What about all those AI photo descriptions?

Hi Franco, I agree with you, I quite enjoy reading a write-up of the acquisition, editing, problems that were encountered along the way etc. In a way it's like I'm reading part of the person's journey within this hobby/profession and it gives me a chance to learn about equipment that I don't currently have or techniques that I haven't tried yet.

Astrophotography is quite a difficult and niche activity and in taking the images, we know what we're imaging (roughly speaking) so when published onto a platform like this to similar people, the chances are that they too roughly know what the image is. For example, I'd be surprised if a user on here came across an image of the Horsehead Nebula and didn't know that "The Horsehead Nebula (also known as Barnard 33 or B33) is a small dark nebula in the constellation Orion" (taken from Wikipedia), so putting that in the description seems a bit redundant to me.

But I suppose that's just my opinion. I get some people might use their gallery on here as a means of sharing their images with others that perhaps don't know as much about these objects as the average astrophotographer might do.

I have often wondered what to do with the description box. Do I add details for the AP community or for the general public? Maybe we need two text boxes, one for object information and another for technical comments.

Tony Gondola:
Maybe we need two text boxes, one for object information and another for technical comments.

That is a good idea. I tend more toward presenting object information but some folks (Kevin Morefield comes to mind) often present lots of processing detail. Depending on the interests of the viewer, they might be more interested in one or the other so having separate sections would make things easier.

Dan R:
Hi Franco, I agree with you, I quite enjoy reading a write-up of the acquisition, editing, problems that were encountered along the way etc. In a way it's like I'm reading part of the person's journey within this hobby/profession and it gives me a chance to learn about equipment that I don't currently have or techniques that I haven't tried yet.

Astrophotography is quite a difficult and niche activity and in taking the images, we know what we're imaging (roughly speaking) so when published onto a platform like this to similar people, the chances are that they too roughly know what the image is. For example, I'd be surprised if a user on here came across an image of the Horsehead Nebula and didn't know that "The Horsehead Nebula (also known as Barnard 33 or B33) is a small dark nebula in the constellation Orion" (taken from Wikipedia), so putting that in the description seems a bit redundant to me.

But I suppose that's just my opinion. I get some people might use their gallery on here as a means of sharing their images with others that perhaps don't know as much about these objects as the average astrophotographer might do.

Totally agree with the ´journey´aspect - literally at the moment -  as I am using a travel rig in Peru right now, and considering the change from OSC to mono. I like to add processing notes that will be useful for me (and others) in years to come both for planetary and DSO work. No detailed workflow, but just a few notes as reminder about a trick or process I used at the time. I´ve also found a few inconsistencies in the literature about nomenclature and magnitude - so I try and point these out to.

AstroShed:
I don’t really think using AI is an issue for writing the image description, I am more concerned about the fast pace of AI processing tools, that will make any cheap telescope with not so great optics, produce “image of the day” quality images, is this not a worry to you guys..?
more than just using to write the scientific details of an object..
Lol, maybe it’s just me 😂

It's an interesting question really. There is no question in my mind that we will very soon have software that can correct for all common optical aberrations that can be mathematically described. No optical system is perfect so I really see that as a positive as long as the software isn't creating detail that would never have been there in the first place. Now if you've invested in a very expensive OTA, or even worse, you are a producer of the product, you're probably not going to be very happy. On the other hand, if you own an average OTA, you'll be very supportive of the technology. It will level the playing field a lot, that's for sure. I do rather like the idea that the best images in the future will be the result of careful planning, tasteful composition and processing rather than just the result of how much you can afford.

AstroShed:

Tony Gondola:

AstroShed:
I don’t really think using AI is an issue for writing the image description, I am more concerned about the fast pace of AI processing tools, that will make any cheap telescope with not so great optics, produce “image of the day” quality images, is this not a worry to you guys..?
more than just using to write the scientific details of an object..
Lol, maybe it’s just me 😂

It's an interesting question really. There is no question in my mind that we will very soon have software that can correct for all common optical aberrations that can be mathematically described. No optical system is perfect so I really see that as a positive as long as the software isn't creating detail that would never have been there in the first place. Now if you've invested in a very expensive OTA, or even worse, you are a producer of the product, you're probably not going to be very happy. On the other hand, if you own an average OTA, you'll be very supportive of the technology. It will level the playing field a lot, that's for sure. I do rather like the idea that the best images in the future will be the result of careful planning, tasteful composition and processing rather than just the result of how much you can afford.

Yes, it all really depends on how far the AI will go in correcting for optical errors, will Autoguiding be needed at all, no bother correcting for tilt, backspacing will be no issue at all, close enough will do, does that not take some of the skills out of it all…I don’t know which way I lean really on this, I have a modest set up, probably about £7000 worth, and i spent a long time getting it all dialled in as best I can, then BX came a long and I felt I had wasted hours of adjusting and dialing in the tilt and backspacing issues I had, also any slight guiding errors, no problem BX will sort them…where does it end..🤷🏼‍♂️
But certain there will be many people loving the AI for much more budget set ups.
what will manufactures do now, will they even bother producing top notch optics, or like WO just tell people to use BX

It's getting easier all the time and I think that trend will continue. Look at smart telescopes, some of those images with external processing, what they lack is mostly due to choices made on the hardware design end, small apertures, no field de-rotation. I suspect that the real skill will shift out of the capture end and into the processing end. I'm with you, I love to tinker and tweak to get the best possible results, I think a lot of us here do. Photography used to involve glass plates and manual guiding for hours and hours in a freezing observatory. I would wager that the images we get today are better than those and push the limits far beyond anything thought possible 100 years ago. Maybe we need to just realize that it's a new frontier and see what all this new technology makes possible.

Franco Grimoldi:
Well...

My concern is not really about the AI-ness of the description but the fact that it's copy-pasted from somewhere else, with no apparent personal input or added value. FWIW, it could be verbatim from Wikipedia, same effect.

Personally, every time I see such descriptions I just skip them, and I tend to comment less on those images as well. On the contrary, a personal note encourages interaction and discussion.

I found this to be a fine subject for discussion, until that is, the first statement that referred to the IOTD.  The question posed here was about the Description text, and not about the image whatsoever!  The IOTD topic, regardless of nuances stated here have been beat to death on many other Forum threads, some as recent as last week.  Referring to the IOTD in a Forum almost always devolves into negative territory, regardless of how well meaning the original question was.  So I hope this thread can avoid that.

I agee with what everyone here has stated regarding the topic at hand.  Personally, I think it is a nice gesture from any poster of images to add something about the natural history of the object being posted.  Even if it is a cut and paste from Wiki or other.  Everyone should recognize that not everyone here on AstroBin are astrophysicists and capable of writing detailed descriptions.  Also not everyone here cares, either.  But for those interested, as I said, it is a nice gesture.  Understand, that not everyone that sees your image is an astrophotographer, but may be seeking visual understanding of what an object is.  This point covers a natural history type of Description.

For the astrophotography details, again, I think it is a nice gesture.  However, understand that not everyone cares about how the image is acquired or processed.  Not everyone who will see it is an astrophotographer.  Also, note that much of the information that can be stated in a Description is included already in other sections, if the poster chooses to do so.  I think acquisition is better handled in AstroBin than processing details.  However, as an astrophotographer, I often really do not care about the processing details so much.  But sometimes I do!  If a poster chooses not to share those details, I respect that.  (Recently I saw a Forum thread that centered on the fact that one or more individuals failed to fully disclose how they achieved their result!)  As for the natural history Descriptions, I feel that photographic details are also a nice gesture and require no essential disclosures.  This is not science, where Method details would be required to even publish.  This is art and some of those details may be an artist's secret formula, so to speak. 

I would add that it would be proper for those who include a natural history description of the subject reference the source of that information.  I believe that AB does not have a strict rule that we do that (correct me if I am wrong), but it is a sensible thing to do.  If one uses Wikipedia, then it is a good to recognize the service that Wikipedia provides.  Also, to the savvy, it is well known that Wikipedia can suffer from errors on occasion, so if the description provided can lead a reader of your post to further research to confirm or deny the Wiki post, then that is how things should work.  Even information published in reviewed scientific journals can have faults.  But for multiple reasons, it is good to give references where sensible.

There is no reference/mention to the word IOTD as far I can see. Am I suffering from selective blindness?

AstroShed:
I am more concerned about the fast pace of AI processing tools, that will make any cheap telescope with not so great optics, produce “image of the day” quality images

Didn't technically say "IOTD" but close enough.

andrea tasselli:
There is no reference/mention to the word IOTD as far I can see. Am I suffering from selective blindness?

*IOTD is not a word, but Astroshed spelled it out. Each and every post after that mentioned or alluded to that post.   You as well as I know where that sort of mention will take any thread, especially when talk refers to image quality, fairness, etc. 

Certainly the originator made no mention to image quality of any sort. It was just about how we post information on the subject and where that info comes from.

Alan Brunelle:
Certainly the originator made no mention to image quality of any sort. It was just about how we post information on the subject and where that info comes from.

Well, I'm good in that regard as I never post anything other than what I know about the subject, if ever. But mostly never.

Careful, now. How do you know it’s AI? 

I have a little bit of a following, so I write somewhat elaborate descriptions that (I hope) are casual and fun to read. Every paragraph is rooted in my own research of verified sites , like NASA, and scientists’ own views into new data from Hubble and Webb. I add an attribution if the info is highly specific or extremely new, but otherwise don’t if I’ve oversimplified (then just saying, “many scientists now think…”) because it’s a generalized essay, not a research paper. It’s supposed to be entertaining.  

I guess now someone is going to now accuse me of using a machine’s words merely because the prose has attitude and is grammatically correct (most of the time).