Would anyone like a crack at my data? PLEASE!!

[cthorpey]

Hi Folks,

I am fairly new to imaging and could do with some help.

I have a SW 130p modified to allow my Canon 1100D to achieve prime focus and a HEQ5 Pro

The image below is the result of 30 x 45 sec ISO1600 lights and 5 darks stacked in DSS and "processed" in Startools.

I am wondering if there is any more I could get out of the data (ie I am really bad at processing) or is it a case of garbage in, garbage out?

I have attached the output from DSS and would very much appreciate someone else having a tinker with it.

Thanks in advance...

Chris

M31 DSS Output1.tif

[ollypenrice]

Here you go.

You need to kill the gradient by equalizing the background sky (EG Pixinsight's DBE or ABE or Gradient Xterminator.) Then hit it with some noise reduction selectively applied. At this point you can push local contrasts a bit and then stretch the brighter data.

I may be wrong but I think this data was non-linear, ie it had been partially stretched in DSS. This isn't a good idea. Longer subs would make all the difference.

Bear in mind that I did this very quickly, as in 5 minutes. It isn't a careful processing job.

There is no colour data in the image. What happened there? I don't use DSS so I don't know where it was lost.

Olly

http://ollypenrice.s...39556&k=FGgG233

[cthorpey]

5 Mins!!! WOW took me all afternoon

Not sure what I did to kill the colour, and I just aligned the RGB peaks in DSS.

I think my main problem is that I dont understand the terminology here.

When you say gradient am I right in thinking that you are referring to the whitish "box" that seems to be surrounding the galaxy in my image?

Apologies for the dumb questions!

[ollypenrice]

A gradient is a drift in brightness across an image from one side or corner to the other. You are right that yours had a kind of 'box of brightness' around the object.

You mentioned darks but not flats. Flats are absolutely essential. You image a flat light source such as a luminescent panel of even brightness until the histogram peak is about a third of the way to saturation (ie a third of the way to the right.) You take about 20 of these and combine them, calibrating them with a master bias used as a dark. (A bias is the shortest dark frame your camera can take. Again, average at least 20 of these to make a master.)

Some take flats in the twilight using a white T shirt over the scope. This has rarely worked for me and is ulikely to work for a Newtonian where stray light can enter the scope from below the mirror and from the focuser, etc. I do them in the dark with a panel.

Olly

[lensman57]

Hi Folks,

I am fairly new to imaging and could do with some help.

I have a SW 130p modified to allow my Canon 1100D to achieve prime focus and a HEQ5 Pro

The image below is the result of 30 x 45 sec ISO1600 lights and 5 darks stacked in DSS and "processed" in Startools.

I am wondering if there is any more I could get out of the data (ie I am really bad at processing) or is it a case of garbage in, garbage out?

I have attached the output from DSS and would very much appreciate someone else having a tinker with it.

Thanks in advance...

Chris

Hi Chris,

If possible please supply a none stretched version of the stacked file ( without adjustments in DSS applied ), the data as Olly has pointed out appears to have been manipulated.

Regards,

A.G

[StuartJPP]

Had a play with it in PixInsight (DBE) and then PS6, as already mentioned there is no colour data. I also rotated it 180 degrees, to me it looks a lot better this way...just a personal preference. As Olly says, flats are really important.

[steppenwolf]

I can't find any colour data in here but here's my seven minutes worth!:-

[cthorpey]

Blimey!

I am guessing it is just me not knowing what I am doing then.

heheheh, every day is a school day. I am learning.

I have attached the raw unadulterated DSS output as requested.

Does anybody know of a tutorial that goes along the line of..

Look at this-------> this is an example of a gradient, to get rid of this we can do

a...........

b.............

or

c..............

Now look at this ----->

etc

I know my data is flawed and I know I missed stuff but being able to see what could be done is inspiring

Thanks

Chris

DSS straight.TIF

[lensman57]

Blimey!

I am guessing it is just me not knowing what I am doing then.

heheheh, every day is a school day. I am learning.

I have attached the raw unadulterated DSS output as requested.

Does anybody know of a tutorial that goes along the line of..

Look at this-------> this is an example of a gradient, to get rid of this we can do

a...........

b.............

or

c..............

Now look at this ----->

etc

I know my data is flawed and I know I missed stuff but being able to see what could be done is inspiring

Thanks

Chris

Hi Chris,

your effort is a lot better than my last of the M31 and that was the result of nearly 3 1/2 hours of exposure, I ended up with worst gradients than yours and so much noise in the DSLR image that in the end I just deleted the data out of frustration.

A.G

[steppenwolf]

The key to this is getting the data right early and Olly is quite right here, you've missed out the most important calibration data - Flats. These are a pain to take but they eradicate the first part of your gradients - vignetting.

My workflow for your image was as follows, carried out in PhotoShop CS3:-

1. Draw a generous selection line around the galaxy edges

2. Feather the selection by 15 pixels

3. Invert the selection

4. Run Russel Croman's Gradient Xterminator on the image

5. Adjust levels to darken background and lighten the galaxy

6. Copy the image to a new file and sharpen it

7. Return to the original image, make a duplicate layer and smooth it

8. Copy the sharpened image to the layer underneath the new smoothed layer

9. Use the eraser to remove the smoothness over the dust lanes

10. Final tweak in Curves

[cfpendock]

And what Steve (above) quite rightly omitted to say was to buy a copy of his book "making every photon count". This will provide an excellent grounding on what the different tasks of processing involve.

Chris

[Davey-T]

Does anybody know of a tutorial that goes along the line of..

Look at this-------> this is an example of a gradient, to get rid of this we can do

a...........

b.............

or

c..............

Now look at this ----->

etc

Chris

Have a search on Youtube for vid's on removing gradient etc.

Dave

[StuartJPP]

I started off with Doug's guides on using DSS to stack and then processing in PS with GradientXTerminator. I found the best way to eliminate most though was by taking and using flats....

[ollypenrice]

I think you have to accept that learning astro image processing is a long journey and one without end, since we all get better at it over time and that never changes. I find two year old images fairly embarrassing sometimes. Very embarrassing, in fact! But the fun is in the learning.

For me the two hardest parts to get right are background sky and stars. The main object is usually the easiest part.

While there are other options, absolutely nothing beats Dynamic Background Extraction (or sometimes the one click Automatic Background Extraction) in Pixinsight. I image at one of the darkest sites in mainland Europe but I still make this my first operation in post-processing. It also sets your colour balance to rights in the same operation. Impossible to over emphasize its virtues. If you shoot from LP then this has to be a number one.

DSS is admirably free but would never be my first choice of stacking/calibrating programme. I use AstroArt 5.0 which is good value and highly controllable. Many of our guests see it in action and decide that they can't live without it. (I'm not on commission, by the way!! Far from it, unfortunately.)

There are good web tutorials and bad ones. If you want to start from a sound base with someone who understands the principles in depth (and whose images prove this) start with Rob. http://www.middlehil...d Tutorials.htm

Olly

[cthorpey]

Hi,

Thank you all so much for your help. I will work through the tutorials suggested and MEPC is being read and inwardly digested. I suppose many newbies like myself find the learning curve more like a learning overhanging cliff but the sense of community on here is awesome

Thanks again folks

Chris

[Quatermass]

It does take some time to learn the processing side of things that is for certain, suggest you keep a note book as you try out the many methods and let this become your processing bible. I have kept one for the last 2 years and would be lost with out it now. Some more tutorials on my utube channel as well if you want more homework.

Have fun.QM