Jump to content

  •  

CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Photo

OSC vs MONO image challenge

  • Please log in to reply
16 replies to this topic

#1 sn2006gy

sn2006gy

    Vendor - Rockchuck Summit Observatory

  • ****-
  • Vendors
  • topic starter
  • Posts: 1,542
  • Joined: 04 May 2020
  • Loc: Austin, TX

Posted 01 September 2020 - 02:20 PM

I'd love to do an OSC vs Mono challenge on here.  I have a 2600mc, i know there is no (currently available) mono variation of it, but i'm happy to image something (similar objects) with anyone running any mono and compare results.

 

Let's say we pick a handful of targets in NB and Broadband and share our results (limit targets to keep this from just being an astrobin mirror)

 

Anyone interested?

 

Not even worried about the methods used... use what you do natively for osc and natively for mono.  Just want photos to speak for themselves.

 

Any targets that would be good for this?  


  • elmiko likes this

#2 Huangdi

Huangdi

    Apollo

  • *****
  • Posts: 1,105
  • Joined: 24 Jul 2019

Posted 01 September 2020 - 06:53 PM

I think the reason that these challenges don't really exist is because the outcome is easy to foretell. Given proper acquisition, calibration, integration and processing, the mono image will have increased resolution, no matter what. That's really the end of it.. I don't see why you'd need a comparison for that. But I'd be curious to see some results anyway. Sadly I don't have access to either at the moment.

Edited by Huangdi, 01 September 2020 - 06:53 PM.

  • Jim Waters and Peter in Reno like this

#3 Jim Waters

Jim Waters

    Fly Me to the Moon

  • *****
  • Posts: 6,472
  • Joined: 21 Oct 2007
  • Loc: Phoenix, AZ USA

Posted 01 September 2020 - 07:15 PM

I think the reason that these challenges don't really exist is because the outcome is easy to foretell. Given proper acquisition, calibration, integration and processing, the mono image will have increased resolution, no matter what. That's really the end of it.. I don't see why you'd need a comparison for that. But I'd be curious to see some results anyway. Sadly I don't have access to either at the moment.

waytogo.gif  +1



#4 Jeff2011

Jeff2011

    Soyuz

  • *****
  • Posts: 3,727
  • Joined: 01 Jan 2013
  • Loc: Sugar Land, TX

Posted 01 September 2020 - 08:34 PM

These comparisons are interesting but provide no definitive conclusions.  There are too many other factors involved than just the cameras.  I have both types of cameras.  Modern backlit OSC using bayer drizzle can hold there own against mono LRGB.  The real power and advantage of mono cameras is narrowband.  The multi-band pass narrowband filters for OSC just don’t give you the control and flexibility that mono narrowband has.


  • dvb, Jim Waters, Peter in Reno and 1 other like this

#5 SeymoreStars

SeymoreStars

    Gemini

  • *****
  • Posts: 3,487
  • Joined: 08 May 2014
  • Loc: Pennsyltucky

Posted 01 September 2020 - 08:53 PM

These comparisons are interesting but provide no definitive conclusions.  There are too many other factors involved than just the cameras.  I have both types of cameras.  Modern backlit OSC using bayer drizzle can hold there own against mono LRGB.  The real power and advantage of mono cameras is narrowband.  The multi-band pass narrowband filters for OSC just don’t give you the control and flexibility that mono narrowband has.

That give me a great idea for a challenge. These two images were created from the same location, same equipment. The only difference is the filters.

sn2006gy try your hand at recreating these with an OSC.

 

First the RGB

Image16
 
Second the Narrowband in HOO palette.
Image12

  • Jim Waters likes this

#6 AtmosFearIC

AtmosFearIC

    Apollo

  • -----
  • Posts: 1,315
  • Joined: 10 Dec 2015
  • Loc: Melbourne

Posted 01 September 2020 - 08:54 PM

It's really not that simple I'm afraid. A vast majority of the imaging I've been doing for the last few years has been a mixture of mono and OSC with wide field images. Now with wide fields there is a definite resolution increase with mono over OSC. Drizzle integrating does help with colour fidelity but nothing beats the raw resolution of mono with wider fields. 

I've personally found that getting a FWHM below 2.4 pixels with a OSC is nigh impossible. Bayer drizzle does help tighten that FWHM a bit as against a normal drizzle or no drizzle at all but it is a limiter. On mono however the lowest FWHM in a stack I've ever managed with 1.3 pixels with a Sigma Art 85mm @ F/2.8 in the centre with an ASI1600. It gets difficult to measure at that point though when a lot of the stars are pretty much single pixels but that would NEVER be possible with a OSC.

 

Now where this gets more complicated is at longer focal lengths and higher image scales. If you can sample at 3 pixels then interpolation should no longer cause a blur and you end up with just raw resolution. I've gone and had a look at a whole bunch of higher resolution images of Eta Carina on AstroBin.

 

OSC:

This first one I'd say is the cheapest and most basic setup, actually kinda admire the result taken with 2000+ 3s

https://www.astrobin.../full/f5xgzx/0/

 

The second has potentially the best fine structure detail out of all the OSC and RGB images.

https://www.astrobin.../full/tlj3q7/0/

 

The third shows where the resolution can be stumped with wider field and the bayer matrix. A mono could have better details.

https://www.astrobin.../full/vv8yjn/B/

 

Mono:

The first one shows what small aperture and mono can achieve. Given the image scale it could potentially be done with a OSC too but absolute resolution is likely aperture halted.

https://www.astrobin.../full/pgp2ee/0/

 

The second is a HaRGB and one that I thought was going to show more than it does, lots of Ha showing but lacking fine structure.

https://www.astrobin.../full/he247b/0/

 

The third has lots of structure but lacks finer details compared to the second OSC

https://www.astrobin.../full/401472/0/

 

The forth is the beginning of the narrowband, softer than I was expecting from such a large instrument.

https://www.astrobin.../full/3saz7q/0/

 

The fifth I think is the same instrument (different data set) but appears to have a higher contrast than the forth.

https://www.astrobin.../full/7jop5a/C/

 

The sixth has the highest resolution/detail of the whole lot.

https://www.astrobin.../full/7sed56/0/

 

So, what I am getting from this list is that the highest resolution images aren't always going to be from a mono. When it comes to picking up global detail (nebulosity) being able to do narrowband helps considerably, whether it be full narrowband or a HaRGB. Out of this list of 9 images the best on a resolution stand point is a SHO image taken with a 16" RCOS from Chile, shock horror haha

What did surprise me a little was that the best non narrowband image to my eye was taken with a OSC on a Mewlon. It came close to the last on the list but not quite but it bested all of the others in my opinion.


  • Mert and Jim Waters like this

#7 SeymoreStars

SeymoreStars

    Gemini

  • *****
  • Posts: 3,487
  • Joined: 08 May 2014
  • Loc: Pennsyltucky

Posted 01 September 2020 - 08:59 PM

Gorgeous images AtmosFearIC.



#8 jdupton

jdupton

    Soyuz

  • *****
  • Posts: 3,959
  • Joined: 21 Nov 2010
  • Loc: Central Texas, USA

Posted 01 September 2020 - 09:01 PM

Julian,

 

I think the reason that these challenges don't really exist is because the outcome is easy to foretell. Given proper acquisition, calibration, integration and processing, the mono image will have increased resolution, no matter what. That's really the end of it.. I don't see why you'd need a comparison for that. But I'd be curious to see some results anyway. Sadly I don't have access to either at the moment.

   Is CFA Drizzle not a proper integration method? If you use CFA Drizzle on an OSC it will have the same resolution as its mono counterpart (if the same sensor is available in both varieties).

 

   Just playing devil's advocate here...

 

 

John


Edited by jdupton, 01 September 2020 - 09:04 PM.


#9 Alex McConahay

Alex McConahay

    Voyager 1

  • *****
  • Posts: 11,006
  • Joined: 11 Aug 2008
  • Loc: Moreno Valley, CA

Posted 01 September 2020 - 09:45 PM

If you call it a challenge, yeah, it could be fun.

 

But if you call it an experiment, there is just no way it will really prove anything. Too many variables changing. An experiment changes only one variable at a time. 

 

And the biggest variable would be processing. We did several shows on The Astro Imaging Channel where we provided identical sets of data. (Everybody processed the same data.) A dozen imagers processed the data. The resulting shots were dramatically different. Just not the same. Star size was different, color was way different, sharpness, noise reduction, everything. They were just different shots. If you can get that much difference with identical data, what could you possibly prove with different data?

 

Alex



#10 noodle

noodle

    Vostok 1

  • -----
  • Posts: 169
  • Joined: 20 Feb 2006
  • Loc: Hong Kong

Posted 01 September 2020 - 10:22 PM

I did a similar test  a few years ago  :)
STF8300MM (mono CCD)  vs  CentralDs 5Dii (modified  OSC)

both images are calibrated with PI  and non-linear stretched with auto stf/ histogram transf.  , no futher processing .

 

M101_A_B_1.jpg

shooting info. :

https://astro-noodle...l?view=flipcard


  • bugbit likes this

#11 AtmosFearIC

AtmosFearIC

    Apollo

  • -----
  • Posts: 1,315
  • Joined: 10 Dec 2015
  • Loc: Melbourne

Posted 01 September 2020 - 10:38 PM

Gorgeous images AtmosFearIC.

 

I wish I took that last one! The resolution and image smoothness is pretty fantastic.

 

 

Julian,

 

   Is CFA Drizzle not a proper integration method? If you use CFA Drizzle on an OSC it will have the same resolution as its mono counterpart (if the same sensor is available in both varieties).

 

   Just playing devil's advocate here...

 

 

John

 

Theoretically that may be the case but it has never been my experience in practice except for when sampling allows.

 

 

If you call it a challenge, yeah, it could be fun.

 

But if you call it an experiment, there is just no way it will really prove anything. Too many variables changing. An experiment changes only one variable at a time. 

 

And the biggest variable would be processing. We did several shows on The Astro Imaging Channel where we provided identical sets of data. (Everybody processed the same data.) A dozen imagers processed the data. The resulting shots were dramatically different. Just not the same. Star size was different, color was way different, sharpness, noise reduction, everything. They were just different shots. If you can get that much difference with identical data, what could you possibly prove with different data?

 

Alex

 

I completely agree to this. The fourth and fifth image that I found on AB are taken with the same instrument (500mm ASA astrograph) and 16803 but have quite different images. A part of this could very well be from different processing rather than anything else.



#12 GA-HAMAL

GA-HAMAL

    Viking 1

  • -----
  • Posts: 547
  • Joined: 18 Aug 2014
  • Loc: Poland

Posted 02 September 2020 - 01:43 AM

I did some interesting tests once.
Plexiglass strip with lines from the printer.

IMG_2901x.jpg.12b9052e0652ae258d79385ca4

Atik Titan mono and color camera.
Adjustment: pixel line - no pixel line - pixel line

mono

MC-ATM.png.3b429160714c2b6968f1501d51601

color

stripes set on a bayer mask GRGRGR...

MC-ATC1.png.ba091841d157a35db8c8da4fbd7d

MC-ATC1a.jpg.0d1a27675aefe33aca04906bffe

stripes set on a bayer mask BGBGBG...
MC-ATC2.png.341302a2c4bc58655f1234753a46

MC-ATC2a.jpg.e5d5967fce7e43689e607004484

Edited by GA-HAMAL, 02 September 2020 - 01:54 AM.

  • Lead_Weight likes this

#13 GA-HAMAL

GA-HAMAL

    Viking 1

  • -----
  • Posts: 547
  • Joined: 18 Aug 2014
  • Loc: Poland

Posted 02 September 2020 - 01:45 AM

test not subject to weather and seeing fluctuations

Atik Titan mono
xxx-mono.png.af1ab8ec01851b622c8d36ed242

Atik Titan color - debayer
xxx-color-color.png.9f1d4247d3d35d0442cb

Atik Titan color - lum
xxx-color-lum.png.41075c7854263b78f8b98d

Atik Titan color - fit
xxx-color-RAW.png.ae68cc357c1e60ada49f97



and an interesting final comparison of mono and color - stack 1000 frames

ATIK-TITAN-MONO.png.8766d30ba4d45f6a13d2ATIK-TITAN-COLOR.png.4f210457fa4707575c1

Edited by GA-HAMAL, 02 September 2020 - 02:15 AM.

  • Dean J. likes this

#14 AtmosFearIC

AtmosFearIC

    Apollo

  • -----
  • Posts: 1,315
  • Joined: 10 Dec 2015
  • Loc: Melbourne

Posted 02 September 2020 - 02:09 AM

The OSC is holding up better than I was anticipating against the mono being pushed at the edge of the resolution limit.


  • GA-HAMAL likes this

#15 GA-HAMAL

GA-HAMAL

    Viking 1

  • -----
  • Posts: 547
  • Joined: 18 Aug 2014
  • Loc: Poland

Posted 02 September 2020 - 02:13 AM

Well, I was surprised too.
Each stack is 1000 frames. I used dithering by pressing the base on which the optical assembly stood so that the image migrated at the subpixel level.

Edited by GA-HAMAL, 02 September 2020 - 02:20 AM.

  • Dean J. likes this

#16 rockstarbill

rockstarbill

    Voyager 1

  • *****
  • Posts: 10,531
  • Joined: 16 Jul 2013
  • Loc: Snohomish, WA

Posted 02 September 2020 - 03:53 AM

Sure bro, lets go.

 

PacMan Nebula is the target. HA vs HA on OSC. Get prepared to be wrecked. You will use an HA filter right? If not -- the test is flawed.

 

Image scale is 0.9"/px. AGO 10" IDK, ASI6200MM-Pro camera, Chroma 3nm HA filter.


Edited by rockstarbill, 02 September 2020 - 04:02 AM.

  • idclimber likes this

#17 freestar8n

freestar8n

    Vendor - MetaGuide

  • *****
  • Vendors
  • Posts: 11,235
  • Joined: 12 Oct 2007
  • Loc: Melbourne, Australia

Posted 02 September 2020 - 06:00 AM

I think the reason that these challenges don't really exist is because the outcome is easy to foretell. Given proper acquisition, calibration, integration and processing, the mono image will have increased resolution, no matter what. That's really the end of it.. I don't see why you'd need a comparison for that. But I'd be curious to see some results anyway. Sadly I don't have access to either at the moment.

There is absolutely no reason to expect a loss of resolution with OSC if Bayer drizzle is used and if the exposures are dithered.  Both mono and osc involve stacking individual pixels - individually - with alignment into a final image.

 

There is very little difference to expect a loss of efficiency either - since the filter bands are broader with OSC - and they are a better match to the eye.  This is particularly true if the mono filters have a big gap in them for light pollution - and you either image from a dark site or your sky background has no emission peaks in the gap.  You would then equally lose light from the sky background and any continuum objects you are imaging such as galaxies.

 

There is a doubling of G pixels in OSC - but that acts in a similar way to combining L with RGB - since a dominant component of L is from G.

 

These topics have been gone over many times.  Mono and OSC are fundamentally different - and as mentioned above there is a great variation of results due to processing even when given the same data.

 

If people try both and try processing different ways they may come to have a strong preference.  But there is no simple explanation for it that I'm aware of.  Loss of resolution and loss of photon count simply don't apply.  They are just different.

 

Frank


  • Joe G, jdupton, Dean J. and 1 other like this


CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.


Recent Topics






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics