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

New ZWO ASI 183 Mono - Uncorrectable amp glow

  • Please log in to reply
206 replies to this topic

#1 andysea

andysea

    Soyuz

  • *****
  • topic starter
  • Posts: 3547
  • Joined: 03 Sep 2010
  • Loc: Seattle, WA

Posted 06 January 2018 - 03:27 AM

I know this is a new camera and very few people are probably using it but I want to shout out at the group and see if I am the only one with this issue.

The camera has a pretty prominent amp glow, which we all knew and it should calibrate out. However in my sample I am unable to calibrate out the amp glow.

I have attached a stack of 13x5min subs with an Astrodon 3nm HA filter. The subs have been calibrated with flats and darks. No bias as I didn't scale the darks. 

They are shot at gain zero with the sensor at -20.

 

Are other people having this issue?

I typically get decent images out of the other ZWO camera that I own and I think I went through the correct calibration steps. I also tried calibration with darks only and the result was just about the same s the glow is just not calibrated out by the darks.

 

Thanks for any feedback!

Attached Thumbnails

  • ic434 copy.jpg


#2 einarin

einarin

    Surveyor 1

  • *****
  • Posts: 1987
  • Joined: 23 Dec 2016

Posted 06 January 2018 - 04:37 AM

Have noticed that with my older ASI camera.

And to get amp glow better removed I have simply taken longer darks - ie  for 3 min lights used 4 or 5 min darks.

Don't know if this is a real cure but it has worked for me and maybe worth a try.



#3 vdb

vdb

    Apollo

  • -----
  • Posts: 1469
  • Joined: 08 Dec 2009

Posted 06 January 2018 - 04:41 AM

Frankly I think this chip is not usable, the only way to make it work is that you shoot darks in the same run as the lights. I know every CMOS chip does a calibration, which can result in different characteristics ... 

I have the QHY163 which also exhibits this problem, but, AMP glow is less and 2, the recalibration doesn't seem to either differ much or maybe less frequent (don't know if this is the case)

 

Look at this thread for the discussion of the issue observed with different results with same settings, you have by no doubt already read it, it's a major red flag, I was really considering buying one but now it's off the table, or QHY is "more" stable in handling the Ampglow issue ...

https://www.cloudyni...release-thread/


Edited by vdb, 06 January 2018 - 04:42 AM.

  • andysea likes this

#4 Suavi

Suavi

    Messenger

  • -----
  • Posts: 448
  • Joined: 05 May 2017
  • Loc: 20.25 deg South of the Equator

Posted 06 January 2018 - 05:22 AM

Perhaps I am saying something obvious, but were the darks and lights taken at the same ambient temperature? Even with the same set sensor temperature, camera's electronics may create varied internal unwanted signal when ambient temperature varies significantly, for example taking darks at a warmer indoor environment and lights at cooler outdoor temperature.


  • GA-HAMAL likes this

#5 vdb

vdb

    Apollo

  • -----
  • Posts: 1469
  • Joined: 08 Dec 2009

Posted 06 January 2018 - 06:01 AM

In the past there where some camera's that had the problem, it's almost impossible to correct for and really makes you waste a lot of time matching not only ccd temp but also ambient. In this case I'm pretty sure its the auto calibration of the sensor ... which is not correctable, not even by Sony apparently ...

There is a way to disable the circuit that causes the amp glow but it apparently does not completely solves the issue and is something QHY is doing ... but I'm speculating ...

 

/Yves



#6 sg6

sg6

    Fly Me to the Moon

  • *****
  • Posts: 6068
  • Joined: 14 Feb 2010
  • Loc: Norfolk, UK.

Posted 06 January 2018 - 08:30 AM

Silly question but you are using USB3 ? Totally expect you are.

There is or was a "Pro" ugrade for the 071 camera, agreed not yours, and in the one reading of the upgrade description, and reason for it, I read there was mention that the upgrade added memory and so sped up the data extraction. USB3 was faster but the bigger sensors seemed to negate the faster USB3.

 

Although you are not using the 071 just wondering if the communications is USB3. As USB2 may be too slow and so introducing greater amp glow.

 

Searching around it seems the Pro upgrade,as an aftermarket option, is (was?) only available for the 071 camera.


  • t-ara-fan likes this

#7 buckeyestargazer

buckeyestargazer

    Vendor - Buckeyestargazer.net

  • *****
  • Vendors
  • Posts: 4749
  • Joined: 12 Jan 2008
  • Loc: IN, USA

Posted 06 January 2018 - 09:12 AM

I thought it was only the ASI1600/QHY163M that shouldn't use bias?  Have you tried using Bias just to see what it does?



#8 andysea

andysea

    Soyuz

  • *****
  • topic starter
  • Posts: 3547
  • Joined: 03 Sep 2010
  • Loc: Seattle, WA

Posted 06 January 2018 - 11:02 AM

Thanks guys for the feedback!

 

- The darks were taken at a slightly higher ambient temperature than my darks. Perhaps 4c~5c higher. The darks were obviously matched for sensor temperature length and gain.

- The 183 is a "Pro" version I believe this means that it has the internal buffer make it work with USB2. I was actually using USB2.

- I have not tried to use bias. I corrected my lights with what I thought were perfectly matched darks.

 

I will try to return the camera as I have no time to debug a system given the small amount of clear skies that we gat.


  • arivas likes this

#9 Jon Rista

Jon Rista

    ISS

  • *****
  • Posts: 23803
  • Joined: 10 Jan 2014
  • Loc: Colorado

Posted 06 January 2018 - 11:06 AM

Andy, before you return the camera, can you provide some of your darks and lights? I can do a quick check to see what the differences are. You may simply need them to Team View into your computer and apply a firmware update, which I have had them do, and it is pretty painless and quick. Since you mentioned corner glows, you probably need a firmware update.

#10 andysea

andysea

    Soyuz

  • *****
  • topic starter
  • Posts: 3547
  • Joined: 03 Sep 2010
  • Loc: Seattle, WA

Posted 06 January 2018 - 11:08 AM

Thank you Jon I can do that. I will setup a DB share and send you the link.



#11 rockstarbill

rockstarbill

    Fly Me to the Moon

  • *****
  • Posts: 6351
  • Joined: 16 Jul 2013
  • Loc: Snohomish, WA

Posted 06 January 2018 - 03:13 PM

Yeah a single 5 minute dark frame could easily confirm if this is related to the firmware. I see glows in the corners of the image posted here. The firmware was supposed to fix that.


  • BeltofOrion likes this

#12 Konihlav

Konihlav

    Surveyor 1

  • -----
  • Posts: 1670
  • Joined: 05 Mar 2009
  • Loc: Germany

Posted 06 January 2018 - 03:44 PM

I had same issue (uncorrectable amp-glow) on my QHY163 which I returned to China for 2nd time. Now they just sent me another one, a new one, and this time it looks OK to me (after a year+ of issuing)...

 

today I have just tested two brand new ASI1600MM which were both OK.

 

There's simple test to do, as I mentioned somewhere, just take 10 dark frames. Create a master dark (median, integer value). In PixelMath in PI do just: singledark - masterdark and max STF stretch the image. The amp-glow must be removed. This is a very simple check for every CMOS based camera that people have at home... quick to start with. If this is OK and later light calibration fails, typically the issue is in wrong flats or wrong dark for flats (may need to be 3 and more seconds long).

 

my $0.02


  • dragracingdan, scopenitout, GA-HAMAL and 1 other like this

#13 Jon Rista

Jon Rista

    ISS

  • *****
  • Posts: 23803
  • Joined: 10 Jan 2014
  • Loc: Colorado

Posted 06 January 2018 - 04:22 PM

I had same issue (uncorrectable amp-glow) on my QHY163 which I returned to China for 2nd time. Now they just sent me another one, a new one, and this time it looks OK to me (after a year+ of issuing)...

 

today I have just tested two brand new ASI1600MM which were both OK.

 

There's simple test to do, as I mentioned somewhere, just take 10 dark frames. Create a master dark (median, integer value). In PixelMath in PI do just: singledark - masterdark and max STF stretch the image. The amp-glow must be removed. This is a very simple check for every CMOS based camera that people have at home... quick to start with. If this is OK and later light calibration fails, typically the issue is in wrong flats or wrong dark for flats (may need to be 3 and more seconds long).

 

my $0.02

I would enhance that PM just a bit:

 

K: Dark + median(Dark) - MasterDark

 

This will effectively restore the original offset in the Dark, and avoid any clipping, which can help you see if subtracting the master dark results in an inversion of the amp glow...which is another indication of a problem (it would mean the master had a brighter amp glow than the single frame, which could indicate some kind of cross-frame inconsistencies.)


  • NorthField likes this

#14 andysea

andysea

    Soyuz

  • *****
  • topic starter
  • Posts: 3547
  • Joined: 03 Sep 2010
  • Loc: Seattle, WA

Posted 06 January 2018 - 05:41 PM

Awesome thanks! I will do that test right away.

I am also just about to upload everything in DB and I will post the link.



#15 rockstarbill

rockstarbill

    Fly Me to the Moon

  • *****
  • Posts: 6351
  • Joined: 16 Jul 2013
  • Loc: Snohomish, WA

Posted 06 January 2018 - 05:47 PM

I'll try this same test in a few as well

#16 rockstarbill

rockstarbill

    Fly Me to the Moon

  • *****
  • Posts: 6351
  • Joined: 16 Jul 2013
  • Loc: Snohomish, WA

Posted 06 January 2018 - 05:59 PM

Did this test, here is the result. 5 min single dark, 5 minute dark master. 

Attached Thumbnails

  • DarkSubtractPixelMath.JPG


#17 rockstarbill

rockstarbill

    Fly Me to the Moon

  • *****
  • Posts: 6351
  • Joined: 16 Jul 2013
  • Loc: Snohomish, WA

Posted 06 January 2018 - 06:07 PM

Data uploaded here:

 

https://1drv.ms/f/s!...kofoJMgQbZOpmSA



#18 andysea

andysea

    Soyuz

  • *****
  • topic starter
  • Posts: 3547
  • Joined: 03 Sep 2010
  • Loc: Seattle, WA

Posted 06 January 2018 - 06:08 PM

That looks pretty much like my result.

There is a pretty strong horizontal banding, shouldn't that be eliminated by the master dark? I wonder if the bias is unstable.

I used gain zero - highest dynamic range in the driver setting.

 

The file attached has an aggressive stretch to exacerbate  the banding.

Attached Thumbnails

  • Image06.jpg


#19 andysea

andysea

    Soyuz

  • *****
  • topic starter
  • Posts: 3547
  • Joined: 03 Sep 2010
  • Loc: Seattle, WA

Posted 06 January 2018 - 06:40 PM

My flats were 1.5s long and I took dark flats. Pixinsight was not able to stack my dark flats and reported this error "Zero or insignificant signal detected (empty image?)"

I then took 40 bias and Pixinsight was able to stack those.

I redid the 1.5s darks twice and both times PIxisight refused to stack them.

Here is the link to the raw data.

 

https://www.dropbox....H1uaLwguPa?dl=0

 

I didn't find ZWO's return policy, does anyone know what that is?



#20 Jon Rista

Jon Rista

    ISS

  • *****
  • Posts: 23803
  • Joined: 10 Jan 2014
  • Loc: Colorado

Posted 06 January 2018 - 06:41 PM

There is more banding at Gain 0. That is why I ended up using Gain 53...much lower banding, but still good dynamic range (~11.8 stops). It is worth a try seeing what Gain 53 looks like. The horizontal banding is random frame to frame, but it takes a decent number of frames to average out. At Gain 0 and very long subs you might not get enough to average out the banding. That is another reason to use Gain 53.



#21 rockstarbill

rockstarbill

    Fly Me to the Moon

  • *****
  • Posts: 6351
  • Joined: 16 Jul 2013
  • Loc: Snohomish, WA

Posted 06 January 2018 - 06:51 PM

My flats were 1.5s long and I took dark flats. Pixinsight was not able to stack my dark flats and reported this error "Zero or insignificant signal detected (empty image?)"

I then took 40 bias and Pixinsight was able to stack those.

I redid the 1.5s darks twice and both times PIxisight refused to stack them.

Here is the link to the raw data.

 

https://www.dropbox....H1uaLwguPa?dl=0

 

I didn't find ZWO's return policy, does anyone know what that is?

Well it looks like this camera didn't get the firmware update, as the single 300 sec dark shows glow in the corners that should not be there. My guess here is that if the firmware were updated, the problem would go away.



#22 andysea

andysea

    Soyuz

  • *****
  • topic starter
  • Posts: 3547
  • Joined: 03 Sep 2010
  • Loc: Seattle, WA

Posted 06 January 2018 - 07:18 PM

I will try at gain 53.

There is also a very weird artifact in my flats. It is not caused by the filter as I tried rotating it. I also tried with a different telescope and the artifact is the same. I also tried with and without the off axis guider with no difference.

the flats seem to correct the 300s subs but they completely mess up a set of 600s subs. The latter are not in the DB share.

If I shine a light on the sensor at a certain angle I can actually see a reflection in the shape of the artifact that I see in my HA flats.
There is definitely something going on but I can't pinpoint the issue.


Edited by andysea, 06 January 2018 - 07:19 PM.


#23 rockstarbill

rockstarbill

    Fly Me to the Moon

  • *****
  • Posts: 6351
  • Joined: 16 Jul 2013
  • Loc: Snohomish, WA

Posted 06 January 2018 - 07:26 PM

What artifact and was it the HA filter?

#24 andysea

andysea

    Soyuz

  • *****
  • topic starter
  • Posts: 3547
  • Joined: 03 Sep 2010
  • Loc: Seattle, WA

Posted 06 January 2018 - 07:44 PM

Yes it's the HA filter. I think you can see it in the attached image. It's the crescent shaped shadow going across the middle of the sensor more or less horizontally.

 

Attached Thumbnails

  • flat.jpg


#25 Ken Sturrock

Ken Sturrock

    Cardinal Ximénez - NO ONE EXPECTS THE SPANISH INQUISITION!

  • *****
  • Administrators
  • Posts: 7970
  • Joined: 26 Sep 2009
  • Loc: Denver, CO

Posted 06 January 2018 - 09:13 PM

Andy, I was shooting the same object last night, although in LRGB and the amp glow seemed to calibrate out just fine with matching darks (120 seconds, -10c).

I did notice that any frames taken at less than half-a-second had that really nasty horizontal banding (Gain 50). In my case, my flat panel was a little bright and the luminance flat showed the banding, but for example, the red did not. Naturally, the banding would propagate to the sub during calibration. I cleaned up the striped flats with the "Canon de-bander" in PI. Mine also had a bit of corner glow in the upper & lower left sides of the frame, but I'm not sure if it was an internal source to the camera or just my usual light gradients...




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