Jump to content

Banner.jpg.b83b14cd4142fe10848741bb2a14c66b.jpg

LRGB with horrible LP


ayoussef

Recommended Posts

A question for all the experts, enthusiastic here . . .

LP in my area became so intense that I cannot  take more than 30 seconds LRGB images anymore even when I lowered the gain, I am using QSI 6120 with baader LRGB filters, so I tried to stack 3 LP filters ahead of the filter wheel, now I can do 5 minutes before the LP overwhelm the sensor.

The question is, how accurate is each sub coming from the LRGB filters ? because if I used a color dslr for example with the LP filters the result image will be leaning heavily toward the blue, do I need to process the stack of each filter individually first to produce correct colors ?

I tried this technique against the bubble nebula, the red filter subs actually showed the red emission of the nebula, not as good as the Ha of course due to the less exposure time and being broadband.

I also tried RVB filters before and I still couldn't beat the LP.

Link to comment
Share on other sites

Light pollution filters will tint an image however good colour calibration routines will correct this.

Alternatively bite the bullet and go Narrowband, there is plenty of Ha targets in the later months of the year.

Link to comment
Share on other sites

Thanks for the tip Earl, I already have Ha,OIII,SII filters set, but for some targets like the wizard or the bubble I wanted to give the stars natural colors by taking rbg subs and combine it with the Ha.

So how do I calibrate the rbg subs to correct the effect of LP filters ? do I do that for each filter or combine them first into rbg then calibrate the result image?

Thanks!

Link to comment
Share on other sites

Here is an example of NGC281 image, I tried giving the stars real rgb but couldn't. 

<html><body><a href="http://www.astrobin.com/full/198413/0/?real=&mod="><img src="http://cdn.astrobin.com/images/thumbs/01fe221740c90a180636b3a48295c34b.16536x16536_q100_watermark.jpg" alt="NGC281" style="width:600px;height:420px;border:0"></a></body></html>

15x1200" Ha bin 1x1

7x120" R bin 2x2

7x120" G bin 2x2

7x120" B bin 2x2

I used the stacked Ha for luminance too.

Link to comment
Share on other sites

Thanks for the tip Earl, I already have Ha,OIII,SII filters set, but for some targets like the wizard or the bubble I wanted to give the stars natural colors by taking rbg subs and combine it with the Ha. So how do I calibrate the rbg subs to correct the effect of LP filters ? do I do that for each filter or combine them first into rbg then calibrate the result image? Thanks!

I use PI to do all colour calibration, its a simple process which works very well.

Link to comment
Share on other sites

I use PI to do all colour calibration, its a simple process which works very well.

The histogram in Photoshop / Paint.NET / GIMP is your friend.

Change the levels of each channel until the balance is right, although if you don't have enough data it will be impossible (as I learned with my orion pic, which is now perminently green because that was the only channel that got 'enough' data) you can use a colour picking tool to help. (I think it's a plugin in  Paint.net, a built-in feature in Photoshop and I'm not sure about GIMP)

Aditionally, and this is just making sure, are you familiar with histogram stretching? Because that can even cut through a full moon to some extent!

My quick picture of andromeda (8 mins unmodded DSLR under a >80% moon.) http://i.imgur.com/cNhP1aw.png

My aforementioned orion pic (9 mins under I believe ~20% moon) http://i.imgur.com/OSXFO28.jpg

Neither of those images would have been as "Good" without histogram stretching. Some details in M42 would have been lost entirely and the couple of grey pixels that make up the flame would not be visible.

Good luck!

    ~pip

Link to comment
Share on other sites

I think it's very simple. You need Pixinsight. I've been sent LP images to play with on many occasions and my initial reaction is one of horror. Bright orange, bright blue!?!?! What on earth am I supposed to do with that? Well, a quick dose of DBE or even the automated ABE zaps the sky back to neutral and gives a perfectly reasonable colour balance which may hold up untouched for the rest of the processing.

I doubt that I will ever want to process an entire image in PI (in fact I know I won't!) but ABE, DBE, SCNR green and LHE make it the bargain of the century even if I never touch any other tool in it.

Olly

Link to comment
Share on other sites

I doubt that I will ever want to process an entire image in PI (in fact I know I won't!) but ABE, DBE, SCNR green and LHE make it the bargain of the century even if I never touch any other tool in it.

Olly

Apologies in advance for a little tangent, but what would you not do in PI out of curiosity?

Sorry again

Matt

Link to comment
Share on other sites

Apologies in advance for a little tangent, but what would you not do in PI out of curiosity?

Sorry again

Matt

All of those steps are essential prior to applying a histogram stretch, Olly processes that sort of thing in PS, to good effect. I personally prefer to use PI as i am more familiar with it.

Link to comment
Share on other sites

Thats a right faff when its done in a couple of clicks in PI

Well, probably, but Paint.net & GIMP is all I have access to so any of this PixInsight magic is unknown to me.

Also, in the chance he doesn't have PixInsight like me, it would be his only choice AFAIK

Link to comment
Share on other sites

it is worth your time and effort getting top know these routines in PI and if like the OP has made the sizeable investment in a QSI 6120 you really need the software to back up that level of investment in hardware.

The PI / PS combination is the standard to aim for.

Link to comment
Share on other sites

Apologies in advance for a little tangent, but what would you not do in PI out of curiosity?

Sorry again

Matt

I hate masks and love layers. The problem with masks is that it is hard to make them go exactly where you want them to go. I make my own star masks in Ps sometimes but am very careful about how I use them. Mostly I prefer to make a copy layer, carry out an operation on it globally (as a bottom layer) then select what I want to retain from it by using a number of different selection tools and then erasing the top layer to let the bottom one through. I can use the eraser set only to a partial opacity, see how it looks, use it at a higher opacity in places - and where I can't get the selection tools to identify the bits I want to erase I can just do it by eye.

Ps has many selection tools and these make it so easy to work on the bits you want to work on. Trying to talk PI's numbers into giving you the mask you want is, to me, an absurdly remote way of doing something I simply want to do with my hands because I can see just what I want to do in front of my nose! Making masks is like playing charades...

Olly

Link to comment
Share on other sites

Thanks Olly. I have been PI only since I downloaded it. Will certainly try and rehydrate my atrophied PS skills, as there seem to be a few places where it is the tool to use. Apologies again ayoussef for the tangent - hope you got the info and help you needed.

Link to comment
Share on other sites

The histogram in Photoshop / Paint.NET / GIMP is your friend.

Change the levels of each channel until the balance is right, although if you don't have enough data it will be impossible (as I learned with my orion pic, which is now perminently green because that was the only channel that got 'enough' data) you can use a colour picking tool to help. (I think it's a plugin in  Paint.net, a built-in feature in Photoshop and I'm not sure about GIMP)

Aditionally, and this is just making sure, are you familiar with histogram stretching? Because that can even cut through a full moon to some extent!

My quick picture of andromeda (8 mins unmodded DSLR under a >80% moon.) http://i.imgur.com/cNhP1aw.png

My aforementioned orion pic (9 mins under I believe ~20% moon) http://i.imgur.com/OSXFO28.jpg

Neither of those images would have been as "Good" without histogram stretching. Some details in M42 would have been lost entirely and the couple of grey pixels that make up the flame would not be visible.

Good luck!

    ~pip

So in white lp area, can I image the pinwheel  galaxy for example with subs of 2 minutes for each filter? I am using QSI6120 with both 2.8 F newtonian and 4.25F 100mm esprit (I use reducers to lower the F with both)

Link to comment
Share on other sites

I think it's very simple. You need Pixinsight. I've been sent LP images to play with on many occasions and my initial reaction is one of horror. Bright orange, bright blue!?!?! What on earth am I supposed to do with that? Well, a quick dose of DBE or even the automated ABE zaps the sky back to neutral and gives a perfectly reasonable colour balance which may hold up untouched for the rest of the processing.

I doubt that I will ever want to process an entire image in PI (in fact I know I won't!) but ABE, DBE, SCNR green and LHE make it the bargain of the century even if I never touch any other tool in it.

Olly

And thats why I am trying to stick to mono, but triple the time of imaging is not my favorite thing in the world, specially that for the last 6 weeks I had only 2 clear nights. I am really resisting getting a color ccd.

Link to comment
Share on other sites

I took this image of Hercules cluster over the weekend, 60X30" , 20 for each filter with no calibration data, I used 2.8f 8" newtonian with QSI6120. I used PI for stacking and color calibration then used PS for final touches.

What do you guys think of the color balance ? 

Link to comment
Share on other sites

And thats why I am trying to stick to mono, but triple the time of imaging is not my favorite thing in the world, specially that for the last 6 weeks I had only 2 clear nights. I am really resisting getting a color ccd.

You need the same ammount of exposure whether you're on mono or colour. Each channel individually gets 66% less on colour and that means you need 3x the exposure!

Link to comment
Share on other sites

The colour/mono issue seems to stem from the fact that to get to use a decent integration algorithm, you need over 8 subs per channel. Doing LRGB, that means at least 32 subs for a complete image. It seems a waste to go for small exposure lengths just so you can get an image in a single night when the optimal on nearly every target will be much higher. This all means you are looking at multiple nights to get decent mono images, especially this time of year (summer).

This all doesn't invalidate the maths that you are getting much more equivalent exposure than an OSC, but to get 12 x 10 mins OCS subs (2 hours) is a lot easier to do than 8 x 5 minutes R,G,B with all the extra filter change hassles with no L for something to make an actual image with in those 2 hours.

Link to comment
Share on other sites

And thats why I am trying to stick to mono, but triple the time of imaging is not my favorite thing in the world, specially that for the last 6 weeks I had only 2 clear nights. I am really resisting getting a color ccd.

You don't triple the imaging time. You need longer in OSC than in LRGB as I'll explain in a moment.

You need the same ammount of exposure whether you're on mono or colour. Each channel individually gets 66% less on colour and that means you need 3x the exposure!

No, you'll actually get more light in a given time using mono LRGB than you will get either in RGB or OSC.

1) When you shoot L you are catching R and G and B so you are working, in theory, three times as fast. However, whenever I compare a synthetic luminance taken from a colour image with a true luminance through a Baader L filter I find the synthetic L to be worth only a quarter, not a third, of the real Lum.

2) OSC cameras mostly shoot twice as much green as red or blue. This is OK in daylight where the green is used as the luminance channel (stated in the Bayer patent) but green is not at all optimal as luminance on astronomical targets. (You'd probably do better on many targets wth RRGB. Note that many people use Ha - a special case of red - as partial luminance on emission nebulae.) Given the usefulness of SCNR Green or Hasta La Vista Green it's tempting to say that the last thng we want is extra green!

3) In a mono camera you can shoot the colour in Bin 2, saving time.

4) Many targets will 'pop' in an Ha layer and, suitably processed, will contain more useful data than the red channel would collect in the same time. An Ha filter can't collect as much light as a red but it's the contrast in Ha that is useful in processing.

I've said this before but the term 'One shot colour' is a bit misleading. It leads people to think they are getting something for nothing. Alas they're not, because there is no such thing as a 'one shot colour pixel.' A One Shot Colour camera would be better described as 'A quarter of a shot red, half a shot green and quarter of a shot blue' camera.* Hehheh, no, it isn't going to catch on as a name!!!  :grin: But it is more accurate and would help to dismiss the idea that you can work faster with these cameras. They can be less exasperating but they cannot be faster.

Olly

* We could reduce this to an acronym. QSRHSGQSB. Nah, still not going to sweep into orthodoxy, methinks...  :p

Link to comment
Share on other sites

You can get good results with OSC though so its not all doom and gloom, it does limit your options though.

No doom and gloom at all, but OSC isn't faster. This is so widespread a misconception.

Matt, regarding the need for sufficient subs for the algorithms to get the noise down, I really don't think that's very important in an RGB layer. It's more important in L, certainly. In RGB you can use aggressive hot pixel filters and other noise reduction without getting the 'vaseline on the lens' look. That look will be there in the RGB but the L will put the grain back and restore a natural appearance.

Olly

Link to comment
Share on other sites

Thanks Olly. I have been PI only since I downloaded it. Will certainly try and rehydrate my atrophied PS skills, as there seem to be a few places where it is the tool to use. Apologies again ayoussef for the tangent - hope you got the info and help you needed.

I have to say, you cannot go wrong with both tools even for a beginner like me.

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue. By using this site, you agree to our Terms of Use.