A culture of criticism?

[ollypenrice]

I think there should also be some caution when making a comment on what image a person has produced, just as an example, we all know that a one shot colour camera has an inferior image due to the bayer matrix as opposed to an LRGB or Narrowband image, some people cant afford say a 300/4000 series ATK for example, LRGB/Narrowband filters and a filter wheel and all the other stuff, someone could of put there heart and sole into producing an image with what kit they can afford, and before someone makes a comment they need to research the individuals equipment and comment accordingly.

Si, while the important part of your post is not limited to this technical issue, I'm not sure that we do all know that one shot colour is inferior. When I asked Atik for an OSC to test alongside my mono for Astronomy Now I was so impressed that I bought it. Plenty of APODS have come from one shot colour cameras. Greg and Noel and Dietmar Hager show what can be done with them. I know this was just an example you were giving but I think it's worth addressing anyway.

On another technical point, all my images are processed in Ps7 which is limited to 8 bit for many operations. I must fork out for a full 16 bit version but I just haven't yet bitten that financial bullet.

Olly

[jsigone]

personally I like the critics or suggestions on the image I post. It does help me learn and will help speed up the process if the critic will give some advice or link a tut for me to follow. All my knowledge of processing has been through tuts, trial & error and posting on forums like this one.

I do wish I had better gear, but I got other hobbies and family to support. So I spread my extra wealth on AP gear, bikes and my EVO X. Coming from a racing family, "run what you brung" is my moto and try to get the best I can from the gear I have.

[pete_l]

I think Orion the Hunter makes a good point....none of us know what a good image should look like and, like art in a museum, beauty is in the eye of the beholder. I recently completed a wide field of M8 / M20 and, while processing, went on line to get an idea of what it should look like. Is M20 red and M8 blue? Are they a mix of both? After reviewing 10 or 20 images of the pair...each beautiful in it's own right, but each differently "interpreted"...I came to realize that there was no single, correct image. I finished my photo with what was most pleasing to me...am I wrong?

Is that really the case though, Joe?

I'd say there are certain general points that hold true for any astronomical image. So while we may choose to alter colours and intensities for aesthetic reasons or to illustrate a particular aspect of an image, there are some basics that we all strive for - such as minimising the effects or limitations of our equipment on the final image.

To that end there are some elements that seperate a "good" quality final result from a poor one.

• We know that the sky background is dark, so effects such as light pollution or a noisy background are aspects we don't want in our photos.
  

• We know that stars don't have dark rings around them. They've been added by the processing we do (well, some of us do )
  

• Same with planetary images: features on Moon/Mars/Jupiter or Saturn's rings.
  

• We know that stars are points, so if they are trailed, misshapen or distorted due to coma or field rotation, those are effects we've added ourselves. You could even make the same case for diffraction spikes, but a lot of people seem to like them.
  

I'd reckon from my own images and processing that these factors account for a large proportion of my "early" images. I knew at the time they contained these artifacts and shortcomings and over time I have both learned how to avoid adding some and discovered what I need to buy to reduce some others,

[GordonH]

Hi

I agree on all of the above, if you want to improve then you need "constructive feedback" (a better term than criticism). While it is nice to receive praise from your peers about your images it doesn't tell you if there is anything amiss and therefore you don't learn anything.

I have to say that the improvements that I have made in my imaging/processing have been from the constructive feedback that I have received in various forums and groups. One of the people that I always eagerly await feedback from is a good friend of mine Richard Crisp (especially in relation to narrow band) because he has often been able to suggest little tweaks that make an image really pop and turn a very good image into a fantastic one. Sometimes we miss these things and need somebody else to look at it. It is important to remember, however, that you will only improve if you take on board the feedback and try it

Best wishes

Gordon

[SlipperySquid]

It totally depends on the person receiving the Critique. Constructive feedback or criticism call it what you like I can take it all but some people can't, fact of life.

Matt.

[ollypenrice]

It totally depends on the person receiving the Critique. Constructive feedback or criticism call it what you like I can take it all but some people can't, fact of life.

Matt.

...in which case I guess the thing is to make clear the kind of response you'd like to have to your picture?

Olly

[pietervdv]

Very interesting thread...

To my opinion listening to critisim is the only way you can grow as an imager. The manner how this is interpreted by the imager is an other matter. Some people see this as encouragement, some as a personal attack.

You will be able to distinguish constructive criticism easily, just forget the rest. This is a fact in everyday life aswell, I believe. You will always find negative people if you look close enough.

Fact is that astrophotography is a very complex form of photography, and there is not a "dummy guide book" describing how to do it in all its aspects. Forums like this offer a unique portal among imagers, and only with the remarks/critique we can grow and become better.

On forums like this the very best imagers are right next to the absolute beginners, the way it should be.

I am also on some forums regarding conventional photography, and there the atmosphere is totally different. There is a lot of jealousy and beginners are often torn to shreds there. I have rarely experienced this in AP.

AP is not at all a contest to get the best image, but it is all about how you feel about your work. To me every one of my images is special, and my very first (crappy ones) are the ones I cherish the most. It is only through encouragement from others that I am still persisting.

just my 2 cents...

Pieter

[roundycat]

I have said a lot of this before so please forgive me for repeating some of it.

There are two main aspects of an astro picture that get criticised. The overall colour and the detail in the picture. As far as colour is concerned we can do no worse than to follow the recognised world leaders and try to replicate what they have done. Emission nebulae are reddish, galaxies have a yellowish core and bluish outer region. We all know about the things that muck up colour so remarks about the colour can generally only follow the 'it looks magenta on my monitor' type of comment.

Apart from the basic colour the thing that reaches out from a finished picture and stabs you in the eye is the all-to-often seen halos around stars. Whether they come from over enthusiastic sharpening, De-con, DDP or the monumentally stupid mistake of splitting the picture into layers, one for the nebula and one for the stars and then working on them separately, the end result is there to see.

The best advice, as far as I can see, is to watch what you are doing. Don't just go through the process (or workflow) and do the same thing to each picture. An easy way to see the folly of this is to take two different pictures that have come straight from the stack, no stretches at all. If using Photoshop start a new action and then stretch one of the pictures until it looks 'nice', don't go mad. Then stop recording the action and play it on the other picture.

You will be amazed at what a mess it looks. Each picture needs to be treated as unique.

The thing you should always do, without fail, is to study the history palette and go back and forth after each adjustment. Look carefully at the stars and at the very first sign that all is not well go back and do it again.

You should do this at all stages. For my own work I often spend a lot more time looking at the extremities of the picture and flicking back and forth through history to ensure all is ok. I know my way around Photoshop and can process a picture in minutes if I don't keep checking it. Add in the time taken to make sure I have not mucked it up somewhere and it could take hours.

I'm looking forward to massive improvements after this thread!!!

Dennis

[ollypenrice]

That's all sound Dennis, though I fear I often make the monumentally stupid mistake working on stars in a separate layer! Of course, I only do this where the image has no stars, or virtually no stars, embedded in nebulosity. Then of course, it is impossible to do separate stars layers and I don't try. I used separate star layers for the Fireworks galaxy and the Leo Triplet but not the Swan, for instance.

I know you've said in the past that a single curve can usually handle the full brightness range. OK, not M42/Trapezium, probably. However, I have taken doing separate stretches and carefully layering them together, as in this Triplet.

http://ollypenrice.smugmug.com/Other/Galaxies/i-q9fN3Zv/0/X3/LEO-TRIPLET-in-TEC140-X3.jpg

I just found it easier than trying to get one curve to do it all. I suppose the danger is inventing changes in the final curve which don't reflect nature. I keep the different curves very close for this reason.

In this image only the galaxy is HaLRGB. The stars are simply RGB. (Actually the Ha had hardly any effect on star colour in this one but I removed it anyway.)

http://ollypenrice.smugmug.com/Other/Galaxies/i-GMSRcBf/0/X3/HaLRGB-Fireworks-galaxy-softer-X3.jpg

I totally agree about responding to each image individually. Each seems to need its own curve, maybe after the first iteration anyway, though you do more in Curves and less in Levels than I do. I need to look at this.

I also agree that flitting back and forth after any adjustment is vital. Where some part of the image has benefitted and another not, I will create a 'with' layer and a 'without' layer and erase what I don't want to keep.

Some people may feel that treating different parts of an image in different ways is not something they want to do. I don't mind doing it because I don't see my pictures as scientific, but I wouldn't want the selective differences to be fundamentally different. (Bad sentence!)

Although I haven't yet tried it, Rob offered this tip on another thread;

As regards sharpening and avoiding halos.....dead easy.

Run unsharp mask, then before you do anything else at all.....go to...

EDIT/FADE/FADE UNSHARP MASK and set the mode to LIGHTEN.

You may find that you initially have to sharpen slightly more than normal, as this process softens it a bit, but it loses all of the halos.

I'd love to see more processing routines carried out by others. Many people don't do much with their data while staying with me so I don't see as much as you might suppose.

Olly

[philj]

I'm a lapsed amateur artist and if I had works exhibited in galleries, shops etc I sometimes used to purposely lurk in the background during normal opening hours for the odd day after the opening exhibition and earwig comments. I considered this to be valuable feedback and would enable me to improve.

This feedback good or bad would be unsolicited and not pressured, the worst critics IMHO are family and friends because they will say its good to avoid hurting your feelings, similarly identifying yourself as the artist and asking opinions I felt would put alot of folks under pressure to say nice things even though they couldnt stand it.

I learnt to accept all comments even rude ones, yep there were the occasional ones:) and I accepted them and used this information to improve on my work. I must admit sometimes it was difficult to accept some comments but invariably there were reasons for the comments. Remaining anonymous prevented me from trying to justify myself and taught me how to accept criticism.

I enjoy imaging but I am in no way expert at it and there is tons of room for improvement in my images. I post them for the same reasons, feedback good or bad in order to improve and get hints and tips that I may have missed.

Therefore my take on this is if you are easily upset and offended by criticism constructive or otherwise dont put your work out there.

Philj

[steppenwolf]

Although there is a general 'workflow' that one can go through, every image is different, even your own images of the same object taken at different times and each deserves its own specific treatment. Sky conditions, altitude, darkness levels and number of subframes all conspire together to make a fixed workflow a little constraining.

I like to work in layers because although I could rely on the history tool, I love the ability to disable any layer in the stack from view to compare it with what has gone before as for me, a certain amount of trial and error works best as I always have in mind a specific end goal that I am working towards - I always keep this multi-layer stack on file so that I can re-visit it at any time in the future.

Obviously, it isn't all trial and error and my general workflow (I'll use my one shot colour version here) is as follows:-

MaximDL

1. Calibrate, RGB convert and stack

2. Preliminary stretch ensuring no clipping

3. Digital Development Process (DDP) for some galaxies but never for nebulae

4. Save as IEEE Float FITS file

5. Save again as 16bit TIF file using 'Screen Stretch' as reference for 'levels'

PhotoShop

1. Import TIF file, duplicate layer and make fine adjustments in 'levels' to each channel to remove any residual light pollution (I use a Hutech IDAS LP filter for OSC work so this is not normally and issue)

2. Duplicate layer and check for and eradicate any colour cast

3. Duplicate layer and boost colour saturation iteratively

4. Duplicate layer and adjust curves (normally at two points - left hand of slope for black point, right hand of slope for brightness and contrast)

5. Duplicate layer twice and sharpen top layer

6. Either

a) Adjust blend until just the right amount of sharpening is achieved or

Convert top layer to a layer mask and 'paint in' the transitional edges if required so that only those are sharpened.

7. Flatten top two layers

That is my basic workflow and I could stop there a lot of the time but I don't, I then do hours of fiddling to get the image just as I want it!

[Psychobilly]

Doesn't this sub-board already provide a mechanism for "critique" and comes with a "health warning"....?

Imaging - Image Processing, Help and Techniques Post an image in here for specific help on what may have gone wrong. Do not post an image in here then moan that people said it needs improvement . Don't forget to include as many details as possible about the equipment used, settings and observing conditions.

Peter...

[ollypenrice]

Doesn't this sub-board already provide a mechanism for "critique" and comes with a "health warning"....?

Imaging - Image Processing, Help and Techniques Post an image in here for specific help on what may have gone wrong. Do not post an image in here then moan that people said it needs improvement . Don't forget to include as many details as possible about the equipment used, settings and observing conditions.

Peter...

Good poinht. I thnk the board is under used and will post more stuff on it in the future.

Olly

[reddoss]

PhotoShop

1. Import TIF file, duplicate layer and make fine adjustments in 'levels' to each channel to remove any residual light pollution (I use a Hutech IDAS LP filter for OSC work so this is not normally and issue)

2. Duplicate layer and check for and eradicate any colour cast

3. Duplicate layer and boost colour saturation iteratively

4. Duplicate layer and adjust curves (normally at two points - left hand of slope for black point, right hand of slope for brightness and contrast)

5. Duplicate layer twice and sharpen top layer

6. Either

a) Adjust blend until just the right amount of sharpening is achieved or

Convert top layer to a layer mask and 'paint in' the transitional edges if required so that only those are sharpened.

7. Flatten top two layers

Steve,

I know there is a section in your book about image processing using Photoshop etc, but have you concidered writing another just on the processing? I think this where alot of us fall down. Even though I have the video tutorial CD's by Jerry Lodriguss and Adam Bloc, I feel I am still pretty clueless and spend hours messing about in Photoshop getting no where!

[steppenwolf]

I know there is a section in your book about image processing using Photoshop etc, but have you concidered writing another just on the processing?

Lots of people have asked me this and I have already started to write it - and another title too! Early days right now but I will get there.

[reddoss]

Lots of people have asked me this and I have already started to write it - and another title too! Early days right now but I will get there.

I look forward to it being published.

Cheers,

Justyn

[shaunster]

Theres already a book available on photoshop, now in its second edition it features everything but with a cost to match at £42, well worth it though if you want a photoshop 'bible'

Photoshop Astronomy - SCS Astro

[JoLo]

Pete, point well taken. I agree there are basic standards that make up a solid astrophoto and are prerequisites for processing that follows. I think most of us can compare two photos and realize one is superior to the other and whether the difference lies in acquisition or processing is not really important. The only point I was trying to make (obviously not very well) is that two solid photos of the same object that adhere to the basics for acquisition and processing can still be very different interpretations. Is one right or one wrong, is one better than the other? IMHO, only the person doing the work can make that call...

[Sp@ce_d]

Doesn't this sub-board already provide a mechanism for "critique" and comes with a "health warning"....?

Imaging - Image Processing, Help and Techniques Post an image in here for specific help on what may have gone wrong. Do not post an image in here then moan that people said it needs improvement . Don't forget to include as many details as possible about the equipment used, settings and observing conditions.

Peter...

yeah, that's where I thought we post stuff for full on critique. At least that's how I interpreted it.

[roundycat]

Hi Olly,

two quotes from your post. (I don't really know how to do this).

I know you've said in the past that a single curve can usually handle the full brightness range. OK, not M42/Trapezium, probably.

I am fairly sure I have read something like this before but I can't think where. Just for the record, I have never said that a single curve can handle any processing requirement. Simply because it is not possible. It is far more usual for me to use between four and ten curves for different purposes. If I have any kind of secret with Curves it is little and often.

As regards sharpening and avoiding halos.....dead easy.

Run unsharp mask, then before you do anything else at all.....go to...

EDIT/FADE/FADE UNSHARP MASK and set the mode to LIGHTEN.

You may find that you initially have to sharpen slightly more than normal, as this process softens it a bit, but it loses all of the halos.

One area of Photoshop that has been a mystery to me for years is the 'Fade' item. I believe this acts on what has been described as a floating layer. That is, a layer that can be worked on (once?) but doesn't show in the layers palette.

You can use the 'Lighten' layer mode simply by duplicating the layer and then sharpen. Then set the layer mode to lighten and any dark halos or other by-products will go. I think this is a bit of a cheat because the artefacts that Lighten gets rid of should not be there in the first place. I used this trick about ten years ago to get rid of dust in the sky when I was scanning slides and negs.

Dennis

[ollypenrice]

Hi Olly,

two quotes from your post. (I don't really know how to do this).

I know you've said in the past that a single curve can usually handle the full brightness range. OK, not M42/Trapezium, probably.

I am fairly sure I have read something like this before but I can't think where. Just for the record, I have never said that a single curve can handle any processing requirement. Simply because it is not possible. It is far more usual for me to use between four and ten curves for different purposes. If I have any kind of secret with Curves it is little and often.

As regards sharpening and avoiding halos.....dead easy.

Run unsharp mask, then before you do anything else at all.....go to...

EDIT/FADE/FADE UNSHARP MASK and set the mode to LIGHTEN.

You may find that you initially have to sharpen slightly more than normal, as this process softens it a bit, but it loses all of the halos.

One area of Photoshop that has been a mystery to me for years is the 'Fade' item. I believe this acts on what has been described as a floating layer. That is, a layer that can be worked on (once?) but doesn't show in the layers palette.

You can use the 'Lighten' layer mode simply by duplicating the layer and then sharpen. Then set the layer mode to lighten and any dark halos or other by-products will go. I think this is a bit of a cheat because the artefacts that Lighten gets rid of should not be there in the first place. I used this trick about ten years ago to get rid of dust in the sky when I was scanning slides and negs.

Dennis

Ah, sorry, I'm mis-representing you then Dennis! My apologies.

I'm glad to hear this because I've been busting my little head over how to make one curve do all. But I'm sure someone did say that they did it that way.

Olly

[RobH]

One area of Photoshop that has been a mystery to me for years is the 'Fade' item. I believe this acts on what has been described as a floating layer. That is, a layer that can be worked on (once?) but doesn't show in the layers palette.

You can use the 'Lighten' layer mode simply by duplicating the layer and then sharpen. Then set the layer mode to lighten and any dark halos or other by-products will go. I think this is a bit of a cheat because the artefacts that Lighten gets rid of should not be there in the first place. I used this trick about ten years ago to get rid of dust in the sky when I was scanning slides and negs.

Dennis

I didn't realise that the 'EDIT' command was a floating layer Dennis....what you say about sharpening a duplicate and then setting the blend to LIGHTEN makes sense.

I disagree about it being a 'cheat' though as it's not like there's a competition or something....if it works, it works, and that's that

I use a combination of that technique and deconvolution filters often....works for me

Cheers

Rob

[Tom OD]

Tricky question this, and I was just popping in to see the equipment section!! Ha ha.

Generally I don't want to offer negative criticism, so I might not post a reply, this hobby is hard enough as it is without knocking people back un-intentially a step or two.

Beginners to the subject need help and ask for it, then its good to offer as much advice as you can. If you've producing images that wow, then I think its only fair to expect a serious critic when asked, right down to the corners / found a faint satelite trail etc....

I want to get better and better too, this is a learning and sharing group.

The aforementioned monitor calibration point is a huge white elephant though (Never used that phrase before so hope its correct). I think those who post and ask for advice (especially on any colour subjects) should mention if their monitor is calibrated, and how, or with what device, ie The Spyder, Windows own calibration step by step, which is an eyeball procedure etc..

Tom

[roundycat]

Rob, I only meant cheat in the sense that it is a lazy workaround. I'm with you really, if it helps the pretty picture then do it.

Tom, do you mean red herring? I have found that monitor calibration is not as good as it is made out to be. OK if you are talking about your monitor, your camera, your scanner etc but once you bring another monitor into the mix it often goes wrong. Often that monitor is in someone else's house so under different lighting. The light falling on the screen is the biggest offender. You have no control at all as to how I view the picture, calibrated monitor or not.

Dennis

[RobH]

Rob, I only meant cheat in the sense that it is a lazy workaround. I'm with you really, if it helps the pretty picture then do it.

Tom, do you mean red herring? I have found that monitor calibration is not as good as it is made out to be. OK if you are talking about your monitor, your camera, your scanner etc but once you bring another monitor into the mix it often goes wrong. Often that monitor is in someone else's house so under different lighting. The light falling on the screen is the biggest offender. You have no control at all as to how I view the picture, calibrated monitor or not.

Dennis

I know you weren't being critical Dennis....no worries

I agree with you regarding monitor calibration.

I tried using Spyder 3's on my computers (all laptops), but had no success at all, so what I do now, at the final processing stages, is check out the web for other images of the same object, and if my colour looks similar to most of them, then I know I'm on the right track.

I also use the info tool in photoshop to actaully get exact readings of things like background levels...that way, it doesn't matter what the monitor says, I know that, for example, black is really black, and white is really white.

Cheers

Rob