Saturn & Jupiter last night SkyMax 180Pro

[Chrb1985]

Hi guys. Here are some from last night. Hope someone can help me out. I can't seem to get much color in my final images.

But anyway, I'm very pleased with the results. The Saturn one is with a 1.6x barlow Jupiter is without barlow.

[astroman001]

The planets should look yellow at low altitude due to scattering of blue. How did you correct the colour? Saturn looks blue which is strange. Jupiter is not far off, have you tried boosting saturation in photoshop or Gimp?

Do you have the white balance set to auto on your colour cam, assuming this is taken with a colour cam.

Normally I colour correct in Photoshop. Go to Levels, blue channel and move the RH slider left until the white areas look about right. That's all you should have to do.

Peter

[JeremyS]

Nice result @Chrb1985 - they must be very low down for you

[knobby]

Nice result, the colour is down to personal taste I guess, I use auto colour in PS6 ... this is how your Saturn turns out.

 

Edited August 3, 2020 by knobby

[Chrb1985]

5 hours ago, knobby said:

Nice result, the colour is down to personal taste I guess, I use auto colour in PS6 ... this is how your Saturn turns out.

 

Wow! Thanks so much buddy. I just downloaded PS on saturday

[vlaiv]

3 hours ago, knobby said:

the colour is down to personal taste I guess

Never fully understood this stance.

Color is what it is - take an object and image of that object and look at them and if color is the same - it is proper color rendition, if it is not the same - color is off, simple as that.

Fact that we can't take Jupiter and place it next to our image of Jupiter does not change the fact that Jupiter has certain color and we should be able to capture it properly. Take two images of Jupiter and if you process color properly - they should look the same color wise - have same hue, same saturation, etc ...

We don't even have issue of lighting here - planets are always illuminated with same light source and we know spectrum of that light source.

 

[michael.h.f.wilkinson]

3 minutes ago, vlaiv said:

Never fully understood this stance.

Color is what it is - take an object and image of that object and look at them and if color is the same - it is proper color rendition, if it is not the same - color is off, simple as that.

Fact that we can't take Jupiter and place it next to our image of Jupiter does not change the fact that Jupiter has certain color and we should be able to capture it properly. Take two images of Jupiter and if you process color properly - they should look the same color wise - have same hue, same saturation, etc ...

We don't even have issue of lighting here - planets are always illuminated with same light source and we know spectrum of that light source.

 

Colour perception is highly dependant on whoever is looking at it, and viewing conditions. A good exampe is the apparent "steel blue" colour of sunlight towards totality of an eclipse. Objectively, the colour is still ordinary daylight, but because light is dimming, and normally that means the sunlight turns more yellow/orange, the brain interprets this as a more blue colour of the light. The white balance of the eye when looking at Jupiter might be very different of that when looking at the finished image under daylight conditions. I always perceive Jupiter as pretty yellowish, but often got comments that images where I hadn't used the auto-white balance of Registax were too yellow, even though they matched my perception of the image through the telescope. Using auto white balance turned the images much whiter

Which one is correct? You would have to measure the spectral content through the telescope (with all calibration problems that entails), and compare it to the spectral content of the images, which would depend heavily on the monitor or printing procedure

[knobby]

1 hour ago, vlaiv said:

Never fully understood this stance.

Color is what it is - take an object and image of that object and look at them and if color is the same - it is proper color rendition, if it is not the same - color is off, simple as that.

Fact that we can't take Jupiter and place it next to our image of Jupiter does not change the fact that Jupiter has certain color and we should be able to capture it properly. Take two images of Jupiter and if you process color properly - they should look the same color wise - have same hue, same saturation, etc ...

We don't even have issue of lighting here - planets are always illuminated with same light source and we know spectrum of that light source.

 

I guess I was thinking more along the lines of different monitors / eyes, I'm sure every planetary image starts off fairly washed out till we adjust saturation etc. hence the personal taste bit.

[vlaiv]

54 minutes ago, michael.h.f.wilkinson said:

Colour perception is highly dependant on whoever is looking at it, and viewing conditions. A good exampe is the apparent "steel blue" colour of sunlight towards totality of an eclipse. Objectively, the colour is still ordinary daylight, but because light is dimming, and normally that means the sunlight turns more yellow/orange, the brain interprets this as a more blue colour of the light. The white balance of the eye when looking at Jupiter might be very different of that when looking at the finished image under daylight conditions. I always perceive Jupiter as pretty yellowish, but often got comments that images where I hadn't used the auto-white balance of Registax were too yellow, even though they matched my perception of the image through the telescope. Using auto white balance turned the images much whiter

I agree that color perception will be different under difference circumstances and that is why we have standards - like sRGB.

Images are supposed to be encoded in sRGB standard when viewed at computer monitor because that will render most closely resembling color as if object was viewed in conditions defined by sRGB standard.

These parameters correspond to viewing conditions most people use computer monitor in. Dimly lit room with neutral background (white wall within dimly lit room), etc ...

This is irrespective of perception and it is related to spectrum of recorded light.

You are partially right that one needs to examine spectral content of the light - but luckily our vision is based on tri chromatic system and we don't need a full spectrum in order to reproduce color. We just need to provide three values in particular color space. For CIE XYZ space, here are matching functions:

If you compare that to your camera response, you'll notice that these functions are different. Even if we take CIE RGB space (not to be confused with sRGB space):

Still not the same.

First step would be to create color transform matrix from camera space to some absolute color space like XYZ. After that we have standardized transforms to common color spaces like sRGB.

1 hour ago, michael.h.f.wilkinson said:

Which one is correct? You would have to measure the spectral content through the telescope (with all calibration problems that entails), and compare it to the spectral content of the images, which would depend heavily on the monitor or printing procedure

Depends on what your image represents.

Does it represent an image that observer would see at an eyepiece of a telescope somewhere on the earth (thus looking thru earth's atmosphere) or does it represent image that observer would see when looking thru a telescope in orbit - or floating in space some distance away from Jupiter?

Both cases can be represented in such way that "observer" sitting in their dimly lit living room and looking at properly calibrated computer screen would see the same color as their respective counterpart.

One just needs to apply proper set of transforms for their camera and wanted outcome (in case of telescope on earth - simple camera -> XYZ -> sRGB linear -> sRGB gamma encoded, while in case of observer in outer space camera -> remove atmosphere influence (blue scatter attenuation) -> XYZ -> sRGB linear -> sRGB gamma encoded).

Most people do following to their image:

raw_camera_values -> null transform to sRGB -> color balance using some some automatic algorithm

No proper encoding of gamma in sRGB, treating camera RGB values as if they were sRGB linear values and doing automatic color balance based on algorithm (usual grey world or similar algorithm that just assumes that all RGB components should be equally represented in regular image and that on average world is "grey" - very poor assumption when you have subject like Planet that can have distinct dominating color).

Now we have a huge problem - problem of what people are used to. If you do yellowish Jupiter and explain to people that this is what Jupiter looks like thru the eyepiece - they will say that it "does not look good" because it is different than most Jupiter images that they've seen in their life - all being wrong because all of them are done with same wrong processing.

 

 

[michael.h.f.wilkinson]

3 minutes ago, vlaiv said:

Depends on what your image represents.

 

This is precisely my point. Standards like sRGB don't solve this. Even behind a telescope, it depends heavily on the exact white balance of the eye (very different in the dark winter nights compared to the grey summer nights here in the north). Furthermore, an experienced observer will spot many subtle details a novice might miss, so the perception is different again. As a third variant of my Jupiter shots, what about this one:

[Chrb1985]

5 hours ago, michael.h.f.wilkinson said:

This is precisely my point. Standards like sRGB don't solve this. Even behind a telescope, it depends heavily on the exact white balance of the eye (very different in the dark winter nights compared to the grey summer nights here in the north). Furthermore, an experienced observer will spot many subtle details a novice might miss, so the perception is different again. As a third variant of my Jupiter shots, what about this one:

I liked the third one best .

[Chrb1985]

.

 

 

Edited August 3, 2020 by Chrb1985