ASI2600MC Success

[rockenrock]

Hi Robin,
You know how much I struggled, from the beginning, to get my QHY168C downloading consistently. I could never get Sensor Analysis to work with it. Thank you for finding the reason Sensor Analysis could not complete -- changing exposure did not change the ADU level, then it would suddenly jump to a new level. QHY has confirmed my camera is working correctly. Hmmmm. Thank you for sticking with me for my painful journey the last 6 weeks!

Now I have the ASI2600MC. It connected and downloaded images flawlessly using a 2 meter USB to powered hub, and a 3m cable from hub to laptop. Next I ran sensor analysis 3 times without problem. Linearity was 67% on first run because the sensor temp was still going down. Last 2 runs were 99.4% and 99.5%.

After plotting the data same as ZWO user manual, it is very consistent and close to spec.

For this camera the big step down in read noise is at gain = 100. In sensor analysis there is a data point at gain 61, and the next is 100. Is the 61 by design, or could on another run could it also have a data point just below 100?

In the AppData/Sharpcap/Sensor Analysis directory I can see one ASI2600 json file, not three. Does the latest run overwrite the previous run?

Thanks,
Roger

[rockenrock]

As a reminder, this is the problem with the QHY168c. Same results from loaner 168C as my 168C. SKD version did not change results. Light panel vs natural light did not change results.
Roger

[admin]

Hi Roger,

Good to hear that the ZWO camera works flawlessly

The early ZWO cameras that had the step change in read noise all stepped down between gain 60 and 61, which is why SharpCap adds extra points to measure there. Now there are so many different cameras (and many different points where the read noise steps down for different models) that it is not realistic to add all the possible points as measurements points.

Fortunately you can add additional gain measurement points if you wish using SharpCap's scripting language - see this thread : viewtopic.php?t=5054

As to the single output file - yes, previous measurements usign the same camera/bit depth/read mode are overwritten, so only the most recent is preserved.

cheers,

Robin

[rockenrock]

Robin,
I am surprised that you set up the software that I can copy this script into the scripts pull down...

SharpCap.SelectedCamera.SetGainsOfInterestForSensorAnalysis([200,210,220,230,240,250,260,270,280,290,300,31
0,320,330,340,350])

and then it will insert these test points into the main sensor analysis. Of course I need to change to my 'gains of interest' into the scripts pull down.
If I keep learning so much from you, soon I will know enough to be dangerous!

The quest continues.
Roger

[rockenrock]

Robin,
I am very surprised. My first script, and I had great success to run the script!!! I copied above into the lower half of the Scripting Console, changed the gains I wanted, hit run button, then started Sensor Analysis. Took a lot longer to run, but still reasonable.
Thanks for your advice!
Roger

[admin]

Hi Roger,

good to hear that it all worked nicely for you

cheers,

Robin

[rockenrock]

Hi Robin,
I really like the Brain measurement of sky brightness. I could see the effect of adding my dew shield (blocks local lights), and elevation. Very nice. I am still learning it, so my questions below may have answers in the manual. I am using version 4.0.9246. I know there is a newer version, which I will install soon.

When I run Brain the choices are to target for unity gain, or maximum dynamic gain. For the ASI2600MC, the closest gain setting to unity gain is zero and this is the same gain setting for maximum dynamic range. So does this choice have any meaning for this camera?

The solution always was to set gain to minimum, no matter I manually input some different sky values, or change the maximum exposure time. So what conditions would result in a higher gain, ie the magical 100? My guess is to go to a really low sky value which you discussed in your Sharpcap presentation.

When I run Brain, and select red, or green, or blue channel it will not solve (stuck at a very low msec). But it works with darkest. I had ignore edges turned on. So what is darkest? Does it matter that it would not complete with R/G/B?

None of above has me stuck from imaging. I just want to know more.
Thanks,
Roger

[admin]

Hi Roger,

the things that will push the brain to suggest higher gains are either

1) darker sky brightness measurement - either darker skies or perhaps a light pollution or narrowband filter
or
2) A shorter maximum exposure length set in the brain window - for instance if you had a wobbly old mount that could only track nicely for 10s then setting a max of 10s might prompt a higher gain.

The ASI2600 does not have a 'unity gain', so when that is selected SharpCap aims for the closest to unity that is available - zero gain for this camera.

I just had a look at the sensor analysis data I have here for the 2600MC, and the data has a higher read noise (about 2.2) at 100 gain than the value of 1.5 shown on the ZWO product page. Now I am not sure if that is a correct measurement or not (I don't have that camera). Certainly the higher read noise than expected at gain 100 will tend to prevent that value from being suggested when choosing 'max dynamic range'. Have you done your own analysis of the camera?

Not sure about why the R/G/B channels get stuck - I will investigate. It's odd, since normal choice is green or lum.

cheers,

Robin

[rockenrock]

Hi Robin,
Thanks for your reply. I will check again later if brain will recommend 100 gain. I am going to blue zone farm in USA West Virginia. I have no idea what the sky value will be there.
My camera had a read noise of 1.5 at 100 gain, and 3.3 at 99 gain.
Attached is my sensor analysis data I ran a ambient (27C) without cooling, and at 0.0C with cooling. I hope it useful for you.
I ran the script adding additional points around the magic 100 gain setting position.
I think it closely agrees with ZWO published data.

Another question: The ASI2600 is 16 bit sensor readout. So it reads 0-65K. If I have a gain setting of 100, the full well is 16,800 ADU per the charts. So if using Sharpcap to measure the ADU of a pixel, what number will mean it is saturated? If you tell me 16,800 maybe I should forget 100 gain setting and go to 0 gain with >50k full well as brain recommends!

Thanks,
Roger

[admin]

Hi Roger,

thanks for sharing the data - the figures you have look pretty much identical to the published ones - I guess the data samply I have was a bad measurement in some way.

You are quite right - the maximum number of electrons that you can measure without saturating is about 17000 at gain 100 - ie 3x less than minimum gain. That isn't always an issue though - if the read noise reduction is enough then maybe using 100 gain will let you take 15s exposures vs 60s at zero gain. That hypothetical situation would mean that you could take 240 x 15s exposures in an hour at 100 gain or 60 x 60s at zero gain. In that case, 100 gain wins even with the lower full well capacity as you are taking four times as many frames but the full well is only 3x smaller.

The brain takes account of all of these things, so if it is recomending zero gain then the low gain, high FWD, longer exposure but fewer frames must be coming out as the best option.

cheers,

Robin