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

Opinions on Grade 1 vs. Grade 2 CCD?

  • Please log in to reply
160 replies to this topic

#1 Jon Rista

Jon Rista

    ISS

  • *****
  • topic starter
  • Posts: 24,994
  • Joined: 10 Jan 2014
  • Loc: Colorado

Posted 12 March 2016 - 01:57 PM

I am working on putting together a Moravian G3-16200 camera package. This will be my first CCD, I had almost pulled the trigger on a QSI 683 last year, but when the KAF-16200 cameras were announced it sounded like a much better compromise in terms of field of view between my existing 5D III and a smaller CCD sensor. Anyway, I have most of the details figured out. I chose Moravian because the cost is good, they have a 7-pos FW option, an OAG option, and both M48 and Canon EF mount options The one thing I am not sure on is whether I should spend the extra $830 or so on a Grade 1 sensor or not, or settle for a Grade 2 and put that money towards filters (which, being 2", are very expensive...the Astrodon E-series LRGB 50mm are $950, and EACH of the Astrodon 3nm NB filters are $995 (although I may start with a 5nm Ha which is $725). So the extra cost for Grade 1 covers the vast majority of the cost for either LRGB or an NB filter. I'd like to start out with both Ha and OIII, so where to invest that $830 is basically my key decision.

 

Does anyone have experience with both Grade 1 and Grade 2 KAF sensors? I've read some reports that bad columns on Grade 2 CCDs can increase in count over time...however the question of whether that can also happen to a Grade 1 CCD has never actually been answered (although I have seen it asked.) If a CCD sensor can actually degrade over time, I feel investing in a Grade 1 that could likely develop bad columns would be a waste of money.

 

I am also curious if anyone has used CosmeticCorrection in PixInsight to fix bad columns? I have been working with CosmeticCorrection for a while now, and I have found it is actually quite effective at replacing dead pixels with meaningful content from it's neighbors. It even does it properly with CFA data if you enable the CFA option, referencing the like-colored pixels primarily to fill in for hot or cold pixels. I know that CosmeticCorrection has the ability to map out rows or columns, either entirely, or with a limiting range (in case only part of the column is dead). I just don't know exactly how it works as far as filling in the missing content for those rows or columns. 

 

Anyway, real-world insights into whether getting a Class 1 sensor really matters for astrophotography would be very welcome. I did find this on Atik's web site, which is interesting:

CCD Grades

 

The Atik 11000 is fitted with a KAI 11002 class II sensor as standard. This allows for a small number of column defects. Alternatively, a class I sensor with no column defects can be specified. Does the CCD grade matter? Well, we feel the price premium placed on the class I sensor is not justified when the camera is going to be used for astro-imaging.

 

In short, this is because Kodak’s grading system uses a test in which the CCD is not cooled; exposures are for fractions of a second; and the sensor is read out several times a second. But conditions for the sensors are very different in astro-imaging because the sensor is cooled; exposures are typically for many tens of seconds; and readout takes several seconds. It is also expected that the images will be corrected with a dark frame. The upshot is that Kodak’s grading tests do not reflect astro-imaging performance in a useful way.

 

 

However the "growth" of column defects that some have mentioned with Grade 2 CCDs is the primary concern. If column defects can form over time (due to powerful cosmic ray strikes? I've seen new hot pixels appear on my 5D III due to that a few times.) then I am very curious whether it can happen with a Grade 1 CCD as well.


  • Charlespl likes this

#2 orlyandico

orlyandico

    Cosmos

  • *****
  • Posts: 9,881
  • Joined: 10 Aug 2009
  • Loc: Singapore

Posted 12 March 2016 - 02:33 PM

if you've seen the raw images from some of the "professional" astronomical CCD's (some of which can be obtained on arXiv) you will not be concerned at all about a few column defects...   :p

 

but it is your money..


  • Jon Rista likes this

#3 WesC

WesC

    Fly Me to the Moon

  • *****
  • Posts: 5,323
  • Joined: 06 Feb 2013
  • Loc: La Crescenta, CA

Posted 12 March 2016 - 02:41 PM

I use cosmetic correction to remove the "bad" column in my QSI 683... Most of the time that takes care of it completely. You do have to have a decent amount of subs though, at least 10.

 

You can also dither, which will supposedly help a lot more. I haven't tried that yet though. 


Edited by WesC, 12 March 2016 - 02:43 PM.


#4 Ricky

Ricky

    Gemini

  • *****
  • Posts: 3,321
  • Joined: 09 Sep 2003
  • Loc: Nor Cal

Posted 12 March 2016 - 06:03 PM

It seems that there are many more class 2 versus class 1 images floating around and to be honest...I can't tell the difference (with proper post processing).  The cost difference, to me, wouldn't be worth it.  But if I got the $$$ i'd probably get the class 1...


Edited by Ricky, 12 March 2016 - 06:03 PM.


#5 Scott Sloka

Scott Sloka

    Vostok 1

  • -----
  • Posts: 138
  • Joined: 03 Sep 2011

Posted 12 March 2016 - 08:05 PM

Hi Jon,

 

I use cosmetic correction as I have a few column defects on my 16803 sensor. I use the "master dark" option as it seems to work well, although I have yet to explore the column mapping option. I love that it seems to map the defect pixels to the regional background very effectively. If I forget to use it while stacking, the column defects are very obvious. Very cool function.

 

Please let me know if you would like a sample image to play with, if that helps you make your decision.

 

Scott



#6 Jon Rista

Jon Rista

    ISS

  • *****
  • topic starter
  • Posts: 24,994
  • Joined: 10 Jan 2014
  • Loc: Colorado

Posted 13 March 2016 - 01:30 AM

Thanks for the input, everyone! I appreciate it. I think I'll go with a Class 2, as it sounds like CosmeticCorrection will handle the problem (if it even exists) quite well. Going to save my money and put it towards filters. 



#7 WesC

WesC

    Fly Me to the Moon

  • *****
  • Posts: 5,323
  • Joined: 06 Feb 2013
  • Loc: La Crescenta, CA

Posted 13 March 2016 - 02:52 PM

Hey Jon,

Here's an example of how well Cosmetic Correction cleans up my column defect. Keep in mind my seeing was terrible for this image and so my processing is pretty aggressive...  however you're looking at a 300% crop here, so it looks pretty bad.

 

But, anyway, in the final stack, the column defect is long gone. And I'm not dithering for this either.

 

CD_cleanup.JPG

 

 


  • Hobby Astronomer likes this

#8 Madratter

Madratter

    Voyager 1

  • *****
  • Posts: 13,238
  • Joined: 14 Jan 2013

Posted 13 March 2016 - 02:54 PM

Personally, I would go Grade 1. I know myself and any column defects would really bother me, even though I know how to process them out.


Edited by Madratter, 13 March 2016 - 02:58 PM.


#9 Jon Rista

Jon Rista

    ISS

  • *****
  • topic starter
  • Posts: 24,994
  • Joined: 10 Jan 2014
  • Loc: Colorado

Posted 13 March 2016 - 05:47 PM

Thanks, Wes! Exactly the kind of info I was looking for. I've had good luck with CosmeticCorrection so far, so I don't think it will be a huge problem. I am already using it to correct stuck pixels in my flats, so correcting any potentially bad columns shouldn't be a problem (especially without the CFA.) 

 

I've done some more research, and it sounds like column defects can form on Grade 1 and even Grade 0 sensors. I guess it's the fundamental difference in design between CCD and CMOS sensors that allows it to happen. The whole charge shifting thing. Seems like a powerful enough cosmic ray strike could potentially kill off either a whole column or part of a column. Who knows what else might cause it. Anyway...given that, I would rather not spend the extra money buying a CCD that might not remain column-defect free forever. I know for a fact that my 5D III has had whole small clusters of pixels go dead or end up stuck due to cosmic ray strikes (was blinking through a set from late last year...all the subs before a particular strike were fine, all the subs after had a 2x3 pixel matrix that looked mostly dead. O_o)



#10 pfile

pfile

    Fly Me to the Moon

  • -----
  • Posts: 5,779
  • Joined: 14 Jun 2009

Posted 14 March 2016 - 12:14 PM

i don't know if there's a correlation between class1/class2 and the rate at which a sensor degrades, but degradation is the big PITA here. if a class2 sensor develops bad columns at a higher rate, then you have to get new darks and bias all the time. and if you choose to fix the bad columns with CC (which works great), you're having to go in and tweak the settings all the time.

 

i think my STT-8300M has a class1 sensor and it has zero column defects and the bias and dark signal has been very consistent for years. there are cosmic ray hits all over my darks/lights and it seems none the worse after 3 years.

 

rob



#11 Jon Rista

Jon Rista

    ISS

  • *****
  • topic starter
  • Posts: 24,994
  • Joined: 10 Jan 2014
  • Loc: Colorado

Posted 14 March 2016 - 02:16 PM

From what I was reading, it takes a powerful ray hit to cause damage. I've had countless in my frames over the last couple of years, only the one in particular seemed to do any real damage, and another that resulted in a stuck pixel (which even shows up in my bias). I haven't found any information about the rate of degradation. That is an interesting point...I wonder if there is information about that somewhere?



#12 WesC

WesC

    Fly Me to the Moon

  • *****
  • Posts: 5,323
  • Joined: 06 Feb 2013
  • Loc: La Crescenta, CA

Posted 14 March 2016 - 07:43 PM

My QSI developed its defect after only 1 month. I was REALLY bummed out.



#13 Jon Rista

Jon Rista

    ISS

  • *****
  • topic starter
  • Posts: 24,994
  • Joined: 10 Jan 2014
  • Loc: Colorado

Posted 14 March 2016 - 11:06 PM

Has it developed any more since?



#14 WesC

WesC

    Fly Me to the Moon

  • *****
  • Posts: 5,323
  • Joined: 06 Feb 2013
  • Loc: La Crescenta, CA

Posted 15 March 2016 - 11:33 AM

Nope, just that one. Honestly I think it might have been heat related. Last summer was very warm, even in the evenings. somewhere around the mid 90s.

 

Even so I was able to keep the cooler running at 80% to reach -20C. Pretty impressive. I tried to push it to -30C, but the coolest it would go was -23.8C.



#15 tianzi9088

tianzi9088

    Explorer 1

  • *****
  • Posts: 92
  • Joined: 13 Jan 2015
  • Loc: Newton, MA

Posted 15 March 2016 - 12:01 PM

Wes,

I've heard it's highly unrecommended to over cooling qsi camera beyond -20C, which would result in building vertical banding issues.



#16 JJK

JJK

    Gemini

  • *****
  • Posts: 3,466
  • Joined: 27 Apr 2008

Posted 15 March 2016 - 02:04 PM

Personally, I would go Grade 1. I know myself and any column defects would really bother me, even though I know how to process them out.

Me too, and I ordered one from FLI.

 

For the 16200, there could be a significant difference in the number of defect between Grade 1 and Grade 3 CCD chips.



#17 JJK

JJK

    Gemini

  • *****
  • Posts: 3,466
  • Joined: 27 Apr 2008

Posted 15 March 2016 - 02:09 PM

Wes,

I've heard it's highly unrecommended to over cooling qsi camera beyond -20C, which would result in building vertical banding issues.

If that's true, what is the cause?

 

I've cooled my FLI MicroLine 8300 (zero defects) to -30*C and beyond w/o any issues.



#18 WesC

WesC

    Fly Me to the Moon

  • *****
  • Posts: 5,323
  • Joined: 06 Feb 2013
  • Loc: La Crescenta, CA

Posted 15 March 2016 - 02:12 PM

Wes,

I've heard it's highly unrecommended to over cooling qsi camera beyond -20C, which would result in building vertical banding issues.

 

QSI specs this camera to cool to -45C. Who is making this recommendation? QSI isn't... they've told me that cooling below -20C in the summer is unlikely, but going lower in the winter is certainly achievable. But I stick to -20C because I can use the same darks year round.



#19 Madratter

Madratter

    Voyager 1

  • *****
  • Posts: 13,238
  • Joined: 14 Jan 2013

Posted 15 March 2016 - 03:01 PM

My guess is the reason for the recommendation is possible RBI. I would simply test and see if it is a real problem or not.

#20 JJK

JJK

    Gemini

  • *****
  • Posts: 3,466
  • Joined: 27 Apr 2008

Posted 15 March 2016 - 03:02 PM

My guess is the reason for the recommendation is possible RBI. I would simply test and see if it is a real problem or not.

But RBI can be mitigated with flooding.



#21 Madratter

Madratter

    Voyager 1

  • *****
  • Posts: 13,238
  • Joined: 14 Jan 2013

Posted 15 March 2016 - 03:10 PM

True. But QSI probably doesn't have that built in. I know that SBIG only has flooding implemented for some of their cameras.

#22 tianzi9088

tianzi9088

    Explorer 1

  • *****
  • Posts: 92
  • Joined: 13 Jan 2015
  • Loc: Newton, MA

Posted 15 March 2016 - 03:24 PM

I asked QSI about the precautions of operating QSI camera in typical New England winter last november when I sent my 683 back for repair, and I've got a nice written-up from Andy who was working on my problem at that time. I have to say it's quite a long written up and show how responsible and professional the staffs at QSI are. There are some quotes:

". When the camera body is very cold, we recommend powering on the camera, with the CCD

cooler running at the set point for at least 10 to 15 minutes prior to starting your imaging
session.
. The fans should be set to automatic/quiet, not full. Running the fans on the full setting in cold
conditions could drive the camera below the operating range.
. If you're using the KAF-8300, the lowest operating sensor temperature specified by the CCD
manufacturer is -20C. We do not recommend operating the CCD cooler below this
recommended set point
."

 

And there is also a nice summary for that:

". CCDs and electronics change behavior with temperature and over time. In general, QSI does not

recommend using significantly older Master calibration images for new Light images, especially
if the ambient temperatures vary significantly, or if calibration anomalies appear.

 

. Avoid over-cooling the camera body – don’t use the fans if they aren’t needed. The heat
generated by the thermionic cooler actually helps keep the electronics in the required
temperature range. Ideally, the cooler power should run between 20 and 80 percent. Running it
outside of this range can limit headroom and interfere with proper temperature regulation.

 

. The KAF-8300 sensor's minimum temperature set point is specified by OnSemi at -20C. QSI
recommends not running the cooler below this temperature. If it is extremely cold out, you may
need to warm the camera body to achieve proper temperature regulation. If necessary, a dew
heater or some type of insulation will raise the body temperature.
"


Edited by tianzi9088, 15 March 2016 - 03:26 PM.


#23 Jon Rista

Jon Rista

    ISS

  • *****
  • topic starter
  • Posts: 24,994
  • Joined: 10 Jan 2014
  • Loc: Colorado

Posted 15 March 2016 - 03:35 PM

True. But QSI probably doesn't have that built in. I know that SBIG only has flooding implemented for some of their cameras.

 

SBIG and FLI do support flooding, but only for the deeper cooled chips, like the KAF-16803. I've not seen flooding on any KAF-8300 camera from any manufacturer. 



#24 Jon Rista

Jon Rista

    ISS

  • *****
  • topic starter
  • Posts: 24,994
  • Joined: 10 Jan 2014
  • Loc: Colorado

Posted 15 March 2016 - 03:37 PM

 

Wes,

I've heard it's highly unrecommended to over cooling qsi camera beyond -20C, which would result in building vertical banding issues.

 

QSI specs this camera to cool to -45C. Who is making this recommendation? QSI isn't... they've told me that cooling below -20C in the summer is unlikely, but going lower in the winter is certainly achievable. But I stick to -20C because I can use the same darks year round.

 

 

Technically speaking, the QSI specs the camera to cool by a delta-T of -45°C. That does not mean it is intended to actually be cooled TO -45°C. With deep cooling with FSI sensors, you can have problems with RBI. From what I understand, and I don't actually have practical experience here just the theory, is that cooling beyond about -20-25°C sensor temp can start allowing RBI issues. The bigger cameras often have -60-65°C delta-T cooling, so they can potentially reach much colder temperatures than most KAF-8300 parts. 



#25 Jon Rista

Jon Rista

    ISS

  • *****
  • topic starter
  • Posts: 24,994
  • Joined: 10 Jan 2014
  • Loc: Colorado

Posted 15 March 2016 - 03:41 PM

 

Personally, I would go Grade 1. I know myself and any column defects would really bother me, even though I know how to process them out.

Me too, and I ordered one from FLI.

 

For the 16200, there could be a significant difference in the number of defect between Grade 1 and Grade 3 CCD chips.

 

 

There could be a lot of defects with a Grade 3. As I understand, and I'm at a loss to find the reference now, the KAF sensors are rated for a normative ~2 column defects with Grade 2, with a potential maximum of I believe it was 15. I don't know what the rating is for Grade 3...but I wouldn't get a grade 3 sensor, that's basically the defect bin.




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