Image Quality

[Alanj49]

I have come back to the hobby after a VERY long lay off, and purchased a Starwatcher 2i Pro and a ZWO ASI Air for guiding. On the Starwatcher I am using a Canon EOS 70D with a CLS clip filter. I did a session last night and on the back of the camera the images taken for stacking looked fine, except that there seemed to be a lot less stars showing than there actually were. I did a number of sets of 10 x 60 seconds and one of 10 x 180 seconds. There was a blue cast which I believe comes from the CLS filter, but that did not worry me as it could be processed out, the images both .jpg and .fit have a magenta cast in the top left and more of a greenish cast in the bottom right. The fits are of better quality than the jpg's which are quite small and pixelated.

My questions are, what would be causing the cast? why would the images show less stars than there actually were? and what effect would the CLS filter have on all of this.

I have attached a jpg and a matching fit as an example ( excuse the antenna and antenna wire I was after what was behind them).

Many thanks in advance

Light_60.0s_Bin2_0004_1.fit

[Astro Noodles]

Do you have a powerful source of light nearby?

What ISO setting do have?

[vlaiv]

Hi and welcome to SGL

I'm not sure where do you get that red cast in upper left corner. I loaded fits into imageJ, did quick debayering, scaled channels, converted to composite image (RGB) and got this:

There is a bit of light pollution but otherwise image looks ok.

As far as number of stars - well, you mist focus a bit. When you don't focus properly on stars - their light gets spread out and they are less visible. They really need to be pin point in order to show properly.

Here is non debayered raw data crop - you can see a lot of stars - but since they are little circles instead of dots - they are rather faint.

[Alanj49]

Many thanks for your quick reply, I have no idea either where the cast came from, I thought that the CLS filter would get rid of all of that....I have downloaded ImageJ as that is a new one on me. I thought that I had focussed, but seemingly not well enough. The lens I was using was a Canon 24-70 set to nearer the 24 end which made it a tad difficult to see things on the live view screen. I will get stuck into ImageJ and see if I can do some stacking, will have plenty of time as last night was the last (we are told) clear night for a while. I am quite pleased though that the guiding seems to have worked quite well. 

I have a lot to pick up on again, but will hopefully get there...

Once again many thanks.

[Alanj49]

58 minutes ago, Astro Noodles said:

Do you have a powerful source of light nearby?

What ISO setting do have?

Many thanks fore the reply, the iso was set at 100, and there wasn't a particularly bright area nearby, it was the same on other images taken in a different direction. The CLS filter I would have thought would get rid of that..

[Astro Noodles]

7 minutes ago, Alanj49 said:

Many thanks fore the reply, the iso was set at 100, and there wasn't a particularly bright area nearby, it was the same on other images taken in a different direction. The CLS filter I would have thought would get rid of that..

It's a very low ISO setting. 800 or 1600 would be better. That way you won't have to stretch the image as far.

See this site Home | DSLR Astrophotography (dslr-astrophotography.com) which @Alien 13 just put me onto.

Edited July 23, 2021 by Astro Noodles

[ollypenrice]

Where do colour gradients come from? I have no idea but they come from somewhere. I live at a very dark site and still get them. When I used an OSC CCD camera I also got, quite routinely, a diagonal red-green imbalance, always in that chip orientation, similar to yours. I concluded that it came from the camera, not the sky, since it was always diagonal much as yours is. I don't say I'm certain about this, it was a long time ago, but it makes a strong memory.

In any event, everyone gets colour gradients across images and you need to find a gradient removal tool you can live with. I use Pixinsight's Dynamic Background Extraction but there are others in AstroArt, Astro Pixel Processor and Gradient Xterminator. I wouldn't bust a gut over why gradients appear. They just do and you need to get rid of them.

Olly

Edited July 23, 2021 by ollypenrice
Typo

[vlaiv]

2 hours ago, Alanj49 said:

I have downloaded ImageJ as that is a new one on me.

Give it a go, but it is not as user friendly as some other software as it is primarily for scientific image manipulation.

Check out Fiji - it is distribution loaded with plugins. In principle - you can perform every step of processing with it, but I tend to finish off things in Gimp as it is much easier. You can however do stacking and everything else in ImageJ and even code your own stuff if you know how to program - it accepts macros and you can also use Java to write plugins.

[Alanj49]

2 hours ago, Astro Noodles said:

It's a very low ISO setting. 800 or 1600 would be better. That way you won't have to stretch the image as far.

See this site Home | DSLR Astrophotography (dslr-astrophotography.com) which @Alien 13 just put me onto.

I kept the ISO down to try and reduce the noise, which is always worse in dark areas.  I will have a good look at that site- thank you..

[powerlord]

as someone with 6 months experience, I feel a bit under qualified to say this, but you've got the same '40 years of photography' knowledge i have holding you back.

Don't worry about high ISO - thats what stacking is for. if you can't see it (well.. accurately..didn't record it) no amount of stacking will help.

kick that ISO up.

https://dslr-astrophotography.com/iso-values-canon-cameras/

suggests best iso is 1600 for your 70d.

that, and get a bahtinov mask for a tenner for focus and you'll be cooking on gas

[vlaiv]

1 hour ago, Alanj49 said:

I kept the ISO down to try and reduce the noise, which is always worse in dark areas.  I will have a good look at that site- thank you..

Noise by itself is not really that important - it is signal to noise ratio that is important.

ISO is just multiplier and as such does not impact SNR at all (it does a bit but in different ways). If you multiply ratio of two numbers with same coefficient - well, that ratio does not change 5:1 is the same as 5*2.23 : 1*2.23 (you can simply cancel out 2.23 from ratio).

What ISO does for astrophotography is two fold:

1. It affects full well capacity and your ability to record bright stuff like star cores properly.

This does not mean you should use low ISO because of this - as there is another better way to deal with that - you just take a few short exposures that you use to replace over exposed parts in long exposure image (just scale data properly before blending).

2. It affects read noise - high ISO means lower read noise.

Again - read noise is not something that should overly concern you if you know how to handle it. It is only important as long as it is largest source of noise - but with DSLRs in light polluted areas - it very quickly becomes unimportant. In any case - read noise should be determining factor for your exposure length - you expose until you swamp read noise with other noise source (like thermal or LP noise - both of which grow with time while read noise constant per exposure).

All this means that ISO is really not that important - go with some nice middle value and if you want higher ISO - you can expose for shorter and if you want lower ISO - expose for longer so that read noise is swamped by LP noise for example. Simple as that

[Alanj49]

11 hours ago, powerlord said:

as someone with 6 months experience, I feel a bit under qualified to say this, but you've got the same '40 years of photography' knowledge i have holding you back.

Don't worry about high ISO - thats what stacking is for. if you can't see it (well.. accurately..didn't record it) no amount of stacking will help.

kick that ISO up.

https://dslr-astrophotography.com/iso-values-canon-cameras/

suggests best iso is 1600 for your 70d.

that, and get a bahtinov mask for a tenner for focus and you'll be cooking on gas

You are right, I have 54 years of photography behind me, where you are always told to keep the ISO low. I saw from the link that the prime OSO is 1600, so I will give that a go. Another reason I was using a low ISO was to try and keep the histogram as much as possible to the left due to the filtration of the CLS filter, through which it also made focussing a bit more difficult, and I didn't really want to start taking it out and back into the camera body in the dark.

[Alanj49]

11 hours ago, vlaiv said:

Noise by itself is not really that important - it is signal to noise ratio that is important.

ISO is just multiplier and as such does not impact SNR at all (it does a bit but in different ways). If you multiply ratio of two numbers with same coefficient - well, that ratio does not change 5:1 is the same as 5*2.23 : 1*2.23 (you can simply cancel out 2.23 from ratio).

What ISO does for astrophotography is two fold:

1. It affects full well capacity and your ability to record bright stuff like star cores properly.

This does not mean you should use low ISO because of this - as there is another better way to deal with that - you just take a few short exposures that you use to replace over exposed parts in long exposure image (just scale data properly before blending).

2. It affects read noise - high ISO means lower read noise.

Again - read noise is not something that should overly concern you if you know how to handle it. It is only important as long as it is largest source of noise - but with DSLRs in light polluted areas - it very quickly becomes unimportant. In any case - read noise should be determining factor for your exposure length - you expose until you swamp read noise with other noise source (like thermal or LP noise - both of which grow with time while read noise constant per exposure).

All this means that ISO is really not that important - go with some nice middle value and if you want higher ISO - you can expose for shorter and if you want lower ISO - expose for longer so that read noise is swamped by LP noise for example. Simple as that

K read a lot of this in the link that was provided above, a lot of years in photography where you are 'trained' to keep the ISO is a hard habit to break. I have set the camera to ISO 1600, and I will leave it here and work round it and use the variable part of the exposure triangle - shutter speed and take multiple exposres.

[rickwayne]

In terrestrial photography we don't have to stretch as we do in astro, so you don't notice the phenomenon. But if you shoot two images at the same shutter speed and aperture, one at low ISO and one at high, and then pull the exposure and contrast up on the low-ISO one to yield the same histogram as the high one's,  you'll find pretty identical noise. In terrestrial work you'd just open the iris or extend the shutter speed or add light to the scene, and say that the low-ISO image was just underexposed.

So was the high-ISO one -- it's just that the camera's electronics did the stretching for you. In either case, you'll wind up with a noisier picture.

In astro, of course, we're nearly always working at the maximum useful aperture, and we can't add light to the scene. So we extend the "shutter speed" by increasing the integration time, and turn up the gain by increasing the ISO and by stretching.