M16 Pillars of Creation in mono

[swag72]

So if one uses PS to do astrophotography one can't produce an image that the PI people (and those imagers that regurgitate ad nauseam the same stuff) consider worthy ........... Interesting.

[Scott]

photoshop may well have been designed for terrestial image manipulation but remember, there are many astro specific add-ons/plug-ins that have been created soley for AP. I don't have PI, but I wish I did. It won't replace PS but its got a few steps that I'd like to add to my processing arsenal 

[Scott]

3 hours ago, swag72 said:

If you hate processing... perhaps this hobby is not for you.... I say this in all sincerity based on the fact that I will spend in excess of perhaps 8 to 10 hours on processing ONE image. One has to enjoy the processing side in my opinion as sadly there's not a one button to press to get the perfect image.

Then there's the bit about manipulating data and imaging what is there .......... how do you feel about Narrowband imaging? As the colours you see are most definitely NOT there and are manipulated in one form or another. If you look at 20 images of M16 for example, you will see 20 different images, none will be the same as there just isn't that level of objectivity in imaging...... it's all subjective in my opinion. 

 

Its not the processing I hate... It's my inabilities that test my patience

[ChrisLX200]

I have PixInsight and I use it for some things but there are so many other features I neither understand nor want TBH. Like most new users of PixInsight I scoured the web for tutorials (because there is no manual for it - not even the individual functions are well documented and what is really missing is how all these tools work together as a whole). Anyway, I ended up following recipes of those that knew better than me, and for the example given it sort of worked OK. However, moving on to my own data and it just fell apart   You cannot just slavishly follow someone else's workflow because your own data is unique.

ChrisH

[Rodd]

10 minutes ago, johnrt said:

Crivens! If I'd read that utter rubbish before I bought Pixinsight, I wouldn't have touched it with a bargepole!! And for what it's worth I'm yet to find a single magic recipe in any version of photoshop.

 

14 minutes ago, ollypenrice said:

Steaming tosh! Utter steaming tosh. And both pretentious and insulting.

Oooooh, that feels better. I've been wanting to say that for a long time. . Here's why:

PI uses masks rather than layers. The hubris of the (autistic?) PI mentality is that it can somehow derive from the data everything that is needed to manipulate the data into 'correctness.' They say it is scientific. The mask, or the algorithm, will be perfect. 

Will it heck. PI is a graphics programme. In DBE the user defines what should be neutral background. Well that's scientific!!!  Same as Background Neutralization. Equally scientific.  Tell it your subjective impressions and it will scientifically confirm them. Which is just what you do in Photoshop. The mask parameters are these. Who made them? I did.

Scientific?  Get outta here!

I'm not knocking PI and I use it myself, but I'll have none of this 'scientific' claptrap. If you don't want to get involved, personally, with your images and do your best to shape them into what you believe to be true or informative then you are in the wrong job. Don't imagine that some kind of spurious 'objectivity' in PI will lead you to the truth. It won't.

Our data is imperfect. It's damned good but it's imperfect. How do we know? We look at it! Our eyes are remarkably sophisticated. ANd we do use them to look at the finished picture, remember...

Olly

 

 

 

 

So the idea of "staying true to the data" is hogwash?  That is a PI mantra.  If I was convinced I would have an easier go at it in PS, I would jump ship--but so far, no software I have tried (AIP4Win, PI, Nebulosity, CCDops) seems to work toward that end.  Calibration, alignment, stacking, stretching--is done in all processing platforms.  I like PI better than Nebulosity for this.  

[Rodd]

18 minutes ago, swag72 said:

So if one uses PS to do astrophotography one can't produce an image that the PI people (and those imagers that regurgitate ad nauseam the same stuff) consider worthy ........... Interesting.

The PI forum will not let you post an image unless "ALL" processing was done in PI.  That is a rule they have.

[Rodd]

18 minutes ago, Scott said:

Its not the processing I hate... It's my inabilities that test my patience

Finally....someone who understands!!

[swag72]

2 minutes ago, Rodd said:

So the idea of "staying true to the data" is hogwash? ...........

I will only speak for myself here and say that my main aim is to produce a pretty picture....... There, I've said it and people can damn me if they so wish. To that end I experimented with my MCM process where I created a hubble palette style image from only Ha data. You can see it here .

When I said about the PI folks considering it worthy, I wasn't only talking about whether anyone can post an image on the PI forum that has had a tickle with anything else .... I was talking more about the people who tell you that if you do not use PI then your images are really not as good as those created in PI. Perhaps one should tell that to Robert Gendler who still uses CS2 and is frankly one of THE best imagers out there, or Perhaps Adam Block as well............ Clearly their images would be so much better if they jumped ship ............ if you believed the PI rhetoric of course

[Rodd]

16 minutes ago, ChrisLX200 said:

I have PixInsight and I use it for some things but there are so many other features I neither understand nor want TBH. Like most new users of PixInsight I scoured the web for tutorials (because there is no manual for it - not even the individual functions are well documented and what is really missing is how all these tools work together as a whole). Anyway, I ended up following recipes of those that knew better than me, and for the example given it sort of worked OK. However, moving on to my own data and it just fell apart   You cannot just slavishly follow someone else's workflow because your own data is unique.

ChrisH

 

4 minutes ago, swag72 said:

I will only speak for myself here and say that my main aim is to produce a pretty picture....... There, I've said it and people can damn me if they so wish. To that end I experimented with my MCM process where I created a hubble palette style image from only Ha data. You can see it here .

When I said about the PI folks considering it worthy, I wasn't only talking about whether anyone can post an image on the PI forum that has had a tickle with anything else .... I was talking more about the people who tell you that if you do not use PI then your images are really not as good as those created in PI. Perhaps one should tell that to Robert Gendler who still uses CS2 and is frankly one of THE best imagers out there, or Perhaps Adam Block as well............ Clearly their images would be so much better if they jumped ship ............ if you believed the PI rhetoric of course

Quite frankly--I don't know what to believe anymore.  Not that I ever did.  I am a firm believer in "taking what is good (or what you can use), and leaving the rest".  Someday I will get PS and maybe between the two I will be better off.  But--until then, even if PI is bad.....there are people who produce cracking images with it.    I should be able to do better than I am able to do with PI.

[ollypenrice]

14 minutes ago, Rodd said:

 

So the idea of "staying true to the data" is hogwash?  That is a PI mantra.  If I was convinced I would have an easier go at it in PS, I would jump ship--but so far, no software I have tried (AIP4Win, PI, Nebulosity, CCDops) seems to work toward that end.  Calibration, alignment, stacking, stretching--is done in all processing platforms.  I like PI better than Nebulosity for this.  

You want to stay true to the data? Fine. Here's a real thriller of an image which is as true to the data as I can make it. Stacked, calibrated, pure as the driven snow:

So why do we need graphics programmes at all?

In case you were too overwhelmed by the sheer excitement of the true data that's Deneb at the top.

Olly

 

 

 

 

[Rodd]

19 minutes ago, swag72 said:

I will only speak for myself here and say that my main aim is to produce a pretty picture....... There, I've said it and people can damn me if they so wish. To that end I experimented with my MCM process where I created a hubble palette style image from only Ha data. You can see it here .

When I said about the PI folks considering it worthy, I wasn't only talking about whether anyone can post an image on the PI forum that has had a tickle with anything else .... I was talking more about the people who tell you that if you do not use PI then your images are really not as good as those created in PI. Perhaps one should tell that to Robert Gendler who still uses CS2 and is frankly one of THE best imagers out there, or Perhaps Adam Block as well............ Clearly their images would be so much better if they jumped ship ............ if you believed the PI rhetoric of course

Oh yeah--I love the MCM images.  But There is no verbiage on HOW it was done. (sort of like processing in general!).  HI get the why.

[Rodd]

11 minutes ago, ollypenrice said:

You want to stay true to the data? Fine. Here's a real thriller of an image which is as true to the data as I can make it. Stacked, calibrated, pure as the driven snow:

So why do we need graphics programmes at all?

In case you were too overwhelmed by the sheer excitement of the true data that's Deneb at the top.

Olly

 

 

 

 

Though I get your point, I would say that there are aspects of ALL processing platforms that are the same-The image above certainly needs to be modified (stretched) using techniques common to all software, that would be considered "staying true to the data" regardless of which software was used.  I am not familiar enough with the science of processing to really know what "Staying true to the data" means.  

[johnrt]

3 hours ago, Rodd said:

But certain processing techniques increase noise as well as signal, so one must either only target signal, or reduce noise after it is increased by whatever method one uses to "process".  Do you deactivate all "noise reduction" elements when using various PI tools?

I do not process in PI beyond calibration, stacking, combining and using the background neutralisation and colour calibration (and maybe DBE) tools. Everything else is done is CS5, from stretching using curves & levels onwards.

I think of my images in zones, high, medium and low signal and seek to only apply processing steps where they are needed - colour saturation and adjustment to just the nebula or galaxy, star reduction to only the stars, sharpening and contrast adjestment to only the highest signal areas. No need to use any noise reduction of you can achieve this, which you can with ease in photoshop.

[Rodd]

7 minutes ago, johnrt said:

I do not process in PI beyond calibration, stacking, combining and using the background neutralisation and colour calibration (and maybe DBE) tools. Everything else is done is CS5, from stretching using curves & levels onwards.

I think of my images in zones, high, medium and low signal and seek to only apply processing steps where they are needed - colour saturation and adjustment to just the nebula or galaxy, star reduction to only the stars, sharpening and contrast adjestment to only the highest signal areas. No need to use any noise reduction of you can achieve this, which you can with ease in photoshop.

Your supposed to be able to do these things in PI as well.  That's what I meant when I said "Use PI he way it was designed to be used".  You only use it for linear processing.  The PI folks insist that anything that can be done in PS can be done in PI--with more control, not less (their words, not mine).

[johnrt]

3 minutes ago, Rodd said:

Your supposed to be able to do these things in PI as well.  That's what I meant when I said "Use PI he way it was designed to be used".  You only use it for linear processing.  The PI folks insist that anything that can be done in PS can be done in PI--with more control, not less (their words, not mine).

And my point is that it cannot, and does not give you more control, rather less, bludgeoning data with overbearing results, or increasing noise where you do not wish to see the effects of processing - not the deft subtle touch that can be easily created in photoshop.

I have created stupidly complicated masks in pixel math, tweaked every slider there is, experimented with processes new & old and examined every image in statistics and still was left wanting. In what other way was this software designed to be used?

As I said before I used to process 100% in PI but gradually moved further and further away as I could create the processing *I* wanted, that I either couldn't in PI or was overly complicated to achieve.

 

[Rodd]

29 minutes ago, johnrt said:

And my point is that it cannot, and does not give you more control, rather less, bludgeoning data with overbearing results, or increasing noise where you do not wish to see the effects of processing - not the deft subtle touch that can be easily created in photoshop.

I have created stupidly complicated masks in pixel math, tweaked every slider there is, experimented with processes new & old and examined every image in statistics and still was left wanting. In what other way was this software designed to be used?

As I said before I used to process 100% in PI but gradually moved further and further away as I could create the processing *I* wanted, that I either couldn't in PI or was overly complicated to achieve.

 

I have no clue how it was intended to be used.  All I know is some people use it to make tremendous images.  I find it hard to believe that PS is the only platform that will work for me.  I know you are not saying this, but if I switch to PS and I can then render a decent image, it will prove to be true.  It is much more likely that I will find myself in the same boat as I am now, scratching my head saying "others can do it so it MUST be possible".  No...I fear I am missing a key element needed to process successfully, and until I find out what it is, no software will work for me.  I hate the idea of a monthly subscription.

[ollypenrice]

There are some tremendous PI images. They don't look like PI images at all, they look like the object. I think our own Barry Wilson is a PI man and he does lovely stuff. And there's the rub: Only the best imagers get really natural looking results from 'all PI' processing. I'm not going to link to PI images I don't like because I'm not in the business of insulting the hard work and passion of my fellow imagers, but there are lots of PI images I don't like at all. They have a big following, have appeared as APODs and so on. But they look like graphic art and design to me and have an instantly familiar aspect. 

I remember finding an image I did with Yves to be just lacking in attitude compared with a well known PI rendition. I then thought, hang on, the hard blue edge to this bit of nebulosity does not exist. A pure log stretch of our bue data showed no edge at all, just a gradient. I decided - ahem - to remain true to our data...

Olly

[ChrisLX200]

Yes, far too many PI images look to be lacking in homogeneity. Galaxies that appear as though cut from paper and stuck onto a black starry background (for example), seemingly isolated from their celestial environment. They look artificial to my eyes with faint features (which really are faint and should remain so) enhanced to the point of being detached from the rest of the image.

ChrisH

[johnrt]

1 hour ago, Rodd said:

I have no clue how it was intended to be used.  All I know is some people use it to make tremendous images.  I find it hard to believe that PS is the only platform that will work for me.  I know you are not saying this, but if I switch to PS and I can then render a decent image, it will prove to be true.  It is much more likely that I will find myself in the same boat as I am now, scratching my head saying "others can do it so it MUST be possible".  No...I fear I am missing a key element needed to process successfully, and until I find out what it is, no software will work for me.  I hate the idea of a monthly subscription.

Why don't you have a go at this and see if you can recreate what others have done in different software, that might answer some questions for you?

 

[Rodd]

1 hour ago, ollypenrice said:

There are some tremendous PI images. They don't look like PI images at all, they look like the object. I think our own Barry Wilson is a PI man and he does lovely stuff. And there's the rub: Only the best imagers get really natural looking results from 'all PI' processing. I'm not going to link to PI images I don't like because I'm not in the business of insulting the hard work and passion of my fellow imagers, but there are lots of PI images I don't like at all. They have a big following, have appeared as APODs and so on. But they look like graphic art and design to me and have an instantly familiar aspect. 

I remember finding an image I did with Yves to be just lacking in attitude compared with a well known PI rendition. I then thought, hang on, the hard blue edge to this bit of nebulosity does not exist. A pure log stretch of our bue data showed no edge at all, just a gradient. I decided - ahem - to remain true to our data...

Olly

I reckon that no matter what software one uses, producing results as good as Barry is a rare and difficult talent, especially for individuals with little experience.  A side by side comparison done by the same individual skilled in both platforms would be interesting--the only difference would be the software--same data, same person (same habits, tendencies, likes, dislikes etc). 

[ollypenrice]

I think we have rather seriously derailed Sara's superb M16 thread recently, but let this discussion stand as a testament to her perfect processing which I, for one, cannot fault at all.

As I just said to her elsewhere, I feel the PI claim about being true to the data reminds me of those historical monks who, fobidden by their order to handle moiney or draw blood, wore gloves and went into battle with clubs...

Olly

[PhotoGav]

Mmmm, an interesting discussion. Personally, + 1 for PhotoShop, but then I have been using it since version 1, so have developed my skills with it. I simply can't face the learning curve of PI and have so far managed without it.

Back to the original post - Sara, this is another corking image. How is the processing going? When will you be brave enough to release the colour version to the forum?!

 

[swag72]

Thanks Gav ........ Do you think one needs to be brave to post up a Hubble style image of M16?

[PhotoGav]

Brave, just because it's you and I know how many hours an image must sit on your screen before you are brave enough to do anything with it!

As for Hubble's pic of M16, it took them years to just complete the mosaic and it is full of dodgy magenta hues..... This isn't a competition and I'm confident that yours will be a lovely, full, rendition!

[swag72]

Gav - Yes you've sussed me totally!! It's been sat for days and days on the laptop!