The "No EQ" DSO Challenge!

[Filroden]

It's a tough call. I've waited until now to comment as I wanted to see the colour on different screens. I really love the graduation between the lighter core and the darker edges, and this shows really well in your first image. The Rosette does have a lot of OIII, so it should not be just red, so I think the shading is valid (the green/blue of the OIII would soften the strong Ha into being more pink/magenta). After that, it's really just to taste.

It's a good composition, nicely cropped though I think you've missed a stacking artifact in the lower right corner which probably just needs a gradient adding in Lightroom to balance it (rather than crop it out). It's a sharp image without looking sharpened  Your stars are nice too. 

[rotatux]

I'd like to see what Regim B-V color calibration can give on that one. Also, Isn't there a similar feature in Pixinsight ?

If you are willing to try it youself, it's free and only requires a JRE (8 for >=3.4, or 7 for <=3.3.1), but I admit it's yet another software to learn to use; And, better be used to the "star finder" of Regim to set its parameters... Or, would you share the unstretched tiff for us to try ? (trying to develop someone else' image seems an interesting exercise )

[Filroden]

13 minutes ago, rotatux said:

I'd like to see what Regim B-V color calibration can give on that one. Also, Isn't there a similar feature in Pixinsight ?

Yes, the ColourCalibration module balances the background (a user defined area) based on the star colours (either the whole image or another user defined area if there is a lot of nebulosity). I think there is an assumption over the total integrated star colour and it balances to that. However, given stars are not always randomly distributed, I don't find this any more a valid calibration than balancing the histrogram in Photoshop, using the colour temperature slider in Lightroom or any other method.

[The Admiral]

1 hour ago, Filroden said:

It's a tough call. I've waited until now to comment as I wanted to see the colour on different screens. I really love the graduation between the lighter core and the darker edges, and this shows really well in your first image. The Rosette does have a lot of OIII, so it should not be just red, so I think the shading is valid (the green/blue of the OIII would soften the strong Ha into being more pink/magenta). After that, it's really just to taste.

It's a good composition, nicely cropped though I think you've missed a stacking artifact in the lower right corner which probably just needs a gradient adding in Lightroom to balance it (rather than crop it out). It's a sharp image without looking sharpened  Your stars are nice too. 

Thanks Ken. Just to be clear, when you say "..in your first image" are you referring to the one I posted yesterday, or to the first of the two images I posted today? I'm coming round to preferring the first one I posted today, at least at the moment

I'm not convinced about a stacking artefact in the lower right, I can't say it's obvious to me. Pixel values in the 'black' areas in the lower right appear to be very similar to others in the image, at least within the variation. Also, in say this example image, there is a trace of nebula in that region.

20 minutes ago, rotatux said:

I'd like to see what Regim B-V color calibration can give on that one. Also, Isn't there a similar feature in Pixinsight ?

I don't use either of those but I suspect that there is a feature in AstroArt. I've already trialled PI so wouldn't be able to try that again.

Ian

[The Admiral]

5 minutes ago, Filroden said:

Yes, the ColourCalibration module balances the background (a user defined area) based on the star colours (either the whole image or another user defined area if there is a lot of nebulosity). I think there is an assumption over the total integrated star colour and it balances to that. However, given stars are not always randomly distributed, I don't find this any more a valid calibration than balancing the histrogram in Photoshop, using the colour temperature slider in Lightroom or any other method.

I was thinking something along those lines too as I was replying, but not in such a clear way! I don't think that there is any way of precisely colour balancing, it's not as though one can introduce a grey card! There would have to be some assumption made about what constitutes a neutral reference.

Ian

[Filroden]

4 minutes ago, The Admiral said:

Just to be clear, when you say "..in your first image" are you referring to the one I posted yesterday, or to the first of the two images I posted today?

The one you posted yesterday, though of the two today, I prefer the first of those too.

It was this area (I've cropped and boosted the contrast/brightness). It just looks too straight and suggests a stacking artifact rather than the edge of the nebula.

[Filroden]

3 minutes ago, The Admiral said:

I was thinking something along those lines too as I was replying, but not in such a clear way! I don't think that there is any way of precisely colour balancing, it's not as though one can introduce a grey card! There would have to be some assumption made about what constitutes a neutral reference.

Ian

I think it's a reasonable assumption to make in general. We know the total distribution of stars in the galaxy (and in others), so the total integration of this distribution should define the correct colour given enough stars. However, at large scale, there are uneven distribution of stars (clusters spring to mind or whether you are imaging towards the core or an arm of the galaxy or into the halo) where one colour may dominate. So applied with caution it's my preferred method of balancing the colour (after background removal) but I wouldn't hold it up as being accurate. If the image was plate solved and each star's colour positively identified from a database taken from spectroscopic measuresments and this value was used to balance the colour, then I'd think it's closer to a true colour.

[Filroden]

5 minutes ago, Filroden said:

It just looks too straight and suggests a stacking artifact rather than the edge of the nebula.

The more I look at it the more I think you could be right. It's not quite straight (at 1:1 scale it looks very straight but zoomed in it does show some variation). I think its position in the corner has thrown me.

[rotatux]

21 minutes ago, The Admiral said:

I don't think that there is any way of precisely colour balancing, it's not as though one can introduce a grey card! There would have to be some assumption made about what constitutes a neutral reference.

Actually Regim find your stars and does a plate solve with them. Then it loads reference colors from an Internet catalog for the matching stars, and computes a color balance adjustement from the reference and your image. So I believe it's quite physically correct and precise, but you need an image with sufficiently precise star colors and gradients (i.e. not saturated too much) as input.

Edit: And as you see it's not based on your image-only statistics of stars or background distribution. Oops, I'm turning into Regim advocate...

Edited January 23, 2017 by rotatux

[The Admiral]

4 minutes ago, Filroden said:

The more I look at it the more I think you could be right. It's not quite straight (at 1:1 scale it looks very straight but zoomed in it does show some variation). I think its position in the corner has thrown me.

No, I think you are right, it's the sort of artefact I might expect, but I wouldn't normally push things that far to check! It's obvious in that crop, too linear to be anything else I think. I just kept cropping away until the odd colours in the corners disappeared after stretching. I can just about make something out in the original now that I know what I'm looking for.

Ian

[The Admiral]

6 minutes ago, rotatux said:

Actually Regim find your stars and does a plate solve with them. Then it loads reference colors from an Internet catalog for the matching stars, and computes a color balance adjustement from the reference and your image. So I believe it's quite physically correct and precise, but you need an image with sufficiently precise star colors and gradients (i.e. not saturated too much) as input.

Interesting Fabien, neat. But I think that most of my stars are saturated though.

Ian

[Filroden]

3 minutes ago, The Admiral said:

But I think that most of my stars are saturated though.

In Photoshop, if you have Noel's actions, I think it can bring the colour from the star's halo (which shouldn't be saturated) into the centre. It's pretty neat and I think I found a way to do it in PI but I've not been able to replicate it again!

[Filroden]

11 minutes ago, The Admiral said:

I can just about make something out in the original now that I know what I'm looking for.

It helps that my main monitor is a 40" 4k TV though sometimes I think images have more contrast and richer colours on the iPad.

[The Admiral]

18 minutes ago, Filroden said:

In Photoshop, if you have Noel's actions, I think it can bring the colour from the star's halo (which shouldn't be saturated) into the centre. It's pretty neat and I think I found a way to do it in PI but I've not been able to replicate it again!

I don't have Photoshop (but I wonder if the plug-in works with Lightroom). Then again, I think ST has one option for that, would need to check out the manual!

17 minutes ago, Filroden said:

It helps that my main monitor is a 40" 4k TV though sometimes I think images have more contrast and richer colours on the iPad.

Yeah, well!! Is it colour calibrated?

Ian

[Filroden]

1 hour ago, The Admiral said:

Yeah, well!! Is it colour calibrated?

Err...maybe...quickly finding my Spyder after the house move and setting it running!

[happy-kat]

Ipad viewing images I find always better, brighter, sharper just better it's an iOs thing make things look better. Auto enhances.

[Filroden]

I remember calibrating my Mac and laptop but I hadn't calibrated the new TV. However, having run it now, it calibrates and now has an overly warm hue, making whites appears a pale red. When I bring up a colour chart, I have to say the uncalibrated looks much better than the calibrated.

[The Admiral]

What's your target colour temperature? I set my monitor to 6500K, and I assume you can do the same with TVs, or do you use native. TVs generally are much brighter than you'd use as a monitor, so you will see all the stuff going on in the shadows I guess.

Ian

[Nigel G]

Photoshop CS2 is a free download.

Ian I looked on my phone while at work today and see a different shade of colours the difference between Tablets PC's and phones is quite a bit, especially with contrast.

I think I prefer the first one of the 2 posted today,

Colour is a personal preference most of the time, especially with nebula. I'm finding it fairly difficult to tame reds from the modified camera, by the time I have toned down the red, my stars are sometimes very blue or all white, I often spent a fair bit of time in colour module with ST or in PS. Having to bring the saturation down to 50% is common.

Nige.

 

 

 

[Filroden]

My first Ha sub. A single sub of 60s at 300 gain, 50 offset and -20C. No calibration, just a quick levels and curves stretch and no noise reduction. Going to go lie down now...

Actually, I've set my first sequence to be 30 x 60s Ha, 10 x 30s RGB and 30 x 30s L. Hopefully I can repeat that sequence a few times to add to my original data.

[Filroden]

Well, that didn't last long. The target is already too close to my neighbour's heating vent and there is no other easy target to its east. So I've moved to M81/82 to see what I can get. Just doing LRGB as the targets are small in my fov and any Ha data will be lost.

Edited January 23, 2017 by Filroden

[Stub Mandrel]

Just a thought, I use gradient exterminator in Photoshop as a 'first operation' it usually does a good job of balancing overall colour as well as removing gradients - certainly better than auto colour.

[The Admiral]

3 hours ago, Filroden said:

My first Ha sub. A single sub of 60s at 300 gain, 50 offset and -20C. No calibration, just a quick levels and curves stretch and no noise reduction. Going to go lie down now...

Actually, I've set my first sequence to be 30 x 60s Ha, 10 x 30s RGB and 30 x 30s L. Hopefully I can repeat that sequence a few times to add to my original data.

That's looking very promising Ken, plenty of detail in just a single sub., and you have the benefit of a FoV which gives good coverage. Looking forward to see the finished image.

By the way, what version of Skysafari are you using, and does it use the iCloud to provide proper operation? There doesn't seem to be a decent comparison of versions.

Ian

[Filroden]

Just now, The Admiral said:

That's looking very promising Ken, plenty of detail in just a single sub., and you have the benefit of a FoV which gives good coverage. Looking forward to see the finished image.

By the way, what version of Skysafari are you using, and does it use the iCloud to provide proper operation? There doesn't seem to be a decent comparison of versions.

Ian

The developer website shows the comparison, mainly differences in object databases. http://skysafariastronomy.com/

I went with the Pro but bought it when they do there 50% discounts, which are quite frequent (maybe once every month or two). You can sync your settings across iDevices but otherwise there is no need to use the cloud.

[The Admiral]

5 hours ago, Nigel G said:

Photoshop CS2 is a free download.

Ian I looked on my phone while at work today and see a different shade of colours the difference between Tablets PC's and phones is quite a bit, especially with contrast.

I think I prefer the first one of the 2 posted 

Nige.

Thanks Nige. The variation between screens is a perennial problem when you want to put your conventional artwork onto the web, as you never know what the observer is seeing. One goes to great lengths to use a colour managed workflow, but once it's out in digi-space all control is lost.

Ian

Edited January 23, 2017 by The Admiral