What am i doing wrong?

[ollypenrice]

Any comment I may have made about not doing anything other than stacking-calibrating in DSS would assume that the image would then be post processed in a 16 bit software. (For all that, I was a long time user of Ps7 which is largely 8 bit and I found the move to 16 bit CS3 less than overwhelming.)

Personally I don't like or use DSS but a) I don't know much about it so I may be under-estimating it and b ) it is, to its eternal credit, free! So I'm not going to bash it. I've tried it, though, and do the things it does in the reasonably priced AstroArt which is, to my mind, way faster and considerably more transparent.

Reading the thread I wonder if the OP is fully aware of the stretching and black-point-setting process known generally as 'Levels and Curves?' This is the process by which the data is taken from linear to non-linear form in terms of the relationships of brightness within it. The idea is to hold down the dark signal while progressively exaggerating the bright without performing any reversals (in which point A which was darker than point B now becomes brighter than point B.)

Olly

[dph1nm]

Processing in DSS to make the image 'look nice' then importing into 8-bit GIMP for further tweaking works just fine. I do all my processing this way  (I also use 8-bit GIMP for tweaking my professional images without any problem). Trying to do this without adjusting in DSS first results in a GIMP image with highly quantised levels, which isn't really stretchable. I presume (although I haven't tried it yet) that Gimp 2.9 gets round this problem.

NIgelM

[gnomus]

To follow on from what Olly said,  I downloaded your image and looked at the levels in Photoshop.  Here is what I got:

 

Do you notice how far over to the right your histogram peak is?  You would normally drag the black point in to the start of the columns (don't clip anything).  I wonder if you got this result because you were trying to image M51 when there was a full moon.  Is that possible?  Or perhaps you have quite bad light pollution.  If you were shooting during Moon-time then this will have had an impact upon your ability to discern what is already a faint-ish target.  The light from the Moon will reflect off the dust and all the other particles in the air and 'swamp' the faint galaxy.  (You cannot see stars during the daylight - a full Moon is not daylight but it is not dark either).  The other thing is that you only have 5 x 10 minute exposures.  That is not a lot of data.   I would aim for 3-4 times as much.   A useful resource is to look at your target on Astrobin and see what others have done exposure wise.      

[gnomus]

One other thought - you are using the RAW files from the camera and not JPGs aren't you?

[MARS1960]

1 hour ago, ollypenrice said:

Any comment I may have made about not doing anything other than stacking-calibrating in DSS would assume that the image would then be post processed in a 16 bit software. (For all that, I was a long time user of Ps7 which is largely 8 bit and I found the move to 16 bit CS3 less than overwhelming.)

Personally I don't like or use DSS but a) I don't know much about it so I may be under-estimating it and b ) it is, to its eternal credit, free! So I'm not going to bash it. I've tried it, though, and do the things it does in the reasonably priced AstroArt which is, to my mind, way faster and considerably more transparent.

Reading the thread I wonder if the OP is fully aware of the stretching and black-point-setting process known generally as 'Levels and Curves?' This is the process by which the data is taken from linear to non-linear form in terms of the relationships of brightness within it. The idea is to hold down the dark signal while progressively exaggerating the bright without performing any reversals (in which point A which was darker than point B now becomes brighter than point B.)

Olly

Sorry Ollie, it wasn't actually a comment it was a post explaining not to do any processing in DSS that you liked.

I use levels and curves in PS, i do need more practice though, these cloudy skies are a good time to practice on some images i think.

[MARS1960]

18 minutes ago, gnomus said:

One other thought - you are using the RAW files from the camera and not JPGs aren't you?

Yes, i always shoot RAW.

i think you are right about the full moon. It was actually my first night guiding and i was determined to get some nice long subs instead of my usual 30-60secs unguided.

I thought this would give me a better image anyway as i assumed that it being a dim object that long subs would be ideal and totally forgot about the moon behind me. 

I guess this is a target that needs at least 3-4 sessions (without the moon) as the most i can manage before it goes behind the house is around 90 mins.

[gnomus]

1 minute ago, MARS1960 said:

Yes, i always shoot RAW.

i think you are right about the full moon. It was actually my first night guiding and i was determined to get some nice long subs instead of my usual 30-60secs unguided.

I thought this would give me a better image anyway as i assumed that it being a dim object that long subs would be ideal and totally forgot about the moon behind me. 

I guess this is a target that needs at least 3-4 sessions (without the moon) as the most i can manage before it goes behind the house is around 90 mins.

OK - sorry if I offended you - you never know what stage folks are at in their imaging.  

Yes - that Moon will mess you up every time.....  I'm not sure that there's much to be done during full moon time.  Perhaps some open clusters or globulars would be a better target - but you're still going to struggle. 

[ollypenrice]

1 hour ago, gnomus said:

OK - sorry if I offended you - you never know what stage folks are at in their imaging.  

Yes - that Moon will mess you up every time.....  I'm not sure that there's much to be done during full moon time.  Perhaps some open clusters or globulars would be a better target - but you're still going to struggle. 

Even using a 3 nanometer Ha filter I regard the full moon period as a scrapper for at least a week. Unless people are twisting my arm it's more than a week!!!

Olly

[Thalestris24]

Hi

The current 16-bit Gimp 2.9 dev build might be of interest  Looks like a Windows 2.10 isn't too far off

Louise

[Alien 13]

I do 90% or more of my processing in DSS now it does a far better job than PS in my view,  it took me a while to come to this conclusion but I could never match the image a quick DSS stretch could do and the final image has much lower noise for me.

Alan