The Eagle Nebula is part of a diffuse emission nebula, or Hydrogen alpha region, which is catalogued as IC 4703. This region of active current star formation is about 7000 light-years distant.
This image consists of 10 hours Hydrogen Alpha data processed blended with 6 hours of red data processed with a blending of 50%:50%. This Red Ha was then used as the red channel to create an RGB image consisting of 27 hours in total.
Imaging Scope TEC 140 on AP 1100GTO with STT8300 self guiding.
Any advice would be greatly appreciated and if poosible some steps to support what you are saying. Once everything is stacked and aligned I really have no set workflow and would desperately like some advice on that.
You did a nice job, Mark. Maybe a little clipped in the top right or perhaps you just don't have enough Ha for the dimmer stuff. You could probably push the star colours a little harder too? What do you use for processing?
Looks good Mark overall. But then I read you did 27 hours in total and that surprised me.
As Rick pointed out the dimmer areas should be able to lifted considerably with 27 hours of data.
I am not a fan (admittedly I haven't tried it much) of Ha blending with the red channel. Once you do that its a one shot go as there
is no further control over the blend.
To retain control you either use the Don Goldman clipped mask method for narrowband (he has a free video tutorial on his astrodon.com website)
or use Ha as a lighten blend only. With a lighten layer any pixel that is lighter than the one in the layer above gets shown. So if you boost curves in your Ha in lighten blend it will go up into the image. So you have control.
Try watching Don's video and then using his technique (its easy enough) and see if you can get a better result.
Another pointer is with Ha its a bit of a beast that needs to be whipped into line. It easily takes over an image. In your case the overall image is mainly
red which is slightly unappealing. It would be good to boost the RGB layer without Ha and have the Ha as a layer on top as per the Don Goldman video.
Now you can control how much Ha, what shade Ha (345 rather than full red in Photoshop). You can add several layers of Ha using different blending modes to create an effect. Soft light is a nice blending mode, so is lighten. Screen and others are a bit harsh. Don says to use screen but I often use soft light instead.
Thankyou all. I use photoshop Rick for no other reason than thats all I've ever used. I should head down the PI path. I would love someone to give me some recommended tutorials. I've read a number of times zone system and makimg every pixel count but I just don't seem to be getting it. It's a bit frustratimg honestly. Thanks Greg as always for you feedback and I'll go and have a look and try again. I agree I have many hours worth but will try again.
Nothing wrong with using PS, Mark, but I'm pretty clueless with it so I can't offer advice. If you're interested in PI take a look at Warren Keller's book or video tutorials (http://www.ip4ap.com/)
Mark, forgive my impudence, but for 27hrs data, there was obviously more hiding "in the shadows". I took your jpeg and ran it through a few actions in PS CS6 and some simple levels and curves, and got this (just from the one jpeg!)
AWESOME data - wish I had the base FIT files to fiddle with!
Lewis I'll set up a drop box and put them in there for anyone to have a fiddle with but if you do please scribble some basic suggestions on what people are doing. I would love to see what is possible with my data because I almost feel as if I have no idea.
Here is a quick go at your data Mark - would be better to have all the FITS files though, NOT TIFFs (you can stretch a FIT much further). Lossy-compressed JPEG:
I like the original one best. It may not show every bit of nebulosity but to me is a better 'more dramatic' image. After all these are all subjective interpretations...
Unfortunately, the files presented are TIFF, which loses linearity fast (well, a lot faster than the raw FITS do). I had a VERY hard time stretching the red as well, and had to put them into CCDStack and use DDP, balanced with Gamma and Background, then restretched again with DDP rebalanced with Gamma and Background (2 in total I think).
Then adding in the H-alpha using an action by Annie (Annie's Astro Actions for PS) after also having stretched the H-alpha frame.
Balancing the blue in the colour was problematic actually, and you can see it in the end result. That's after a lot of blue curve fiddling, backgraound, midtone and highlight balancing etc etc.
There is also a registration misalignment with I think the H-a frame, so you need to run that through CCDStack or DSS to align with the others, then save that as a scaled image - its a fair bit off in rotation and you can still see a slanted line near the bottom of my image. Also remember to sharpen the H-a prior to stacking it - makes a HUGE difference.
It's nice data - Mark just needs to hand us the FITS for some real playing (HINT HINT!)
Rick could have done it all in The Headache Machine...aka PixInsight
I did actually use PI lewis , what does DDP do exactly?(im guessing bring out the faint regions) would be good to know if i could do something similar in PI - also, i cropped out that line
...would be better to have all the FITS files though, NOT TIFFs (you can stretch a FIT much further). Lossy-compressed JPEG:
Sorry to hijack the thread but Lewis, why can FITS be stretched further than TIFFs? I have seen lots of claims like that but no real explanation anywhere.
Apart from the FITS header which contains the capture information, neither format uses lossy compression and, if saved with the same bit depth, should contain identical information.
Luka, not sure, but I know that the files could not be stretched very far before blowing out, unlike any FITS I have used previously. The TIF's were 16bit vs FITS which are 32bit and I think therefore have less 'Depth" than the 32bit FIT files.
