NGC 7000 with CMOS camera


Robert Chozick
 

I really like this CMOS camera.  It is so sensitive.  I added Ha I took with this OSC and it works well.  I added the Ha in Photoshop in a Layer with a Linear Dodge Add mode with all color channels. I should have turned the camera for this shot.  I overexposed the stars with only 3m exposures.  I will try 2 or 2.5 minute exposures next time out.

Cropped:


Full image:


OSC Ha with Astrodon 5nm filter:



Robert Chozick




Wayne Hixson
 

Hi Robert, nice work. I have had trouble with getting good calibration. Some have said cooling is uneven and font go below -10. What is your calibration process and experience?

Thanks,

Wayne 




On Saturday, August 24, 2019, 8:59 AM, Robert Chozick rchozick@... [ap-gto] wrote:

 

I really like this CMOS camera.  It is so sensitive.  I added Ha I took with this OSC and it works well.  I added the Ha in Photoshop in a Layer with a Linear Dodge Add mode with all color channels. I should have turned the camera for this shot.  I overexposed the stars with only 3m exposures.  I will try 2 or 2.5 minute exposures next time out.

Cropped:


Full image:


OSC Ha with Astrodon 5nm filter:



Robert Chozick




Mike Dodd
 

On 8/24/2019 12:06 PM, Wayne Hixson wayneh9026@yahoo.com [ap-gto] wrote:


Hi Robert, nice work.
I agree.... Very nicely done, Robert.

I have the same ASI1600MC Pro camera, and am very pleased with it. I'm still finding my way in PixInsight, which is a very powerful processing tool.

I have had trouble with getting good calibration.

What kind of problems have you experienced, Wayne?

Some have said cooling is uneven and font go below -10. What is your
calibration process and experience?
On cool nights, around 12C, I have no problem taking my '1600 down to -20C with the cooler running 50%-60%. On warm nights, when the ambient temperature is around 22C, it can reach -10C with the cooler running 70%-80%.

Cooling the '1600 reduces noise, but the sensor is so quiet that there isn't much noise to reduce. Without cooling, subtracting a good master dark frame eliminates most of it.

----------

You didn't ask me, and I am by no means a PixInsight expert, but here's my fairly standard PI workflow:

1. Blink to eliminate poor subs.

2. SubframeSelector to identify best subs for registration (FHWM) and integration (SNR).

3. ImageCalibration. Includes:
* Subtract dark. (I have a library of master darks made from ~12 frames at various times and temperatures).
* Apply flat. (I make master flats at the image's camera angle and 180 degrees from ~5 dusk sky flats). Before I began using flats, I used DynamicBackgroundExtraction to remove vignetting. Flats are better!

4. CosmeticCorrection.

5. Debayer. The pattern for the '1600 is GRBG.

6. StarAlignment.

7. ImageIntegration.

8. MultiscaleLinearTransformation to reduce any remaining background noise.

9. ScreenTransferFunction to adjust intensity and color saturation, then HistorgramTransformation to make the adjustments permanent.

Sometimes I want to reduce star size in nebula images, so:

11. StarMask then MorphologicalTransformation.

A final note: I store each target's images and information in a dedicated folder on my PC's hard drive. That folder contains these subfolders:

* (ACP Plan) - I use ACP automation software to acquire images, and a copy of the imaging plan is stored here so I can get more identical subs at a later date.

* (Darks Used) - The master dark(s) used during PI calibration.

* (Flats Used) - The master flat frame(s) used during PI calibration.

* (PixInsight Project) - The PI project used to process the image. Sometimes I add processes to my workspace, so I want to save everything for future use on this image.

* 1. Camera Subs - The camera subframes.

* 2. Calibrated - PI-calibrated subs.

* 3. Cosmetic Correction - PI-corrected subs.

* 4. Debayered - PI-debayered subs.

* 5. Registered - PI-registered subs.

* 6. Integrated - The final integrated image. Alao contains pre-stretching and post-stretching copies, including MultiscaleLinearTransformation noise-reduction and star reduction. I save everything in .xisf format, plus the final image in .png format for possible further tweaking in Photoshop.

* X. Rejected - inferior subs rejected during SubframeSelection (never know they might be needed).

Subfolder names in parentheses appear first in File Explorer. I added a number to each processed-subs folder so they appear in the correct PI processing order. When processing is complete, I keep the camera subs, but delete the processed subs from their respective folders because I can always reprocess from the start if necessary.

--
Mike

Mike Dodd
Louisa County, Virginia USA
http://astronomy.mdodd.com


Dale Ghent
 

On Aug 24, 2019, at 1:49 PM, Mike Dodd mike@mdodd.com [ap-gto] <ap-gto@yahoogroups.com> wrote:

On cool nights, around 12C, I have no problem taking my '1600 down to
-20C with the cooler running 50%-60%. On warm nights, when the ambient
temperature is around 22C, it can reach -10C with the cooler running
70%-80%.

Cooling the '1600 reduces noise, but the sensor is so quiet that there
isn't much noise to reduce. Without cooling, subtracting a good master
dark frame eliminates most of it.
I run a 1600MM and have settled on cooling at either -15 or -10C, depending on the season. The amount of noise that is removed from the sensor by going colder than -15 is truly negligible. You can use PI's Statistics process to investigate the difference using your darks. Open a master dark of the same exposure time, gain, and offset for each temperature you have. Turn nomalization (and, optionally, scientific notation) off via the checkboxes. Hit the wrench icon on the right side and make sure that Variance and Std. Deviation display are on. Then compare the Variance and StdDev values for your -10C dark against the equivalent -15C dark, and -20C, and so-on. You'll find that after -15C, there really isn't much of an appreciable difference, and hence reduction, in noise on these sensors. This tends to hold true for CMOS in general.

For example, my master dark for an exposure time of 180 seconds @ Gain=139, Offset=21:

-10C:
variance 0.0000016
stdDev 0.0012725

-15C:
variance 0.0000011
stdDev 0.0010689

-20C:
variance 0.0000010
stdDev 0.0009903

I've done test -25C and -30C darks and the differences between those are similarly negligible, but I unfortunately neglected to save the results after deciding I wouldn't bother going that cold.

Given that, I'd much rather have more cooler power headroom instead, preferring to run it at no more than 75% power. I'm also a mobile imager, so there's also some consideration for the amount of juice the TEC is slurping up when running at high duty cycles.

/dale


Wayne Hixson
 

I wasn’t clear, actually talking about the OSC ASI294. I like the 1600 also!


On Aug 24, 2019, at 11:17 AM, Dale Ghent daleg@... [ap-gto] <ap-gto@...> wrote:

 



> On Aug 24, 2019, at 1:49 PM, Mike Dodd mike@... [ap-gto] <ap-gto@...> wrote:
>
> On cool nights, around 12C, I have no problem taking my '1600 down to
> -20C with the cooler running 50%-60%. On warm nights, when the ambient
> temperature is around 22C, it can reach -10C with the cooler running
> 70%-80%.
>
> Cooling the '1600 reduces noise, but the sensor is so quiet that there
> isn't much noise to reduce. Without cooling, subtracting a good master
> dark frame eliminates most of it.

I run a 1600MM and have settled on cooling at either -15 or -10C, depending on the season. The amount of noise that is removed from the sensor by going colder than -15 is truly negligible. You can use PI's Statistics process to investigate the difference using your darks. Open a master dark of the same exposure time, gain, and offset for each temperature you have. Turn nomalization (and, optionally, scientific notation) off via the checkboxes. Hit the wrench icon on the right side and make sure that Variance and Std. Deviation display are on. Then compare the Variance and StdDev values for your -10C dark against the equivalent -15C dark, and -20C, and so-on. You'll find that after -15C, there really isn't much of an appreciable difference, and hence reduction, in noise on these sensors. This tends to hold true for CMOS in general.

For example, my master dark for an exposure time of 180 seconds @ Gain=139, Offset=21:

-10C:
variance 0.0000016
stdDev 0.0012725

-15C:
variance 0.0000011
stdDev 0.0010689

-20C:
variance 0.0000010
stdDev 0.0009903

I've done test -25C and -30C darks and the differences between those are similarly negligible, but I unfortunately neglected to save the results after deciding I wouldn't bother going that cold.

Given that, I'd much rather have more cooler power headroom instead, preferring to run it at no more than 75% power. I'm also a mobile imager, so there's also some consideration for the amount of juice the TEC is slurping up when running at high duty cycles.

/dale


Robert Chozick
 

Thanks Wayne and Mike.  I use Nebulosity for pre-processing.  The lowest I have gone on temp is -10C and in my backyard the lowest I can go in Texas is -5C. The ambient at 10pm is still above 30C and not below 30C all night. I do not suggest ever letting it go higher than 80% cooling.  I have not seen the uneven cooling problem that has been discussed.   I am not doing anything special.  I just make sure my darks are the same time, temp, gain and offset.  One thing I did have to do differently is darken my flat light to make my exposures for flats at least 3-4 seconds and take flat darks.  I placed some neutral density film over my Gerd Neumann Flatfield panel to do this.

I can’t believe my 3 min exposures are too long on my f/5 FSQ.


On Aug 24, 2019, at 2:01 PM, Wayne wayneh9026@... [ap-gto] <ap-gto@...> wrote:

I wasn’t clear, actually talking about the OSC ASI294. I like the 1600 also!


On Aug 24, 2019, at 11:17 AM, Dale Ghent daleg@... [ap-gto] <ap-gto@...> wrote:

 



> On Aug 24, 2019, at 1:49 PM, Mike Dodd mike@... [ap-gto] <ap-gto@...> wrote:
> 
> On cool nights, around 12C, I have no problem taking my '1600 down to 
> -20C with the cooler running 50%-60%. On warm nights, when the ambient 
> temperature is around 22C, it can reach -10C with the cooler running 
> 70%-80%.
> 
> Cooling the '1600 reduces noise, but the sensor is so quiet that there 
> isn't much noise to reduce. Without cooling, subtracting a good master 
> dark frame eliminates most of it.

I run a 1600MM and have settled on cooling at either -15 or -10C, depending on the season. The amount of noise that is removed from the sensor by going colder than -15 is truly negligible. You can use PI's Statistics process to investigate the difference using your darks. Open a master dark of the same exposure time, gain, and offset for each temperature you have. Turn nomalization (and, optionally, scientific notation) off via the checkboxes. Hit the wrench icon on the right side and make sure that Variance and Std. Deviation display are on. Then compare the Variance and StdDev values for your -10C dark against the equivalent -15C dark, and -20C, and so-on. You'll find that after -15C, there really isn't much of an appreciable difference, and hence reduction, in noise on these sensors. This tends to hold true for CMOS in general.

For example, my master dark for an exposure time of 180 seconds @ Gain=139, Offset=21:

-10C:
variance 0.0000016
stdDev 0.0012725

-15C:
variance 0.0000011
stdDev 0.0010689

-20C:
variance 0.0000010
stdDev 0.0009903

I've done test -25C and -30C darks and the differences between those are similarly negligible, but I unfortunately neglected to save the results after deciding I wouldn't bother going that cold.

Given that, I'd much rather have more cooler power headroom instead, preferring to run it at no more than 75% power. I'm also a mobile imager, so there's also some consideration for the amount of juice the TEC is slurping up when running at high duty cycles.

/dale



Robert Chozick




Mike Dodd
 

On 8/24/2019 8:47 PM, Robert Chozick rchozick@aol.com [ap-gto] wrote:

...make my exposures for flats at least 3-4 seconds
and take flat darks.

I don't bother with darks for flats. The noise is small compared to a flat ADU of 20,000-40,000, so why subtract tiny noise from a very bright flat?

PixInsight measures pixel brightness from 0 to 1. I just measured a flat frame and found an average level of 0.32. A dark used for this particular image measured 0.01. So the dark's noise is 3% of the flat's level.

(I could/should increase my flat's level to 0.5 or 0.6. I'm still gaining experience with dusk flats.)

--- Mike


Robert Chozick
 

The problem is that you cannot use bias on a CMOS camera.  Any exposure less than 3 seconds is not advisable (not an engineer so not sure why).  So if you cannot use a bias frame on your flats you have to use flat darks.

On my CCD camera I took flats that were less than 1 second and used bias to subtract.  I think all of these quirks of a CMOS have to be observed to get good calibration.


On Aug 24, 2019, at 8:09 PM, Mike Dodd mike@... [ap-gto] <ap-gto@...> wrote:

On 8/24/2019 8:47 PM, Robert Chozick rchozick@... [ap-gto] wrote:

> ...make my exposures for flats at least 3-4 seconds
> and take flat darks.

I don't bother with darks for flats. The noise is small compared to a 
flat ADU of 20,000-40,000, so why subtract tiny noise from a very bright 
flat?

PixInsight measures pixel brightness from 0 to 1. I just measured a flat 
frame and found an average level of 0.32. A dark used for this 
particular image measured 0.01. So the dark's noise is 3% of the flat's 
level.

(I could/should increase my flat's level to 0.5 or 0.6. I'm still 
gaining experience with dusk flats.)

--- Mike


Robert Chozick




Mike Dodd
 

On 8/24/2019 9:31 PM, Robert Chozick rchozick@aol.com [ap-gto] wrote:


The problem is that you cannot use bias on a CMOS camera. Any exposure
less than 3 seconds is not advisable (not an engineer so not sure why).
So if you cannot use a bias frame on your flats you have to use flat
darks.
I don't understand what's magic about 3 seconds. ZWO specifies an exposure range of 32μs-2000s for the '1600 and the '294. I have seen no issues with flats exposed for less than 1 second with my '1600.

--- Mike


Michael Fulbright <mike.fulbright@...>
 

I routinely use 0.5s to 5s flats.  You just need to match flat darks (darks the same length as your flats) at the same gain value.

Calibration has never been an issue on my 1600.

Michael Fulbright

On 8/24/19 10:16 PM, Mike Dodd mike@... [ap-gto] wrote:
 

On 8/24/2019 9:31 PM, Robert Chozick rchozick@... [ap-gto] wrote:
>
>
> The problem is that you cannot use bias on a CMOS camera. Any exposure
> less than 3 seconds is not advisable (not an engineer so not sure why)..
> So if you cannot use a bias frame on your flats you have to use flat
> darks.

I don't understand what's magic about 3 seconds. ZWO specifies an
exposure range of 32μs-2000s for the '1600 and the '294. I have seen no
issues with flats exposed for less than 1 second with my '1600.

--- Mike



Dale Ghent
 

You may need to have a longer exposure time for flats because LED light panels adjust their brightness via PWM. If the shutter speed of the camera is quicker than what the panel is being driven at for its level of brightness, you will get banding across the image due to how PWM rates interact with rolling shutter sensors.

This need for a longer exposure can lead to an issue where the dimmest setting of a panel is still too bright for that exposure length and the flat is blown out, and one must resort to attenuating the panel light with ND film or other methods.

/dale

On Aug 24, 2019, at 10:16 PM, Mike Dodd mike@mdodd.com [ap-gto] <ap-gto@yahoogroups.com> wrote:

On 8/24/2019 9:31 PM, Robert Chozick rchozick@aol.com [ap-gto] wrote:


The problem is that you cannot use bias on a CMOS camera. Any exposure
less than 3 seconds is not advisable (not an engineer so not sure why).
So if you cannot use a bias frame on your flats you have to use flat
darks.
I don't understand what's magic about 3 seconds. ZWO specifies an
exposure range of 32μs-2000s for the '1600 and the '294. I have seen no
issues with flats exposed for less than 1 second with my '1600.

--- Mike



------------------------------------
Posted by: Mike Dodd <mike@mdodd.com>
------------------------------------

To UNSUBSCRIBE, or for general information on the ap-gto list
see http://groups.yahoo.com/group/ap-gto
------------------------------------

Yahoo Groups Links



Robert Chozick
 

The 3 second rule I have seen is from multiple posts as I go through comments on issues.  I generally try to get the benefit of common experiences and comments whether I totally understand them or not.  I don’t get as much chance as I would like to image in dark skies and don't want to take any chances.  I regularly used .3 sec flats for my STF-8300 CCD without issue.

On Aug 25, 2019, at 12:46 AM, Dale Ghent daleg@... [ap-gto] <ap-gto@...> wrote:


You may need to have a longer exposure time for flats because LED light panels adjust their brightness via PWM. If the shutter speed of the camera is quicker than what the panel is being driven at for its level of brightness, you will get banding across the image due to how PWM rates interact with rolling shutter sensors.

This need for a longer exposure can lead to an issue where the dimmest setting of a panel is still too bright for that exposure length and the flat is blown out, and one must resort to attenuating the panel light with ND film or other methods.

/dale

> On Aug 24, 2019, at 10:16 PM, Mike Dodd mike@... [ap-gto] <ap-gto@...> wrote:
> 
> On 8/24/2019 9:31 PM, Robert Chozick rchozick@... [ap-gto] wrote:
>> 
>> 
>> The problem is that you cannot use bias on a CMOS camera. Any exposure
>> less than 3 seconds is not advisable (not an engineer so not sure why).
>> So if you cannot use a bias frame on your flats you have to use flat
>> darks.
> 
> I don't understand what's magic about 3 seconds. ZWO specifies an 
> exposure range of 32μs-2000s for the '1600 and the '294. I have seen no 
> issues with flats exposed for less than 1 second with my '1600.
> 
> --- Mike
> 
> 
> 
> ------------------------------------
> Posted by: Mike Dodd <mike@...>
> ------------------------------------
> 
> To UNSUBSCRIBE, or for general information on the ap-gto list
> see http://groups.yahoo.com/group/ap-gto
> ------------------------------------
> 
> Yahoo Groups Links
> 
> 
> 


Robert Chozick




Dale Ghent
 

The only goal with flats is to get an illuminated frame from an evenly-lit light source, where the histogram is a nice hump centered at the %50 point. It doesn't have to be perfectly nailed at 50%. in truth, a flat is alright so long as no pixels are clipped one way or the other, but a perfect flat will look like that.

Obviously, getting that kind of result with an correct (or, adequate) exposure time is going to be specific to every setup, from optics speed and sensor settings, to the intensity of the flat light source - as well as considerations for any annoying interactions such as the rolling shutter one I mentioned. So there really isn't (and shouldn't be) a "rule" when it comes to exposure length and flats.

To take the fiddling out of arriving at this kind of result, we put a Flat Wizard into the app I contribute to:
https://nighttime-imaging.eu/docs/develop/site/tabs/flatwizard/

/dale

On Aug 25, 2019, at 2:53 AM, Robert Chozick rchozick@aol.com [ap-gto] <ap-gto@yahoogroups.com> wrote:

The 3 second rule I have seen is from multiple posts as I go through comments on issues. I generally try to get the benefit of common experiences and comments whether I totally understand them or not. I don’t get as much chance as I would like to image in dark skies and don't want to take any chances. I regularly used .3 sec flats for my STF-8300 CCD without issue.

On Aug 25, 2019, at 12:46 AM, Dale Ghent daleg@elemental.org [ap-gto] <ap-gto@yahoogroups.com> wrote:


You may need to have a longer exposure time for flats because LED light panels adjust their brightness via PWM. If the shutter speed of the camera is quicker than what the panel is being driven at for its level of brightness, you will get banding across the image due to how PWM rates interact with rolling shutter sensors.

This need for a longer exposure can lead to an issue where the dimmest setting of a panel is still too bright for that exposure length and the flat is blown out, and one must resort to attenuating the panel light with ND film or other methods.

/dale

On Aug 24, 2019, at 10:16 PM, Mike Dodd mike@mdodd.com [ap-gto] <ap-gto@yahoogroups.com> wrote:

On 8/24/2019 9:31 PM, Robert Chozick rchozick@aol.com [ap-gto] wrote:


The problem is that you cannot use bias on a CMOS camera. Any exposure
less than 3 seconds is not advisable (not an engineer so not sure why).
So if you cannot use a bias frame on your flats you have to use flat
darks.
I don't understand what's magic about 3 seconds. ZWO specifies an
exposure range of 32μs-2000s for the '1600 and the '294. I have seen no
issues with flats exposed for less than 1 second with my '1600.

--- Mike



------------------------------------
Posted by: Mike Dodd <mike@mdodd.com>
------------------------------------

To UNSUBSCRIBE, or for general information on the ap-gto list
see http://groups.yahoo.com/group/ap-gto
------------------------------------

Yahoo Groups Links




Robert Chozick
rchozick@aol.com






Robert Chozick
 

The issue is not the length of time for the flat for flat purposes. The issue is that it is my understanding that the CMOS camera itself should not ever have an exposure less than three seconds.   I always take my flats at around 30,000 ADU.  I adjust the exposure to get there.  To get there with my flat machine I have to put in neutral density film. 


On Aug 25, 2019, at 6:41 AM, Dale Ghent daleg@... [ap-gto] <ap-gto@...> wrote:

 


The only goal with flats is to get an illuminated frame from an evenly-lit light source, where the histogram is a nice hump centered at the %50 point. It doesn't have to be perfectly nailed at 50%. in truth, a flat is alright so long as no pixels are clipped one way or the other, but a perfect flat will look like that.

Obviously, getting that kind of result with an correct (or, adequate) exposure time is going to be specific to every setup, from optics speed and sensor settings, to the intensity of the flat light source - as well as considerations for any annoying interactions such as the rolling shutter one I mentioned. So there really isn't (and shouldn't be) a "rule" when it comes to exposure length and flats.

To take the fiddling out of arriving at this kind of result, we put a Flat Wizard into the app I contribute to:
https://nighttime-imaging.eu/docs/develop/site/tabs/flatwizard/

/dale

> On Aug 25, 2019, at 2:53 AM, Robert Chozick rchozick@... [ap-gto] <ap-gto@...> wrote:
>
> The 3 second rule I have seen is from multiple posts as I go through comments on issues. I generally try to get the benefit of common experiences and comments whether I totally understand them or not. I don’t get as much chance as I would like to image in dark skies and don't want to take any chances. I regularly used .3 sec flats for my STF-8300 CCD without issue.
>
> On Aug 25, 2019, at 12:46 AM, Dale Ghent daleg@... [ap-gto] <ap-gto@...> wrote:
>
>
> You may need to have a longer exposure time for flats because LED light panels adjust their brightness via PWM. If the shutter speed of the camera is quicker than what the panel is being driven at for its level of brightness, you will get banding across the image due to how PWM rates interact with rolling shutter sensors.
>
> This need for a longer exposure can lead to an issue where the dimmest setting of a panel is still too bright for that exposure length and the flat is blown out, and one must resort to attenuating the panel light with ND film or other methods.
>
> /dale
>
> > On Aug 24, 2019, at 10:16 PM, Mike Dodd mike@... [ap-gto] <ap-gto@...> wrote:
> >
> > On 8/24/2019 9:31 PM, Robert Chozick rchozick@... [ap-gto] wrote:
> >>
> >>
> >> The problem is that you cannot use bias on a CMOS camera. Any exposure
> >> less than 3 seconds is not advisable (not an engineer so not sure why).
> >> So if you cannot use a bias frame on your flats you have to use flat
> >> darks.
> >
> > I don't understand what's magic about 3 seconds. ZWO specifies an
> > exposure range of 32μs-2000s for the '1600 and the '294. I have seen no
> > issues with flats exposed for less than 1 second with my '1600.
> >
> > --- Mike
> >
> >
> >
> > ------------------------------------
> > Posted by: Mike Dodd <mike@...>
> > ------------------------------------
> >
> > To UNSUBSCRIBE, or for general information on the ap-gto list
> > see http://groups.yahoo.com/group/ap-gto
> > ------------------------------------
> >
> > Yahoo Groups Links
> >
> >
> >
>
>
>
> Robert Chozick
> rchozick@...
>
>
>
>
>
>


Mike Dodd
 

On 8/25/2019 12:54 PM, rchozick rchozick@aol.com [ap-gto] wrote:


The issue is not the length of time for the flat for flat purposes. The
issue is that it is my understanding that the CMOS camera itself should
not ever have an exposure less than three seconds.
OK; whatever works for you. ZWO specifies the cameras as good for exposures as short as 32 microseconds, which is a lot shorter than 3 seconds.

I think this thread has reached its conclusion, so I will not post further to it.

--- Mike


Robert Chozick
 

I agee. Enjoy your camera. 

Robert 


On Aug 25, 2019, at 12:03 PM, Mike Dodd mike@... [ap-gto] <ap-gto@...> wrote:

 

On 8/25/2019 12:54 PM, rchozick rchozick@... [ap-gto] wrote:
>
>
> The issue is not the length of time for the flat for flat purposes. The
> issue is that it is my understanding that the CMOS camera itself should
> not ever have an exposure less than three seconds.

OK; whatever works for you. ZWO specifies the cameras as good for
exposures as short as 32 microseconds, which is a lot shorter than 3
seconds.

I think this thread has reached its conclusion, so I will not post
further to it.

--- Mike


Dale Ghent
 


I have to say that makes Zero sense. I think you may be confusing that with something else, but I can’t think of what that might be. There is nothing inherent to CMOS sensors that dictates a minimum exposure time. In fact CMOS sensors are regularly used for high speed captures in the astronomy world and outside of it. CMOS sensors are the domain of consumer and professional cameras of all sorts and are used all the time with millisecond exposures. 

On Aug 25, 2019, at 12:54 PM, rchozick rchozick@... [ap-gto] <ap-gto@...> wrote:

The issue is not the length of time for the flat for flat purposes. The issue is that it is my understanding that the CMOS camera itself should not ever have an exposure less than three seconds.   I always take my flats at around 30,000 ADU.  I adjust the exposure to get there.  To get there with my flat machine I have to put in neutral density film. 


On Aug 25, 2019, at 6:41 AM, Dale Ghent daleg@... [ap-gto] <ap-gto@...> wrote:

 


The only goal with flats is to get an illuminated frame from an evenly-lit light source, where the histogram is a nice hump centered at the %50 point. It doesn't have to be perfectly nailed at 50%. in truth, a flat is alright so long as no pixels are clipped one way or the other, but a perfect flat will look like that.

Obviously, getting that kind of result with an correct (or, adequate) exposure time is going to be specific to every setup, from optics speed and sensor settings, to the intensity of the flat light source - as well as considerations for any annoying interactions such as the rolling shutter one I mentioned. So there really isn't (and shouldn't be) a "rule" when it comes to exposure length and flats.

To take the fiddling out of arriving at this kind of result, we put a Flat Wizard into the app I contribute to:
https://nighttime-imaging.eu/docs/develop/site/tabs/flatwizard/

/dale

> On Aug 25, 2019, at 2:53 AM, Robert Chozick rchozick@... [ap-gto] <ap-gto@...> wrote:
>
> The 3 second rule I have seen is from multiple posts as I go through comments on issues. I generally try to get the benefit of common experiences and comments whether I totally understand them or not. I don’t get as much chance as I would like to image in dark skies and don't want to take any chances. I regularly used .3 sec flats for my STF-8300 CCD without issue.
>
> On Aug 25, 2019, at 12:46 AM, Dale Ghent daleg@... [ap-gto] <ap-gto@...> wrote:
>
>
> You may need to have a longer exposure time for flats because LED light panels adjust their brightness via PWM. If the shutter speed of the camera is quicker than what the panel is being driven at for its level of brightness, you will get banding across the image due to how PWM rates interact with rolling shutter sensors.
>
> This need for a longer exposure can lead to an issue where the dimmest setting of a panel is still too bright for that exposure length and the flat is blown out, and one must resort to attenuating the panel light with ND film or other methods.
>
> /dale
>
> > On Aug 24, 2019, at 10:16 PM, Mike Dodd mike@... [ap-gto] <ap-gto@...> wrote:
> >
> > On 8/24/2019 9:31 PM, Robert Chozick rchozick@... [ap-gto] wrote:
> >>
> >>
> >> The problem is that you cannot use bias on a CMOS camera. Any exposure
> >> less than 3 seconds is not advisable (not an engineer so not sure why).
> >> So if you cannot use a bias frame on your flats you have to use flat
> >> darks.
> >
> > I don't understand what's magic about 3 seconds. ZWO specifies an
> > exposure range of 32μs-2000s for the '1600 and the '294. I have seen no
> > issues with flats exposed for less than 1 second with my '1600.
> >
> > --- Mike
> >
> >
> >
> > ------------------------------------
> > Posted by: Mike Dodd <mike@...>
> > ------------------------------------
> >
> > To UNSUBSCRIBE, or for general information on the ap-gto list
> > see http://groups.yahoo.com/group/ap-gto
> > ------------------------------------
> >
> > Yahoo Groups Links
> >
> >
> >
>
>
>
> Robert Chozick
> rchozick@...
>
>
>
>
>
>


Robert Chozick
 

I will ask my tech guru friend why he thinks that is the case and report his answer.  I am an attorney and like science but only learn through reading about it now with no formal background.


On Aug 25, 2019, at 1:40 PM, Dale Ghent daleg@... [ap-gto] <ap-gto@...> wrote:


I have to say that makes Zero sense. I think you may be confusing that with something else, but I can’t think of what that might be. There is nothing inherent to CMOS sensors that dictates a minimum exposure time. In fact CMOS sensors are regularly used for high speed captures in the astronomy world and outside of it. CMOS sensors are the domain of consumer and professional cameras of all sorts and are used all the time with millisecond exposures. 

On Aug 25, 2019, at 12:54 PM, rchozick rchozick@... [ap-gto] <ap-gto@...> wrote:

The issue is not the length of time for the flat for flat purposes. The issue is that it is my understanding that the CMOS camera itself should not ever have an exposure less than three seconds.   I always take my flats at around 30,000 ADU.  I adjust the exposure to get there.  To get there with my flat machine I have to put in neutral density film. 


On Aug 25, 2019, at 6:41 AM, Dale Ghent daleg@... [ap-gto] <ap-gto@...> wrote:

 


The only goal with flats is to get an illuminated frame from an evenly-lit light source, where the histogram is a nice hump centered at the %50 point. It doesn't have to be perfectly nailed at 50%. in truth, a flat is alright so long as no pixels are clipped one way or the other, but a perfect flat will look like that.

Obviously, getting that kind of result with an correct (or, adequate) exposure time is going to be specific to every setup, from optics speed and sensor settings, to the intensity of the flat light source - as well as considerations for any annoying interactions such as the rolling shutter one I mentioned. So there really isn't (and shouldn't be) a "rule" when it comes to exposure length and flats.

To take the fiddling out of arriving at this kind of result, we put a Flat Wizard into the app I contribute to:
https://nighttime-imaging.eu/docs/develop/site/tabs/flatwizard/

/dale

> On Aug 25, 2019, at 2:53 AM, Robert Chozick rchozick@... [ap-gto] <ap-gto@...> wrote:
> 
> The 3 second rule I have seen is from multiple posts as I go through comments on issues. I generally try to get the benefit of common experiences and comments whether I totally understand them or not. I don’t get as much chance as I would like to image in dark skies and don't want to take any chances. I regularly used .3 sec flats for my STF-8300 CCD without issue.
> 
> On Aug 25, 2019, at 12:46 AM, Dale Ghent daleg@... [ap-gto] <ap-gto@...> wrote:
> 
> 
> You may need to have a longer exposure time for flats because LED light panels adjust their brightness via PWM. If the shutter speed of the camera is quicker than what the panel is being driven at for its level of brightness, you will get banding across the image due to how PWM rates interact with rolling shutter sensors.
> 
> This need for a longer exposure can lead to an issue where the dimmest setting of a panel is still too bright for that exposure length and the flat is blown out, and one must resort to attenuating the panel light with ND film or other methods.
> 
> /dale
> 
> > On Aug 24, 2019, at 10:16 PM, Mike Dodd mike@... [ap-gto] <ap-gto@...> wrote:
> > 
> > On 8/24/2019 9:31 PM, Robert Chozick rchozick@... [ap-gto] wrote:
> >> 
> >> 
> >> The problem is that you cannot use bias on a CMOS camera. Any exposure
> >> less than 3 seconds is not advisable (not an engineer so not sure why).
> >> So if you cannot use a bias frame on your flats you have to use flat
> >> darks.
> > 
> > I don't understand what's magic about 3 seconds. ZWO specifies an 
> > exposure range of 32μs-2000s for the '1600 and the '294. I have seen no 
> > issues with flats exposed for less than 1 second with my '1600.
> > 
> > --- Mike
> > 
> > 
> > 
> > ------------------------------------
> > Posted by: Mike Dodd <mike@...>
> > ------------------------------------
> > 
> > To UNSUBSCRIBE, or for general information on the ap-gto list
> > see http://groups.yahoo.com/group/ap-gto
> > ------------------------------------
> > 
> > Yahoo Groups Links
> > 
> > 
> > 
> 
> 
> 
> Robert Chozick
> rchozick@...
> 
> 
> 
> 
> 
> 



Robert Chozick




W Hilmo
 

I’ve been using an ASI1600MM-cool (3rd gen) for a few years now.

 

I routinely shoot flats that are between 0.1 and 1.0 seconds by shooting directly into a clear, blue sky in broad daylight and then bias calibrating them.  Flats made this way work great – even better that t-shirt flats or twilight flat, where I can get longer exposures.  Here are the factors that I think about, in terms of flat duration and bias vs dark calibration:

 

I don’t know of any sensor that benefits from a 3 second exposure.  My older CCD camera benefits from a flat that is 3 seconds or longer because that mitigates issues with the mechanical shutter and even illumination.  My ASI1600 does not have a mechanical shutter, only an electronic one.  It can take very, very short exposures with no illumination issues.

 

It does turn out that on very, very short exposures (somewhere shorter than 0.1 milliseconds, but I don’t remember the exact length), the division of labor between the hardware and the driver changes.  For zero second exposures, the hardware manages the exposure time.  I believe that for slightly longer exposures, the driver manages the exposure time.  This results in slight but consistent differences in the readout values between 0 second frames with zero illumination and 0.1 seconds with zero illumination.  For that reason, my “bias” frames are taken at 0.1 seconds, and I stay above 0.1 seconds for all other exposures as well.

 

Finally, regarding calibration of the flats themselves, this is important if you want the best correction.  The issue is that mathematically, you want the operation to include only the illumination data when the light frames are flat calibrated.  If you don’t do something to correct for the bias or pedestal, then the math is slightly off.  For this reason, I bias calibrate my flat frames.  Some people use flat darks, but I doubt that there is enough dark current in a sub second exposure to worry about the difference between a dark and a bias.

 

I hope that this helps,

-Wade

 

From: ap-gto@...
Sent: Sunday, August 25, 2019 12:07 PM
To: ap-gto@...
Subject: Re: [ap-gto] NGC 7000 with CMOS camera

 

 

I will ask my tech guru friend why he thinks that is the case and report his answer.  I am an attorney and like science but only learn through reading about it now with no formal background.

 

 

On Aug 25, 2019, at 1:40 PM, Dale Ghent daleg@... [ap-gto] <ap-gto@yahoogroups..com> wrote:

 


I have to say that makes Zero sense. I think you may be confusing that with something else, but I can’t think of what that might be. There is nothing inherent to CMOS sensors that dictates a minimum exposure time. In fact CMOS sensors are regularly used for high speed captures in the astronomy world and outside of it. CMOS sensors are the domain of consumer and professional cameras of all sorts and are used all the time with millisecond exposures. 


On Aug 25, 2019, at 12:54 PM, rchozick rchozick@... [ap-gto] <ap-gto@...> wrote:

The issue is not the length of time for the flat for flat purposes. The issue is that it is my understanding that the CMOS camera itself should not ever have an exposure less than three seconds.   I always take my flats at around 30,000 ADU..  I adjust the exposure to get there.  To get there with my flat machine I have to put in neutral density film. 


On Aug 25, 2019, at 6:41 AM, Dale Ghent daleg@... [ap-gto] <ap-gto@...> wrote:

 


The only goal with flats is to get an illuminated frame from an evenly-lit light source, where the histogram is a nice hump centered at the %50 point. It doesn't have to be perfectly nailed at 50%. in truth, a flat is alright so long as no pixels are clipped one way or the other, but a perfect flat will look like that.

Obviously, getting that kind of result with an correct (or, adequate) exposure time is going to be specific to every setup, from optics speed and sensor settings, to the intensity of the flat light source - as well as considerations for any annoying interactions such as the rolling shutter one I mentioned. So there really isn't (and shouldn't be) a "rule" when it comes to exposure length and flats.

To take the fiddling out of arriving at this kind of result, we put a Flat Wizard into the app I contribute to:
https://nighttime-imaging.eu/docs/develop/site/tabs/flatwizard/

/dale

> On Aug 25, 2019, at 2:53 AM, Robert Chozick rchozick@... [ap-gto] <ap-gto@...> wrote:
> 
> The 3 second rule I have seen is from multiple posts as I go through comments on issues. I generally try to get the benefit of common experiences and comments whether I totally understand them or not. I don’t get as much chance as I would like to image in dark skies and don't want to take any chances. I regularly used .3 sec flats for my STF-8300 CCD without issue.
> 
> On Aug 25, 2019, at 12:46 AM, Dale Ghent daleg@... [ap-gto] <ap-gto@...> wrote:
> 
> 
> You may need to have a longer exposure time for flats because LED light panels adjust their brightness via PWM. If the shutter speed of the camera is quicker than what the panel is being driven at for its level of brightness, you will get banding across the image due to how PWM rates interact with rolling shutter sensors.
> 
> This need for a longer exposure can lead to an issue where the dimmest setting of a panel is still too bright for that exposure length and the flat is blown out, and one must resort to attenuating the panel light with ND film or other methods.
> 
> /dale
> 
> > On Aug 24, 2019, at 10:16 PM, Mike Dodd mike@... [ap-gto] <ap-gto@...> wrote:
> > 
> > On 8/24/2019 9:31 PM, Robert Chozick rchozick@... [ap-gto] wrote:
> >> 
> >> 
> >> The problem is that you cannot use bias on a CMOS camera. Any exposure
> >> less than 3 seconds is not advisable (not an engineer so not sure why).
> >> So if you cannot use a bias frame on your flats you have to use flat
> >> darks.
> > 
> > I don't understand what's magic about 3 seconds. ZWO specifies an 
> > exposure range of 32μs-2000s for the '1600 and the '294. I have seen no 
> > issues with flats exposed for less than 1 second with my '1600.
> > 
> > --- Mike
> > 
> > 
> > 
> > ------------------------------------
> > Posted by: Mike Dodd <mike@...>
> > ------------------------------------
> > 
> > To UNSUBSCRIBE, or for general information on the ap-gto list
> > see http://groups.yahoo.com/group/ap-gto
> > ------------------------------------
> > 
> > Yahoo Groups Links
> > 
> > 
> > 
> 
> 
> 
> Robert Chozick
> rchozick@...
> 
> 
> 
> 
> 
> 

 

 

Robert Chozick

 

 

 


Robert Chozick
 

I did some research and talked to my friend about it.  I think the issue is that bias in a CMOS is more variable than in a CCD.  I saw a graph, shown below, that shows you need at least a .3 sec exposure to make bias closer to the base bias of longer exposures.  What this says to me is that it is not a 3 sec minimum but a .3 sec minimum and it appears as if maybe 1.5 sec is better.  The comments state that due to some base bias change throughout the range in a CMOS that flat darks would always be better than bias, even for short flat exposures.  It is really easy to cover the scope and just fire off 20 more exposures at the same settings that you just took your flats.  If this graph shows the repeatable bias each time maybe the only lesson here is to take flat darks instead of bias for any exposure flat.  

The other comment about flat darks vs bias is in regard to the amp glow.  The amp glow may be brighter in the short exposure vs bias and be enough to make a difference in calibrating flats.

Take the above for whatever its worth and you can do your own research on the various posts.  I’m just the messenger :-)

Robert




On Aug 26, 2019, at 1:59 PM, 'Wade Hilmo' y.groups@... [ap-gto] <ap-gto@...> wrote:


I’ve been using an ASI1600MM-cool (3rd gen) for a few years now.

 

I routinely shoot flats that are between 0.1 and 1.0 seconds by shooting directly into a clear, blue sky in broad daylight and then bias calibrating them.  Flats made this way work great – even better that t-shirt flats or twilight flat, where I can get longer exposures.  Here are the factors that I think about, in terms of flat duration and bias vs dark calibration:

 

I don’t know of any sensor that benefits from a 3 second exposure.  My older CCD camera benefits from a flat that is 3 seconds or longer because that mitigates issues with the mechanical shutter and even illumination.  My ASI1600 does not have a mechanical shutter, only an electronic one.  It can take very, very short exposures with no illumination issues.

 

It does turn out that on very, very short exposures (somewhere shorter than 0.1 milliseconds, but I don’t remember the exact length), the division of labor between the hardware and the driver changes.  For zero second exposures, the hardware manages the exposure time.  I believe that for slightly longer exposures, the driver manages the exposure time.  This results in slight but consistent differences in the readout values between 0 second frames with zero illumination and 0.1 seconds with zero illumination.  For that reason, my “bias” frames are taken at 0.1 seconds, and I stay above 0.1 seconds for all other exposures as well.

 

Finally, regarding calibration of the flats themselves, this is important if you want the best correction.  The issue is that mathematically, you want the operation to include only the illumination data when the light frames are flat calibrated.  If you don’t do something to correct for the bias or pedestal, then the math is slightly off.  For this reason, I bias calibrate my flat frames.  Some people use flat darks, but I doubt that there is enough dark current in a sub second exposure to worry about the difference between a dark and a bias.

 

I hope that this helps,

-Wade

 

From: ap-gto@... <ap-gto@...> 
Sent: Sunday, August 25, 2019 12:07 PM
To: ap-gto@...
Subject: Re: [ap-gto] NGC 7000 with CMOS camera

 

  

I will ask my tech guru friend why he thinks that is the case and report his answer..  I am an attorney and like science but only learn through reading about it now with no formal background.

 

 

On Aug 25, 2019, at 1:40 PM, Dale Ghent daleg@... [ap-gto] <ap-gto@yahoogroups..com> wrote:

 


I have to say that makes Zero sense. I think you may be confusing that with something else, but I can’t think of what that might be. There is nothing inherent to CMOS sensors that dictates a minimum exposure time. In fact CMOS sensors are regularly used for high speed captures in the astronomy world and outside of it. CMOS sensors are the domain of consumer and professional cameras of all sorts and are used all the time with millisecond exposures. 


On Aug 25, 2019, at 12:54 PM, rchozick rchozick@... [ap-gto] <ap-gto@...> wrote:

The issue is not the length of time for the flat for flat purposes. The issue is that it is my understanding that the CMOS camera itself should not ever have an exposure less than three seconds.   I always take my flats at around 30,000 ADU..  I adjust the exposure to get there.  To get there with my flat machine I have to put in neutral density film.. 


On Aug 25, 2019, at 6:41 AM, Dale Ghent daleg@... [ap-gto] <ap-gto@...> wrote:

 


The only goal with flats is to get an illuminated frame from an evenly-lit light source, where the histogram is a nice hump centered at the %50 point. It doesn't have to be perfectly nailed at 50%. in truth, a flat is alright so long as no pixels are clipped one way or the other, but a perfect flat will look like that.

Obviously, getting that kind of result with an correct (or, adequate) exposure time is going to be specific to every setup, from optics speed and sensor settings, to the intensity of the flat light source - as well as considerations for any annoying interactions such as the rolling shutter one I mentioned. So there really isn't (and shouldn't be) a "rule" when it comes to exposure length and flats.

To take the fiddling out of arriving at this kind of result, we put a Flat Wizard into the app I contribute to:
https://nighttime-imaging.eu/docs/develop/site/tabs/flatwizard/

/dale

> On Aug 25, 2019, at 2:53 AM, Robert Chozick rchozick@... [ap-gto] <ap-gto@...> wrote:
> 
> The 3 second rule I have seen is from multiple posts as I go through comments on issues. I generally try to get the benefit of common experiences and comments whether I totally understand them or not. I don’t get as much chance as I would like to image in dark skies and don't want to take any chances. I regularly used .3 sec flats for my STF-8300 CCD without issue.
> 
> On Aug 25, 2019, at 12:46 AM, Dale Ghent daleg@... [ap-gto] <ap-gto@...> wrote:
> 
> 
> You may need to have a longer exposure time for flats because LED light panels adjust their brightness via PWM. If the shutter speed of the camera is quicker than what the panel is being driven at for its level of brightness, you will get banding across the image due to how PWM rates interact with rolling shutter sensors.
> 
> This need for a longer exposure can lead to an issue where the dimmest setting of a panel is still too bright for that exposure length and the flat is blown out, and one must resort to attenuating the panel light with ND film or other methods.
> 
> /dale
> 
> > On Aug 24, 2019, at 10:16 PM, Mike Dodd mike@... [ap-gto] <ap-gto@...> wrote:
> > 
> > On 8/24/2019 9:31 PM, Robert Chozick rchozick@... [ap-gto] wrote:
> >> 
> >> 
> >> The problem is that you cannot use bias on a CMOS camera. Any exposure
> >> less than 3 seconds is not advisable (not an engineer so not sure why).
> >> So if you cannot use a bias frame on your flats you have to use flat
> >> darks.
> > 
> > I don't understand what's magic about 3 seconds. ZWO specifies an 
> > exposure range of 32μs-2000s for the '1600 and the '294. I have seen no 
> > issues with flats exposed for less than 1 second with my '1600.
> > 
> > --- Mike
> > 
> > 
> > 
> > ------------------------------------
> > Posted by: Mike Dodd <mike@...>
> > ------------------------------------
> > 
> > To UNSUBSCRIBE, or for general information on the ap-gto list
> > see http://groups.yahoo.com/group/ap-gto
> > ------------------------------------
> > 
> > Yahoo Groups Links
> > 
> > 
> > 
> 
> 
> 
> Robert Chozick
> rchozick@...
> 
> 
> 
> 
> 
> 

 

 

Robert Chozick

 

 

 



Robert Chozick