I think the "pop" you're referring to is usually associated with contrast, and "flat" images usually have a lack of contrast. After all other editing, almost last thing before image resize and saving for web etc, maybe try a very slight S curve adjustment.

[OzEclipse]

Lewis

In order to create the impression of more three-dimensionality of an image to a viewer, you need to increase contrast in mid-tones without increasing contrast between the highlights and shadows so that the image doesn't look over worked. So stretch the mid-tones leaving shadows and highlights anchored. Highlights and shadows should of course be appropriately adjusted for white and black point before the mid-tone enhancement.

Joe

[Stonius]

Is that basically a gamma adjustment?

Best

markus

[OzEclipse]

Markus,

Gamma is basically a crude contrast bump although the exact effect of gamma you refer to would depend on which program you are referring to.

Not sure about programs like PI.

In Photoshop there are numerous ways to do it, each a little different in effect.
I like to keep all in my toolbox.


a midtone selection mask using a blendif layer and layer blending
LAHR filtering
ACR (Lightroom or Photoshop) carefully using the clarity slider.


I am currently preparing a new photography workshop scheduled for July on (non-astro) photography.

INSERT SHAMELESS PLUG HERE :D
https://www.photoaccess.org.au/learn/classes/camera-skills-intermediate/

So I don't have time to write a tutorial. If you're coming to SPSP, catch me there during the day and I'll be happy to show you.

regards

Joe

[Paul Haese]

Globalised adjustment to images tend to have many unwanted effects. I use a lot of masking and work on contrast to give the perception of depth to my images. For the most part this works well. I don’t use PI and never likely to either but there are some good tutorials on the use of photoshop for astro imaging. These teach the basic skills which you perfect with experience. I imagine there must be similar with PI. As they said practise makes perfect.

[Stonius]

This may be too broad for your specific enquiry, but the trick to getting a sense of depth into any 2D image relies on monocular depth cues. It's worth reading through these (https://en.wikipedia.org/wiki/Depth_perception). From that you will be able to derive the sorts of things that will help that effect, but it will also assist in everyday terrestrial photography composition.

Cheers

Markus

[gregbradley]

An interesting question. Image processing is a difficult art and even doing it for many years it can still pose regular challenges.

Not every image works out. As Rick said a lot of small things done well add up to a good image. Also given the variable nature of sky, cloud, wind, fog and technical issues these all pose barriers to getting all the raw ingredients. I've always had the opinion with photography in general that you need to take lots of images as some are simply not that great.I find that is true of astrophotography as well. Some images really come together well and others are a pain and will never amount to much. The skills we learn are an attempt to get that keeper rate higher and higher.

I have noticed though that when you get excellent data image processing falls in your lap. Its when things are not 100% ideal that we have problems that need to be solved.

So concentrate on the basics first to capture as close to ideal data as possible.

Then work out your image processing workflow. It tends to be the much the same image after image.

But little things cause issues like a not perfect alignment of images, or the red subs being more bloated than the other colours giving red rings around stars and so it goes on.

Having said that there are tricks you develop along the way to bring life and vitality to an image. Keeping it more luminous, brightish without
too bright and exposing a lot of noise.

There is no reason to not get a pop from Pixinsight but it seems in my limited use of PI that its great for the basic put together of an image but colour processing tweaks is where it seems to be weakest and where Photoshop is so strong.

Perhaps use the strengths of the various softwares to your advantage.

Greg.

[archioptic]

In addition to what everyone else has said I find that a Morphological Transformation with a good star mask can do wonders to help bring some of that detail forward.

[RickS]

Hi Greg,

PI is actually very strong on colour manipulation but, like many things in PI, it's not immediately obvious how to do it ;) A good mask and CurvesTransformation (R, G, B, a*, b* and Hue curves) are all you need. Creating the masks can be tricky but I like to think that my contribution of the ColorMask script has largely solved that problem.

Cheers,
Rick.

[RickS]

It certainly can help to emphasise the main subject, Nathan. Of course, you may be accused of having dull stars... just kidding, Mike :lol: