Hi, I’m using Siril & Sirilic for about a year now and really appreciate this great piece of software. I recently switched to a cooled OSC astronomy camera and started using Duo-NB filters.
I just finished collecting data for the first project with this setup but now ran into issues trying to stack this data.
Since I collected data over 9 sessions (29h+) I’m using Sirilic to stack.
My camera provides images with 16bit but when I look at the extracted Ha & Oiii images, they somewhat look like 8bit or lower. I even tried to stack without any calibration frames but the result is the same (so the issue is not related to calibration frames).
I also tried the 32bits float mode resulting in the same outcome.
I guess this is not what it’s supposed to look like and would greatly appreciate any advice what I do wrong and how to fix this.
The attached image is the HaGroup zoomed in by 144%.
Hello and welcome!
Is that a single image shown, after Ha - OIII split? Or what other process has been done to it?
What does the statistics say about it and about the original image? (you can copy and paste the output of the stat command for simplicity)
Thanks for the reply!
The screenshot shows the final result after stacking all the data I collected, so not just a single frame.
It shows the Ha opened up right after stacking (displayed is the HD-Autostretch). No processing applied at all. I will add the stats later when I’m back home.
Yeah, on a single frame its no problem. Its just with the complete data set.
Could it be a workaround to stack the data as RGB and apply the HaOiii extraction afterwards? (If so, is there a console command for that?)
Of course not. Once demosaiced you cannot retrieve CFA channels.
But I don’t see how you can obtain a low bit image with a dataset of images I just saw.
I know it’s super weird and I don’t understand why it happens. All the images should look somewhat like the single frame I shared. I checked them all one by one
I think I know why. What is the exposure of one sub? 240s? For me it is too low.
if the sigma of the background sky is at least three times the sigma of the bias, then the minimum unit exposure time is reached, as this allows the weakest useful signal to be output.
And I think this is not the case here and the autostretch is too strong here for your data.
I had the same issue with my setup when I was shooting 300s in OIII (I have monochrome camera). Now I shoot 600s and no more issues.
Nvm, I forgot to clear the folder. With just the 10min subs the image is obviously more noisy but I can’t see the weird pattern yet. The question is what happens if I add more data.
Maybe another thought …
I was using the High Full Well Mode on all the images, maybe that is the reason for the low mean (it increases readnoise)?
And since I have full data sets including lights, darks, biases, flats and darkflats I don’t need to tick any of the bias subtract options in sirilic right?
