Ok wvreeven now is more clear for me
Thanks and compliments for your nice photos posted in your profile.

Today will be a clear night without moon, so I'll take my first photos with my new camera... always if will not occur problems.

Paolo

you should almost certainly not use gain 0 with these new Sony sensors, the read noise is much higher at gain 0, but at gain 100 the read noise drops significantly with a slight gain in dynamic range. it is something unique to these Sony sensors, this logic does not apply to the asi1600.

Thanks for pointing that out. I failed to take the bottom diagram into account. I will start using gain 100 from now on. Luckily the lack of amp glow means that I only need to re-shoot one series of dark and that it!

@suvowner thanks a lot for your explantion.
So I'll use always gan 100 for better results.

Paolo

Using 100 gain is perfect if you're doing short subs where you can't take advantage of full well capacity, and need to image with light pollution.

However, if you have dark skies, and want to do long subs, you're better off at gain 0 so you can use the full well capacity of the sensor. This means you won't saturate bright areas as quickly, like getting blown out stars.

Hi Lead_weight,
from my asi quick user guide i s written that using the Highest Dynamic Range setting (so with gain at zero) gives better results (as you also said) with longer exposures, but as well as with brighter (light polluted) skies.
So following this guide I'll need to use gain zero under light polluted skie... strange....

Thanks in advance!
Paolo

What leadweight is saying is generally true, however because of the huge drop in read noise at gain 100, which is unique to these new Sony sensors , his statement is most likely not correct.

Just know that this topic has been debated at length in various threads on cloudy nights, with often a lot of disagreement , so don’t take it from me, please watch the YouTube video by dr robin glover, the developer of sharp cap, also be sure and watch the follow video on gain settings, which is linked in the description

I've watched this video before, it's great. But what he basically says is there's no reason to ever take long subs with CMOS cameras. There's many advantages to doing short subs, like less reliance on tracking. But there is one good reason to do long subs, and I don't remember if he mentions or not. Cameras like the 2600 and 6200 have come out since this video was made. 55 MB subs on the 2600 and 100 MB subs on the 6200 require tons of processing power and hard drive space. You basically need a super computer to process large stacks of images with huge file sizes.

Yesterday I stacked an image that had 1,000 frames of 60s each. It took half the day. The longest I've done was 1,600 frames which took just over 8 hours on a 8 core iMac Pro. And that's images from the ASI1600 which are only 32 MB each. The largest stack I've done with the 2600 was around 300-400 images of 2 min subs. Going to 2 minutes of course cuts the number of subs in half but I'm hitting the limit of sub length in my bortle 7 skies with that time.

In his follow Up where he discusses optimum gain that he had to skip over in the original talk, my takeaway is the lower read noise at gain 100 will get you a lot more faint detail , because the internal electronics of the sensor changes its conversion gain and drops the read noise with almost no penalty in dynamic range.

I agree that a ridiculous number of subs is counter productive, but you can’t get away from the read noise change at gain 100 with this camera, so at gain 0 your total imaging time would need to increase by about 9 times Give or take to get the same faint detail than at gain 100. That is really the main point I was trying to convey by referencing robin’s talk.

Doing some quick numbers using the tools.sharpcap.co.uk data for my sky (SQM 19.0) and telescope F-ratio (F/5) I get recommended exposure times of 3.16, 9.49 and 132 sec for L, RGB and 7 nm narrow band respectively when using a gain of 100 (read noise of 1.5) and of 17.2, 51.7 and 721 seconds when using gain 0. So to keep things practical, as Lead_weight also indicates, I should consider using gain 0 for L and RGB and gain 100 (or rather 110 because it should be slightly above the gain at which the change happens) for my sky and telescope. Very interesting...

Not sure if the tool you used accounts for the multiplication factor for color/rgb if I remember was factor of 3, and factor of 25 for 12nm and 100 for 3nm narrowband.

It does. Check the end of the video from which you took that still. The tool is mentioned there (which is why I used it in the first place). When you go to the webpage and select Mono-camera then you can see how selecting L or RGB or Narrowband (7 nm) modifies the computed sky electron rate.

By the way, the whole computation is done in JavaScript so someone could grab it and transform it in an offline tool