First takes and questions on dedicated CMOS after 10y with DSLRs (dithering, flat, flattener) - Forums

Topic starter

Jan 20, 2021

Dear Friends,

I am in Astrophotography for almost a decade and was using DSLRs only but changed to a dedicated CMOS camera lately. So here are some thoughts and questions regarding my first experiences.

The very first image I took I was struck with major image calibration issues which I never experienced with DSLRs before (and of course the points should also apply to DSLRs since these are also CMOS cameras but I found that so far they did not matter as majorly as with dedicated CMOS cameras).

After reading up on some basics like, "USE FLAT-Darks!" (which I didn't at first, but i used "optimize" which appears to be a big nono but never mattered with my DSLR imaging).

Second, in my first try I had extrem "rainy noise" and was told DITHER OR DIE! which I also never did with DSLRs...
This particular fact I just could not take for granted as I bought a super accurate Lightrack II so I would not need guiding. Therefore I "refuse" to invest yet another junk of money, cables and weight for only shifting my images, especially since I only have tracking in one axis which I don't know if would be enough anyways (first question that I would be interested in)?

So my approach was to rotate my camera every bunch if images for 90 deg. and capture around a meridian flip which also introduces more variance (image drift into the other direction, I think). Since I am using a square sensor ASI533 I didn't lose any frame area.
Long blabla-short: I processed my first 160min (80x2) of exposure and tada there was not a single sign of "rainy noise".

Since I am new to this I wonder if you can tell me that what I got out of only 160min is "normal" for my setup (40/180mm Askar + ASI533 + l-eNhance and a Bortle 7 Sky)?

I personally have to admit that I expected more, especially with the higher Ha sensitivity compared to my DSLRs, but on the other hand, I never used Filters as I am spoilt with a Bortle3-Alpine Sky

https://astrob.in/l859re/0/

If you take a closer look you might realise that I have terrible stars towards the corners. This is not the Askars fault (which I used also on full-frame witch almost perfect stars) but mine as I didn't have the fitting adapters and was 11mm off the "sweet spot" for the flattener.

But there was clear skies for the first time in ages so I just had to collect them photons

This is where my third question arises. If I get the distance right and continue to collect data, is there a chance that I can combine both data, or will I have trouble mixing "round and triangle star" data?

A fourth question is regarding my flat with the filter on. I am getting a weird diagonal red-green color gradiant when capturing them. I tried different techniques and all lead to these kind of flats, and the line on the bottom is also weird?
The calibrated images and the stack look ok, I guess

I already appreciate your input and am curious about your answers.
CS Robert

andrea tasselli

Frequent contributor

Jan 20, 2021

The stars aren't just terrible (HUGE coma) around the corners they are so everywhere, if I can be so blunt. Well, maybe a tiny spot around lower middle is OK-ish. Can you re-use this exposure? Sure you can but it won't do much good mixing good data with bad. You might remove the stars and try to blend in with the new improved shots (without stars) to get back some of the nebulosity but I cannot guaranteed it. The weirdness of the flat might be related to the interference resonance in the filter at f/4.5. It is so there's nothing you can do about it apart from mounting it in front of the lens. Hard to say without knowing more on how you take flats and how you post-process it. Additionally, you are doing yourself a disservice not binning the camera as you don't need that kind of sampling. The image colour is OK and overall it isn't that bad at all. Finally, I don't dither. Never did and never will. Natural drift should take care of everything and DSLR are notoriously difficult to post-process correctly as issues around darks and how you take them can affect the results quite significantly.

dkamen

Jan 21, 2021

Hi Robert,

When people say "dither" usually they mean move the sensor in any way so that the same stuff lands on different pixels. It is not carved in stone that you must dither a few pixels with the autoguider, it's just sort of a nice thing to be able to do since you do not have to crop too much of the field. If that trick of rotating the camera 90 degrees worked for you, then great, for all intents and purposes you are dithering

What you have captured does look normal for 180 minutes. Quite okay in fact, for such a small aperture. Do no forget, narrowband signal is very weak because it represents a tiny proportion of incoming light. Just because the filter leaves out unwanted stuff does not mean the signal that gets through is made stronger in any way. High sensitivity will result in a larger percentage being captured, but a larger percentage of something small is still small. On fainter targets you might find that you need to increase the gain or offset to get anything meaningful of a 2 minute exposure (or just expose more than 2 minutes), but in this specific image I'd say you did very well.

I wouldn't agree with Andrea that the stars are terrible all over the image. They are acceptable (to me at least) at the center. But the take home lesson here is spacing matters a lot when a flattener is involved, and if you have the filter between the flattener and the lens this *does* affect distancing by a few millimeters (I don't know exactly how much and in what direction). In other words, once you plug that 11mm spacer it might not work as well as you hope, although of course being 2-3mm off is not the same thing at all as being 1.1cm off.

That leaves your question about the flat. First of all, an easy way to tell if it's because of the filter is to remove it and see if flats are still showing the same pattern. If they do, then it is probably down to different response of parts of your sensor to incoming light. Red pixels are a bit more sensitive here, green pixels are a bit more sensitive there and the end result is what you are seeing. Nothing to worry about in that case, it's among the things flats are supposed to correct.

The horizontal line at the bottom is more problematic. It is too straight to be an optical artifact IMO, it must be something electronic or digital in nature. It reminds me of an identical line I see in 80-90% of my subs taken with the ASI178MC and 178MM, except mine is at the top. It has proved impossible to calibrate away and I usually just crop the top 50 pixels or so.
Cheers,
Dimitris

Helpful Respectful Engaging Supportive

Nadir Astro

Jan 21, 2021

Hi Robert,

"Dither or die" is equally valid to DSLR or CMOS cameras, especially under Bortle 7 skies. If you used to image under Bortle 3 before, that explains why you didn't dither before.

Regardless, your first light of California nebula is quite good ! I guess you can combine new images after you make perfect backfocus distance and old images, at least on nebula if you process nebula and stars separately: combine both nebulas and take stars only from correct backfocus sum.

Helpful

DarnitsCloudy

Feb 24, 2021

Hi Robert,

First, I certainly would not bin at 180mm FL!
I've had issues too, going from DSLR to a 2600MC. I think the smaller pixels -> Higher resolution leads to some calibration issues being easier to spot… But most of the issues I've had came down to post-processing issues. Going to FITS files changed my workflow a bit, and introduced a number of errors that took some experimenting to fix.

The color of your flats can be quite different depending on whether it's a skyflat (often weak in red channel) or from a more uniform flat-device… In the end it shouldn't matter, as long as you color balance the final result at some point. If I tell Pixinsight to treat each channel seperately in my flat, I get a green frame before background extraction, if I don't I often get a red frame.

By the way, be prepared to work with adapters and spacer rings to find the optimum flatfield point - After weeks of experimenting, I'm still not getting round stars at the edge - My FF DSLR had better edge stars on the same scope.

-J

Helpful Insightful Respectful Engaging Supportive