Imaging small galaxies with focal length under 800mm

[ollypenrice]

7 hours ago, Ships and Stars said:

Thanks Vlaiv, Stephan's Quintet is incredible...👍

It's tricky visually even with a 20 inch Dob, at least with my eyesioght.

Olly

[tomato]

I’ve not been heeding the sound imaging scale advice by attempting to capture small galaxies using an Esprit 150 (FL 1050mm) and an ASI 178 (0.47 arcsecs per pixel), with mixed results. I can guide at 0.5 arcsecs total RMS, about half of the required performance, at this imaging scale. It is no surprise therefore that my best result to date is when I binned the camera 2 x 2, the image is only cropped to remove stacking artefacts.

[StargeezerTim]

Here is a couple with a Star 71 and 100D dslr. Not ideal for galaxies at all but you can get some nice results.

 

 

 

 

 

[Anne S]

I've done the Whale and Hockeystick in the last couple of weeks. Only 2 hours integration as cloud stopped me getting any more. Imaged at 564mm with a SX694.

 

[AstroArgy]

On 23/04/2020 at 23:01, JamesF said:

I have more data for this and need to brighten the colours up a bit as it looks quite dull compared with my laptop, but this is my M51 with an ED80:

And whilst I haven't processed the data yet, this is a crop from the same rig:

Tonight I am working on NGC2403, which also occupies the frame quite nicely.

James

Hey James,

Could I ask you what camera are you using to get these results?

Ta

[JamesF]

3 minutes ago, AstroArgy said:

Hey James,

Could I ask you what camera are you using to get these results?

Ta

It's an Atik 314L+ (mono).

James

Edited April 30, 2020 by JamesF

[AstroArgy]

On 27/04/2020 at 01:13, StargeezerTim said:

Here is a couple with a Star 71 and 100D dslr. Not ideal for galaxies at all but you can get some nice results.

 

 

 

 

 

Hey Tim,

With DSLR's and Galaxy season what kind of exposure time & Iso do you find works best as to get the level of detail you are getting?

Hope you dont mind me asking 😊

[StargeezerTim]

9 hours ago, AstroArgy said:

Hey Tim,

With DSLR's and Galaxy season what kind of exposure time & Iso do you find works best as to get the level of detail you are getting?

Hope you dont mind me asking 😊

I usually start at 5 mins, iso 400, no filters, and if the histo is too far to the right, reduce it.

[Xiga]

Imaging small galaxies at short FL is not something i've done a lot of myself, but it can certainly be done, so definitely have a go!

All shot with a D5300 and an 80ED. None more than 2-3 hrs each:

Leo Triplet:

M81 & M82:

And a 100% Crop:

M101:

NGC 4565:

[AstroArgy]

On 02/05/2020 at 01:19, Xiga said:

Imaging small galaxies at short FL is not something i've done a lot of myself, but it can certainly be done, so definitely have a go!

All shot with a D5300 and an 80ED. None more than 2-3 hrs each:

Leo Triplet:

M81 & M82:

And a 100% Crop:

M101:

NGC 4565:

Nice work Xiga!!!

Question when you say 2-3 hours you mean of total integration time?

How long do you run your exposures / ISO and so on? 

I'm curious to understand what it takes (other than post-prod) to achieve such lovely pics 😃

[tingting44]

These photons were captured many years ago using a SW 130pds telescope (650mm) & Canon 7d sensor

 

 

 

 

 

 

i really must make the effort to re visit this old data for a fresh run through pixinsight

Edited May 4, 2020 by tingting44

[Xiga]

1 hour ago, AstroArgy said:

Nice work Xiga!!!

Question when you say 2-3 hours you mean of total integration time?

How long do you run your exposures / ISO and so on? 

I'm curious to understand what it takes (other than post-prod) to achieve such lovely pics 😃

Thanks AstroArgy.

Yes each one had roughly 2.5 hrs of total exposure time. I can't remember exactly how much but they would be very close to that. Taken in Bortle 4 skies. If you look me up on Astrobin you can see the exact capture details. I only use ISO 200 on the D5300 and for broadband, the exposures are usually 6 mins long. 

[Datalord]

On 24/04/2020 at 10:37, ollypenrice said:

Being plug and play they save time while reflectors are arguing with their owners!

Fighting words! I have a few reflector images to compare. And the RC is very maintenance light.

First M106 with my RASA at 620mm. This is with a QHY247, so small pixels to Vlaivs point.

Here's M106 at 2361mm with my 12" RC.

So, when we talk about the "larger" galaxies, I don't think you win much by going up in aperture and FL. Like Olly says, you can crop your way to a lot. But my interest is very much in galaxies, especially the clusters farther away. I think this is where the longer FL has benefits.

Abell 779.

The long FL is what I love, but omg, making it work cost sweat and a small luxury car in financial outlay.

[ollypenrice]

7 hours ago, Datalord said:

 

So, when we talk about the "larger" galaxies, I don't think you win much by going up in aperture and FL. Like Olly says, you can crop your way to a lot. But my interest is very much in galaxies, especially the clusters farther away. I think this is where the longer FL has benefits.

Abell 779.

 

 

The long FL is what I love, but omg, making it work cost sweat and a small luxury car in financial outlay.

The question is, 'What are the active ingredients in high resolution amateur imaging?'

Focal length? Pixel size? Aperture? Seeing? Guiding?

We have to put seeing at the top because even quite modest amateur high res setups can offer, on paper, the other four ingredients in a measure which the seeing will always limit. The only way to beat your home seeing is to set up a remote rig in a dedicated location.

So what will your seeing support? 0.8 arsecs or 0.9, or whatever? It really isn't likely to be any less. In this case there will be a remarkable number of ways in which to get to that resolution by juxtaposing assorted focal lengths and pixel sizes. The combination I've chosen, small aperture and small pixels, can get very close indeed to what I used to get from a much bigger reflector and bigger pixels, so close that I'm not, personally, chasing more. However, a little more resolution is possible for those who want it.  What would really tempt me in a big reflector would be the opportunity to work in bin 2 or even bin 3 while retaining a pixel scale of just under an arcsec per effective pixel.  It would be fast. Unfortunately our time with a big reflector was affected by the fact that the camera we had would not bin satisfactorily so we were obliged to over-sample. It was never possible to present an image at full size because it just revealed 'empty resolution.' The price you pay is a drastically reduced FOV in exchange for a very small increase in final resolution. Yer pays yer money... etc.

Olly

[Datalord]

38 minutes ago, ollypenrice said:

The question is, 'What are the active ingredients in high resolution amateur imaging?'

Focal length? Pixel size? Aperture? Seeing? Guiding?

We have to put seeing at the top because even quite modest amateur high res setups can offer, on paper, the other four ingredients in a measure which the seeing will always limit. The only way to beat your home seeing is to set up a remote rig in a dedicated location.

So what will your seeing support? 0.8 arsecs or 0.9, or whatever? It really isn't likely to be any less. In this case there will be a remarkable number of ways in which to get to that resolution by juxtaposing assorted focal lengths and pixel sizes. The combination I've chosen, small aperture and small pixels, can get very close indeed to what I used to get from a much bigger reflector and bigger pixels, so close that I'm not, personally, chasing more. However, a little more resolution is possible for those who want it.  What would really tempt me in a big reflector would be the opportunity to work in bin 2 or even bin 3 while retaining a pixel scale of just under an arcsec per effective pixel.  It would be fast. Unfortunately our time with a big reflector was affected by the fact that the camera we had would not bin satisfactorily so we were obliged to over-sample. It was never possible to present an image at full size because it just revealed 'empty resolution.' The price you pay is a drastically reduced FOV in exchange for a very small increase in final resolution. Yer pays yer money... etc.

Olly

I think we agree violently.

For reference, I use bin2 for everything colour related, so only lum and Ha is bin1. I'm going to argue, by having looked enough at subs, that there is a definite difference in the subs between bin1 and bin2, so having the rig at E-Eye, with that mount, enables me to benefit from 0.54" per pixel. But colour, yes, 1.04".

[andrew s]

@ollypenrice the only addition I would make to what you say above is the with modern, low read noise, CMOS cameras over sampling is not an issue.

Regards Andrew 

[ollypenrice]

42 minutes ago, Datalord said:

I think we agree violently.

For reference, I use bin2 for everything colour related, so only lum and Ha is bin1. I'm going to argue, by having looked enough at subs, that there is a definite difference in the subs between bin1 and bin2, so having the rig at E-Eye, with that mount, enables me to benefit from 0.54" per pixel. But colour, yes, 1.04".

I like 'agree violently!' 😁

I also like, ' I'm going to argue, by having looked enough at subs, that there is a definite difference in the subs between bin1 and bin2...'  If you can see a difference there is a difference! I can sometimes see a difference between 0.9"PP and 1.1"PP on our dual rig, too. Depends on the seeing.

Olly

[vlaiv]

56 minutes ago, Datalord said:

For reference, I use bin2 for everything colour related, so only lum and Ha is bin1. I'm going to argue, by having looked enough at subs, that there is a definite difference in the subs between bin1 and bin2, so having the rig at E-Eye, with that mount, enables me to benefit from 0.54" per pixel. But colour, yes, 1.04".

 

10 minutes ago, ollypenrice said:

I also like, ' I'm going to argue, by having looked enough at subs, that there is a definite difference in the subs between bin1 and bin2...'  If you can see a difference there is a difference! I can sometimes see a difference between 0.9"PP and 1.1"PP on our dual rig, too. Depends on the seeing.

I would also like to see the difference - do you mind showing me the difference?

[ollypenrice]

5 minutes ago, vlaiv said:

 

I would also like to see the difference - do you mind showing me the difference?

Difficult in my case because, having decided that the difference was sometimes visible, I took to doing Lum at 0.9 and RGB at 1.1 systematically. (It's a dual rig so this is logical anyway.) I'll try to shoot a comparison shot on a night of good seeing, though.

Olly

[Datalord]

24 minutes ago, vlaiv said:

 

I would also like to see the difference - do you mind showing me the difference?

Same as Olly, I do different filters in bin1 and 2, as well as different exposure lengths. But I guess I could do a few shots to compare. 

[vlaiv]

17 minutes ago, ollypenrice said:

Difficult in my case because, having decided that the difference was sometimes visible, I took to doing Lum at 0.9 and RGB at 1.1 systematically. (It's a dual rig so this is logical anyway.) I'll try to shoot a comparison shot on a night of good seeing, though.

Olly

 

1 minute ago, Datalord said:

Same as Olly, I do different filters in bin1 and 2, as well as different exposure lengths. But I guess I could do a few shots to compare. 

I found it rather interesting that you both actually saw difference between 0.9" and 1.1". I think that most people would be hard pressed to see difference between double sampling rates - for example 0.8"/px and 1.6"/px.

Could you at least tell me what sort of difference did you see?

Here is some text that has been blurred so that we hit optimum sampling rate - one has been sampled at 1.1"/px and other at 0.9"/px.

Do you see any difference - apart from left image obviously being slightly smaller. To my eye it just seems a bit sharper than image sampled at 0.9"/px (larger one), although smaller letters might be harder to read without glasses

 

[RayD]

Crikey, I had trouble seeing the boxes, never mind reading the text, so there is definitely no chance of me ever seeing the difference.  Glad I image at 1.3"/p 😂

[andrew s]

Visual acuity varies considerably  and spotting subtle differences is strongly impacted by training.

I don't  know if these differences can be spotted or not, but if to identical scale, maybe subtracting them and looking at the residual might be informative.

Regards Andrew 

[vlaiv]

8 minutes ago, andrew s said:

Visual acuity varies considerably  and spotting subtle differences is strongly impacted by training.

I don't  know if these differences can be spotted or not, but if to identical scale, maybe subtracting them and looking at the residual might be informative.

Regards Andrew 

I've done that and residual is usually smaller than one in thousand or so - and usually contains noise, more than anything else.

Here is example done with x2 resolution reduction - strong oversampling case. Original text was sampled at ideal resolution (1"/px, FWHM at 1.6"), then image was created at lower resolution by binning x2 (2"/px) and up scaled back to 1"/px using cubic b-spline resampling. Third image was made as difference of the two. Histogram is shown on the left and it shows aliasing artifacts. Difference is in -0.2-0.2 range and actual histogram of difference is:

As you see - standard deviation of residual is 0.0284 - and that is for x2 lower resolution with under sampling.

Here is case for oversampling - so ideal resolution is at 2"/px and we have image at 1"/px and 2"/px:

Again histogram of the difference:

Now values are +/- 0.06 and StdDev is 0.0097.

Histogram is not centered and symmetric - because there is probably sub pixel shift due to resampling method used (not very sophisticated - nor did I align subs in any way).

As you see - even x2 difference in resolution is likely to cause difference that is buried in noise - less than a few percent if over sampling.

[andrew s]

@vlaiv that looks conclusive. 

Thanks Andrew