Anyone Fancy Processing Some Data ?

[Newforestgimp]

Hi All,

Im after a favour, below are a couple of drop box links to two sets of data captured last night, I cant get DSS to stack them, it sees no stars.
I wondered if any of you processing gurus could have a go at one or both sets and see why you think DSS wont stack them ?
they come from my new big rig as i call it, ive been having some teething problems shall we say and these are the first captures for me. I appreciate there might not be enough data to get a decent image i was experimenting with 1min and 3 min exposures on two different targets. I would like to see if anyone can determine why i cant stack, no stars found etc. captured on RC8" with ASI071MC Pro 90 gain 0 Offset.

They are only Lights, no darks or flats as yet captured.

M27         https://www.dropbox.com/sh/z6ncux1ucblhuil/AACm_3lbRj5dQP7x5-oA7rvva?dl=0

Crescent   https://www.dropbox.com/sh/hne3yqytxj4kgdr/AAB9VPI1mssHVCed_9ut0TCHa?dl=0

See what you think....

Andy

[vlaiv]

You need to collimate your scope. Stars are rather poor looking.

DSS can't make them out as they look more like small DSOs than pin point stars. One way of tricking the DSS or other software is to bin your data (in a certain way to keep color information) - so that star size reduces.

For example - here is x6 reduced image (single sub):

Now stars look more point like.

BTW - you really need flats and darks, and using offset of 0 is not good idea.

 

[Nosieman]

Adjusting the threshold for detection will yield a better detection.

I just tried 2 of your files with the detection down to 2% and got DSS to "see" 61 stars.

[Lee_P]

I've just taken a look at your Crescent data using PixInsight. Collimation / focus is off.

Still, here's your data straight out of integration:

And with a very quick edit:

[Newforestgimp]

Thanks Vlaiv,

Im getting a bit tired of the autofocus routines, i thought my eyesight was bad but.....
Which software are you using to do the binning ? im not familiar with doing that.

[Newforestgimp]

46 minutes ago, Nosieman said:

Adjusting the threshold for detection will yield a better detection.

I just tried 2 of your files with the detection down to 2% and got DSS to "see" 61 stars.

 

yeah but it only stacks about 4 from the entire list.

[Newforestgimp]

45 minutes ago, Lee_P said:

I've just taken a look at your Crescent data using PixInsight. Collimation / focus is off.

Still, here's your data straight out of integration:

And with a very quick edit:

thank you,

I dont doubt it on both points, collimation and focus, the collimation was much worse when i first attempted taking any captures. I just binned the lot. Do you think focus or collimation is the primary reason ?

The focusing is annoying me that the autofocus routines always give different results after multiple passes, leaving me unsure which is the best focus position.

[Lee_P]

4 minutes ago, Newforestgimp said:

thank you,

I dont doubt it on both points, collimation and focus, the collimation was much worse when i first attempted taking any captures. I just binned the lot. Do you think focus or collimation is the primary reason ?

The focusing is annoying me that the autofocus routines always give different results after multiple passes, leaving me unsure which is the best focus position.

I'm a refractor guy so am not that clued up on collimation, but my guess is that your collimation is out of whack to such an extent that your autofocus routine can't work properly. So if I were you I'd first spend time getting the collimation spot-on; then I expect your autofocus will work much better.  

[Newforestgimp]

20 minutes ago, Lee_P said:

I'm a refractor guy so am not that clued up on collimation, but my guess is that your collimation is out of whack to such an extent that your autofocus routine can't work properly. So if I were you I'd first spend time getting the collimation spot-on; then I expect your autofocus will work much better.  

I'm starting to think I'm a refractor guy too 🙂

[Newforestgimp]

1 hour ago, vlaiv said:

You need to collimate your scope. Stars are rather poor looking.

DSS can't make them out as they look more like small DSOs than pin point stars. One way of tricking the DSS or other software is to bin your data (in a certain way to keep color information) - so that star size reduces.

For example - here is x6 reduced image (single sub):

Now stars look more point like.

BTW - you really need flats and darks, and using offset of 0 is not good idea.

 

I was advised to use 90 and 0 for the ASI071 not used a dedicated astro cam before, what settings would you suggest ?

[Xilman]

2 hours ago, Newforestgimp said:

Hi All,

Im after a favour, below are a couple of drop box links to two sets of data captured last night, I cant get DSS to stack them, it sees no stars.

I stack differently. First I run the subs through the astrometry.net plate solver. For just a few you could use the on-line facility at nova.astrometry.net but I found it worthwhile to install the software locally.

After the solver has done its job you end up with a FITS file containing a world coordinate system, or WCS. The WCS enables other software to convert between pixel locations and RA/Dec coordinates. In partticular, it allows SWarp to stack images.

Works very well for me but YMMV.

Oh, one thing: best to split 3-colour images into separate files and then stack independently. Re-combine the three or four stacks to regenerate a final colour image.

Others have already comented on your focus and collimation issues.  The astrometry.net solver is remarkably tolerant of such things.

[Xilman]

1 hour ago, Newforestgimp said:

Thanks Vlaiv,

Im getting a bit tired of the autofocus routines, i thought my eyesight was bad but...

Get a Bahtinov mask. Seriously. It makes focusing a doddle. I also have bad eyesight with myopia and astigmatism and had enormous trouble focusing by eye.

[The Lazy Astronomer]

3 hours ago, Newforestgimp said:

The focusing is annoying me that the autofocus routines always give different results after multiple passes, leaving me unsure which is the best focus position.

An autofocus routine should be highly repeatable if the settings are properly configured.

As has been mentioned, it may be the miscollimation throwing the focus measurement off, but if you get that sorted and still have autofocus issues it's likely that your focusser step size or backlash compensation settings aren't quite right.

[vlaiv]

5 hours ago, Newforestgimp said:

Thanks Vlaiv,

Im getting a bit tired of the autofocus routines, i thought my eyesight was bad but.....
Which software are you using to do the binning ? im not familiar with doing that.

I'm using ImageJ for that sort of processing. You can use super pixel mode as first step in DSS, but first look at that collimation issue.

You want your stars to be nice and round.

Here are my test shots when I was in process of collimation - corner stars (same scope - RC8"):

See - similar to your stars - ill defined not pin point

After one round of collimation - things got better (center of the field):

But I was still not happy with corner stars:

I the end, I managed to get even corner stars to look nice:

4 hours ago, Newforestgimp said:

I was advised to use 90 and 0 for the ASI071 not used a dedicated astro cam before, what settings would you suggest ?

Gain setting depends on your exposure length. Lower the gain (or slower the scope) - longer exposure you'll need for your subs. With said scope, I'd say that unity gain is ok - so keep that at 90.

As far as offset - well, you need to measure that.

Point of offset is that you avoid clipping to the left. Way you measure it is rather simple - you need to take bias subs.

Set your working gain and start with some offset - you can start at offset zero, and take several bias subs (camera covered like for darks but exposure length set to minimum). Take say 10 or 16 of subs. Stack those subs using minimum as stacking operation and after that do statistic on resulting stack. Look at minimum pixel value in the image. If it is equal to zero - raise offset and repeat procedure.

If it's not equal to zero - you are done. From now on, use that offset with gain you used for the measurement.

[Newforestgimp]

I will have another round or 3 of collimation, I’m even considering one of those electronic Collimators like the Ocal as I’m just not sure how I would tell if it was any better as it looked pretty good by eye.

I guess the positive on this set of subs is the stars are better than before first collimation.

[vlaiv]

3 minutes ago, Newforestgimp said:

I will have another round or 3 of collimation, I’m even considering one of those electronic Collimators like the Ocal as I’m just not sure how I would tell if it was any better as it looked pretty good by eye.

I guess the positive on this set of subs is the stars are better than before first collimation.

I followed this tutorial for collimation of my RC:

https://deepspaceplace.com/gso8rccollimate.php

except I did not use Bahtinov mask (which I'm finding useless), but rather FWHM readings in SharpCap (there is focus help utility that will read FWHM values)

[Newforestgimp]

Ooh I’ll take a look at that, I was using the Orion User Manual.

[Newforestgimp]

18 minutes ago, vlaiv said:

I followed this tutorial for collimation of my RC:

https://deepspaceplace.com/gso8rccollimate.php

except I did not use Bahtinov mask (which I'm finding useless), but rather FWHM readings in SharpCap (there is focus help utility that will read FWHM values)

Interesting article, the corner stars might take a few goes to understand the method fully though.

[Xilman]

1 hour ago, Newforestgimp said:

I will have another round or 3 of collimation, I’m even considering one of those electronic Collimators like the Ocal as I’m just not sure how I would tell if it was any better as it looked pretty good by eye.

I use the Bahtinov Grabber app. Really does make auto-focussing a doddle, not least because it is essentially independent of the seeing. Whether your stars are 2 arcsec or 15 arcsec across (and I have experienced both extremes) really doesn't matter.

With seeing as bad as 10-15 arcsec, about all that can be usefully done is photometry unless you have a very wide FOV ...

[tomato]

I took your M27 data through Astro Pixel Processor, it successfully analysed all 36 subs. I took the best 18 and it registered, normalised and stacked these, with no errors. I took this stack through the StarTools default workflow and used the star shrink and repair tools to improve the star shapes.

I would say the biggest areas for improvement is scope collimation and focus, and taking calibration frames, that will make the image a whole lot sharper and remove the bunnies from the background.👍

 

[Newforestgimp]

Just now, tomato said:

I took your M27 data through Astro Pixel Processor, it successfully analysed all 36 subs. I took the best 18 and it registered, normalised and stacked these, with no errors. I took this stack through the StarTools default workflow and used the star shrink and repair tools to improve the star shapes.

I would say the biggest areas for improvement is scope collimation and focus, and taking calibration frames, that will make the image a whole lot sharper and remove the bunnies from the background.👍

 

Holy moly !! I really didn’t expect anything like that certainly not without darks or flats, I was just curious if anyone could stack them as DSS didn’t want to know.

What work flow do you use in startools ? I find sometimes my results are less that pleasant 😂 

[tomato]

Oh, and based on the fits header entries, APP reported a  camera gain of 90 on the M27 subs.

In StarTools I just follow the elements from left to right on the tool panel on the left hand side, I didn't move the default settings a lot except for the Wipe tool (had to increase the aggressiveness) and the second auto dev where I altered the parameters to achieve a balance between background brightness and areas of maximum exposure. ST runs lightning quick on my machine so I can see what effect the changes make within a few seconds, so its easy to home in on an optimum setting.

If you want help on the optimum workflow Ivo Jager, the creator of StarTools,  has taken  SGL member's data set and provides you with a detailed workflow after applying his considerable skill. It certainly put me on the right track towards improving my processing with the package. Once you have sorted the capture issues identified, I would PM him with some data if, (like me) you like the way the software works.

[Newforestgimp]

1 hour ago, vlaiv said:

I followed this tutorial for collimation of my RC:

https://deepspaceplace.com/gso8rccollimate.php

except I did not use Bahtinov mask (which I'm finding useless), but rather FWHM readings in SharpCap (there is focus help utility that will read FWHM values)

Interesting article, the corner stars might take a few goes to understand the method fully though.

[Newforestgimp]

4 minutes ago, tomato said:

Oh, and based on the fits header entries, APP reported a  camera gain of 90 on the M27 subs.

In StarTools I just follow the elements from left to right on the tool panel on the left hand side, I didn't move the default settings a lot except for the Wipe tool (had to increase the aggressiveness) and the second auto dev where I altered the parameters to achieve a balance between background brightness and areas of maximum exposure. ST runs lightning quick on my machine so I can see what effect the changes make within a few seconds, so its easy to home in on an optimum setting.

If you want help on the optimum workflow Ivo Jager, the creator of StarTools,  has taken  SGL member's data set and provides you with a detailed workflow after applying his considerable skill. It certainly put me on the right track towards improving my processing with the package. Once you have sorted the capture issues identified, I would PM him with some data if, (like me) you like the way the software works.

I do like the way ST works, I’m definitely not after the last word in tweaking layers and masks. So I bought the full license, but my APP trial has lapsed, which I liked the stacking process.

thanks for that effort it’s given me some confidence im on the right track.

I did set gain to 90 👍

[happy-kat]

It is possible to manually mark where stars are on your reference frame, this I've done when not enough stars were found. It's the same tool that you'd use to mark a comet.