[RCC] Two years in, how do I take the next step to improve my astrophotography? - Forums

I am now over 2 years into this hobby and it has given me a great deal.

Through my images I now have a new role in community education and dark sky protection.

Astrobin has been a tremendous resource for me in my journey, and I really value all the comments and advice I have received from my Astrobin friends.

For me, the biggest issue is how to take the next step. My Astrobin index has plateaued, and top picks/nomination for top picks have dried up. Recently an image (The Helping Hand Nebula) using data I had requested from Roboscopes was awarded IOTD, yet my image using the same data merited not even a nomination.

So clearly, I need to improve.

Recently I have focussed on wide-field, shorter focal length imaging (200mm and 640mm) to better match the poor seeing (4+ arcsec) at my relatively dark site (Bortle 3). I still follow a "standard" post-processing recipe that I learned from Warren Keller's excellent book, mostly using PI [although APP or PTGui for mosaics].

Based on my recent portfolio, I would be really grateful if my Astrobin friends to give me some pointers of where I should focus the most. Perhaps many have experience the same "APpause" before breaking through to the next level.

Some things I have been pondering include:

Better use of equipment (improve collimation, focussing)
Longer integration times (mostly I image for upwards of 7hours/target now)
Improving post-processing (I guess this is always true)
Reverting to long focal lengths
Limiting my images to only the very best conditions
Getting more followers on Astrobin

As ever, I would be very grateful for any advice.

Clear skies

Brian

Hi Brian,

You asked a tough question. Honestly, I don't have answers. Your images are already very good, actually, exceptionally good for a two-yearer. Any further improvement on top of your already very steep growing curve is challenging.

At the same time, I would suggest against using getting IOTD/TP or not to judge your achievements. There are too many factors that can determine the outcome of IOTD/TP selections in addition to the quality of the image. Don't let it take the fun away from you.

All that being said, I do have a feeling that your last two images taken with the 8" Hypergraph are shallower than what I would expect. More than 7 hr of integration is a lot for this aperture and the two relatively bright targets. Maybe it's the light pollution (I doubt) that degrades the efficiency of the exposure, or it is something else that makes these two image not as deep (or smooth) as they should be. This is just for the last two images though. Your other images did not give me a similar (not as deep as it should be) impression.

Cheers,
Wei-Hao

Some of my best images have no awards. If I based their worth on what a few random people I don't even know think is worthy, I would quit this hobby altogether.

Please stop judging the worth of your images by whether or not they are awarded some badge. It is largely outside your control. Judge them by whether they are improved versus your efforts from a year ago and also by how they look compare to images taken by others that you appreciate. Then study the differences between your images and those - is it equipment, processing, integration time or sky quality. That will provide the best direction for improvement.

Wei-Hao Wang:
Hi Brian,

You asked a tough question. Honestly, I don't have answers. Your images are already very good, actually, exceptionally good for a two-yearer. Any further improvement on top of your already very steep growing curve is challenging.

At the same time, I would suggest against using getting IOTD/TP or not to judge your achievements. There are too many factors that can determine the outcome of IOTD/TP selections in addition to the quality of the image. Don't let it take the fun away from you.

All that being said, I do have a feeling that your last two images taken with the 8" Hypergraph are shallower than what I would expect. More than 7 hr of integration is a lot for this aperture and the two relatively bright targets. Maybe it's the light pollution (I doubt) that degrades the efficiency of the exposure, or it is something else that makes these two image not as deep (or smooth) as they should be. This is just for the last two images though. Your other images did not give me a similar (not as deep as it should be) impression.

Cheers,
Wei-Hao

Hi Wei-Hao

Funny you should mention that - but I had the same feeling. I don’t think its light pollution.

There are one or two other thoughts I had.

1) Dewed optics. Its been very frosty here for the past few weeks. Regularly sub-zero in the mornings. I do keep the dew heaters going full blast, and I can’t see anything on the mirrors or corrector. But I needed to disassemble the filter wheel to see if I have anything on the filters. I suspect in this case, I would have an issue with flat fielding.

2) Extinction

We have had a lot of spectacular sunrises/sunsets here lately (e.g. the image of Marariki rising) which is widely attirbuted to the Tongan volcano. Indded, in taking morning flats I noted that the R flat appear to be much ‘quicker’ than the B or G.

On the basis, that it is most likely operator error, I favour 1) but 2) isn’t totally crazy.

Thanks again

Brian

Hi Brian,

Ditch IOTD! You know from your professional life: how many scientists produce results that are worth the nobel prize and how many receive the award? I guess that answer is obvious.
Besides some would argue that some things weren’t adequate for a nobel prize.

Regarding FL: it doesn’t matter. The most mind blowing images that I’ve seen were made with a DSLR and wide-field lens. But that’s a personal matter of taste. And for all focal lengths we can find superb results.

An important aspect: the journey matters. To me it’s not always the results but the journey to a decent image.

…. and fiddling with equipment

CS, Björn

Would have to agree with Arun H. What does the IOTD thing matter? As nice at it is to get recognition for your work; it's still a hobby you do for your own enjoyment, not to collect a badge.

Furthermore, Personally I don't see a value in achieving such a badge if the data comes from remote sources/rigs like in your example. Unless it's an "Edit of the day" or why not download some Hubble data then?... - but, each to their own.

I don't know exactly how the IOTD stuff is done, I believe it's done through nominations and judges, meaning it's still a personal preference thing from the people that nominate it and judge it. That doesn't mean that an image is bad, just because it didn't get nominated. It's not a rating system.

As for what to improve, it's hard to say.
- Bigger projects always seem to do well on Astrobin. Massively long integration, big mosaics, etc.
- Long focal length might be an "ok" idea. As it's done a little less often and a lot focus on wide field, so you would stand out a bit more, perhaps.. - but again, I wouldn't do that to collect badges.. And only do that if you get more enjoyment out of the hobby by going for the "longer focal length targets"

As a beginner I can't give you any advice on improving - I would actually be enthusiastic if I would get even close to your level within 2 years. What I can offer as advice and what helped me a lot to enjoy this hobby even more was to relax a little with respect to my own expectations and to appreciate each small step of progress and the learning experience. This doesn't mean to give up to try improving but to have faith that it will get better over time and with some persistence. Looking at your list I would skip the last three topics. Clear skies can be very rare - maybe you're blessed with better conditions - then even skipping clear nights because conditions could be better doesn't sound like a success story - in particular if you intend to aim at longer integration times. Changing gear towards longer focal lengths is very tempting but basically you throw away a lot of the experience with your current gear and it sort of contradicts the "better use of equipment" point. You're just starting from zero. Personally I committed myself to my current gear for at least one year because I believe using the same gear over and over again is the way to get better.

Clear skies
Wolfgang

Brian, we've been doing this for around the same length of time, and to some extent I know what you mean. I felt I'd plateaued around 6 months ago. So I changed things up a bit, doing and focussing more on post processing in PS on starless SHO layers and then adding RGB stars back in place of the original SHO… and I think it's made a difference and I can now see improvement on the same targets I imaged a year ago… this self observed improvement hasn't really been matched in the top pick department with around a 20 image drought in nominations. I think others have already given good advice to not focus too much on the top picks. The judges and nominators I feel do a very good job and most images I see picked I agree with but there's always some random and some subjective elements to it. And with one's own images it's hard to be objective… I'll often think something I put up is worthy and then nothing and then something I did as a mere filler image with limited integration time gets nominated.

For what it's worth I think some of your recent 200mm images that didn't get nominated are obvious Top Pick quality.

And of course I hanker to try longer focal length too ;-)

Cheers

Rob

I don't want to rain on the parade but if you want some RCC, I think you are being sunk by the technical side of the hobby. That is to say your hypergraph images show some wacky reflections and/or backfocus/tilt/collimation issues. You can clearly process data to the point of getting top picks, so now you just need to take some time and learn more of the other, boring side of the hobby.

No one likes to waste nights playing with allen keys and shims but its gotta be done if you want the best images you can get.

As for top picks and all that, well, the new changes certainly didn't make things any better. I think as paying members we need to demand better of the system.

I see lots of photos I that do not garner any recognition that I would put over many that do garner recognition.
I think many of the IOTD are over saturated for my tastes.
Looking at you pics, one suggestion I would make is star reduction.
The background stars can overwhelm the target.
Check out Pix MLT and google StarNet star reduction. StarNet removes stars altogether. But there is a very cool recipe using StarNet and a mask for reducing the number of stars.

Hi Brian,

the answer to the question "how to improve" really depends on what is your goal, why you do this. And this is something only you can know.

For example, my goal was to get deep, detailed images comparable to SDSS. So I invested my time into learning how my equipment works and fine tuning it. I employ a minimal post processing workflow in order to minimize the amount of artifacts, only essential steps are included. Integration time is at least 20+ hours for most targets. All of this because I simply want to see what is up there and I like to think about an image as of a page from a sky atlas

And now some real advice

: If you want to get an IOTD, then you should definitely improve post processing and choose a target, which is likely to get an IOTD. This is not to say that your images are badly processed! It is just they definitely can be improved with more sophisticated processing methods. In your last image (Twiddlebug Nebula & Aotearoa Dark Nebula) I see dark rings around stars, which are especially visible in IC1283, etc. Looking at the background pixel values I see that R is often zero. R, G, B = 0.00, 0.05, 0.06 looks wrong. I do not believe that your camera received no photons in red channel

BTW I agree with people here who say that IOTD is not an objective measure of image quality. The process is random and subjective.

Hope this helps.

Best regards,
Vitali

Hi Brian, congratulations on your 2 years here on Astrobin, it might be possible that you feel like you have plateaued in that your processing workflow has settled in a "comfort zone" so to speak as @robonrome alluded too. It took more than a year to settle on a flow that I could be comfortable with that fit my skills, but there are also new scripts and processes happening all the time so my workflow seems to constantly evolve. For example the newish Mike Cranfield/ David Payne GHS script as well as the brand new AstroSwell Star De-emphasizer (both in PI) have given my images a new life at least as far as OSC and crappy skies can allow.

As far as awards go let me add a bit of insight. For those that are not familiar, and as far as any of us can tell, the system was set up to be blind, without bias, partiality or prejudice, quite noble intentions. The IOTD/TP process is comprised of 22 Submitters, 23 Reviewers and 8 judges and I believe these numbers have not changed even though the membership of Astrobin has probably more than tripled since you joined. This might go far to explain why your and everyone's awards have dried up, except for perhaps the awards to Amateur Hosting, Dark sites, and data set processors. The Reviewers can only choose from what the Submitters pick and the judges can only choose from what the Submitters give them... and just 2 images are nominated a day... and Random and Subjective are indeed key words, thanks @Vitali

https://welcome.astrobin.com/iotd

The numbers of nominators seems small to accommodate the vast number of entries, yet I understand that the Reviewers are some times overwhelmed as the images add up. The most frustrating aspect for many of us is, unfortunately, the aspect of the blind nomination. The on going evidence confirms that there is a clear advantage to dark skies and great equipment in AP and that advantage is made pretty clear to all of us every day if we look at it, and don't forget processing is king and over saturation is the current queen.

In the end Astrobin is a great space and Salvatore has done a fantastic job with the web site mixing user friendly interchange and community with Astro Photography.

scott

The IOTD process may be subjective (since humans are involved) but not random. And it is a kind of feedback. In contrast with other platforms/social media, we are happy to belong to a specialized community where our fellows are knowledgeable of the difficulties our hobby involve and with different aesthetic criteria than the average Facebook user. That's why this feedback is gold, whatever the objections you might have regarding the IOTD process.

Speaking of feedback, the first two things I notice when I look at your gallery (at the broadband images since I am not familiar with narrowband imaging) are the lack of midtones in some of your broadband images, and the uneven background. Looking more closely, in some other images the stars have processing artifacts (for example the stars in the Lagoon Nebula, and the stars in the Centaurus A closeup).

On the other hand I find the image of Sh2-63 to be splendid.

You are blessed with great skies, a great variety of high-quality equipment, and a pension! I think for two years in the hobby you are doing extremely well. This is my humble opinion, I wish you to be well and to enjoy the hobby! Clear skies!

Many thanks to all that have responded so far. Some very helpful comments and suggestions.

I accept that the IOTD/Top Pick process may be imperfect (like all processes) but I still find it a good indicator of how one's skills are progressing, according to your peers.

Based on the feedback, I think my greatest area for improvement is in post-processing.

1) Lack of mid-tones
2) Processing artifacts around bright stars
3) Uneven background
4) Star reduction

At the risk of prevailing upon my AB friends' wisdom yet further, this brings up a few questions

1) Lack of mid-tones. My normal stretch is done with Histogram or CurvesTransformation in PI. Should I be looking at a Arcsinh stretch?
2) Processing artifacts around bright stars. Is this primarily caused by Deconvolution? I may be a little heavy-handed here, simply because my intrinsic seeing is so poor.
3) Uneven background. I always take sky flats for use during processing and then use AutomaticBackgroundExtractor with a low order (1) to extract any remaining background. Not sure what else I can do. I know may people using DynamicBackgroundExtractor, but I alway find that it can overcorrect - unless one uses a very low order fitting, and then one is almost back toABE. Perhaps the problem is due to the reflections etc. in the Hypergraph (see below)?
4) Star reduction. This has been a constant thorn in my post-processing. I am now using EZ-StarReduction but, after an initial honeymoon period, I am not so enamoured of it. I look forward to trying out some the other suggestions. Like @robonrome, I have moved to star removal (and then star reinsertion) using StarExterminator to improve things. I guess the message here is that there are new tools becoming available all the time.

I also note the comment about better collimation for my Hypergraph. No doubt some images taken with it were out of collimation [I post everything - warts and all] but most we taken with what appeared (to me at least) a well-collimated scope. Including the use of cheshire/autocollimator, artificial star and collimation on real star. Lots of tiny iterative tweaks with an Allen key. One issue is that the Hypergraph arrived with one locking screw for the 1ary mirror lying in the box. I have not been able to reinsert it into the screw hole at the back of the 1ary. Shortlty I will be moving location, and I think an opportunity to take the Hypergrpah apart and re-tune it.

CS Brian

Björn Arnold:
Hi Brian,

Ditch IOTD! You know from your professional life: how many scientists produce results that are worth the nobel prize and how many receive the award? I guess that answer is obvious.
Besides some would argue that some things weren’t adequate for a nobel prize.

Regarding FL: it doesn’t matter. The most mind blowing images that I’ve seen were made with a DSLR and wide-field lens. But that’s a personal matter of taste. And for all focal lengths we can find superb results.

An important aspect: the journey matters. To me it’s not always the results but the journey to a decent image.

…. and fiddling with equipment

CS, Björn

This is such a good anwer IMHO. Copied for truth - kind of nails it for me too. Thanks,

Ian

If you aren't doing this for yourself in this hobby I would quit.. You get hundreds of likes and awards, why do you care?
I know a lot of photographers of all disciplines and unless they are Pros and sell books and Workshops, they do it for enjoyment.
You must or you would not spend the time you did to get your excellent images.

Brian Boyle:
1) Lack of mid-tones. My normal stretch is done with Histogram or CurvesTransformation in PI. Should I be looking at a Arcsinh stretch?
2) Processing artifacts around bright stars. Is this primarily caused by Deconvolution? I may be a little heavy-handed here, simply because my intrinsic seeing is so poor.
3) Uneven background. I always take sky flats for use during processing and then use AutomaticBackgroundExtractor with a low order (1) to extract any remaining background. Not sure what else I can do. I know may people using DynamicBackgroundExtractor, but I alway find that it can overcorrect - unless one uses a very low order fitting, and then one is almost back toABE. Perhaps the problem is due to the reflections etc. in the Hypergraph (see below)?
4) Star reduction. This has been a constant thorn in my post-processing. I am now using EZ-StarReduction but, after an initial honeymoon period, I am not so enamoured of it. I look forward to trying out some the other suggestions. Like @robonrome, I have moved to star removal (and then star reinsertion) using StarExterminator to improve things. I guess the message here is that there are new tools becoming available all the time.

To answer your questions above:

1. You should really use a combination of both, plus Curves. The sequence depends on the subject matter so it may different from time to time but could be construed as HT -> ASINH -> Curves

2. Avoid Deconv if at all possible. If you want to reduce the star's size then use star reduction methods not Deconv.

3. ABE is NOT the way to go in most cases. Use DBE with high density mesh and low threshold deselecting sample points in areas overlapping the subject. I can show how is done if you have a problematic frame to work with.

4. Use sparingly and then only with light hand. Only when the field is really crowded and the subject is drowned out. I use EZSR too and with the right parameters it can be used in most situations. I wish PI had a proper routine for this but I might be hoping in vain.

Brian,

you may give ghsastro – astrophotography a try for stretching. I find it quite powerful although I'm far from reaching the end of the learning curve. In case of questions you get excellent support in this thread Pixinsight image stretching – GHS Version 2 - AstroBin.

Clear skies
Wolfgang

Brian Boyle:
4) Star reduction. This has been a constant thorn in my post-processing. I am now using EZ-StarReduction but, after an initial honeymoon period, I am not so enamoured of it. I look forward to trying out some the other suggestions. Like @robonrome, I have moved to star removal (and then star reinsertion) using StarExterminator to improve things. I guess the message here is that there are new tools becoming available all the time.

Have you tried shorter sub exposure times on the hypergraph? I learned my lesson of star bloat when I did the lagoon last summer doing 600s subs at F/4.72 and have since dropped it down to 180s for dark skies.

When you shoot at 300s at F/2.8 its like 852s at F/4.72. Perhaps your stars will not bloat as much with shorter subs.

Thanks for the amazingly helpful suggestions. I have tried a few out on a re-process of the Lagoon [Its a bit like trying to change your golf swing - you can only change a few things at any time, otherwise it all falls apart].

Overall, I am happier with the result. The big differences are in the star reduction and dynamic range. Still a lot to learn here, but I am enjoying the journey thanks to my AB friends.

CS Brian

Brian Boyle:
Thanks for the amazingly helpful suggestions. I have tried a few out on a re-process of the Lagoon [Its a bit like trying to change your golf swing - you can only change a few things at any time, otherwise it all falls apart].

Overall, I am happier with the result. The big differences are in the star reduction and dynamic range. Still a lot to learn here, but I am enjoying the journey thanks to my AB friends.

CS Brian

Looks great to my eyes. The example showed me that I have the same issues with my mid tones. I was aware of my star bloat problems but not of the mid tone issue. Thank you for sharing, I learned a lot from it.

Clear skies
Wolfgang

Brian Boyle:
Thanks for the amazingly helpful suggestions. I have tried a few out on a re-process of the Lagoon [Its a bit like trying to change your golf swing - you can only change a few things at any time, otherwise it all falls apart].

Overall, I am happier with the result. The big differences are in the star reduction and dynamic range. Still a lot to learn here, but I am enjoying the journey thanks to my AB friends.

CS Brian

Looks great to my eyes. The example showed me that I have the same issues with my mid tones. I was aware of my star bloat problems but not of the mid tone issue. Thank you for sharing, I learned a lot from it.

Clear skies
Wolfgang

Hi Wolfgang,

So glad you found this forum useful. For my Lagoon do-over, I used StarNet to create a starmask and then use HT to change black and mid-points of the stars. I then used EZ Star Reduction to further reduce the stars. I think EZ StarReduction was having issues simply because my stars are so bloated under my conditions of poor seeing. The image of the Lagoon was taken in 6arcsec seeing. As far as the mid-tones are concerned, II have only just scratched the surface with the Asinh stretch in PI - and I am aware there are better versions from AB users. So lots to learn there. For my first attempt, I used a stretch factor of 1.5 and a black point (estimated by the process) of 0.1. This was after first setting black point in HT at 0.5% on the data histogram.

CS Brian

Hi Brian,

I think your images are beautiful and for processing 2 years, outstanding. I have a few suggestions/ideas that my mentor taught me over the last year:
1. Use Adobe Photoshop if you do not already do so in addition to Pixinsight, the two complement each other.
2. I typically do the preprocessing and stretching in Pixinsight, I also use AutoHistogram to do my initial stretch and refine with HT.
3. For narrowband and also broadband, I sometimes, depending on the image, separate the stars and process the RGB/SHO and starless separately and combine using Photoshop blend techniques. I can go into detail if you would like too, on this.
4. Usually I finish the final LRGB image using Camera Raw or tools like Shape or Nik Pro or Tonal contrast.

I found each image has different demands depending on the data, I will try to add a detailed description to my images on how I processed them in case it is of help.

CS! Gowri

Brian, the IOTD process ebbs and flows and it seems (to me) that right now the judges are curating the awards and focusing recognition more on backyard imagers, which occasionally leads to some really wonky picks. But that doesn't matter. The real value of the IOTD process is that it lays out a good list of (mostly) good examples for how to improve your own work. It's hard to give a specific thing that you can do to improve other than to say that when you process an image, spend some time looking at how others have done it; but not for the purpose of copying what they do. Look up close at the quality of the stars. Check out up close how the noise is handled. Examine how (and where) sharpening is applied. And finally, step back to gauge the colors and over all impact of the image. Then put away the samples that you've looked at and try to do it on your own. When you have it just right, close the image and sleep on it for a day. Then go back to your examples and study them again. Then, put them away and open your image. It should hit you like a ton of bricks if you have it wrong. You know the basics so training your eyes and sleeping on it are the best pieces of advice that I can give you. And...if I ever figure out how to get it right every time, I'll be sure to let you know. Just don't hold your breath. Achieving the perfect image is hard...REALLY hard.

John