Level of details in prime focus vs eyepiece images

[PhysicoRaj]

You'll probably want go go from RAW to 16 bit TIF for stacking, 16 bit TIF to 16 bit TIF for histogram stretching and then 16 bit TIF to 8 bit JPG for final display.

I would prefer to process in 16b TIFF, but AS!3 would not process them (or maybe I am doing something wrong), which I have to figure out. Thus, I am working with 16b PNG now. I read that PNG is a lossless compression format so I hope it is not a bad alternative to TIFF?

And what do you think of the CR3 to TIFF conversion - are there different TIFF formats that could store more or less information depending on the way it is converted?

 

[PhysicoRaj]

Update:

As @sophiecentaur suggested, I did some terrestrial shooting and compared images before and after stacking and post processing with the EP view. Here are the results.

Experiment #1:
1) Shoot multiple exposures of a terrestrial target at prime focus in RAW
2) Process ONE of the RAW images to get a stretched and contrasty uncompressed image
3) Digitally magnify the image to nearly the size it is in EP view
4) Compare both the image and live EP view

Result #1: EP view is better than the processed single exposure.

Experiment #2:
1) Stack and align ALL of the raw exposures in AS!3 and obtain an uncompressed image
2) Process the image to stretch histogram and adjust curves
3) Digitally magnify the image to nearly the size it is in EP view
4) Compare both the image and live EP view

Result #2: Stacked and processed image is close enough in detail (or maybe even a bit more detailed) than live EP view!

Here are the images: [please note that images here are not full size but I have zoomed in on the region of interest to understand the difference]

The single exposure processed prime focus image with region of interest for details marked in red:

Single exposure: the red part zoomed in digitally to view the details:

Stacked image: red part zoomed in digitally to view the details:

So, this means, the resolution was enough in the prime focus image, only that a single RAW exposure could not capture the resolvable details. Nor was the movie used to stack the frames able to capture it as the movie was a compressed format.

I would like to hear your thoughts on the results of this experiment.

 

[sophiecentaur]

I did some terrestrial shooting and compared images before and after stacking and post processing

I've read about some terrestrial photographers who stack their pictures but not using a movie sequence. (Image enhancement for reading crooks' car number plates seems to work brilliantly on the Cinema so it may be used more than you'd think) The Seeing, if the the weather is half decent wouldn't benefit from it. I was just suggesting you could find any inadequacies of your basic system (60mm Objective etc.) from pictures of small, contrasty objects. I was clutching at straws, mainly.
You'll just have to wait for a good day and take a dozen or so Raw shots of the Sun and see how much improvement you can get. Your experience just goes to show how good our brains can be at dragging out important details from 'average' quality images.

*But the improvement in apparent resolution you got from stacking was notable on the printing on that sign.

 

[PhysicoRaj]

The Seeing, if the the weather is half decent wouldn't benefit from it. I was just suggesting you could find any inadequacies of your basic system (60mm Objective etc.) from pictures of small, contrasty objects. I was clutching at straws, mainly.

I get it. But it still went ahead to prove that the EP view can be replicated. Not that I have not seen stacked images with improved quality, but nowhere could I find enough evidence that I can see something through the EP and also capture it (without EP projection) at the same quality, particularly after I saw a blatant difference in my sun image.

I could've easily tried this on the Sun, but its 100% overcast here, forecasted to remain the same for the next 3 days .

My main concern was that the details I saw in the EP view were lost either in between the sensor pixels or lost by the capability of the objective optics, but this experiment proved otherwise (correct me if wrong).

Your experience just goes to show how good our brains can be at dragging out important details from 'average' quality images.

It has certainly got my perspective changed on astrophotography. When I started the thread I took the EP view for granted, without giving credits to the brain. Rather, thought that the sensor at prime focus was picking up a really bad image.

I think I suspect my optics much less now thanks to this.

I still need to do the same process to the Sun and compare with the picture in #1 to be sure. And then also work on my mechanical issues in focus and balance, that should help me extract the last drop!

 

[sophiecentaur]

but nowhere could I find enough evidence that I can see something through the EP and also capture it (without EP projection) at the same quality,

This would go against "everyone else's" experience over history and nowadays. Since the earliest photos were taken we have seen so much more than we used to through EPs (me in my back garden too), so it's not just better objective lenses.

I guess it must all be down to your elven ancestry; you're just not like us ordinary mortals.

 

[PhysicoRaj]

This would go against "everyone else's" experience over history and nowadays. Since the earliest photos were taken we have seen so much more than we used to through EPs (me in my back garden too)

That makes a ton of sense and I think I was a bit too worried why it wouldn't work that way with my setup!
Mostly because I have no way to experience what others 'see' with their eyes.

I may have now eliminated some variables by convincing myself of obvious things

 

[sophiecentaur]

I have no way to experience what others 'see' with their eyes.

You do, however, have access to many astrophotographs from home astronomers, using (probably) better optics than you. You can easily observe the differences in image quality. The longer you can put up with the inadequacies of your present gear, the more likely it is that you will either drop the hobby or spend your money in a fruitful direction, once you have decided what to specialise on.
I think I have spent far too much on my kit without having though things through about what I actually want to do. Lock away your wallet for a while, at least!

 

[Devin-M]

You’ll lose most of your quality when you upload to this site. That’s why I host my images on a different site before linking to them here because if you upload them here they will be unceremoniously yet drastically downsized and compressed prior to display by the upload algorithm. Also you won’t see the full details on your images on your computer unless you view them at 100% to see the individual pixels because the screen does not typically have nearly enough detail to show you the full image quality when you can see the entire image on your screen.

Also keep in mind your brain does a lot of post processing to the imagery that comes in from your eyes before you “see” but you usually have to do that same post processing manually with a camera and photo editing software to make it look like what you “saw”… we call that histogram stretching / white balance adjustment / etc and there is a lot of intricacy to that step which may take time and trial error to get better at.

In my experience the camera “sees worse” than the human eye before histogram stretching but the camera “sees better” than the human eye after histogram stretching.

Keep in mind if you’re seeing the full image and it’s photo printed less than 20 inches wide at 300 dpi (your resolution is 24MP - 6000x4000), or if you aren’t zoomed in 100% on the screen where you can see the individual pixels, then you aren’t seeing the full image quality.

 

[Devin-M]

Also keep in mind your brain does a lot of post processing to the imagery that comes in from your eyes before you “see” but you usually have to do that same post processing manually with a camera and photo editing software to make it look like what you “saw”… we call that histogram stretching / white balance adjustment / etc and there is a lot of intricacy to that step which may take time and trial error to get better at.

See this post…

Post in thread 'Our Beautiful Universe - Photos and Videos'
https://www.physicsforums.com/threads/our-beautiful-universe-photos-and-videos.800540/post-6474670

https://www.speakev.com/attachments/2c88c68a-cc18-4865-88c8-327ce5da8699-gif.149749/

 

[PhysicoRaj]

You do, however, have access to many astrophotographs from home astronomers, using (probably) better optics than you. You can easily observe the differences in image quality.

Eyepiece projection, yes I can compare EP projection images from others with my EP view.

The longer you can put up with the inadequacies of your present gear, the more likely it is that you will either drop the hobby or spend your money in a fruitful direction, once you have decided what to specialise on.
I think I have spent far too much on my kit without having though things through about what I actually want to do. Lock away your wallet for a while, at least!

I think I now understand why you were focused on debugging my scope gear. It might not be a factor here but if it is for what I want to do next, that will be a problem.

My scope is a native f/12 700mm. If I am right this mostly suits imaging the Sun, Moon, Jupiter, Saturn and some bright and narrow fov DSO's. That's all. Enough for a beginner like me if it works out.

Among the SSO's I've listed, Saturn will be the ultimate test of adequacy. I will try Saturn in the coming weeks and I can understand my scope's weaknesses better. If it doesn't work out, I will keep it for Solar and Lunar while the camera kit lenses do MW and Andromeda. I will save up for a 'calculated' scope in the future. Wallet is sealed both ways.

See this post…

Post in thread 'Our Beautiful Universe - Photos and Videos'
https://www.physicsforums.com/threads/our-beautiful-universe-photos-and-videos.800540/post-6474670

Seems like hist-stretching is the single most powerful technique in astro-imaging.

What is your Bortle?

 

[Devin-M]

I predict your next major purchase will be an equatorial mount that can support your telescope and camera. That opens up photographing deep sky objects like nebulae and galaxies but you need to be tracking the stars for that because you will have to do 2 minute exposures and the stars will just look like lines across the image with no equatorial mount / tracking.

 

[sophiecentaur]

That opens up photographing deep sky objects like nebulae and galaxies

You have to make a choice between planetary and other subjects, though because a really useful planetary scope needs a narrow field of view, which will not squeeze in a lot of structures like nebulae and asterisms.

Planets tend to be the obvious choice when starting but there are far more other subjects to image if you go for DSOs. Then, if you have any decent lenses on your DSLR, there are big things like the Milky Way (choose the right focal length) which will benefit a lot from stacking and 'levels and curves'.

Enjoy

 

[PhysicoRaj]

I predict your next major purchase will be an equatorial mount that can support your telescope and camera. That opens up photographing deep sky objects like nebulae and galaxies but you need to be tracking the stars for that because you will have to do 2 minute exposures and the stars will just look like lines across the image with no equatorial mount / tracking.

Yes. But still a long way to go. I will have to get a bigger aperture scope for smaller and fainter DSO's. Those trackers are pricey enough and also being able to handle a fat scope will give them a hefty price tag.

You have to make a choice between planetary and other subjects, though because a really useful planetary scope needs a narrow field of view, which will not squeeze in a lot of structures like nebulae and asterisms.

Planets tend to be the obvious choice when starting but there are far more other subjects to image if you go for DSOs. Then, if you have any decent lenses on your DSLR, there are big things like the Milky Way (choose the right focal length) which will benefit a lot from stacking and 'levels and curves'.

Enjoy

If I get decent shots of Saturn (Cassini Division?) with my existing scope I'd be overjoyed.

The Milky Way should be a good target for any short-medium focal length DSLR lenses. The longest I have is 250mm which can fit Andromeda in frame perfectly. Asking for more is asking for a scope as well as a decent tracker / computerized scope mount - my budget would not agree.

I will likely be trying some easy DSO's untracked - learning while I save up.

 

[sophiecentaur]

If I get decent shots of Saturn (Cassini Division?) with my existing scope I'd be overjoyed.

With just 600mm(?) objective, the area of the DSLR sensor covered will likely be quite small. People tend to use Catadiopteric scopes like Schmidt Cassegrain to give themselves an objective focal length well in excess of 1m and small format, high res sensors - followed by all the usual tricks of course.
To get a good idea of image size, use Stellarium (free and easy to use, if you haven't yet come across it) Top right of the screen are buttons to show sizes of objects both in EP and on sensor. In the setup section, you can add your own data if there's not a ready made version of all your dimensions. Gone are the days when you had to do calculations to work out what you could see with all the combinations. Ifaik, it won't show you the sort of pixellation you will get with a coarse sensor but it's still a useful tool.

 

[PhysicoRaj]

With just 600mm(?) objective, the area of the DSLR sensor covered will likely be quite small. People tend to use Catadiopteric scopes like Schmidt Cassegrain to give themselves an objective focal length well in excess of 1m and small format, high res sensors - followed by all the usual tricks of course.
To get a good idea of image size, use Stellarium (free and easy to use, if you haven't yet come across it) Top right of the screen are buttons to show sizes of objects both in EP and on sensor. In the setup section, you can add your own data if there's not a ready made version of all your dimensions. Gone are the days when you had to do calculations to work out what you could see with all the combinations. Ifaik, it won't show you the sort of pixellation you will get with a coarse sensor but it's still a useful tool.

700mm objective and I have a 3x barlow. I heard that people use an 'extender' instead of a barlow, but I want to try with what I have got with the scope and without buy anything new.

I have stellarium, I have my objective and eyepiece loaded there, but I don't yet have my sensor loaded. I am going to enter my sensor details and check. But I think there is still one catch, Stellarium will not do ISO and exposure simulation, which is important and decides whether I only see glare or some rings / surface details.

 

[sophiecentaur]

there is still one catch,

You can sort out stuff like exposure but there's nothing you can do about an image that's too big to fit or so small, it only occupies a dozen pixels. Stellarium is pretty magic for a free package, imo.

 

[PhysicoRaj]

You can sort out stuff like exposure but there's nothing you can do about an image that's too big to fit or so small, it only occupies a dozen pixels. Stellarium is pretty magic for a free package, imo.

I checked my objective with my sensor on Stellarium, and it shows Saturn like a star, with those glare spikes and all which does not really show how much the actual image (disc) size is w.r.t the sensor outline. It does seem like the rings would be visible. This would appear properly on a DSLR when the exposure and gain are set properly. Now how do I measure what fraction of the sensor size / how many pixels the image is?

 

[sophiecentaur]

I checked my objective with my sensor on Stellarium, and it shows Saturn like a star, with those glare spikes and all which does not really show how much the actual image (disc) size is w.r.t the sensor outline.

You can probably get rid of that simulated effect by just zooming in. But the point is made that Saturn is pretty small on your sensor. There is no solution to that, other than a better objective lens, a clearer atmosphere and a higher res sensor - at least twice the focal length of your present scope. But, as I have already commented, this is where you need to decide what subjects you want to look at - or which subjects are best suited to your scope.

Basically, your scope is very 'entry level' and no one would seriously try for astro images on such a scope - for a start, the mount doesn't appear to have tracking.

This leaves your problem reconciling what. you see with what the DSLR records. You have to compare like with like and it's easy to zoom in with photoshop and reveal things that you would never see in the EP. Your brain is making a good job of finding features in the EP image that are 'only just there'. That's been happening for centuries - for example the Martian Canals which you can see on early sketch maps of Mars. For a while, everyone could see them because they expected to find them.

You will get more satisfaction with your scope looking at groups and clusters of stars. Resolving double stars is good sport and so is spotting features in the Orion Nebula. When you have become expert enough in that part of astronomy and you can stand the expense (several hundred quid), then do your homework and choose appropriate gear.

 

[PhysicoRaj]

You can probably get rid of that simulated effect by just zooming in.

I will also lose track of the sensor outline. I can actually see the planet's rings when the sensor outline is pretty much my PC screen, which means if I digitally zoom into full screen on my PC I should be able to see the rings - provided it has been recorded with a good number of pixels in that area. Which is not guaranteed.

But the point is made that Saturn is pretty small on your sensor. There is no solution to that, other than a better objective lens, a clearer atmosphere and a higher res sensor - at least twice the focal length of your present scope. But, as I have already commented, this is where you need to decide what subjects you want to look at - or which subjects are best suited to your scope.

Basically, your scope is very 'entry level' and no one would seriously try for astro images on such a scope - for a start, the mount doesn't appear to have tracking.

Well then, I will keep this for Solar, Lunar and if possible Jovian imaging. Atleast I can resolve the larger moons of the gas giants. Seems I should be satisfied with what I get while I learn.

I will probably stick to DSLR kit lenses for now for astro imaging, like the Milky Way on wide angle and the Andromeda on the telefoto. That should teach me enough.

This leaves your problem reconciling what. you see with what the DSLR records. You have to compare like with like and it's easy to zoom in with photoshop and reveal things that you would never see in the EP. Your brain is making a good job of finding features in the EP image that are 'only just there'. That's been happening for centuries - for example the Martian Canals which you can see on early sketch maps of Mars. For a while, everyone could see them because they expected to find them.

You will get more satisfaction with your scope looking at groups and clusters of stars. Resolving double stars is good sport and so is spotting features in the Orion Nebula. When you have become expert enough in that part of astronomy and you can stand the expense (several hundred quid), then do your homework and choose appropriate gear.

My scope neither has the aperture for deep sky imaging, nor the focal length for planetary
Stellarium says the sensor+objective can perfectly fit Orion, Triangulum, Pleiades, etc. but taking exposures on this mount and aperture will be a pain. More than using the scope, using Stellarium will give me a better idea on what scope I should buy in the future.

Edit:
I dis-assembled the focus rack case on my scope and drilled a hole into the casing, through which I can now screw in a thumbscrew to hold the rack from sliding after focusing.

 

[sophiecentaur]

My scope neither has the aperture for deep sky imaging, nor the focal length for planetary

Which shows you why (well-off) enthusiasts spend loads and loads on their equipment. It's not so that they can look proudly at it and stroke it. Astrophotography is not like HiFi, in that the results you get from spending a lot of money are actually tangible.

But the good news is that 'the other side' of astronomy, namely the facilities for learning about it and appreciating what's out there, are cheaper and much more accessible than they have ever been. Plus, the images you and I are ever likely to produce are not going to be as high quality or as rare as those you can see on the Web.

 

[Devin-M]

My scope is a native f/12 700mm. If I am right this mostly suits imaging the Sun, Moon, Jupiter, Saturn and some bright and narrow fov DSO's. That's all. Enough for a beginner like me if it works out.

Among the SSO's I've listed, Saturn will be the ultimate test of adequacy. I will try Saturn in the coming weeks and I can understand my scope's weaknesses better. If it doesn't work out, I will keep it for Solar and Lunar while the camera kit lenses do MW and Andromeda. I will save up for a 'calculated' scope in the future. Wallet is sealed both ways.

With just 600mm(?) objective, the area of the DSLR sensor covered will likely be quite small. People tend to use Catadiopteric scopes like Schmidt Cassegrain to give themselves an objective focal length well in excess of 1m and small format, high res sensors - followed by all the usual tricks of course.
To get a good idea of image size, use Stellarium (free and easy to use, if you haven't yet come across it) Top right of the screen are buttons to show sizes of objects both in EP and on sensor. In the setup section, you can add your own data if there's not a ready made version of all your dimensions. Gone are the days when you had to do calculations to work out what you could see with all the combinations. Ifaik, it won't show you the sort of pixellation you will get with a coarse sensor but it's still a useful tool.

I was able to capture this after converting just the green channel to monochrome with 20 stacked 16-bit tifs (600mm f/9 1/160th sec 1000iso, shot in RAW format and histogram stretched in Adobe Lightroom). What I used was a 300mm f/4.5 nikon lens + tc-301 nikon 2x teleconverter for effective 600mm f/9 with 1000iso and 1/160th second exposures on a nikon d800 full frame sensor dslr, no tracking, 20 exposures 7 seconds apart in interval shooting mode with 3 second exposure delay after mirror flip up, image upscaled via interpolation, final conversion was to gif format to avoid jpg compression "blockiness" and hosted on separate server to avoid compression by this site's host:

https://www.speakev.com/attachments/saturn_stacked_mono_green2-gif.150147/

https://www.speakev.com/attachments/dsc_3510-2-gif.150133/

https://www.speakev.com/attachments/dsc_3510-gif.150134/

https://www.speakev.com/cdn-cgi/image/format=auto,onerror=redirect,width=1920,height=1920,fit=scale-down/https://www.speakev.com/attachments/saturn_stacked-jpg.150142/

 

[sophiecentaur]

I was able to capture this

Good effort, considering the scope. You have done the business with the software and there's no doubt what the image shows!

The Nikon lens would be expected to be better optically than your low cost scope. If you could manage to do some tracking then you would get much better results. In fact, the first / next thing you could consider would be an equatorial mount. There are many different levels of what you could go for. There are simple tracking mounts which will allow you much longer exposures with light equipment.

I said this before but it's really worth while going for second hand equipment. I can't recall where you live but there are many astro forums all over the world and people are always wanting to upgrade their gear and you can get things at really good prices.

 

[Devin-M]

this was the rig I used, but the equatorial mount / tracker was turned off…

 

[Devin-M]

Thinking about it a bit more why only the green channel was decent quality, I suspect that since my lens is not an apochromatic refractor, you can’t get all the different colors perfectly in focus at the same time. I used a bahtinov mask on the lens during the focusing process so theoretically achieved near ideal focus. Since green is in between red and blue on the color spectrum, it sort of makes since that it would end up being the color most in focus. The magenta halo around the inner green part is presumably the out of focus red and blue wavelengths…

https://www.speakev.com/cdn-cgi/image/format=auto,onerror=redirect,width=1920,height=1920,fit=scale-down/https://www.speakev.com/attachments/saturn_stacked-jpg.150142/

https://www.speakev.com/attachments/saturn_stacked_mono_green2-gif.150147/

 

[sophiecentaur]

Photoshop has a filter for Chromatic Aberration so you could use RG and B channels. That could help the noise. But I think it could be pretty Labour intensive.

 

[Devin-M]

I was thinking for best quality with the equipment it would be necessary to basically take the pictures 3 times and re-focus for each of the 3 color channels (RGB) and then merge them together.

 

[sophiecentaur]

Not a bad idea. But, as they say “you can’t make a silk purse out of a sow’s ear”.

otoh, getting what you can get out of inadequate data is what astronomy is all about. None of us live on a mountain too in Chile.

 

[PhysicoRaj]

Amazing results for the gear you used!

My scope is a 100mm more, but as sophiecentaur said the optics quality is lower than a DSLR lens. I shot the moon with my scope. But I am waiting to have perfect skies for trying Jupiter and Saturn.
The moon at 700mm on my sensor, stacked 85 TIFs and sharpened:

The EP view is 35x magnification and my 'brain' does see more details, but I feel stacking and sharpening is giving me satisfactory details for the optics I have. I particularly think the 60mm aperture is limiting a bit (?)

Even though I will wait a bit more before I extract everything out of my scope and go for a different gear, I will mention here that I'm kind of stuck in between an EQ mount vs a star tracker.

An EQ mount will give me the easier tracking ability (even an un-driven one) and will take either cameras or telescopes. But on the flip-side its comparatively less portable, requires a new scope (can't use the existing one) and I don't know how user-friendly it is to use an 'un-driven' EQ mount??

A star tracker is portable, light and easy to setup and cheaper than a driven EQ mount but will not take a hefty scope. But thinking that the planets are just few in front of the many DSO's, maybe star tracker makes sense?

Maybe the answer to this depends on what my long term objects are? Star tracker + DSLR lenses I think will do fine on most of the nebulae. The planets and galaxies are the ones that require focal length (and aperture for the faint galaxies). If I go for second hand equipment, I could actually first acquire a star-tracker and then try a driven EQ mount. That will let me explore what kind of objects I want to keep shooting long term.

 

[sophiecentaur]

I don't know how user-friendly it is to use an 'un-driven' EQ mount??

There is still one member at my Astro Club who does it all manually. Results pretty fair too. Reminds me of the joke with the punch line "And before you ask, standing up in a hammock"

 

[Devin-M]

When I’m shooting invisible things in the night sky, since I don’t use an equatorial mount that automatically slews to the target, I have a cord that transfers the pictures to my phone and I upload a test shot to http://nova.astrometry.net/upload to see where I’m pointing and then adjust… I was able to capture supernova remnant Simeis 147 a few months ago on my dslr with a composite RGB image and a Ha narrowband image through a clip-in 6nm Ha filter…