Beginner DSLR tests: noise, gradients and low-contrast results – looking for advice

Igor Fulvi · Dec 31, 2025 at 07:27 AM

Description

These images are early test shots made to understand DSLR astrophotography workflow and limitations.

Equipment & settings

• Canon EOS R8

• Sky-Watcher Star Adventurer GTi (polar alignment ~1.5′)

• No autoguiding

• Acquisition via StellarMate / INDI

• 30 s exposures

• ISO 1600

• Dark frames applied

Both targets were very low above the horizon (~20°).

Image 1 – M42 (no filter)

📷 VeraLux_Stretch (Grandi).png

• 97 × 30 s, ISO 1600

The nebula is visible but looks very washed out and lacking contrast or “character”.
I would appreciate advice on whether this is mainly:

• an exposure strategy issue,

• a processing workflow issue (contrast, HDR, local enhancement),

• or simply due to target altitude and sky conditions.

Image 2 – NGC 2238 (Optolong L-eNhance)

📷 NGC 2238 (Grandi).png

• 194 × 30 s, ISO 1600

Strong red background, visible diagonal banding and very weak nebula signal.

I’m trying to understand:

• whether 30 s subs are too short for a dual-band filter

• how much total integration time is realistically needed

• optimal ISO choice for the EOS R8

• whether the diagonal pattern could be walking noise

Additional issue
When capturing RAW FITS with StellarMate, the sensor is detected as ~6188×4100 px but files are saved as 6000×4000 px, producing black borders after stacking.
Any insight on INDI cropping or whether CR3/cRAW would be preferable is welcome.

📷 NGC_2238_Light_001_c (Grandi).pngNote on optics
I’m aware that I’m currently imaging without a field flattener, and I’m already evaluating the purchase of one.
I understand the expected consequences on the frame edges (field curvature, star elongation and vignetting), so those issues are known and not the main focus of this post.

Have you tried using the EKOS internal guider for StellarMate? It’s very powerful and will allow you to dither.

Also, I would avoid shooting objects that low in the sky, you’ll be shooting through atmospheric soup. I usually shoot objects that are at least 35 degrees in altitude.

Alright, to answer your questions in order:

Calibration frames are always going to be an important part of your imaging session. When imaging with a DSLR, it’s even more important to have great calibration frames since the sensor isn’t cooled and Canon cameras also tend to have horizontal banding. The general guidance for calibration frames has been to take at least 20 frames each for flats, darks, and biases. I usually shoot 40 of each, but that’s my preference only.

You don’t need a guide scope to use the EKOS internal guider. It uses virtual GPS in order to guide. It should be fairly straight forward to implement. This is a feature I wish NINA had.

Dithering is something you might hear debate on whether you need to or not. Personally, I’ve always had to dither and have never had success without it. Essentially, dithering is just moving your camera a few pixels to randomize your fixed pattern noise (e.g. hot pixels, etc). Dithering doesn’t get rid of noise—it helps to balance out noise so it can be averaged out during stacking. The rain-like noise on one of your example frames exhibits the classic walking noise pattern. Now, this is where the debate comes in since you can avoid dithering by having excellent polar alignment, superior calibration frames, and a high quality cooled camera. You can also do dithering the old fashioned way by moving the mount a little in each direction between frames.

Obviously, they were just test shots and I was aware that shooting low on the horizon would create a mess, and for M42 in the end, it was exactly what I expected, but the bitter surprise was with the Rosette Nebula; I didn't expect such a mess.

I only took 8 darks, do you think that's too few? So flats are important? And to think I hadn't given them much thought.

I don't have a guide telescope, but I still enabled dithering (only for the Rosette Nebula, M42 doesn't have dithering).

What causes it and what is 'walking noise'?

I understand the things I was underestimating. Thank you.

To make flats on the spot, I read that some people use tablets with a white screen with a few sheets of paper in front of them. Is that a good idea? How many should I take?

As for the Rosette Nebula, I had set dithering every two shots, whereas for M42 I hadn't, yet it seems to have been much more disastrous, especially in terms of noise, with the Rosette in particular.

Actually, in NINA for the guider you can select “Direct Guider” and NINA will dither the mount without autoguiding software running. I did that on a regular basis in the early days and it seemed to work.

Arun H · Dec 31, 2025 at 09:16 PM

Tony Gondola:
Actually, in NINA for the guider you can select “Direct Guider” and NINA will dither the mount without autoguiding software running. I did that on a regular basis in the early days and it seemed to work.

yes - that was the exception I was referring to. You use direct guiding in NINA if you are not guiding but using a mount with a sky model and no guiding. I use direct guiding when I am using my Mach2 with a sky model, and in this case, guiding is unnecessary.

Unless the OP explicitly enabled the equivalent of direct guiding, if such a thing is even supported in the software he is using, my guess is that no dithering actually happened.

I believe in EKOS (the OP is using StellarMate) it’s a checkbox option called “non-guide dither pulse” in the guide tab.

Astroberry had the same option since it used EKOS as well, so I would assume StellarMate has this option too.

Let's also take the topic from a different angle. Since I already have a DSLR and had read online that 'it could be done' with astrophotography as well, I went for it, also encouraged by the successes with the simple Seestar S50. But now I realize that maybe a DSLR isn't the best choice. Would all these problems I've encountered with DSLRs be resolved with astronomical cameras like yours (perhaps less expensive for now)? I was looking at something like the ASI585MC (maybe the Air series) or the ASI533... what do you think?

Obviously, I don't think (having now understood the world behind astrophotography) that changing cameras is the solution to all problems, but at least it might help avoid the ones I've encountered so far.

I shoot with a Nikon Z5, a ZWO ASI858MC Pro and a ZWO ASI533MC Pro and a SW SA GTi and a ZWO AM3. Here are my observations:

1. It is no use getting darks or biases from the Z5. Even when I get raw files, the bias and the darks are statistically the same. Since the Z5 does not have a BSI sensor, I am certain the EXPEED system is not giving me raw data. So I have stopped using darks and biases. The first black point correction takes care of the statistical noise floor.

2. There is a faint circular pattern in my red and blue data and a faint vertical line when I use biases/flat darks in calibration. Looks like the capital letter Phi. It is very faint. I am assuming the line is the divide between the 2 registers where the sensor data is being simultaneously read. Not sure what the circular pattern is. Generally, unless I am shooting dust clouds like the Mandel Wilson 3, this is lost in denoising. Also another reason I stick to lights and flats only now with the Z5.

3. If the payload is more than 8-8.5 lbs, no use dithering with the SW SA GTi. You will lose too much time waiting for it to settle. But I can get <1.2” guiding with proper balancing. Look into off axis weights. I use a small rig clamp and 100/200g weights to counter the asymmetric shape of the MILC.

4. If you are using clip on filters and Z system lenses or telescopes without field flatteners/reducers, you will get severe blooming like the rosette pic. The flange distance between the last element of the Z system lens and the clip on filter is <16mm. The filter is not designed to handle such a large light cone. Probably the same issue without a field flattener. A field flattener on a telescope will collimate the light cone to a much tighter angle of incidence, making the filter work better.

5. Do not trust ASIAir or similar to decide flat exposure times with DSLR/MILC. I go by the camera histogram and make sure most of the histogram is between the 25% and 50% lines (ideally 33% peak is best), and the histogram is not touching the y axis. If you have vignetting issues, and need to go past the 50% line to move the histogram off the y axis, make sure not to go past the 75% line in any channel. As many people have said above, proper flat correction is the most important issue with calibration.

6. Again with cooled cameras with BSI sensors (no amp glow), build a library of darks and biases and take flats with ASIAir or similar. Happy hunting. and Happy New Year!

CS,

GD

I did a quick test with the same processing in PixInsight, but using only the lights, without darks and flats. 🫣

📷 integration_autocrop_ABE (Grandi).pngAt this point, the suspects are actually the darks, as he said AstroÅmazer

Actually, now that the image is more acceptable, I've noticed that the brightest stars aren't as well defined as they should be.

No, I haven't activated EKOS's internal guide.

Do you think the internal guiding of EKOS is an advantage or not? What could be the reason that the brightest stars are so blurry?