M33 with the RASA8

[Datalord]

Nope, no collimation problems ever. I have some issues with star shapes in 3 corners, which I also strongly suspect is down to tilt. Focus otherwise stays through the night with the Celestron motor focuser installed.

[gorann]

11 minutes ago, vlaiv said:

Why are you trying to compare these two cameras? They don't have same sensor.

QHY247C has IMX193 sensor while ASI2600 has IMX571 sensor.

QHY has 36K full well and 14bit ADC while ASI one has 50K and 16bit ADC.

ASI has rather simple gain scheme - gain is 0.1db units. This means that for 60 gain increase (or closer to 61) - e/ADU value doubles (or rather halves). This makes it easy to calculate any e/ADU value if you know unity gain or any other gain value.

In this particular case, since ADC is 16 bit and FW is 50K - there will be no unity gain - initial gain is close to 50K/64K = 50000 / 65535 = ~0.763e/ADU.

On the other hand QHY seems to have linear gain - or gain translates into e/ADU linearly (look at first graph - it is line).

It starts somewhere around 3.4 at above graph - which is a bit strange since 36K /16K = ~2.13. Initial gain should be somewhere around 2.13e/ADU.  There is error with above charts.

Indeed, here is chart from QHY website:

Other charts can be found here:

https://www.qhyccd.com/index.php?m=content&c=index&a=show&catid=94&id=14&cut=1

In any case - you can't just compare gain 0 and gain 100 with doubling exposure and expect it to be the same.

Only thing that you get by changing gain is read noise decrease and smaller full well capacity. If you want to compare two gains you need to compare these two values.

Let's compare gain 0 and gain 100 for ASI2600 from this perspective:

Gain 0 has e/ADU of 0.763e/ADU and gain 100 has e/ADU of about 0.25. Difference in FW will be about 3 times. You need x3 times shorter exposure with gain 100 to match full well capacity when you sum those three exposures.

Read noise at 0 gain is about 3.4e while at 100 gain it is about 1.48e (very close to 1.5 but a bit less). When you stack 3 of these smaller read noises - you get 1.48 * sqrt(3) = ~2.5634352e

This is less than 3.4e.

There fore it is better to use 1/3 exposure with gain 100 than whole exposure at gain 0 - for ASI2600.

 

 

Thanks Vlaiv for clarifying things as always! Then I stick to gain 100 with my ASI2600 and live with the large amount of subs. We were only comparing the camera settings because Datalord has one camera and I have the other (ASI) but as I expected they are not really comparable since the chips are different.

[gorann]

18 minutes ago, Datalord said:

Nope, no collimation problems ever. I have some issues with star shapes in 3 corners, which I also strongly suspect is down to tilt. Focus otherwise stays through the night with the Celestron motor focuser installed.

Sound very promising, except for the tilt issues.

[vlaiv]

9 minutes ago, gorann said:

Thanks Vlaiv for clarifying things as always! Then I stick to gain 100 with my ASI2600 and live with the large amount of subs. We were only comparing the camera settings because Datalord has one camera and I have the other (ASI) but as I expected they are not really comparable since the chips are different.

If you don't like it - you don't have to go with very large number of subs. Use gain 100 and long exposure. In the end of your session - shoot just a couple of short exposure (like really short - couple of seconds each).

Use these short exposures to fill in saturated parts of the image. Although you have only a few very short exposures - you'll be using them in places where signal is already very strong (saturated sensor in long exposure) - so SNR won't be an issue there.

Only "problem" is deciding how to blend short and long data. I prefer to blend in linear stage, but you can blend in stacking phase or later in post processing stage in PS.

[Datalord]

5 minutes ago, gorann said:

except for the tilt issues

yeah. Unfortunately I have the V1, so I suspect it has something to do with the main mirror in mine. 😞

[gorann]

2 minutes ago, vlaiv said:

If you don't like it - you don't have to go with very large number of subs. Use gain 100 and long exposure. In the end of your session - shoot just a couple of short exposure (like really short - couple of seconds each).

Use these short exposures to fill in saturated parts of the image. Although you have only a few very short exposures - you'll be using them in places where signal is already very strong (saturated sensor in long exposure) - so SNR won't be an issue there.

Only "problem" is deciding how to blend short and long data. I prefer to blend in linear stage, but you can blend in stacking phase or later in post processing stage in PS.

Thanks for that advice Vlaiv. I am running 2 min exposures tonight with the RASA f/2 so it should be managible since I will probably only get about 3 hours before dawn up here.

[gorann]

12 minutes ago, Datalord said:

yeah. Unfortunately I have the V1, so I suspect it has something to do with the main mirror in mine. 😞

What is the difference between V1 and V2 of the RASA11? I assume it has to do with the holding/focusing of the primary. I think the RASA 8 came last so it was maybe V2 from the start.

[Datalord]

3 minutes ago, gorann said:

What is the difference between V1 and V2 of the RASA11? I assume it has to do with the holding/focusing of the primary. I think the RASA 8 came last so it was maybe V2 from the start.

Yeah, V2 has "New Ultra-Stable Focus System (USFS)" which is pretty much just a better primary hold. I don't know what the 8 has, but I assume it is already better in that regard.

[gorann]

11 hours ago, Datalord said:

Yeah, V2 has "New Ultra-Stable Focus System (USFS)" which is pretty much just a better primary hold. I don't know what the 8 has, but I assume it is already better in that regard.

Yes, the RASA 8 also have the USFS, probably adopted from the 11" V2 as the RASA 8 was the last addition to the family.