#1  
Old 22-04-2016, 05:03 PM
Camelopardalis's Avatar
Camelopardalis (Dunk)
Drifting from the pole

Camelopardalis is offline
 
Join Date: Feb 2013
Location: Brisbane
Posts: 5,425
Question Pixel v pixel...who wins???

Folks,

I've recently been fortunate enough to have access to a couple of different sensors and had a go at trying to benchmark them, in the interests of science

The target was the keyhole/wedge area of Eta Carinae nebula, as it has both subtle and higher contrast features. Each image represents the same exposure time.

They are a fraction of the original images but to retain colour information I chose 24-bit PNGs and then had to squeeze them in to 200KB.

So...how to pick a winner from these?

Thoughts welcomed...

Cheers,
Dunk
Attached Thumbnails
Click for full-size image (eta-A.png)
188.7 KB120 views
Click for full-size image (eta-B.png)
199.1 KB120 views
Click for full-size image (eta-C.png)
64.6 KB107 views
Reply With Quote
  #2  
Old 22-04-2016, 05:45 PM
lazjen's Avatar
lazjen (Chris)
PI cult member

lazjen is offline
 
Join Date: Dec 2012
Location: Flaxton, Qld
Posts: 2,064
Just based on the pictures I see here, #2 looks the best to me. After looking at #2, the others seem blurry.
Reply With Quote
  #3  
Old 22-04-2016, 06:30 PM
RickS's Avatar
RickS (Rick)
PI cult recruiter

RickS is offline
 
Join Date: Apr 2010
Location: Brisbane
Posts: 10,584
Post some calibrated FITS files and I'll run a few numbers for you, Dunk. It's difficult to judge processed data. I've made a few silk purses out of sow's ears

Cheers,
Rick.
Reply With Quote
  #4  
Old 22-04-2016, 06:36 PM
Atmos's Avatar
Atmos (Colin)
Ultimate Noob

Atmos is offline
 
Join Date: Aug 2011
Location: Melbourne
Posts: 6,980
#2 has the smallest pixels I assume, followed by #1 and then #3 getting larger.
#2 is the best resolved although I cannot figure out anything more than that
Reply With Quote
  #5  
Old 22-04-2016, 06:53 PM
h0ughy's Avatar
h0ughy (David)
Moderator

h0ughy is offline
 
Join Date: Apr 2005
Location: NEWCASTLE NSW Australia
Posts: 33,156
yep concur - #2
Reply With Quote
  #6  
Old 22-04-2016, 07:07 PM
Nikolas's Avatar
Nikolas (Nik)
Dazed and confused

Nikolas is offline
 
Join Date: May 2012
Location: Melbourne
Posts: 3,254
#2 for me also
Reply With Quote
  #7  
Old 22-04-2016, 07:17 PM
raymo
Registered User

raymo is offline
 
Join Date: Sep 2011
Location: margaret river, western australia
Posts: 6,070
And me.
raymo
Reply With Quote
  #8  
Old 22-04-2016, 08:26 PM
Camelopardalis's Avatar
Camelopardalis (Dunk)
Drifting from the pole

Camelopardalis is offline
 
Join Date: Feb 2013
Location: Brisbane
Posts: 5,425
Thanks all...it's so hard to look at them side by side on a computer screen and tell one from another

Note that these are all unprocessed - just stacked and with PI's STF applied (for its sins).

They were registered onto the mono image (the smallest) so that I'd be looking at the same field...does this introduce any gremlins of its own? As the camera rotation was not the same in each case, there would have been pixel rotation involved in registering.

But indeed, #2 is from the camera with the smallest pixels...5.2 microns...my modded Canon 1100D. The pixels in images 1 and 3 are about the same size at 6.5 microns, 1 being colour, 3 being mono. I was hoping to see some differentiation between the colour and mono at the same pixel size...the Loch Ness monster of increased resolution, or just variance in seeing perhaps?

I should add that these are what you experts call "under sampled", given my scope focal length is 550mm, but when the images were captured I suspect the seeing would have been better than my resolving capabilities.

@RickS - I did the same calibration on all of them...nil
Reply With Quote
  #9  
Old 22-04-2016, 08:36 PM
RickS's Avatar
RickS (Rick)
PI cult recruiter

RickS is offline
 
Join Date: Apr 2010
Location: Brisbane
Posts: 10,584
Quote:
Originally Posted by Camelopardalis View Post
@RickS - I did the same calibration on all of them...nil
Fair enough, Dunk. That's better than doing a half-arsed calibration which will add noise
Reply With Quote
  #10  
Old 22-04-2016, 08:43 PM
Camelopardalis's Avatar
Camelopardalis (Dunk)
Drifting from the pole

Camelopardalis is offline
 
Join Date: Feb 2013
Location: Brisbane
Posts: 5,425
Phew...lucky escape

The mono CCD was cooled but not set point controlled. All three were taken at more or less the same ambient temperature, possibly a degree or two warmer the night the CCD images were taken.
Reply With Quote
  #11  
Old 22-04-2016, 08:50 PM
Atmos's Avatar
Atmos (Colin)
Ultimate Noob

Atmos is offline
 
Join Date: Aug 2011
Location: Melbourne
Posts: 6,980
Looking at them on my laptop instead of my iPhone (didn't even notice that two of them were colour!). #3 looks sharper (tighter) than #1 but #1 also looks like it has some elongation.
With #3 being mono it also has a greater dynamic range than #1 but that is mostly likely to be simply the difference between mono (filter) and any OSC.
Reply With Quote
  #12  
Old 22-04-2016, 09:20 PM
Camelopardalis's Avatar
Camelopardalis (Dunk)
Drifting from the pole

Camelopardalis is offline
 
Join Date: Feb 2013
Location: Brisbane
Posts: 5,425
Agreed...#1 is from a DSLR, as is 2

What I really need are some RGB subs...I (mistakenly) took some at 2x2 binning, and that really doesn't help the situation

Interesting re: elongation. PI judged the eccentricity to be larger. I must get my eyes checked!
Reply With Quote
  #13  
Old 23-04-2016, 06:08 AM
Slawomir's Avatar
Slawomir (Suavi)
Registered User

Slawomir is offline
 
Join Date: Sep 2014
Location: North Queensland
Posts: 3,240
That's a very interesting comparison Dunk.

To my eye, #2 is certainly the noisiest of the three, but also it has the tightest stars. Perhaps it would also be interesting, in the interests of science, to also compare a less bright region in all three?
Reply With Quote
  #14  
Old 23-04-2016, 08:16 AM
Camelopardalis's Avatar
Camelopardalis (Dunk)
Drifting from the pole

Camelopardalis is offline
 
Join Date: Feb 2013
Location: Brisbane
Posts: 5,425
Good question, Suavi...

Let's look at a less bright region of the frame, still with some features, but some darker areas.

Same order as before. STF applied to the full image (crop in the case of the DSLRs) before cropping to tiny images for upload.
Attached Thumbnails
Click for full-size image (eta-A-darker.png)
167.6 KB38 views
Click for full-size image (eta-B-darker.png)
197.0 KB38 views
Click for full-size image (eta-C-darker.png)
52.6 KB41 views
Reply With Quote
  #15  
Old 23-04-2016, 09:49 PM
RickS's Avatar
RickS (Rick)
PI cult recruiter

RickS is offline
 
Join Date: Apr 2010
Location: Brisbane
Posts: 10,584
Here are the stats I derived from single subs and the 30 minute stacks. Note that the single subs are 180s for the OSC cameras and 300s for the mono CCD, so I have estimated the SNR for a 2x180s stack for the OSC camera to give a number that is reasonably comparable.

Cheers,
Rick.
Attached Thumbnails
Click for full-size image (ScreenShot003.jpg)
116.7 KB40 views
Reply With Quote
  #16  
Old 23-04-2016, 09:59 PM
RickS's Avatar
RickS (Rick)
PI cult recruiter

RickS is offline
 
Join Date: Apr 2010
Location: Brisbane
Posts: 10,584
And here are comparison images of the three single subs with colour converted to luminance and image scale equalised.
Attached Thumbnails
Click for full-size image (ScreenShot005.jpg)
69.4 KB52 views
Reply With Quote
  #17  
Old 24-04-2016, 07:59 AM
Merlin66's Avatar
Merlin66 (Ken)
Registered User

Merlin66 is offline
 
Join Date: Oct 2005
Location: Junortoun Vic
Posts: 8,904
Rick,
I think the oversampling in #2 is assisting in improving the resolution.
Great images!
Reply With Quote
  #18  
Old 24-04-2016, 09:34 AM
gregbradley's Avatar
gregbradley
Registered User

gregbradley is offline
 
Join Date: Feb 2006
Location: Sydney
Posts: 17,877
It does show the gain you can expect by correctly matching your pixel size to your focal length and local seeing.

For the bulk of imagers 6 microns is that sweet spot.

Greg.
Reply With Quote
  #19  
Old 24-04-2016, 10:02 AM
Camelopardalis's Avatar
Camelopardalis (Dunk)
Drifting from the pole

Camelopardalis is offline
 
Join Date: Feb 2013
Location: Brisbane
Posts: 5,425
Quote:
Originally Posted by RickS View Post
Here are the stats I derived from single subs and the 30 minute stacks. Note that the single subs are 180s for the OSC cameras and 300s for the mono CCD, so I have estimated the SNR for a 2x180s stack for the OSC camera to give a number that is reasonably comparable.
That's awesome Rick thanks so much for taking the time out to analyse the data.

The old CCD, hereon known as "ol' faithful" really does walk away from the DSLRs. Cooling the DSLRs would make for some more interesting data points.

Quote:
Originally Posted by RickS View Post
And here are comparison images of the three single subs with colour converted to luminance and image scale equalised.
Wow, yeah not only is there a clear difference between the DSLRs, but the CCD still shows its mettle
Reply With Quote
  #20  
Old 24-04-2016, 02:24 PM
CAAD9's Avatar
CAAD9 (Adam)
Registered User

CAAD9 is offline
 
Join Date: Jul 2010
Location: Brisbane
Posts: 48
First of all I'm only into observing so most of the details discussed in this thread are well out of my depth.

But base purely on looking at both sets of 3 pics, I like #2 best for stars and #1 from each set for the gas. The 1s seem to have the sharpest edges between the gas and background sky.

Great effort irrespective.

Cheers

Adam
Reply With Quote
Reply

Bookmarks

Thread Tools
Rate This Thread
Rate This Thread:

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT +10. The time is now 12:31 AM.

Powered by vBulletin Version 3.8.7 | Copyright ©2000 - 2024, Jelsoft Enterprises Ltd.
Advertisement
Bintel
Advertisement
Testar
Advertisement