Group acquisition project

[Rodd]

Just an idea.  I have often wondered just how deep an image can be using the equipment at hand.  I have collected 25-30 hours of data on a target--but never 100 hours (per channel).  It would be interesting if a group of people agree to pool their data on an object to render an image with 500 hours of exposure time.   If anyone is interested, here is my idea:

1) for the sake of simplicity, focal lengths and FOVs should be as similar as possible

2) A mutually agreeable target--preferably one that does not take up the whole FOV so framing is not as critical (a galaxy, or part of a nebula)

3) Narrowband might be easier as gradients will not be a strong, and folks won't have to worry about the Moon as much

4) We need to agree on acceptable sub quality using an agreed limit to FWHM, eccentricity, and  median. (SNR and noise will take care of themselves.).  If you have to throw out sattalite or plane trail subs--do so, but I find PI's integration tool very effectively eliminates them.

5) If we are feeling ethused (and confident) we could also make quite a mosaic.

I suggest each person collects 10 hours of data, calibrates, registers and stacks the subs, but does not crop or do any other processing (5 hours would be good as well).  The stack would be uploaded to the forum as a FITs file and anyone interested can download it and add it to there data for that stack.  If 20 people collect 5 hours of data--we would have a 100 hour stack.  In this way, we can build 100 hour (or more!) stacks and create either an LRGB/HaLRB or HaSHO image that will be as deep and resolved as is possible.  Creating a master stack from the sub stacks would be easy, as integrating 10 50 hour stacks results in the same result as integrating all the subs that make up those stacks.  That's why using stacks will be much easier than trying to use subs. 

If anyone is interested, I will be imaging at 1,000mm for the next month or so.  Before I switch out OTAs I could collect 5-10 hours of data on our target.  I will be switching to 318mm when I finish the project I am on.  However--before I start imaging at 318mm, which I will do for an extended period of time (providing TNR really did fix the collimation), I could throw the reducer on my 1,000mm setup and image at 700mm to collct the data for this project.

Rodd

Edited October 13, 2019 by Rodd

[ollypenrice]

There have been SGL collaborations before and I like the idea. I think the target should be a very faint one, or have very faint components, to maximise the benefit.

Rather than settling on focal length it would be better to set an approximate image scale in arcseconds per pixel.

Olly

[Rodd]

6 minutes ago, ollypenrice said:

There have been SGL collaborations before and I like the idea. I think the target should be a very faint one, or have very faint components, to maximise the benefit.

Rather than settling on focal length it would be better to set an approximate image scale in arcseconds per pixel.

Olly

Thats probably true, though FL and resolution seem to floow each other for the most part.  But I guess if resolution is kept similar it will be easier to process.  We can alwasy rescale within reason.  A faint target would be best I agree.  Who knows, 500 hours may reveal structures previously unseen!

Rodd

[ollypenrice]

14 hours ago, Rodd said:

Thats probably true, though FL and resolution seem to floow each other for the most part.  But I guess if resolution is kept similar it will be easier to process.  We can alwasy rescale within reason.  A faint target would be best I agree.  Who knows, 500 hours may reveal structures previously unseen!

Rodd

I found little to choose between data taken at 2.4 metres/0.6"PP and data taken at 1 metre and 0.9"PP. I do think it's resolution which matters and to be inclusive I'd have thought that something around 2"PP would find lots of takers.

Olly

[Jonk]

8 minutes ago, ollypenrice said:

it's resolution which matters

I would agree with this, and would be happy to contribute, but my resolution is currently 3.16"pp so probably wouldn't be able to offer much.

I have suggested this (data sharing) to a friend of mine down the road and he's up for it, so we may do our own to start with.

Another option would be to setup 3 categories:

3-4"pp

2-3"pp

1-2"pp

The target can be mutually agreed, probably based on the time of year, i.e. if started now, choose a target that is at 30degrees once astro dark starts. Of course this will vary, but for the majority of people in the UK, it would be a good benchmark.

So....who's willing to manage this?!

[Rodd]

1 hour ago, Jonk said:

I would agree with this, and would be happy to contribute, but my resolution is currently 3.16"pp so probably wouldn't be able to offer much.

I have suggested this (data sharing) to a friend of mine down the road and he's up for it, so we may do our own to start with.

Another option would be to setup 3 categories:

3-4"pp

2-3"pp

1-2"pp

The target can be mutually agreed, probably based on the time of year, i.e. if started now, choose a target that is at 30degrees once astro dark starts. Of course this will vary, but for the majority of people in the UK, it would be a good benchmark.

So....who's willing to manage this?!

  I think it will important to agree on a camera orientation so as to minimize the loss of FOV during registration (either 90 degrees of 180 degrees would be easiest).  I have asked the creators of PI how to add long focal length data into smaller focal length, wider FOV images, and have not yet received an answer--it seems not to be as straight forward as it is in Photoshop.  So for my image, I will be restricted to the smallest FOV.  If that will be true for others, it would be helpful to choose a target based on the framing in the smallest FOV.

Rodd

[Pompey Monkey]

4 minutes ago, Rodd said:

  I think it will important to agree on a camera orientation so as to minimize the loss of FOV during registration (either 90 degrees of 180 degrees would be easiest).  I have asked the creators of PI how to add long focal length data into smaller focal length, wider FOV images, and have not yet received an answer--it seems not to be as straight forward as it is in Photoshop.  So for my image, I will be restricted to the smallest FOV.  If that will be true for others, it would be helpful to choose a target based on the framing in the smallest FOV.

Rodd

Star registration will do a <ahem> stellar job of this.

[Rodd]

Just now, Pompey Monkey said:

Star registration will do a <ahem> stellar job of this.

I know, but if one sub is oriented 90 and the other 180, there will be a lot of lost FOV due to cropping required. 

Rodd