Trying my best

[stepping beyond]

I've got a lot on my plate learning SGpro and learning how to process RGB images , this is one hell of a ride on been on for 3 years , that I've set out on  . This test of m42 is taken with asi174MM cooled -10c no calibrations 6 minutes of data and  no guiding . I'm having trouble getting the colors to come out the right way any help is more than welcome to advance me in my adventure. Please don't laugh to hard , that it's green blue and pink. I would hate it if I was responsible for someone having a chronic laughing episode. I'm learning a lil' bit at a time.

[Red Dwarfer]

Good early result !

 

 

[spillage]

it might be worth making you rgb files accessible so we have a look at the data. it does look very blue for a rgb image. 

[stepping beyond]

That is a kick in the right direction , I catured in png format in Sharpcap theen stacked lum and rgb separate and tweaked in adjust and color adjust in ccdstack2 , then used plane align in maximdl , photoshop {astronomy tools} back to ccdstack color adjust and had trouble getting it to  split into tricolor . The stacked image wanted to stay {1}color but, this is {1} of many that I figured is the best attempt that I've had so far after 2 weeks of trying . Here's  the stacked rgb , L 3s , L30s, R30s,g30sand B30s have at it Y'all and let me know what I need to add or change in my workflow. I just got a newer dark library made and I'll add .

[wimvb]

Why would you capture in png format? Best stick with regular Fits.

Calibrate each filter & stack to get three (four) masters: (L,) R, G, B. After stacking, use only one program for processing, until you are familiar with that.

Adhere to the KISS principle, and Keep It Simple.

[stepping beyond]

I was just playing with some data trying to get a handle on what tools does what , it started to look good and then the more I did the worse it got . Instead of dumping the data , I decided to try and get used to how to go about processing rgb . Wim, I'm trying to learn this stuff and it's a real [removed word]. I'm doing just what my buddy is telling me to do but, I'm not as schooled @{ 12 yrs. of rgb }as he is. Why he says to do it this way " he says some software does more and better at different aspects of processing the image." He wants me to get used to the tools in each . He's does some of the finest imaging that I've seen and he's 3000 miles away. It might be oversaturated in red but, I like the detail I'm pulling out. I just got to get a handle on bringing the colors out wim. I'll remind myself to keep it simple lil' bits at a time , right?

[wimvb]

4 hours ago, stepping beyond said:

I'll remind myself to keep it simple lil' bits at a time , right?

That's right: small steps. You'll get there. I like your last image. Even though the colour is very red, you've got lovely detail showing, and a lot less noise. That's definite progress in my opinion.

Since we use the same camera, one more tip. I found it really necessary to take good calibration frames, especially matched darks (same temp, gain and time) and flats. I also take dark flats (darks that match the flats), and don't bother with bias. Getting the darks right is crucial in order to get rid of amp glow. The good thing is, unlike with a dslr, you can shoot darks any time after your lights, as your camera is cooled.

Good luck with it.

[ollypenrice]

Is this a one shot colour camera? I'll assume so and continue: I would say you are using the wrong Bayer offset. The software needs to know which pixels are placed under which colour filters so that it can ascribe the red filtered pixels to the red channel, the green to the green and the blue to the blue.

Since the Orion nebula is mostly red it looks like this isn't happening.

Stacking software usually has a way of choosing the correct allocation, known as the offset.

Olly

[stepping beyond]

I had moved my hand over the touchpad on my Asus-i5 and my dark library was misplaced somewhere in my computer, instead of wasting time trying to locate them. Olly, I've got an asi174MM cooled and now I have sgpro with the frame and mosaic plugin added and learning what does what is a steep climb, with help from a good buddy  . I need to read up on and test using the offset function , that's on the to do list , thank you Olly . Wim , I processed out the ampglow in the 120s and 300s and I'll add it .

[ollypenrice]

There are not many possible options in the Offset selection. When I used an OSC I took a pic of our kitchen and just ran through the options till I got a sensible colour.

On the other hand you might just post a question in the imaging discussion section asking users of the same camera what their offset is, or find it via a net search. It's not a big deal.

Olly

[stepping beyond]

Thanks Olly, I'm trying my best to learn what each tool will do. I'll see what responses I get in Id on the offset . Thank you for your response , I need all the help I can get to get better at imaging. Ron

[stepping beyond]

 What about trying Binning 2x2 on the rgb channels? What yall think?

[Oddsocks]

4 hours ago, stepping beyond said:

Olly, I've got an asi174MM cooled and now I have sgpro with the frame and mosaic plugin added and learning what does what is a steep climb, with help from a good buddy  . I need to read up on and test using the offset function , that's on the to do list , thank you Olly . Wim , I processed out the ampglow in the 120s and 300s and I'll add it .

 

3 hours ago, ollypenrice said:

There are not many possible options in the Offset selection. When I used an OSC I took a pic of our kitchen and just ran through the options till I got a sensible colour.

On the other hand you might just post a question in the imaging discussion section asking users of the same camera what their offset is, or find it via a net search. It's not a big deal.

Olly

 

1 hour ago, stepping beyond said:

Thanks Olly, I'm trying my best to learn what each tool will do. I'll see what responses I get in Id on the offset . Thank you for your response , I need all the help I can get , to get better at imaging. Ron

Hi Ron,

Have been following this topic for a couple of days, have been a bit confused though, as with Olly's comments above regarding an OSC when you said earlier that you were using PNG captures in Sharpcap as your source images for post processing.

You state above that your camera is an ASI174MM, that is a mono camera and can only produce natural colour images when you image through separate red, green and blue filters, with a mono camera no debayering is required and offset is not required, these are only used with OSC or DSLR cameras.

You can produce false colour images with a mono camera and no filters by taking three differing depth exposures and assigning a different RGB colour to the three differing exposures but this is a whole different type of colour imaging.

Assuming you did take separate red, green and blue filtered images and looking at your first image I would have guessed that you had assigned the wrong filtered images to the wrong colour channels in the RGB combine routine.

As far as learning differing software packages by moving between them for different stages of the processing sequence that is only something I would recommend after you have mastered a full processing sequence in a single package, that is the only way you will learn the strengths and weaknesses that each product offers.

You said earlier that you have Maxim and I would recommend using Maxim for capture of long-time exposures with cooled cameras, in preference to Sharpcap which I would only use for video capture of planetary targets. Maxim will capture the images in lossless FIT file format which is superior in every way to PNG. Maxim has its shortcomings of course being only a 32bit program and limited in memory makes it rather slow by todays standards but you should be able to guide, capture Luminance and the three colour filter images for R G B, capture your bias darks and flat calibration frames, all the while Maxim will add the exposure details to the image header so it won't get confused when combining the three colour channels since the filter colour gets recorded with the image data. Non of this is possible when using PNG as your preferred file type. Calibration, alignment, stacking, colour balancing and stretching can all be carried out in Maxim before exporting as a TIF image for Photoshop etc, the important thing is that the image should be as near natural as possible at this stage with the right colours showing.

I have been using Maxim for the best part of thirty years and though I now do all my post processing in PixInsight I still rely on Maxim for capture and guide while under the control of ACP for observatory management.

I think to help you any better you probably need to explain a bit more clearly what it is you are trying to do and your methodology. If you did indeed take separate luminance and red, green, blue filtered images and still have your individual stacked masters of the four groups then consider posting them up in folders on Google Drive or Dropbox and we can look at them and form a better opinion of what is wrong, otherwise we are just guessing in the dark.

William.

[ollypenrice]

Aha, as William says, if this is a mono camera you have to shoot through filters and combine them to get colour.

Olly

[wimvb]

3 hours ago, ollypenrice said:

There are not many possible options in the Offset selection. When I used an OSC I took a pic of our kitchen and just ran through the options till I got a sensible colour.

On the other hand you might just post a question in the imaging discussion section asking users of the same camera what their offset is, or find it via a net search. It's not a big deal.

Olly

Olly, all ASIxxxMM cameras are mono, ASIxxxMC cameras are OSC. The ASI174MM-cool is a small chip cooled mono cmos camera.

(After writing this in all haste, I saw Williams reply ...)

20 minutes ago, stepping beyond said:

 What about trying Binning 2x2 on the rgb channels? What yall think?

This is a cmos camera, not a ccd camera; binning is of little use. Since you're learning the basics, don't complicate things with binning. Think KISS.

As for offset, with this same camera, I use the following:

gain below 100, offset = 10. Gain 100 ... 200, offset = 20. Gain above 200, offset = 50. Ideally you want to have an offset that results in ALL pixel values of a bias frame to be larger than 0. But this results in large offset values for high gain, which reduces the dynamic range. In reality, this is not as critical. The dark current you get with real exposures (several tens of seconds), keeps the pixelvalues above 0. So just start with the values given above, and adjust when you find you need to.

[stepping beyond]

William , I started  using maximdl v5 when I had my Meade  mono tec Pictor and now that I've an up to date Zwo tec camera, My Alanna decided to get  me sgpro to have an up to date software to go with the camera . I was using old imaging equipment on a old xp computer unfortunately they died and I'm learning all over again. I started out doing rgb imaging and  before I could get a handle on it , the computer and then the camera died and my buddy can't seem to get the Pictor fixed after 2 years or my meade osc video/ capture. I tried the OSC cameras and was wanting more detail Osc  cams use a bayer matrix and , I've gone back to mono because mono is the way to learn imaging. I did  post my png stacked individual data but, I take my images now with SGPro in fits  and convert in PScs2 to 16bit . I know that something isn't right , I know I'll eventually get the answer . If I could just ask the question the right way, Knowledge is what I seek and I need to get on the right track , instead of taking 2 steps forward and 3 steps backwards.

[Oddsocks]

Ron,

Are you using a manual filter wheel or an automatic wheel?

I don't know SGPro and have never used it but normally an automatic wheel is configured with which colour filter is in which slot and this information is stored in the fits file header when the image is taken, when the image is colour combined the program looks for that information to place the right colour filter in the right colour channel.

When taking RGB images with a manual wheel in an automation program like SGPro or Maxim you need to be really careful that the right colour filter information is stored in the fits file header otherwise the wrong colour combine will result. 

Try looking at your individual colour images in Maxim, open the fits header viewer and look to see that the right colour filter info is recorded, i.e, if you think the image was recorded through the red filter is that information actually showing in the header. I know from my own experience that it is possible to take a whole sessions images using a manual wheel, moving the filter wheel to the three colours but completely forgetting to update the capture program with the change of colour (automatic) colour combine will then fail as it can't identify which image was taken through which filter. The good thing is that if this is wrong you can edit the fits header and change the colour information so it matches reality.

If you are using an automatic wheel then make sure SGPro is configured for the right colour filter in the right slot.

Even though you take your images in SGPro you can still load the resulting fits files into Maxim for initial post processing and colour combine, 16bit TIF would be a better choice that PNG for export to PS. Unfortunately when you convert to TIF or PNG all the diagnostic data that helps to identify problems is lost, the FITs (information) header is always the first place to look when trying to identify problems.

p.s. just noticed you are in NC, spent quite a few years travelling to Durham for a month or two each year as part of my job, got to quite enjoy the Durham Bulls games. Last time I was there was in 2011 and was surprised how bad the light pollution was right down the coast from Durham to the tip of Florida.

William.

 

[wimvb]

Ron,

unfortunately, the stacked images that you provided in your first post, can't be used for further processing. The SGL server converts them to 8 bit jpg. So even if they were 16 bit png, a lot of information was lost. As I wrote before, it's best to have the images in tif or fits format, since this is uncompressed 16 bit data. The ASI174 outputs images in FITS format as a standard (at least it does with the linux driver), and I wouldn't change those settings in the imaging software. File size is only 4.6 MB per image, which shouldn't be a problem with storage space. Once you've calibrated and stacked all the individual subframes, you can zip them on your harddrive if you need to save space. (But I haven't bothered with that yet.)

[stepping beyond]

LP is pretty bad but, worse near the big cities William , I've got an electronic fw and I live in a small city of 10,000 pop. SW of Charlotte ,NC  5 miles from the upstate of S. Carolina and there huge old trees throughout my town. I'm in an Orange zone  and get some good data in the Fall/ Winter months but, Spring / Summer is Planetary unless I get some great sky from a cold front but, that usually means rain. I do what I can with what I'm dealt. Soccer and college basketball is what I enjoy watching , NC Chapel Hill  are the reigning National college champions . If the skies are clear and there's a game on , I'll remote telescope control from my lift chair while watching the game also. I'll still have to refocus when a filter changes but, a moonlite is on the Xmas list. I'm still building my image kit and one day I'll have an observatory and pier but, I'll take my lashings until then.

[stepping beyond]

 Wim ,  well that's a loss . it's all in the learning experience. I do have a partial sequence with sgp  in8 bit fits like Sharpcap does. I save all my usable data on a flash stick and dump my trash. I don't want to misplace that precious stuff. Data collecting is most of what I'm doing right now and when it rains , I may stack a pile and eventually get around to my 1st run through processing . I just keep adding with my short sessions until , I get my guiding working right. I got a used QHY5-ll M and it doesn't have a nose piece, I'm using my 3x tube and it gets knocked out of whack everytime I have to move it. I hope my buddy can team viewer and figure out my issue . I might get out tonight if this wind will calm down, skies are looking good , it'll be cold tonight and I'll try those settings.

[Oddsocks]

One thing you can try in the day time without losing valuable sky time, to help resolve the colour problem, is just point the telescope, or the camera alone, at a large well lit object of strong primary colour, red, blue or green, placed just a foot or so in front of telescope or camera, then take a single exposure through each colour filter adjusting the exposure time to get a histogram roughly in the mid range for each of the filters. No need for tracking or focusing, you’re just interested in colour. If you can’t reduce the exposure low enough then cover the front of the telescope, or camera, with a piece of kitchen aluminum foil and make a hole in the foil about 1/4” diameter if on the scope, or just a pin-hole if on the camera.

Take the three resulting fits images, no calibration or alignment needed, then carry out a simple rgb combine, you shouldn’t need to stretch the resulting image provided the histogram was around the mid point for the individual filters.

Provided you have the correct filter assigned to the right RGB channels then the resulting image will just be a plain, blurry, blob, but it will be a blob with the same colour as the object you imaged.

If the resulting colour is anything significantly different to the colour of the object you imaged then you can dig deeper and physically check that when SGPro is asked to select the red filter, for example, then the red filter has really moved into the light path.

William.

[stepping beyond]

William ,I'll have to try that .  I'm all set to go for tonight and my buddy from South Carolina is going to get back with me later to Team viewer and find out what's going on with my guiding issue and try to remedy the problem. Also , I'm going to take my efw apart and check the filter position and movement.

[ollypenrice]

I'm sorry to have misunderstood your original post. I'm hopeless at remembering camera model numbers and thought you were using a colour camera.

I would not recommend Moonlite focusers for imaging. I have one and know a good number of other imagers who have tried them. They are nice smooth focusers for visual observing, especially on Dobs where they are always horizontal, but they can slip when asked to hold the weight of a camera. There are better imaging focusers for sure. I'd always go for rack and pinion over Crayford but there are Crayfords with better grip, like the Baader Diamant.

Olly

[Oddsocks]

Ron.

I had a little time free this evening so I took your original reduced quality R, G and B jpgs stacks (only) from your forum post, registered (aligned), converted them to greyscale and colour combined in PixInsight exactly according to your filter assignments as noted in the posted image file titles and the result is a normal colour image for the Orion Nebula.

According to this result your filters are in the correct slots and you have handled them correctly in the stacking routine.

I didn't bother trying to add in the luminance frames since the quality of the forum images is too poor at only 8bit and I haven't done anything else to your data other than a histogram transform, I only was interested to see if this colour problem could be resolved.

You don't need to worry about confirming which filter is in which slot and it appears that the wheel is rotating to the correct filter position.

The problem you are having is something fundamental in the way you are post processing.

Maybe when you have some time, try a post process sequence in Maxim, since you are familiar with that from your early days, if you are still struggling with that put your source image fits stacks into a shared Google drive or Dropbox and post the download link back here to see what some of the more experienced imagers can make of them with a documented walk through for you. I only know how to do colour combine in PixInsight or Maxim 5 and 6, never have tried in Photoshop, but there are plenty of imagers here that do work mostly in PS and can help you more along that path if thats the way you want to follow.

The result of the three RGB channel combine is attached below to show that the colour sequence you are using is correct.

HTH

William.

 

[ollypenrice]

23 hours ago, Oddsocks said:

Ron.

I had a little time free this evening so I took your original reduced quality R, G and B jpgs stacks (only) from your forum post, registered (aligned), converted them to greyscale and colour combined in PixInsight exactly according to your filter assignments as noted in the posted image file titles and the result is a normal colour image for the Orion Nebula.

According to this result your filters are in the correct slots and you have handled them correctly in the stacking routine.

I didn't bother trying to add in the luminance frames since the quality of the forum images is too poor at only 8bit and I haven't done anything else to your data other than a histogram transform, I only was interested to see if this colour problem could be resolved.

You don't need to worry about confirming which filter is in which slot and it appears that the wheel is rotating to the correct filter position.

The problem you are having is something fundamental in the way you are post processing.

Maybe when you have some time, try a post process sequence in Maxim, since you are familiar with that from your early days, if you are still struggling with that put your source image fits stacks into a shared Google drive or Dropbox and post the download link back here to see what some of the more experienced imagers can make of them with a documented walk through for you. I only know how to do colour combine in PixInsight or Maxim 5 and 6, never have tried in Photoshop, but there are plenty of imagers here that do work mostly in PS and can help you more along that path if thats the way you want to follow.

The result of the three RGB channel combine is attached below to show that the colour sequence you are using is correct.

HTH

William.

 

Nothing wrong with that colour. Great stuff.

Olly