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

How my CDK is 13x faster than my f4.5 APO

  • Please log in to reply
41 replies to this topic

#1 Rouzbeh

Rouzbeh

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1,732
  • Joined: 28 Jun 2006
  • Loc: Canada

Posted 14 January 2022 - 04:13 PM

Here is an interesting example of the same target with the same filter and camera.

 

Both resampled to the same scale of 2.29" / pixel

 

 

  • A 76mm Triplet @ f/4.5with 140 mins (28 x 300s) of integration
  •  
  • 318 CDK  @ f/5.3 with 15 minutes of data (3 x 300s)

 

Interestingly the 3 subs from the big CDK have more SNR.

 

SNR = 35.02 db, APO

 

SNR = 39.62 db, CDK

 

The only advantage of the refractor is its wider FOV, however, that can be matched by making a mosaic. (note the Sharpstar probably wont support a FF sensor, the CDK and reducer will).

 

 

 

Below is a blink GIF (need to click on it).

Attached Thumbnails

  • 76mm 300x28 Ha_pipp.gif

Edited by Rouzbeh, 14 January 2022 - 04:22 PM.

  • lambermo, CharLakeAstro, Sasho_Panov and 2 others like this

#2 Rouzbeh

Rouzbeh

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1,732
  • Joined: 28 Jun 2006
  • Loc: Canada

Posted 14 January 2022 - 04:16 PM

318 CDK  @ f/5.3 with 15 minutes of data (3 x 300s)

 

 

Attached Thumbnails

  • 3x300 CN.jpg


#3 Rouzbeh

Rouzbeh

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1,732
  • Joined: 28 Jun 2006
  • Loc: Canada

Posted 14 January 2022 - 04:16 PM

76mm Triplet @ f/4.5with 140 mins (28 x 300s)

 

 

Attached Thumbnails

  • 76mm 300x28 Ha CN.jpg


#4 Rouzbeh

Rouzbeh

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1,732
  • Joined: 28 Jun 2006
  • Loc: Canada

Posted 14 January 2022 - 04:21 PM

The pixel signal number indicate the same ratio.

 

The intrinsic resolution is also much higher in the CDK image.

 

 

 

277k for the CDK

21k for the APO

 

13x  ratio

 

 

Integration time was

 

140min APO

15min CDK

 

9x ratio, but the CDK data has higher SNR. So in theory, only  (140mins/13) = 11 minutes of exposure was needed to match the same scale and SNR of 140 mins with an f/4.5 refractor with 0% central obstruction! 

 

 

Attached Thumbnails

  • numbers.jpg

  • lambermo likes this

#5 Leia

Leia

    Mariner 2

  • -----
  • Posts: 230
  • Joined: 24 May 2021
  • Loc: SE Alaska

Posted 14 January 2022 - 05:10 PM

Is it just inexperienced me, or is the big star image slightly defocused?



#6 gatsbyiv

gatsbyiv

    Apollo

  • *****
  • Vendors
  • Posts: 1,109
  • Joined: 29 Mar 2009
  • Loc: Doylestown, PA

Posted 14 January 2022 - 05:28 PM

Great image!  Impressive how the Sony chips can be resampled with exceptional results even at 5x5!  Hardware binning isn't needed when the read noise is so low.



#7 Midnight Dan

Midnight Dan

    Hubble

  • *****
  • Posts: 14,845
  • Joined: 23 Jan 2008
  • Loc: Hilton, NY, Yellow Zone (Bortle 4.5)

Posted 14 January 2022 - 05:35 PM

I'm assuming you resampled the CDK to match the APO.  If so, then you effectively binned pixels, which will significantly increase your SNR.

 

-Dan

 

** edit **

 

Yes, just saw your chart.  You binned 5x5 which would of course give you huge improvement in SNR.

 

-Dan


Edited by Midnight Dan, 14 January 2022 - 05:38 PM.


#8 Rouzbeh

Rouzbeh

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1,732
  • Joined: 28 Jun 2006
  • Loc: Canada

Posted 14 January 2022 - 07:16 PM

Is it just inexperienced me, or is the big star image slightly defocused?

I believe that's because the the 12 inch ha a lot more resolution, makes the 76mm image look poor in comparison. Both focused with NINAs autofocus V curve routine.

Perhaps I could fine tune the 76mm reducer spacing, but this was fairly on axis, so don't think it will get much sharper. 



#9 Rouzbeh

Rouzbeh

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1,732
  • Joined: 28 Jun 2006
  • Loc: Canada

Posted 14 January 2022 - 07:17 PM

Great image!  Impressive how the Sony chips can be resampled with exceptional results even at 5x5!  Hardware binning isn't needed when the read noise is so low.

Yes the noise at this gain is only 1.56 (e-)

Its easy to simulate any pixel size with these CMOS chips.



#10 Rouzbeh

Rouzbeh

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1,732
  • Joined: 28 Jun 2006
  • Loc: Canada

Posted 14 January 2022 - 07:19 PM

I'm assuming you resampled the CDK to match the APO.  If so, then you effectively binned pixels, which will significantly increase your SNR.

 

-Dan

 

** edit **

 

Yes, just saw your chart.  You binned 5x5 which would of course give you huge improvement in SNR.

 

-Dan

 

Yes bin5x to match the same scale of the APO. To compare apples to apples. 

 

Keep in mind the almost 10x longer integration time of the APO, the faster f/ ratio, no obstruction and no losses from the 2 x reflecting surfaces.

 

Interesting how with all these, the SNR of just 3 x 300s is much higher with the CDK.


  • Midnight Dan likes this

#11 whwang

whwang

    Soyuz

  • *****
  • Posts: 3,877
  • Joined: 20 Mar 2013

Posted 14 January 2022 - 07:59 PM

I have a similar experience. Recently I shoot M33 with my new 12” RC, which is F8 (with large central obstruction). In just 2 hours, I can get the outer spiral arms well imaged, much better than my expectation based on my past experience of M33 using a 6” refractor, which is F7. I was expecting I would need something like 6 to 8 hr to reach the same quality on the outer spirals.

Rebinning or not, the faint stuff is there.

#12 Midnight Dan

Midnight Dan

    Hubble

  • *****
  • Posts: 14,845
  • Joined: 23 Jan 2008
  • Loc: Hilton, NY, Yellow Zone (Bortle 4.5)

Posted 14 January 2022 - 08:56 PM

Yes bin5x to match the same scale of the APO. To compare apples to apples. 

 

Keep in mind the almost 10x longer integration time of the APO, the faster f/ ratio, no obstruction and no losses from the 2 x reflecting surfaces.

 

Interesting how with all these, the SNR of just 3 x 300s is much higher with the CDK.

Well, it's not really comparing apples to apples.  It's comparing two totally different scopes in ways in which they would rarely be used so they are at the same arc-seconds per pixels.  That pretty much eliminates all other factors except aperture.  

 

In other words, it you're focusing the exact same area of the sky onto the same size pixel (or binned pixel) then obviously a scope with a MUCH larger aperture will be gathering MUCH more light for that pixel and be "faster".  And yor etendue calculations bear this out.

 

There's nothing amazing or surprising about your results given your test parameters.  The CDK is not performing any magic, it just has way more aperture.

 

-Dan 


  • psandelle and Borodog like this

#13 Ron (Lubbock)

Ron (Lubbock)

    Apollo

  • -----
  • Posts: 1,250
  • Joined: 17 Aug 2012
  • Loc: Pittstown, NY

Posted 14 January 2022 - 09:04 PM

The refractor isn't performing any magic either - it's just getting its hind end kicked by its big brother.  lol.gif

 

A lot of people might be interested in discussing the difference in star size between the two images.  It's dramatic.  Can you rule out differences in focus/seeing?


  • lambermo and Midnight Dan like this

#14 Rouzbeh

Rouzbeh

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1,732
  • Joined: 28 Jun 2006
  • Loc: Canada

Posted 14 January 2022 - 10:17 PM

Well, it's not really comparing apples to apples.  It's comparing two totally different scopes in ways in which they would rarely be used so they are at the same arc-seconds per pixels.  That pretty much eliminates all other factors except aperture.  

 

In other words, it you're focusing the exact same area of the sky onto the same size pixel (or binned pixel) then obviously a scope with a MUCH larger aperture will be gathering MUCH more light for that pixel and be "faster".  And yor etendue calculations bear this out.

 

There's nothing amazing or surprising about your results given your test parameters.  The CDK is not performing any magic, it just has way more aperture.

 

-Dan 

Comparing the same resolution. Obviously apples to apples would have been compared to another 80mm APO but that wasn't my intention.

 

No its not magic, I had the numbers already and it was just nice see it in practice and thought it would be nice for other members to see the results too.

 

In fact, knowing this is the reason I shoot with large apertures.

 

The CDK12 is being replaced with the larger (same focal length) CDK14.

The FOV penalty is counteracted with the 0.66x reducer and the new QHY600.


  • mrlovt likes this

#15 Rouzbeh

Rouzbeh

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1,732
  • Joined: 28 Jun 2006
  • Loc: Canada

Posted 14 January 2022 - 10:19 PM

The refractor isn't performing any magic either - it's just getting its hind end kicked by its big brother.  lol.gif

 

A lot of people might be interested in discussing the difference in star size between the two images.  It's dramatic.  Can you rule out differences in focus/seeing?

 

I can't rule out seeing, I cant analyze the subs later, they were on different nights but the same Ha filters.

 

My suspicion is that the reducer spacing might need fine tuning or perhaps the Sharpstar simply isn't all that sharp?



#16 Rouzbeh

Rouzbeh

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1,732
  • Joined: 28 Jun 2006
  • Loc: Canada

Posted 14 January 2022 - 10:23 PM

As to the CDK not being used instead of a small APO for widefield:

 

Given it only needs 10 minutes to get more SNR than a small widefield refractor, it should be possible to bin the  CDK and take lots of panels to make up for the large FOV. 

 

So perhaps 3 x 3 or so and then you have a huge field, good SNR, good resolution and no need to have a secondary widefield apo setup :)

 

I might put that to the test when the CDK14 arrives. The FOV is a not too shabby 1.5 degrees with the FF cam.



#17 Rouzbeh

Rouzbeh

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1,732
  • Joined: 28 Jun 2006
  • Loc: Canada

Posted 14 January 2022 - 10:24 PM

I did post this separately, here is the CDK finished image

 

 

Attached Thumbnails

  • Mel15 80pv2 cn.jpg

  • lambermo, giorgio_ne, mrlovt and 3 others like this

#18 ChrisWhite

ChrisWhite

    Cosmos

  • *****
  • Posts: 7,749
  • Joined: 28 Feb 2015
  • Loc: Northeast USA

Posted 14 January 2022 - 11:01 PM

The larger stars on the undersampled frac is exactly what I would expect to see when compared to a seeing limited system. The finer sampling of the cdk will result in a smaller fwhm.
  • Ron (Lubbock) and CharLakeAstro like this

#19 freestar8n

freestar8n

    Vendor - MetaGuide

  • *****
  • Vendors
  • Posts: 11,192
  • Joined: 12 Oct 2007
  • Loc: Melbourne, Australia

Posted 14 January 2022 - 11:09 PM

I think the language in the original posting has been toned down a bit from implying there is no such thing as a "fast" scope - but I think the point being made is that the refractor isn't "really" faster despite its lower f/ratio - because the examples show how much longer it takes to get the same quality of images.

 

I'm not sure if this is considered news or not - but I would hope the result is obvious and has no impact on the many other ways in which lower f/ratio really does mean "faster" - and the term "fast" is used professionally for this reason.

 

Examples like these suggest that "what really matters" for "speed" is how deep you can go in a given angular patch of the sky to collect an image.  Well - my goodness - isn't it obvious it just depends on aperture and has nothing to do with f/ratio?

 

You could drop two scopes off a building and see which hits the ground faster.  Would it be determined by f/ratio?  Boy - that "fast" optics stuff is bogus!

 

The LSST was designed to be optically fast at f/1.234 - and it is described as a fast system.  They won't be taking images of little galaxies and then blowing them up, cropping them, and putting them on the wall.  They will be collecting light from the ski onto a large sensor - and the optical throughput onto a sensor of a given size is determined by f/ratio - with aperture not playing a direct role.

 

Frank


  • Midnight Dan and psandelle like this

#20 Rouzbeh

Rouzbeh

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1,732
  • Joined: 28 Jun 2006
  • Loc: Canada

Posted 14 January 2022 - 11:22 PM

I think the language in the original posting has been toned down a bit from implying there is no such thing as a "fast" scope - but I think the point being made is that the refractor isn't "really" faster despite its lower f/ratio - because the examples show how much longer it takes to get the same quality of images.

 

I'm not sure if this is considered news or not - but I would hope the result is obvious and has no impact on the many other ways in which lower f/ratio really does mean "faster" - and the term "fast" is used professionally for this reason.

 

Examples like these suggest that "what really matters" for "speed" is how deep you can go in a given angular patch of the sky to collect an image.  Well - my goodness - isn't it obvious it just depends on aperture and has nothing to do with f/ratio?

 

You could drop two scopes off a building and see which hits the ground faster.  Would it be determined by f/ratio?  Boy - that "fast" optics stuff is bogus!

 

The LSST was designed to be optically fast at f/1.234 - and it is described as a fast system.  They won't be taking images of little galaxies and then blowing them up, cropping them, and putting them on the wall.  They will be collecting light from the ski onto a large sensor - and the optical throughput onto a sensor of a given size is determined by f/ratio - with aperture not playing a direct role.

 

Frank

 

Well you do lose hair "faster" with a fast scope, that's for sure!

 

I was debating the PW Delta Rho 350 vs the CDK14 a few months ago when placing the order.

 

The DR is a fast f/3 system but I suspect its will be a lot more demanding on the optical quality, collimation, focus, and narrowband filters. Vignetting will also be an issue I suspect.

 

I worked out the "slow" CDK14 with the reducer can be binned 2x and be faster. If the FOV was required a 2 panel mosaic would cover the same field. 

 

I also figured it can be switched back to an f/7.2 system in minutes if needed for long FL work. The 20 something inch length of the DR350 is very attractive for portability though.



#21 Peter in Reno

Peter in Reno

    Voyager 1

  • *****
  • Posts: 12,483
  • Joined: 15 Jul 2008
  • Loc: Reno, NV

Posted 14 January 2022 - 11:45 PM

I am not sure your test is fair in terms of comparing speed in between two very different scopes. Binning 5x5 of large aperture helps reduce exposure times because the size of pixels increased dramatically. Imaging speed is based on focal ratio and pixel size, not the aperture.

 

A more fair test to compare speeds of both scopes is to use same camera and binning and I'm willing to bet the refractor will win due to shorter focal ratio.

 

Peter


Edited by Peter in Reno, 14 January 2022 - 11:45 PM.

  • psandelle likes this

#22 freestar8n

freestar8n

    Vendor - MetaGuide

  • *****
  • Vendors
  • Posts: 11,192
  • Joined: 12 Oct 2007
  • Loc: Melbourne, Australia

Posted 14 January 2022 - 11:46 PM

Well you do lose hair "faster" with a fast scope, that's for sure!

 

I was debating the PW Delta Rho 350 vs the CDK14 a few months ago when placing the order.

 

The DR is a fast f/3 system but I suspect its will be a lot more demanding on the optical quality, collimation, focus, and narrowband filters. Vignetting will also be an issue I suspect.

 

I worked out the "slow" CDK14 with the reducer can be binned 2x and be faster. If the FOV was required a 2 panel mosaic would cover the same field. 

 

I also figured it can be switched back to an f/7.2 system in minutes if needed for long FL work. The 20 something inch length of the DR350 is very attractive for portability though.

If you are comparing scopes without well defined imaging goals in mind, then there is no way for simple metrics to tell you if scope A is better or worse than scope B.

 

Delta Rho is optimized for very different types of imaging compared to CDK14.  The one thing you know right away, though, is the etendue of each system based on its speed and the corrected field radius in mm.  You don't need to know aperture and you don't need to know pixel size - and right away you know which system has higher etendue - and therefore which system is better at pulling down a large flux of photons from the sky into a sharp image.

 

But all that etendue is wasted if you use a small sensor with it - or if you use a large sensor and then crop/zoom to display individual targets.

 

If you say, I want a field this wide in arc-min and I want it sampled at 0.5" per pixel (binned or whatever) then immediately you know any system that can do that for you will be better if it has larger aperture.  You don't need to know f/ratio as long as you have a sensor that is big enough and pixels that are *small enough* to match that after potential binning.

 

But whatever scope and sensor you choose, you can rest assured that the image that lands on the sensor will be "brighter" in terms of irradiance as strictly determined by f/ratio.  So there is no need to prefer a fast or slow scope if you have already dictated those key imaging parameters.  Just get the largest aperture, highest QE etc. combo of scope and sensor that meets your needs.  And with a low noise sensor, small pixels are perfectly fine and you don't need to worry about their etendue.  You can bin or smooth.

 

Frank


  • hahied, Midnight Dan, psandelle and 1 other like this

#23 Rouzbeh

Rouzbeh

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1,732
  • Joined: 28 Jun 2006
  • Loc: Canada

Posted 15 January 2022 - 04:27 PM

I am not sure your test is fair in terms of comparing speed in between two very different scopes. Binning 5x5 of large aperture helps reduce exposure times because the size of pixels increased dramatically. Imaging speed is based on focal ratio and pixel size, not the aperture.

 

A more fair test to compare speeds of both scopes is to use same camera and binning and I'm willing to bet the refractor will win due to shorter focal ratio.

 

Peter

 

The object is want counts for me.

 

I had 2 ways to capture that object at the same scale.  There is no steadfast rule that says you have use bin1 and compare bin1 to bin 1.

 

Its the image I'm am after, and at a specific size (scale) we can see how much fast the reflector get the job done.



#24 Rouzbeh

Rouzbeh

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1,732
  • Joined: 28 Jun 2006
  • Loc: Canada

Posted 15 January 2022 - 04:32 PM

If you are comparing scopes without well defined imaging goals in mind, then there is no way for simple metrics to tell you if scope A is better or worse than scope B.

 

Delta Rho is optimized for very different types of imaging compared to CDK14.  The one thing you know right away, though, is the etendue of each system based on its speed and the corrected field radius in mm.  You don't need to know aperture and you don't need to know pixel size - and right away you know which system has higher etendue - and therefore which system is better at pulling down a large flux of photons from the sky into a sharp image.

 

But all that etendue is wasted if you use a small sensor with it - or if you use a large sensor and then crop/zoom to display individual targets.

 

If you say, I want a field this wide in arc-min and I want it sampled at 0.5" per pixel (binned or whatever) then immediately you know any system that can do that for you will be better if it has larger aperture.  You don't need to know f/ratio as long as you have a sensor that is big enough and pixels that are *small enough* to match that after potential binning.

 

But whatever scope and sensor you choose, you can rest assured that the image that lands on the sensor will be "brighter" in terms of irradiance as strictly determined by f/ratio.  So there is no need to prefer a fast or slow scope if you have already dictated those key imaging parameters.  Just get the largest aperture, highest QE etc. combo of scope and sensor that meets your needs.  And with a low noise sensor, small pixels are perfectly fine and you don't need to worry about their etendue.  You can bin or smooth.

 

Frank

 

That is what I ended up doing, the largest scope that fits in the dome and AP1100 mount is the CDK14 (17 wont fit in/on either).

 

Full frame sensor + 0.66x reducer to use up 64mm or 90% of that 70mm image circle.

 

Also worth noting, the effective area the CDK14 is 15% larger than that of the DR350. 



#25 Peter in Reno

Peter in Reno

    Voyager 1

  • *****
  • Posts: 12,483
  • Joined: 15 Jul 2008
  • Loc: Reno, NV

Posted 15 January 2022 - 04:39 PM

The object is want counts for me.

 

I had 2 ways to capture that object at the same scale.  There is no steadfast rule that says you have use bin1 and compare bin1 to bin 1.

 

Its the image I'm am after, and at a specific size (scale) we can see how much fast the reflector get the job done.

OK, it's the title of your thread compares two different scopes saying one is faster than other was confusing.

 

At same image scale, yes, aperture matters and not the focal ratio. In most cases, it's easier to rely on focal ratio than image scale for imaging speed.

 

Peter


Edited by Peter in Reno, 15 January 2022 - 04:43 PM.

  • Midnight Dan and psandelle 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