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

Comparison Between DSLR and ASI1600mm-Cool Images

  • Please log in to reply
93 replies to this topic

#1 Peter10

Peter10

    Mariner 2

  • -----
  • topic starter
  • Posts: 247
  • Joined: 31 Mar 2007

Posted 17 September 2016 - 07:56 PM

I have been looking at images taken with enhanced  DSLRs (6D ect...) and the ASI1600mm CMOS camera.  As I understand it, most DSLR cameras have 12-bit A/D converters as does the ASI1600mm. From my perspective, the DSLR images generally look better - with the ASI1600mm images looking "thin" and washed out.  I suppose that this could be a function of the processing but I am not convinced of this.  I doubt this is processing related since it appears common to many images taken with this camera - presumably processed independently by different individuals.  Another explanation may be that people are not using the camera correctly and are not maximizing the S/N ratio that is achievable with better exposure / gain settings.    


  • leveye likes this

#2 G. Hatfield

G. Hatfield

    Surveyor 1

  • *****
  • Posts: 1572
  • Joined: 31 Jan 2009
  • Loc: Tucson, Arizona

Posted 17 September 2016 - 08:00 PM

Some examples?

 

George



#3 Ron359

Ron359

    Apollo

  • *****
  • Posts: 1120
  • Joined: 21 Apr 2008
  • Loc: -105 +39

Posted 17 September 2016 - 10:24 PM

The 6D is 14 bit (as are some other DSLRs), 20 meg with 6.55 micron pixels, 43.04 mm on diagonal sensor. There is also more in camera (DIGIC) processing for noise reduction, color balance etc.   So not really comparable to the ASI1600.  It might be more comparable to the older smaller sensor EOS cameras.  



#4 whwang

whwang

    Mercury-Atlas

  • *****
  • Posts: 2660
  • Joined: 20 Mar 2013

Posted 17 September 2016 - 10:36 PM

I don't think you can judge cameras based on some random set of pictures on the internet.  There are too many variables.  The only fair way is to conduct side-by-side test: same location, same weather, same optics, same exposure time, and same (good) processing.  If any one of these requirements is not met, I don't trust the comparison.

 

The resolution of the AD conversion usually only affect the saturation level.  It should not have any impact to the quality in the dark part of the image, if the camera (DSLR or mono camera) is used correctly.  This is a big IF though.  Not everybody knows what he/she is doing.

 

Cheers,

Wei-Hao


  • G. Hatfield and xxTRIPODxx like this

#5 Jerry Lodriguss

Jerry Lodriguss

    Vendor

  • *****
  • Vendors
  • Posts: 6988
  • Joined: 19 Jul 2008
  • Loc: Voorhees, NJ

Posted 17 September 2016 - 11:34 PM

I don't think you can judge cameras based on some random set of pictures on the internet.  There are too many variables.  The only fair way is to conduct side-by-side test: same location, same weather, same optics, same exposure time, and same (good) processing.  If any one of these requirements is not met, I don't trust the comparison.

 

The resolution of the AD conversion usually only affect the saturation level.  It should not have any impact to the quality in the dark part of the image, if the camera (DSLR or mono camera) is used correctly.  This is a big IF though.  Not everybody knows what he/she is doing.

 

Here's my favorite astrophotography lyric: "If I knew what I was doing, I'd be doing it right now."  (In the song "I'm In", written by Radney Foster and George Middleman.) 

 

I have to think that the ASI1600mm just hasn't been around long enough for the masters to have taken it up.

 

Ideally to compare them you would want to shoot them at the same time to account for possible differences in the sky on different nights, or even at different times of the same night (high clouds, haze, etc), and with the same scope.

 

Then you would also have to consider how to weigh the fact that the ASI1600mm is cooled, and DSLRs generally aren't. 

 

I wonder how much signal processing is going on on the sensor of the CMOS chip in the ASI1600mm compared to, say, a Canon 6D, or Nikon D810a. I guess if the signal processing is built into the chip, it would be a similar level of technology, although camera makers like Nikon seem to get more out of the Sony sensors than Sony can.

 

These new cameras that have dark current suppression are great, but that's not removing the noise associated with the thermal signal, so cooling in the ASI1600mm alone is a tremendous benefit in warmer climates.

 

I would think that except for the cooling, the results should be fairly comparable between an ASI1600mm and a new-model DSLR with a low-noise Sony sensor.   I mean, they are not going to be an order of magnitude different.

 

Other than the monochrome part, the cooling is probably the only major difference I would guess. Not that monochrome isn't a big deal if you want to shoot narrowband, it's a huge deal if you do.

 

What do you think Wei-Hao?

 

Jerry


Edited by Jerry Lodriguss, 17 September 2016 - 11:58 PM.


#6 Herra Kuulapaa

Herra Kuulapaa

    Viking 1

  • -----
  • Posts: 509
  • Joined: 10 Dec 2014
  • Loc: Vantaa, Finland

Posted 18 September 2016 - 02:32 AM

Even though I have worked with cooled DSLRs (RGB/mono) for some time now, I find it difficult and time consuming to make objective comparison between them and dedicated astrocameras like ASI1600. Just like Wei-Hao said. I have some test data, but it's not comparable in every way so I rather not publish anything. For example it's not very fair to compare very different sensor sizes (35mm/micro 4/3).

 

BUT, I have now two identical mount/scope setups with guiding so I'd be interested testing same size cooled IMX-071 and ASI1600MC with the most objective way I possible could. Identical gear, exposure time at same time, unity gain, Nikon in true raw mode etc. I'm however lacking the ASI1600MC.

 

Hmmm., I wonder if we could do a comparison with a support from ZWO. What do you think Sam if you're reading this?



#7 whwang

whwang

    Mercury-Atlas

  • *****
  • Posts: 2660
  • Joined: 20 Mar 2013

Posted 18 September 2016 - 03:14 AM

Cooled CMOS cameras are not around for too long, so it is still hard to compare them with DSLRs at every possible level.  Like what Jerry pointed out, even with the same SONY CMOS, different camera companies seem to be able to get different levels of performance out of it.  How it will work on astronomical cooled cameras is still a question mark.  It is too early to tell.

 

Suppose the camera can deliver what the vendor promises, then it comes down to two major differences: cooling and mono.  (People mentioned pixel size and sensor area.  However, even among DSLRs, there are differences in pixel size and sensor area.  So let's just ignore this part for a moment.)

 

In this forum, for many times, I have demonstrated that recent CMOS-based DSLRs do not need cooling when used in places cooler than 10-15 degC, for basic RGB (i.e., not narrow-band) imaging.  On the other hand, if one images in a warm place, warmer than 25 degC, then cooling will definitely help.  There is no doubt about this.  Cooling is also very useful for narrow-band imaging, even in a cool place.  Furthermore, a mono sensor can boost the efficiency of narrow-band imaging by 2x to 4x (depending on which filter you use).  This is very significant.

 

So, if you want to do primarily narrow-band, or if you have to image in a warm place, then a cooled mono CMOS camera is a no-brainer.  

 

This is just theory.  Whether the ASI camera people are talking about can really deliver what it promises, or whether it has any potential issues, is a totally different question.

 

No matter what, I will never worry about the AD conversion.  12bit is not bad at all.

 

Cheers,

Wei-Hao


  • calypsob likes this

#8 akulapanam

akulapanam

    Mercury-Atlas

  • *****
  • Posts: 2515
  • Joined: 26 Aug 2012

Posted 18 September 2016 - 03:20 AM

Have you seen the astrobin group? Some of those shots, especially the galaxies are not achievable with a dslr . Also the lack of noise overall is pretty impressive by comparison.

http://www.astrobin.com/groups/31/


Edited by akulapanam, 18 September 2016 - 03:21 AM.


#9 whwang

whwang

    Mercury-Atlas

  • *****
  • Posts: 2660
  • Joined: 20 Mar 2013

Posted 18 September 2016 - 05:05 AM

I just took a look at the astrobin group. Those images are not too bad, some are actually quite nice. However, except for a few narrow-band images, with respect, I haven't seen anything unachievable with DSLRs.

Edited by whwang, 18 September 2016 - 05:06 AM.

  • Herra Kuulapaa likes this

#10 jgraham

jgraham

    Hubble

  • *****
  • Posts: 19973
  • Joined: 02 Dec 2004
  • Loc: Miami Valley Astronomical Society

Posted 18 September 2016 - 07:23 AM

"...most DSLR cameras have 12-bit A/D converters as does the ASI1600mm..."

 

I'm not sure how many DSLRs still use a 12 bit ADC, but for me one of the biggest improvements in Canon DSLRs was when they moved from a 12 bit ADC with the 400D (XTi) to a 14 bit ADC with the 450D (XSi). I was actually in the process of ordering an ASI 1600 when I noticed that it used a 12 bit ADC. That really surprised me, I would never have thought that a modern CCD would go back to using any less than a 16 bit ADC. There is no way that I'd pay that kind of money for a 12 bit camera. I might be able to use one if I were imaging under dark skies where the wells fill up with starlight, but not from my red zone backyard where the starlight is diluted with skyglow. 12 bits just doesn't give you a lot of room to work with, particularly when you start using curves.

 

For some it may not be a big deal, but for me it was a deal breaker.



#11 mike8888

mike8888

    Viking 1

  • *****
  • Posts: 525
  • Joined: 19 Mar 2015

Posted 18 September 2016 - 01:44 PM

As somewhat of a newbie, could somebody explain 12 bit, ADC vs. 14 bit. Does this refer to the amount of data the 'wells' hold or some other factor/benefit? Thanks!

 

Mike



#12 Charlie B

Charlie B

    Apollo

  • -----
  • Posts: 1422
  • Joined: 22 Mar 2008
  • Loc: Sterling, Virginia

Posted 18 September 2016 - 02:18 PM

As somewhat of a newbie, could somebody explain 12 bit, ADC vs. 14 bit. Does this refer to the amount of data the 'wells' hold or some other factor/benefit? Thanks!

 

Mike

Mike,  A 12-bit Analog to Digital Converter (ADC) samples photon well values into 4096 levels (2^(12)) and a 14-bit ADC has 16384 separate levels.  16-bit cameras can separate the photon data into 65636 levels.  The more levels, the more you can differentiate, by image processing, small amplitude features, for example, a 16 bit ADC can see can see amplitudes 4 times finer than a 12-bit ADC.

 

Regards,

 

Charlie



#13 entilza

entilza

    Soyuz

  • *****
  • Posts: 3695
  • Joined: 06 Oct 2014
  • Loc: Oakville, ON, Canada

Posted 18 September 2016 - 05:45 PM

So far I have been very impressed with the data I've seen with the ASI1600 .. I have quite  a few hours with my Nikon now to have a good judge of the data, and it is better, this is comparing just the colour version as well.  The issue would be if we could cool the Nikon to -20 then I think the Nikon would be just as good if not better.   This winter I will have my Nikon out at lower temperatures to really push it, however in the end I think the ASI1600 is going to be the minimum standard moving forward.  QHY is developing their version, this can only get bigger and better.

 

The Nikon D5300 is still much cheaper and has a larger sensor so that is still a big plus.  I really wish I could have a cooled D5300 sony chip.


  • bsavoie likes this

#14 akulapanam

akulapanam

    Mercury-Atlas

  • *****
  • Posts: 2515
  • Joined: 26 Aug 2012

Posted 18 September 2016 - 06:28 PM

It's not just cool that makes a difference. It's the low noise electronics especially on the read side. There is variance even between scientific camera manufactures here, for example fli.



#15 whwang

whwang

    Mercury-Atlas

  • *****
  • Posts: 2660
  • Joined: 20 Mar 2013

Posted 18 September 2016 - 07:11 PM

Let me remind people that, as long as the gain reaches 1 ADU/e-, you don't need a finer AD conversion.  There is no point going finer than 1e-.   You can't have 0.5 electron, can you?  Under this condition, the only difference between 12 bit and 14 bit is just the saturation level.



#16 jgraham

jgraham

    Hubble

  • *****
  • Posts: 19973
  • Joined: 02 Dec 2004
  • Loc: Miami Valley Astronomical Society

Posted 18 September 2016 - 09:56 PM

I could agree with that if all you were recording was useable signal such as you would get with dark skies. Under urban skies you are gonna lose the first couple of bits (or more) to skyglow. When you subtract that off suddenly your 12 bit ADC is 10 bits or less. Way back when I first started imaging with a 350D and a 400D I was constantly bumping up against the limits of a 12 bit ADC. The fix was to shift over to 16 bit CCDs for a while. I was lured back to DSLRs when they moved up to a 14 bit ADC. I'd still prefer a 16 bit ADC, but 14 bits has worked out fine. I have seen some very good images taken with the ASI 1600, but I just can't imagine spending that kinda money on a 12 bit camera. From the CCD side of things the price of the ASI 1600 may be attractive, but from the DSLR side, not so much.

 

Food for thought.



#17 whwang

whwang

    Mercury-Atlas

  • *****
  • Posts: 2660
  • Joined: 20 Mar 2013

Posted 18 September 2016 - 10:33 PM

Hi John,

 

If the dynamical range is reduced from 14 bit (16383) to 12 bit (4095) because of the floor of sky background, then the sky brightness would be 16383-4096 = 12288.  In other words, in your image, the sky brightness is 75% of the saturation level (12288/16383 = 0.75).  I think what's more logically needed here is just to reduce your exposure time (by paying the price of more subs), not to get a 16 bit camera.  There is really no point exposing so long that the sky is 75% of the saturation level.  Just make the sky several tens times higher than the readout noise should be sufficient.  

 

The readout noise is usually 2-3 electrons in CMOS cameras.  With properly chosen ISO or readout gain, this is is roughly 2 to 10 ADU.  In other words, there is no need to make the sky much brighter than a few hundreds ADU.

 

My 2 cents.

 

Cheers,

Wei-Hao


Edited by whwang, 18 September 2016 - 10:58 PM.

  • Alfredo Beltran and calypsob like this

#18 nicolasM

nicolasM

    Ranger 4

  • *****
  • Posts: 321
  • Joined: 24 May 2015
  • Loc: Landed in Versailles (Fr)

Posted 19 September 2016 - 08:43 AM

Hi John,

 

If the dynamical range is reduced from 14 bit (16383) to 12 bit (4095) because of the floor of sky background, then the sky brightness would be 16383-4096 = 12288.  In other words, in your image, the sky brightness is 75% of the saturation level (12288/16383 = 0.75).  I think what's more logically needed here is just to reduce your exposure time (by paying the price of more subs), not to get a 16 bit camera.  There is really no point exposing so long that the sky is 75% of the saturation level.  Just make the sky several tens times higher than the readout noise should be sufficient.  

 

The readout noise is usually 2-3 electrons in CMOS cameras.  With properly chosen ISO or readout gain, this is is roughly 2 to 10 ADU.  In other words, there is no need to make the sky much brighter than a few hundreds ADU.

 

My 2 cents.

 

Cheers,

Wei-Hao

Hi Wei-Hao, 

 

What you wrote here is very interesting for my case, but I don't understand your explanations. Could you please elaborate?

I am living in a highly polluted environment and I would like to determine the best solution to image in these conditions. 

 

Thank you, 



#19 spokeshave

spokeshave

    Surveyor 1

  • *****
  • Posts: 1901
  • Joined: 08 Apr 2015

Posted 19 September 2016 - 09:51 AM

"...most DSLR cameras have 12-bit A/D converters as does the ASI1600mm..."

 

I'm not sure how many DSLRs still use a 12 bit ADC, but for me one of the biggest improvements in Canon DSLRs was when they moved from a 12 bit ADC with the 400D (XTi) to a 14 bit ADC with the 450D (XSi). I was actually in the process of ordering an ASI 1600 when I noticed that it used a 12 bit ADC. That really surprised me, I would never have thought that a modern CCD would go back to using any less than a 16 bit ADC. There is no way that I'd pay that kind of money for a 12 bit camera. I might be able to use one if I were imaging under dark skies where the wells fill up with starlight, but not from my red zone backyard where the starlight is diluted with skyglow. 12 bits just doesn't give you a lot of room to work with, particularly when you start using curves.

 

For some it may not be a big deal, but for me it was a deal breaker.

If all we did was process single 12-bit images, I would agree with you. But we don't. We collect typically tens (sometimes hundreds or thousands) of sub-exposures and then stack them by averaging. In most cases, the stacking is done in 32-bit space. The averaging process has the effect of legitimately increasing the bit depth to match the stacking space bit depth at a cost of a small amount of noise due to quantization error. The more subs we stack, the less quantization noise is introduced. This noise is far more than made up for by the low read noise in a camera like the ASI1600mm.

 

Tim


  • Alfredo Beltran, whwang and Jon Rista like this

#20 bobzeq25

bobzeq25

    Hubble

  • *****
  • Posts: 15771
  • Joined: 27 Oct 2014

Posted 19 September 2016 - 10:53 AM

Keeping this short.  With stacking any practical differences between 12 and 14 bit are really marginal.

 

Cooling is a large advantage, particularly in warm weather.  I don't do darks with my cooled camera.  If you're talking mono 1600 with filters, it will have a large advantage over a DSLR, with separate filters instead of a matrix, all the pixels are working all the time.  Can't beat that for signal to noise ratio and resolution.

 

It's easy to get lost in the weeds here, when there are big things going on.


Edited by bobzeq25, 19 September 2016 - 10:56 AM.

  • Gucky and AstroCatinfo like this

#21 JukkaP

JukkaP

    Viking 1

  • -----
  • Posts: 529
  • Joined: 14 Aug 2016

Posted 19 September 2016 - 12:57 PM

+1 to bobzeq. Mono camera is totally different then colour. Asi 1600 colour vs. Dslr is interesting. I dont get why ppl would buy large chip cmos "astrocamera" whit the cost of 3 equal dslr. Only the cooling is different. And software and support.... So it's starting to sound interesting after all. My messy text sounds good on my own language so bare whit me. The point is that there is no best option. It allways comes to user preferences. Like me I like dslr for the cost and for the part i can do alot to the camera myself. But to be fair, i think i would get better results whit dedicated astro camera.

I understand colour planetary camera. But would go for mono if i were to buy proper astrocamera.

Herra kuulapaa would tell us the truth. Someone send him a asi 1600 mono :).

I'am in prosess of making my coldfinger cooled nikon d5100 a mono camera. But its just for fun. If i were serious i would get large chip ccd. And i will in some point. The first models are allways upgraded after year or so.

#22 Herra Kuulapaa

Herra Kuulapaa

    Viking 1

  • -----
  • Posts: 509
  • Joined: 10 Dec 2014
  • Loc: Vantaa, Finland

Posted 19 September 2016 - 01:08 PM

Wow, Truth?, well.., my plan was not the mono, but color so we could compare the actual unmodified sensors cooled to same temperature like -10 / -15C / -20C. Hopefully I will be able to conduct such test some point during the coming winter. We'll see..



#23 JukkaP

JukkaP

    Viking 1

  • -----
  • Posts: 529
  • Joined: 14 Aug 2016

Posted 19 September 2016 - 03:41 PM

That will be more accurate comparison. Dont think there can be huge difference between cooled dslr and cooled asi 1600 colour.

Was curious about the differenses in mono modified dslr and the mono version of asi 1600.

#24 Ron359

Ron359

    Apollo

  • *****
  • Posts: 1120
  • Joined: 21 Apr 2008
  • Loc: -105 +39

Posted 19 September 2016 - 04:10 PM

Keeping this short.  With stacking any practical differences between 12 and 14 bit are really marginal.

 

Cooling is a large advantage, particularly in warm weather.  I don't do darks with my cooled camera.  If you're talking mono 1600 with filters, it will have a large advantage over a DSLR, with separate filters instead of a matrix, all the pixels are working all the time.  Can't beat that for signal to noise ratio and resolution.

 

It's easy to get lost in the weeds here, when there are big things going on.

 

That will be more accurate comparison. Dont think there can be huge difference between cooled dslr and cooled asi 1600 colour.

Was curious about the differenses in mono modified dslr and the mono version of asi 1600.

Talk about wandering off into the weeds...  Of course there is no comparing a cooled mono camera w/ filters to a bayer color DSLR, especially for bit depth.  But that wasn't the OP question.    



#25 bobzeq25

bobzeq25

    Hubble

  • *****
  • Posts: 15771
  • Joined: 27 Oct 2014

Posted 19 September 2016 - 05:07 PM

Talk about wandering off into the weeds...  Of course there is no comparing a cooled mono camera w/ filters to a bayer color DSLR, especially for bit depth.  But that wasn't the OP question.    

 

Why not?  The title explicitly said he was interested in the comparison between a DSLR and the ASI1600 mono - cooled.  Those were the salient points.


Edited by bobzeq25, 19 September 2016 - 05:09 PM.



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