Iris Core-25.4 Hours

[Rodd]

I am trying to understand just what direction to take with this image.  Judging from the response to other Irises posted recently, there must be something off about this image (all images have something off--except perhaps those by the greats on this forum who will remain nameless because naming them is unnecessary).  I refer to something being way off--or farther off than other images of the same target.  Yes the dust is a bit noisy (but it is not slathered with noise control), and yes the brightness of the dust is a bit dark (but at least it is not over stretched).  This is as far as I could stretch this data.  I think the dust and stars are passable--but its the core that is the main showcase.  I have studied the Hubble image, and images taken with large instruments, and all the general details and colors are there-perhaps not as sharp in some cases.  But there are no gaping holes.  What more can be expected from a 5" scope?  And yet folks may not agree.  I would be very interested in learning why.  Perhaps then, I will be able to reprocess this data, and collect more if needed, and put it to rest so that I may rest. 

TOA 130 with ASI 1600.  LRGB 25.4 hours of data

 

[PhotoGav]

I know you like to be hyper-critical with yourself and you haven't let us down this time Rodd! You are beating yourself up over a really lovely image. However.........

I kind of know what you mean about there being something that doesn’t quite sit with this image. I am being hyper-critical here, so humour me, to me there appears to be a disconnect between the centre of the image and the outside areas. I like the outer parts - nice subtle processing, delicate colours, not overstretched, not over noise reduced, in fact a nice bit of natural noise, lovely stars... all looking good. It’s the core, ‘the main showcase’, that I struggle a little bit with. It just looks a bit over processed, like it’s just trying too hard to say ‘look at me!’, it’s overdone the lippy before the big night out with the girls and looks like a tart!! Perhaps if the same delicate touch that you have used with the outer areas were applied to the core it would all sit more comfortably together?

On the other hand - what’s wrong with the occasional interlude with a tart?!!?

[Rodd]

11 minutes ago, PhotoGav said:

On the other hand - what’s wrong with the occasional interlude with a tart?!!?

Nothing....provided....(a long list of important considerations!)

Regarding the image--that is a take I was not expecting.   I will consider it over time.  But that, as you say is pretty hyper technical.  Certainly not a "way off" thing.  In fact, if I had to choose between this core, and a core with a white, burned out center such that the central star is not visible--I would choose this one. 

I will no doubt reprocess this data after collecting more lum and I will keep your observation in mind

Rodd

[coatesg]

Likewise, I think this is a good rendition. I'd agree on the dust to some extent - it's perhaps not so much dark but maybe a touch red to my eyes? 

The core itself looks quite blue, but the pink highlights are present, so it really can't be completely off - it maybe is a little too contrasty for some, and it might just sit better in a slightly brighter dust surroundings? 

Still, hyper critical - it's a good image!!

[Rodd]

54 minutes ago, coatesg said:

Likewise, I think this is a good rendition. I'd agree on the dust to some extent - it's perhaps not so much dark but maybe a touch red to my eyes? 

The core itself looks quite blue, but the pink highlights are present, so it really can't be completely off - it maybe is a little too contrasty for some, and it might just sit better in a slightly brighter dust surroundings? 

Still, hyper critical - it's a good image!!

Thanks Graeme.  There will always be variations is preference.  I tried reducing the red in the dust and it turned a flaccid green.  That was the color without small adjustments to the palette.  If you look at deep images of the region that are not over stretched, the dust tends to be reddish-much like the dust in the Perseus region around NGC 1333.  I guess the goal will be to make it more brown than red.

Rodd

Edited October 10, 2019 by Rodd

[coatesg]

I think that brown is what people are generally used to, it's almost a conditioning to having seen other renditions. 

Colour is such a tricky thing (especially when there's no true "right answer"). How did you colour balance the image out of interest - it it using a tool, or "by eye"?

[coatesg]

Also, what's the split of data across the filters?

[Rodd]

13 minutes ago, coatesg said:

I think that brown is what people are generally used to, it's almost a conditioning to having seen other renditions. 

Colour is such a tricky thing (especially when there's no true "right answer"). How did you colour balance the image out of interest - it it using a tool, or "by eye"?

I balance using a histogram.  When the histogram is equal in all colors--the dust looks unhealthy.  Clay like.  So I added a slight amount of red and reduced the stretch.  Any brighter and the noise becomes apparent.

10 minutes ago, coatesg said:

Also, what's the split of data across the filters?

All filters a have enough data to achieve a good signal, but the blue channel has less data.  I posted an earlier version of this image on the forum and the consensus was I had enough blue, so I should concentrate of Lum.  I think I have 3 hours of Lum but 7.5 hours of Red and Green.  Adding more blue will reduce noise in the blue channel, but will it do much for signal.? I collect lots of data mostly to suppress noise.    I plan on collecting another 6-7 hours of lum--maybe more, it depends on the weather.  I could easilly collect another 5 hours of blue first.  But then we would be at 40+ hours.  25 should be more than enough.

Rodd

[Rodd]

1 hour ago, PhotoGav said:

Perhaps if the same delicate touch that you have used with the outer areas were applied to the core it would all sit more comfortably together?

 

1 hour ago, coatesg said:

The core itself looks quite blue, but the pink highlights are present, so it really can't be completely off - it maybe is a little too contrasty for some, and it might just sit better in a slightly brighter dust surroundings? 

Perhaps true....but being the hyper (or even semi hyper) criticalities they are...I do not think they are the reason the image was not well received. 

Rodd

[ollypenrice]

I think the core is tremendous. If some aspects of it are slightly unfamiliar I think that's because they are better than we usually see. The progenitor star is well controlled and the very tricky pinks are clear as day. For me the core is right on the money. If the image has a problem (I stress 'if') it's that noise levels in the brown dust (which I think is just the right colour) are far higher than in the core. For this reason it hasn't been possible to push the brown contrasts as high as would match the core, either.  Rodd is not shooting from a dark site so this is always going to be a problem.

Sure, more exposure beat noise reduction but, when I run into this problem in post processing, I take a look at the noisy area in Ps at close to pixel scale and see what I can do about it. I don't like proprietary NR routines because they use pixel to pixel communication (blurring) which gives the slimy look. I prefer to work on groups of similarly errant pixels in isolation from their neighbours. I'm not above over-smoothing and then adding Gaussian noise in Ps either. I'd rather not but it's been known...

I suspect that really crafty custom NR in the dust could bring it, aesthetically, into line with that splendid core and make the image harmonious with itself from corner to corner.

Olly

[Rodd]

10 minutes ago, ollypenrice said:

I think the core is tremendous. If some aspects of it are slightly unfamiliar I think that's because they are better than we usually see. The progenitor star is well controlled and the very tricky pinks are clear as day. For me the core is right on the money. If the image has a problem (I stress 'if') it's that noise levels in the brown dust (which I think is just the right colour) are far higher than in the core. For this reason it hasn't been possible to push the brown contrasts as high as would match the core, either.  Rodd is not shooting from a dark site so this is always going to be a problem.

Sure, more exposure beat noise reduction but, when I run into this problem in post processing, I take a look at the noisy area in Ps at close to pixel scale and see what I can do about it. I don't like proprietary NR routines because they use pixel to pixel communication (blurring) which gives the slimy look. I prefer to work on groups of similarly errant pixels in isolation from their neighbours. I'm not above over-smoothing and then adding Gaussian noise in Ps either. I'd rather not but it's been known...

I suspect that really crafty custom NR in the dust could bring it, aesthetically, into line with that splendid core and make the image harmonious with itself from corner to corner.

Olly

Thanks Olly--Interesting information about noise control.  I don't know enough about NC or PI to know how TGV Denoise operates.   Meanwhile, more exposure might just be a viable solution (if, as you intone, a solution is needed.  Since I do not want to start another image before I switch OTAs, and the TOA 130 is still set up--it might be a "why not" decision--when else will I have the opportunity to get 40 hours of data on the Iris?).  One thing's for sure.  More exposure will be a lot more fun than fiddling with noise control algorithms! 😎

Rodd

[gorann]

Splendid image Rodd! I agree with Olly that the core is superb and if there is a problem it is the noise in the dust. This is what to be expected when you have a bright core (so more signal and less noise) and outer areas that has to be brought out with stretching, so more noise. I think I would be very happy with the image as it is, but the dust could be a good exercise for some advanced NR experiments (or 50 more hours of data that would maybe help the dust but not the core).

Edited October 10, 2019 by gorann

[gorann]

PS. Or just bring down the brightness of the dust to match the data until you can drive out to a dark place to get some better dust data. But as I said - lovely image that I would be pleased with as it is.

[Rodd]

11 minutes ago, gorann said:

Or just bring down the brightness of the dust to match the data until you can drive out to a dark place to get some better dust data.

Thanks Goran.  The problem is to bring the dust down to a level that completely eliminates the problem renders it too dark for casual viewing.  I have kept the brightness of the dust about as low as possible and still have a viable image (darkness was one of the CCs). 

I am happy with the image...but unless one sets the bar higher, one is less likely to attain such heights.   I think I will collect another solid night of lum (6-7 hours providing no snags).  I do not want to start another image before I switch to the FSQ to see if the collimation issue was fixed (man am I hopeful), so its either collect lum or switch optics.  I fear that if I don't add more lum, I will forever kick myself for not doing it.  (I may balance out teh blue too (that has been a small pebble in my shoe ever since I decided I had enough).

Rodd

[Xiga]

Rodd, your images are always top-notch, and this one is no different. You have good equipment, and you certainly know your way around P.I (from seeing your previous images). Olly has hit the nail on the head in his comment above, I think you're just under-estimating the importance of truly dark skies, especially so when it comes to broadband imaging, and doubly-so when it comes to imaging dark dusty regions.

I'm still just a lowly DSLR imager myself, but when i first started out doing AP i actually jumped straight into Narrowband work, and i loved that i was able to get good results almost instantly. It was a while before i actually even tried some broadband imaging, and when i finally did, it was only then that i realised just how much harder it is compared to NB imaging (at least in my opinion). Narrowband hides virtually all the sins of LP, but broadband imaging really shows up the quality of your skies, and no amount of processing can fix that. Yes, gradient reduction (whatever your chosen flavour) can indeed work wonders, but you have to remember it will Always result in a noisier image. But living in an LP area, we have no choice but to use these algorithms. For those lucky enough to image from dark skies, they still benefit from using these routines, but the negative impact is far less. Out of interest, what does the image look like before and after you apply DBE? (assuming that's what you've ran).

Where you live, do you have the option of travelling to darker skies? If you could, even if only for a short time, i think you would be amazed at the quality of data you would get in such a short amount of time. If travelling is not an option at all, then i'm afraid you will probably just have to accept the upper limits of what you can achieve with the skies available to you, and you may already be pretty close to that. Alternatively, just use more NR and don't let it get to you! lol.

ps - As it is, it's already a great image. I actually really, really like the core just the way it is, although i wouldn't push it any further than this. I do agree though, that it is somewhat out of balance with the outer dusty regions, but the only viable solutions to this are either applying more NR, or getting to darker skies.

Edited October 11, 2019 by Xiga

[ollypenrice]

11 hours ago, Rodd said:

Thanks Goran.  The problem is to bring the dust down to a level that completely eliminates the problem renders it too dark for casual viewing.  I have kept the brightness of the dust about as low as possible and still have a viable image (darkness was one of the CCs). 

I am happy with the image...but unless one sets the bar higher, one is less likely to attain such heights.   I think I will collect another solid night of lum (6-7 hours providing no snags).  I do not want to start another image before I switch to the FSQ to see if the collimation issue was fixed (man am I hopeful), so its either collect lum or switch optics.  I fear that if I don't add more lum, I will forever kick myself for not doing it.  (I may balance out teh blue too (that has been a small pebble in my shoe ever since I decided I had enough).

Rodd

I think lum is a good idea because you can harmlessly use NR on the colour beneath it.

Olly

[coatesg]

7 minutes ago, ollypenrice said:

I think lum is a good idea because you can harmlessly use NR on the colour beneath it.

Agree - I think this will definitely help drag the dust up. TGV in PI is very powerful - it takes a lot of playing around to get the right levels (do it on a preview first - it's much quicker) but it's surprisingly good, esp on the Lum noise - it can denoise the Luminance and colour independently as well.

[ollypenrice]

19 minutes ago, coatesg said:

 it can denoise the Luminance and colour independently as well.

This is vital. I work in Ps and use entirely different NR techniques for L and RGB. It's all very easy in Ps, too. You simply create two layers, apply slightly too much NR to the bottom one, then use the colour select tool on the top layer and erase as much or as little of the noise as you like while looking at the result in real time. You can also select and erase specific kinds of errant individual pixel from the top layer, revealing their denoised counterparts individually. This reduces the effects of pixel to pixel interaction.

Olly

Edited October 11, 2019 by ollypenrice

[astro mick]

Hi Rodd.

Not qualified to add any advice,but a lovely image all the same.

Mick.

[ApophisAstros]

Oh Yes, super,

Roger

[PhotoGav]

3 hours ago, ollypenrice said:

This is vital. I work in Ps and use entirely different NR techniques for L and RGB. It's all very easy in Ps, too. You simply create two layers, apply slightly too much NR to the bottom one, then use the colour select tool on the top layer and erase as much or as little of the noise as you like while looking at the result in real time. You can also select and erase specific kinds of errant individual pixel from the top layer, revealing their denoised counterparts individually. This reduces the effects of pixel to pixel interaction.

Olly

Olly, I'm fascinated by your input to this thread and the techniques that you mention. What a shame that you don't have a YouTube channel with tutorial videos...! I'm inclined to agree with your suggestion that it is the outer bits that aren't 'in sync' with the core and I would love to see what you would do to recitfy the 'issue'.

Just to reiterate, Rodd, this is a lovely image that has stimulated some really good discussion of processing techniques in the pursuit of perfection.

[Rodd]

9 hours ago, Xiga said:

Where you live, do you have the option of travelling to darker skies? If you could, even if only for a short time, i think you would be amazed at the quality of data you would get in such a short amount of time. If travelling is not an option at all, then i'm afraid you will probably just have to accept the upper limits of what you can achieve with the skies available to you, and you may already be pretty close to that. Alternatively, just use more NR and don't let it get to you! lol.

Thanks Xiga.  I totally agree with you about narrowband--I love it.  It took me a long time before I could do broad band.  Now I try to shoot broadband when the Moon is down and Narrowband when the Moon is up.  The problem is, I find myself starting a lot of images because Ha is so rewarding to shoot and when the Moon is up--what else is there to do.  This makes it hard to switch scopes when you want to. 

There is a green zone about an hour away.  And I can go to a bed and breakfast in the Adirondack Mountains about 4 hours away.  the sky in that park is Bortle 1 or 2.  But I long ago stopped breaking my rig down.  When I started I did not know how to image from one night to the next--I was using nebulosity and I just could not get the subs to align.  Now it would not be a problem--but I would still lose imaging time.  I guess I have gotten used to a permanent (semi-permanent in my case) setup.  I do plan on spending a week at the Adirondack Park--but the problem is its really difficult to plan due to the fickle weather.  The place is not an astronomy oriented place, so gI would have to see if the management will allow floating days, or something.

Rodd

Edited October 11, 2019 by Rodd

[Rodd]

4 hours ago, coatesg said:

Agree - I think this will definitely help drag the dust up. TGV in PI is very powerful - it takes a lot of playing around to get the right levels (do it on a preview first - it's much quicker) but it's surprisingly good, esp on the Lum noise - it can denoise the Luminance and colour independently as well.

yeah I am pretty familiar with it--but still learning the really delicate adjustments.  There is a tool called statistics that you are supposed to use (according to the Light Vortex tutorial).  The tool gives you teh standard deviation in exponential format and you put that number in Edge Control (second slider).  But so far it has given me values to the -2 power, which is very aggressive (default is -3).  So so far I have not been able to make it work the way the tutorial suggests.

Rodd

[Rodd]

4 hours ago, ollypenrice said:

This is vital. I work in Ps and use entirely different NR techniques for L and RGB. It's all very easy in Ps, too. You simply create two layers, apply slightly too much NR to the bottom one, then use the colour select tool on the top layer and erase as much or as little of the noise as you like while looking at the result in real time. You can also select and erase specific kinds of errant individual pixel from the top layer, revealing their denoised counterparts individually. This reduces the effects of pixel to pixel interaction.

Olly

This sounds very nice-much easier to control than using TGV Denoise.  Noise control is a funny thing.  When you get it right, its amazing--can't tell any noise control was applied.  Get it wrong, and you experience fatial expressions you didn't know you had.  I think there is a practical limit.  beyond a certain point, a certain amount of noise, noise control will not succeed, no matter how its applied.  Lately, for me noise control is like mercaptan--the odorous gas they put in propane so you can detect it.  It has a distinct, unpleasant smell that can be detected in parts per billion.  When I see noise control, its like getting a big whiff of that smell.

Rodd

[Rodd]

59 minutes ago, astro mick said:

Hi Rodd.

Not qualified to add any advice,but a lovely image all the same.

Mick.

Thanks Mick--don't sell yourself short.  Many docents at museums can't paint or draw if their life depended on it--but they make excellent reviewers.  Just because someone can't do a better job at whatever they are critiquing does not invalidate their critique.   Music is a good example.  Someone might not be able to play an instrument, but they can tell when a musician makes a mistake.

Rodd