Many astrophotographers, one target?

[chubster1302]

Would that even work? 

Let's say, loads of us got together, agreed a target and stacked everything as one 🤔

[WolfieGlos]

I’ve seen images posted on Astrobin with collaborative efforts, so it’s possible. I’m sure I remember seeing one with the comet earlier this year where it was imaged by multiple people.
Whether different types of scopes, cameras, etc need to be considered I don’t know. 

[DirkSteele]

Didn’t Stargazing Live attempt something like this several years ago? As I recall it was a wider field shot of Orion but the results were relatively successful.

[ONIKKINEN]

Sure it would work, however there are some hurdles to overcome first.

All of the data needs to be more or less the same quality taken with similar scopes and similar star FWHM values. Otherwise the end result is worse than the best single contributor (or at the very least only slightly better). This means ideally everyone would be within a few dozen mm in aperture from each other and in similar bortle zones, and with similar cooled cameras. If we mix bortle 1 and 8 we get a result worse than just the b1 data.

Let say all that gets sorted, then there is the issue of stacking. One person would have to stack potentially hundreds of gigabytes or even terabytea. Data sharing could also be an issue.

In the past spring i took part in a collaboration (well, shared my data which was the beginning and end of my contribution unhelpful comments not withstanding). We shared calibrated subs in a google drive folder to the person who stacked it all. It took several days of work to stack everything if i recall correctly (but it was a 400h image...).

A more modest goal of say a hundred hours and if everyone agreed to only take 5min or longer subs would make it a not so difficult project.

[pipnina]

2 hours ago, DirkSteele said:

Didn’t Stargazing Live attempt something like this several years ago? As I recall it was a wider field shot of Orion but the results were relatively successful.

They did and I remember watching it!

Their result was quite good but had obvious troubles as I recall, and given how long it's been since stargazing live was on the TV the cameras in those phones were probably awful, meaning the nebula they pulled out of the image likely came from good data submitted from people's DSLRs or astrocams on camera lenses 😕

I did some googling and apparently this is the image created on stargazing live 10+ years ago:

Hard to say what hardware contributed to the image, but It looks quite clear to me that the horsehead and M42 regions have telescope data pasted over them! Most likely viewers with scopes sent their images in. Either that or the beebs was trying to pull a fast one on us claiming a million images from the phones people had in 2012 were capable of capturing the horsy haha.

I found the image in this thread on some other forum: https://www.flickr.com/groups/1574153@N21/discuss/72157651480273305/

[Neil_104]

It's an interesting idea. I can sense a "SGL Deep Field" coming on 😄

[iantaylor2uk]

I think this has already been tried, with some success - see paper at https://arxiv.org/abs/1406.1528

[WolfieGlos]

Just saw this post over on CN that had 5 people collaborating on the same target 

https://www.cloudynights.com/topic/884028-ngc-3642-in-rgb-and-hargb-a-5-person-collaboration/

I’m intrigued whether you could mix different scopes (ie refractors and reflectors) given the difference with stars from the diffraction spikes? 

[Marvin Jenkins]

Did I see an Astrobuscuit video on YouTube where he was trying to get a large collaboration project up and running.

The most recent collaboration I can think of is the AP guys who claim to have found a huge oxygen target next to The Great Andromeda Spiral.

Marv

 

[tomato]

Astrobiscuit’s project was the Big Amateur Telescope, BAT, he set up a forum on Discord and several targets were attempted, with quite a bit of success. The results of the first target M27, are on here somewhere.

I participated in the first one but the traffic on Discord grew exponentially and I’m afraid I couldn’t keep up, one Astro forum is enough for me.

[DaveS]

As here. I joined Rory's Discord and tried to participate but it just got manic and I had to drop out as I couldn't keep up.

[ONIKKINEN]

17 hours ago, WolfieGlos said:

Just saw this post over on CN that had 5 people collaborating on the same target 

https://www.cloudynights.com/topic/884028-ngc-3642-in-rgb-and-hargb-a-5-person-collaboration/

I’m intrigued whether you could mix different scopes (ie refractors and reflectors) given the difference with stars from the diffraction spikes? 

You can!

Simple processing of the image will yield a result that is a mix of all the diffraction spikes so in most cases a mess of different spikes because there are many ways you can orient your tube and your spider vanes (like below, a dozen scopes mixed).

Alternatively you can do some selective processing by taking bright stars from only one dataset (either with or without diffraction spikes), or some star synthesis voodoo to create diffraction spikeless stars from the input data (not sure how, i just know it can be done).

[pipnina]

10 minutes ago, ONIKKINEN said:

You can!

Simple processing of the image will yield a result that is a mix of all the diffraction spikes so in most cases a mess of different spikes because there are many ways you can orient your tube and your spider vanes (like below, a dozen scopes mixed).

Alternatively you can do some selective processing by taking bright stars from only one dataset (either with or without diffraction spikes), or some star synthesis voodoo to create diffraction spikeless stars from the input data (not sure how, i just know it can be done).

I would hazard that the best method would be to use a sigma clipping method, as this will remove things that are transient between images and datasets like diffraction spikes, satellites, walking noise etc.

Normalisation of the various datasets could be very important though. I'm not sure.

[ONIKKINEN]

2 minutes ago, pipnina said:

I would hazard that the best method would be to use a sigma clipping method, as this will remove things that are transient between images and datasets like diffraction spikes, satellites, walking noise etc.

Normalisation of the various datasets could be very important though. I'm not sure.

Clipping when stacking doesn't remove all the different diffraction spikes if all the datasets have a significant portion in the total integration (or it would have to be turned to such an extreme that actual signal gets destroyed too). If there were 5 refractors and one reflector, then maybe since the reflector is an outlier.

Normalisation, and weighting of the subs in general is the most important thing in a stack where data quality is greatly variable.

[Marvin Jenkins]

Multiple diffraction spikes is the only way to get one up on JWST. The more the merrier I say.

Marv

[LukeSkywatcher]

Wouldn't longitude/latitude be an issue with stacking. I saw 2 side by side images of the Moon, taken from a distance of 900 miles apart. While the rotation was slight, it was still noticeable.

It might be different for DSOs.

Edited July 23, 2023 by LukeSkywatcher

[Astroscot2]

1 hour ago, LukeSkywatcher said:

Wouldn't longitude/latitude be an issue with stacking. I saw 2 side by side images of the Moon, taken from a distance of 900 miles apart. While the rotation was slight, it was still noticeable.

It might be different for DSOs.

DSOs are a bit further away......

[LukeSkywatcher]

17 hours ago, Astroscot2 said:

DSOs are a bit further away......

This is true.