I have managed to get some data on NGC6334 (Cat's Paw) in narrowband. The issue that I have is that when I have calibrated and stacked the SII and OIII data in PI, there is a quite average looking pattern through the image. The Ha data doesn't show this. If you look at the 2 images attached it is quite obvious. Ha consists of 10x5mins and the SII image attached is 16x5mins. Data captured with SBIG STT8300M, Baader NB filters, Stellarvue SVR102T IS scope at f7.5.
Calibrated with darks and flats and all stacked in Pixinsight. Stretch applied with HT and noise reduction done with Multiscale Linear Transform before that.
The "hatching" pattern I am seeing in the data is evident all through the process. I assume it is not due to any issue with the capture as the data was collected over 2 nights and is evident in the SII and OIII on both nights but not in the Ha.
Hi Colin
Yes the subs are dithered. I have done a little more exploration.It seems to really pop out after registration in PI. but again only in the SII and OIII subs.
I redid darks just a couple of weeks ago. Flats were done alongside the lights. But I would have thought if it was a calibration issue it would also affect the Ha data. I may have a redo on the flats just in case.
Sorry about the issues, hopefully these will be resolved quickly.
I feel that the pattern is also in Ha, but buried in the signal. It is noticeable in OIII and SII only perhaps because these usually have much weaker signal and are stretched more, and stronger stretch highlights the pattern more. I hope I'm wrong, but this might be an issue with the camera itself.
I am leaning towards debayering being the issue here. Its too perfectly square it has to be related to pixel/sensor grid and bayer the likely culprit.
But it might bean internal reflection, light bouncing off the sensor surface (giving you the grid pattern) to hit the filter (a flat reflective element) and returning to the sensor to be recorded. might be something a lightbox set of flats can pick up for removal if thats the case. Darks and bias wont.
Its very similar to something I've seen in photography where you get a curved grid pattern caused by a reflection off the shiny sensor surface and back again from the last glass element in a lens thats shiny and curved and has no antiflective coatings. its never noticable in normal photos, but its buried deep and comes out like you've got here.
If its an internal reflection, you might try flipping your filter around?
i dont think its camera problem, maybe cheap filters? still try debayering options to see if thats the cause but otherwies I fear its there in all your shots and next flip filters so front side faces inwards etc or try another set of filters if you can.
Wouldn't be debayering if you're shooting with a mono camera.
It's a bit hard to see at 100% pixel level on your images, because IIS's compression of the JPGs is introducing artefacts as well. But if you're using PI's BatchPreProcessing, maybe try experimenting with the ImageRegistration pixel interpolation methods. I was getting weird artefacts around stars on binned images with "auto" but after some experimenting, bicubic b-spline gave best results and no artefacts.
I tend, however, to agree with Slawomir in that it's there in all filters, just that the Ha signal is dominating, while for the OIII/SII you need to stretch them more. You probably need a lot more OIII/SII signal in terms of sub length - like 15, 20, even 30 min long?
How are you stretching in PI? Maybe try different methods? I know you said you used HT, but was that by copying over the STF settings (ie drag the blue triangle onto HT window etc)? Try masked stretch process and see if better. Try playing with clipping/blackpoint settings. Try stretching using curves instead of HT or masked stretch.