ZWO2600MC issue ??

[Startrek]

I captured 6 hrs ( 3 min dithered guided subs) on M16 last night with my 2600MC , L Extreme and 6” Newt under a descending 48% waxing moon ( used Gain 0 HDR ) Conditions were excellent guiding at around 0.60 total until it deteriorated in the early hours
Camera was cooled to -10C and I captured calibration frames as well
The above procedures I’ve been doing for 2 years now with this camera and had good results on a wide variety of objects at both my imaging sites

Stacked and Processed the data today and the stretched imaged is extremely noisy and the level of nebulosity is like I was using a DSLR with 60 sec subs

I did notice the cooling drop to - 8.3 C st one stage , it’s never ever done that in 2 years I’ve been using the camera

The stretched image is like I was using a DSLR really poor noisy data

Question
The camera is just over 2 years old ( out of warranty) should I perform a ZWO Firmware update ??
It’s as if the camera has lost its performance

Appreciate any advice

I know I’m imaging under Bortle 7/8 but I’ve never had such a noisy image and it’s not as if I’m short of data , there’s plenty of data to stack

I also changed settings in DSS , same outcomes

In addition, I did image the Helix last week and captured 6 hrs ( 3 min subs ) -10C with L Extreme , Cal frames and so on , similar situation, really noisy and stars were all saturated ( I used Gain 100 )

I think there’s definitely something wrong with my camera as it’s been performing well for 2 years at both sites , Bortle 7/8 and Bortle 3

Camera connects fine and operates correctly with my capture software APT

Next clear night I’ll be out testing my backside off trying all scenarios

I’m currently running the camera off the capture Laptop in my Study for 4 or 5 hours to check cooling status every hour

Cheers

[RyanJones]

Hi Martin,

No dew on the first lens of the new coma corrector ? Now that the first lens is at the end of the focuser draw tube, poking out into the OTA, it is far more susceptible to dew compared with the MPCC that had the protection of the draw tube.

Ryan

[Dave882]

Hi Martin- I’m not sure if it’s related but I’ve also been struggling with lots of noise in my images over the last week (same osc). I just assumed it was atmospheric conditions (humidity,dew,high cloud) as these can all contribute. Without seeing your images it’s impossible to say if we’re talking about the same thing tho…
Couple of question tho:
Do you run the cameras anti-dew heater?
Maybe a screenshot or stack showing the noise might be useful to see if there’s any recognisable profile?
Cheers

[Craig_]

Can you upload a .fits sub to have a look at?

[Stephane]

Sorry to hear that Martin. Having a look at the stack would be helpful; a sub too (especially at the time when temperature dropped). The fact that your temperature dropped is relevant. Sometimes with electronics it's just a matter of turning the camera off, then on again, and the issue is gone. Let's hope so.

Any particular noise patterns or vignetting?

[Startrek]

A single fits sub won’t prove anything other than histogram position which was fine
The fits subs look ok individually

[Startrek]

No noise pattern as dithering always takes care of that issue
It’s just a high level of background noise

I’ll post some photos of a stacked initial stretched image and compare it to one I captured a month or so ago under similar conditions and settings etc

[Startrek]

Thanks , good point
Conditions were perfect
Low humidity, slight breeze
No dew all at even at 3am

[Startrek]

Hope it’s just conditions but normally it’s never this bad considering the amount of data and the seeing was excellent, guiding was great for hours
ZWO advise to run the anti dew 24/7
I posted that question a month after I bought the camera
I’ll post some image comparisons

[xelasnave]

My money is on seeing conditions looking good but actually very bad...I notice because of short subs ..stars look like they are breathing and what appears to be noise probably isn't.
ANYWAYSI am experiencing similar...I am dropping 80% of subs and even the ones left are not flash and noise like never before...temp is not a problem

Alex

[Startrek]

I set up the camera tonight in my Study and set a 3 hour plan with 180 sec subs ( darks ) and cooled to -10C to simulate running a capture outside on the mount

Checked the camera every half an hour and the cooling maintained temperature between -9.8C and -10C , so unless I’ve got a suspect 12v cable outside , the camera cooling system is ok for now

These 2 images have been stacked , opened in Startools and auto stretched with just a bin , crop and a basic wipe ( no processing )

M8 Lagoon was capture a month or so ago on a similar night , same set up , 50% Waxing moon , 180 sec subs , similar integration but using Gain 100 or lowest read noise

M16 from last night using Gain 0 highest dynamic range 180 sec subs

By the way compare files sizes after saving as jpegs to upload
M8 805KB
M16 1.98MB ( 50% more noise ) I’ve had to reduce this image file size to load it into IIS

I’m going to have to image M16 again using Gain 100 lowest read noise and same subs etc…., but the moon won’t be the same week after next , heading into the new moon period ( crap weather for the next 8 days ) It’s not going to be a fair comparison

I’ve never had such a noisy image from my 2600MC

Comments welcome

[RyanJones]

There seems to be a streakiness in the noise in the lower left. It would probably help if we had the full frame to look at though so we can see the noise away from the signal of the neb. How come it’s in black and white too ?


Ryan

[Nikolas]

Looks like ice on the sensor, maybe remove and recharge the desiccant

Also try the suspect cable inside. but I reckon icy sensor

[Pepper]

Martin,
How do your darks look.
I think I’ve been having a similar problem.
If I cool the camera and start taking darks for instance, when I autostretch they will come out looking black and noisy. Not grey when auto stretched.
This will go away eventually and start producing cleaner images.
I’m about to send my 533mc back to get looked at as I was getting really clean images after I first bought it.

My theory is the cooler or associated electronics is causing it.

[Startrek]

Startools operates visually in luminance until you hit the colour module towards the end of your processing ( just before noise reduction) entirely different from PI. All the signal data is tracked

[Startrek]

Thanks
I’ll replace the desiccant , I have a spare sachet
I’ve already changed the cable

[Startrek]

Thanks
The 2600MC doesn’t require darks , in fact it’s better to calibrate without them , I use Bias instead and they look ok

Worst case scenario is sending the camera back for inspection and service ( lengthy process at the moment )

[Craig_]

Yeah, wow, I agree that is a pretty bad noise result on M16.

As you say the individual fits files look fine, have you tried stacking and stretching in other software just to rule out your current software as the cause?

[Startrek]

Good suggestion
I’ve already stacked and stretched previous data from June and May again ( eg: M8 attached above ) and DSS and Startools are working fine

[Startrek]

Thanks all for your suggestions and support
Greatly appreciated

I’m going to start from the bottom up and check off item by item until I find an answer on the issue
I don’t think the cause of the noisy image is entirely atmospheric conditions

Checklist
Reload update Ascom 6.6
Reload update Microsoft .Net Framework
Reload update both ZWO camera drivers
Update APT capture software
Clean camera sensor
Clean protective window on camera
Replace desiccant
Replace camera cables ( already done )
Perform cooling test for 3 to 4 hours ( already done )
Clean mirrors on scope
Re collimate scope
Clean L Extreme filter
Next clear night/s perform capture plan on object using Gain 100 with and without L Extreme filter for various exposure lengths 1 minute , 2 minute and 3 minute exposures
Perform capture plan using Gain 0 same as above
Hold off on a ZWO firmware update as this is only required when camera will not connect or operate via Ascom

Hopefully after all the updates , checks and testing I’ll get a definitive answer

Worse case scenario is sending the camera away for servicing and testing

Cheers
Martin

[Startrek]

I just had a reply back from ZWO in my image problem
I’m surprised at such a quick reply
Very much appreciated

Quote …

With the ASI2600, ZWO turns on the Sony sensor's High Conversion Gain (HCG) mode at a gain of 100 or above. You should expect the signal to noise ratio of shorter exposures to be about 8 to 10 dB worse when you use a gain of 0 vs when you use a gain of 100. So, changing the gain to zero does make a huge impact on signal to noise ratio, in the case of the ASI2600.

Remember that by using a constant exposure time, your signal drops dramatically with a gain of 0 (0 dB), while the read noise has popped up by more than 3 dB. With a gain of 100 (10 dB, or factor of about 3.16) your signal has not only gone up by a factor of 3.16, but HCG causes the read noise to go down by a factor of more than 2.

If you use really long exposures (30 minutes to 60 minutes), the read noise contribution is less. But at Bortle 7, I suspect that you won't be able to do that, even with a narrow band filter.

You will loose dynamic range with a gain of 100 (but not by much with Sony's HCG).

If you really need to use gain of 0, try longer exposures (10 to 20 minutes?) to mitigate this read noise problem at gain of 0, especially since you are using a narrow band filter -- although you are probably limited by Bortle 7 skies that prevents you from using very long exposures, even with narrow band filters.

With Bortle 7 skies, you won't really need that much dynamic range for anything but something like M42 and M31. And for those, you can always use high dynamic range techniques (taking images at different exposures). Bortle 2 or 3 skies are different; you have really dark backgrounds, and can make use of the dynamic range.

My recommendation is to bump the gain to 100 to turn on HCG. That uses less time to collect the sub-frames, too.

Personally, I always use gain 100 with my ASI2600 and ASI6200. I usually use 180 second exposures (Bortle 5+ to 6-; different parts of the sky are brighter), but when I use any narrowband filter (Radian Quad for color 2600 and 6200, Chroma 5nm Ha filter for mono 2600), I lengthen the exposures to 300 seconds to mitigate the read noise.

Chen

“Unquote”

I really didn’t think using Gain 0 would be a “Huge” issue using the LExtreme filter
I’ve completed all my procedures listed in previous post so now just waiting for the skies to clear next week

More testing on Gain 100 under Sydney skies

Martin

[RyanJones]

Some very useful information from ZWO and great that you got such a quick response. I was under the impression though from your original post that you had used the same settings with no issues at both of your sites and now you are experiencing the noise issue ?

Ryan

[Startrek]

Yes very surprised about a quick response from ZWO

I’ve mostly used Gain 0 at my Dome under Bortle 3 and have used both Gain 0 and Gain 100 in Sydney under Bortle 7/8

These new generation cameras with the Sony IMX571 sensors or so damn sensitive to changes in LP and sky conditions it’s hard to keep up
I do save my stacks and final images with object name , exposure time , gain setting , location , filters used etc..
Maybe I should go back over my work in the past 18months and create a chart for cross referencing ?

Martin

[RyanJones]

Maybe go back and find one that you did at gain 0 in the same location and see if you had the same noise problem then ? I seem to remember having followed a lot of your work over the years that Gain 0 was what you said you normally used ? It’s certainly all good I for action that you got from ZWO so you can understand why to select certain settings under certain conditions and thanks for sharing it so we can all get an understanding too. I have a feeling your current issue maybe due to something else though ? Maybe ... ?

[Startrek]

Ryan,
I hopeful that it’s not the camera’s internal electronics but you never know
Just need some clear nights to do more testing but we are heading into full moon so better in a week or so
Yes Chen from ZWO was really helpful and informative and hope others with the 2600 series cameras can benefit from his comments

Cheers
Martin

[Craig_]

I have a 2600MM and 533MM, and have imaged previously on the 533MC (so all based on the same sensor design as the 2600MC.) Personally I've found Gain 100 to be ideal for all of these cameras under all sky conditions for all targets, and I adjust the subexposure time only based on target, sky and filter. Maybe I should try Gain 0 a bit more sometime, but I've never really seen the need as I've never been dissatisified with Gain 100.

[Startrek]

I’ve used Gain 0 successfully at my Dome on the South Coast ( Bortle 3 Skies ) pushing 5 min subs and I was aware of all the 2600MC spec charts between Gain 100 and Gain 0 however , I didn’t think the signal loss and noise increase would be so great using Gain 0 under LP Bortle 7/8 skies. The dB figures Chen described are substantial

I’ll be using Gain 100 in Sydney from now on and down at my Dome probably Gain 0 most of the time and Gain 100 depending on sky conditions, moon phase and target type

[gregbradley]

Gain 0 sounds low. Higher gain = lower noise.

I suggest you check what others are using for their setting for gain and offsets.

I use a QHY600m and it is not noisy but some settings have lower noise than others and usually optimum for noise control.

I thought others were using a very high number for the gain with that camera.

Greg.

[Camelopardalis]

I know there are some nay-sayers around here, but looking at their published specs gain 100 is a no-brainer.

Imagine the noise as an error bar. The significant reduction in the size of the error bar when the HCG kicks in is enough to push the dynamic range back up to near maximum.

For the same total exposure time, the SNR of the end result will be higher at gain 100 than gain 0.

[Dave882]

Well I think I may have discovered the reason for my own noisy images over the last few weeks- I pulled the image train apart last night when that rain started coming through (untouched from the previous night) and what I noticed was there was a distinct layer of dew/fog on my filter inside the filter drawer. It disappeared quickly once I had it all apart but was unmistakeable.
May not be the same for you but thought I’d mention it in case.
Cheers

[Startrek]

Greg,
When I bought 2600MC in August 2020 I did post on IIS and CN for info on Gain and offset and the majority recommended Gain 100 as this is when the ZWO set the HCG ( high gain mode ) on the Sony IMX 571 where you benefit from the lowest read noise for the highest dynamic range with a well depth of 20K

You tend to forget this and then jump up to Gain 0 with short subs ( 3min ) and expect a nice clean noise free image under heavy B8 LP skies plus throw the moon in for good measure

The ZWO performance charts on both the 2600MC and 2600MM explain it all

ZWO recommend Gain 0 highest dynamic range for dark skies and long subs

Martin

[Startrek]

Yep the performance charts say it all
I bit of ignorance on my part , you live and learn
Cheers

[Startrek]

Thanks Dave
I’m still not totally convinced it’s just a Gain setting but I’ve done everything possible to my camera , cables , scope and software to rule out my checklist items
If I still get noisy images at Gain 100 with short subs ( 1min to 3min ) the cameras going in for service

Cheers
Martin

[Startrek]

I just browsed on Astrobin under the 2600MC camera and nearly every image ( Galaxy and Nebula) I looked at from all around the world use Gain 100 and 2 and 3 min subs , broadband and narrowband

They were all excellent images, so clean with plenty of depth and detail

No further proof required !!

[AdamJL]

I use Gain 100 as well, Martin :) have used that almost 99% of the time with the 2600MC

[RyanJones]

My money is still on what Dave has said. Either fog or dew on some part.

It’s great that you’ve discovered the benifits of gain 100 over gain 0 but just make sure you’re not getting Segwayed away from your original issue. You have said you have used gain 0 for a long time but the noise issue is new. If you are looking on Astrobin and seeing nice clean images on gain 100, make sure you’re also looking at comparable integration times and not comparing a 5hr one night image with a multi night 20hr image. There are many factors as you know that contribute to noise but don’t lose sight of the fact it’s a new issue to you whist using the same settings you have been using.

Ryan

[Startrek]

Ryan,
If you read my last post to Dave , I mentioned “I’m still not totally convinced it’s just a Gain setting”

And yes on Astrobin I was comparing same camera , Gain 100 , same temp -10C , similar subs , similar total integration time , f5 to f7 fl and so on …..
but in saying that I can’t determine the exact bortle scale of their skies as all they mention is imaged from backyard from a particular main town or City

Cheers
Martin

[Startrek]

Ongoing cloudy nights is making it hard to do any testing let alone imaging
I’ve serviced my camera
Cleaned sensor
Cleaned protective window
Replaced Desiccant
Bench tested operation and cooling , all ok
Updated all drivers and software
Cleaned my L Extreme filter

A 3 hour window last week enabled me to capture some data on M16 again but this time using Gain 100 Lower read noise ( 1.5 e-rms ) vs Gain 0 ( 3.5 e-rms )
It’s certainly not apples for apples but at least I know Gain 100 under heavy LP skies is the way to go in the future ( no more Gain 0 in Sydney)

Attached labeled images of M16 at Gain 100 and Gain 0
Same target
Similar conditions and moon
Same temp
Same length subs
Dithering
Inclusive of Cal frames ( Flats and Bias )
Same integration time approx 1hr 50 mins
Same DSS Stacking process
Just a quick Autostretch in ST
No real processing

Gain 0 is definitely noisier than Gain 100 especially noticeable in the non nebulosity mid outer regions

Further testing required but I’m confident my camera is Ok and it’s not frost or dew etc…

Bortle 7/8 SQM 18.5 is not exactly ideal for AP and a cooled OSC even with an L Extreme filter can only capture so much signal , the rest is skyglow noise and moon glare noise ….,

Cheers
Martin