x6gas

Thank you so much Martin - that's very kind. Thanks for taking the trouble to comment Kluson - really appreciate it.

I have to say that of my own images this target is one of my least favourite. I just couldn't get the colours to look anything other than over-processed but you have done a lovely job here. Kudos.

The 50/50 version gets the thumbs up from me!

Wow - that looks great! Thor's Helmet is still on my wish list but it looks like you have some really good data there Robin!

Hi Robin, Yes that's right, R/G/B 23/23/23 (if all three numbers are the same the 'colour' will be monochrome ranging from black - 0/0/0 - to white - 255/255/255). The way I do it is to place "Color Sampler Tool" points on background sky in the image. You can place up to 4 in my version of PS so I put one towards each corner. The colour sampler tool is one of the tools under the eyedropper circled on the left in red in the image below. Note that there is a tool setting (yellow arrow - apologies for my rubbish drawing!) where you select the sample size - it's best to avoid a single pixel and go for 5 by 5 or 11 by 11 average. Each sample point will give a reading under the info tab (if you can't see the info tab you can activate it from the "window" menu or by pressing f8). The readings for sample point #1 are circled in red on the right below, with the other sample points right and below. To get them all reading the same, I use Levels (Image>Adjustments>Levels or ctrl L) and then select the individual colour channel (see screen shot below), adjusting the black point slider to get the reading I want for that channel before moving on to the next. I still stretch my data in PS and I tend to do a relatively gentle stretch, then check the levels and adjust the R, G, and B channels to 30 if they are above that, then repeat, and repeat, and repeat until I get the stretch I want. Then, as I said above, once I've finished processing I adjust the levels to somewhere between 23 and 27 - in the example above I've chosen 25. Note that if you have a gradient across the image that will show up as different values in the different sample points. I've used Gradient Xterminator on this image but there is often still a little variation across the sample points (you can see my values range from 24 to 26 in fact). A difference of 1 or 2 across the points looks fine in my opinion. I hope this makes sense - please do shout if not! Best wishes, Ian

Thanks Ron, thanks MG - really appreciate it.

Very nice image well done.

My favourite galaxy! I think the uncropped image is the better of the two - it kind of shows the galaxy off as the jewel it is...

Nice result! As Brendan says nothing wrong with those stars, especially at that focal length! I wonder if the sub length is a bit long for good star colour? Do you have a lot of saturated stars in your unstretched data? Also the sky does look a bit dark as I think you alluded in one of your other posts. Many of the great and the good recommend a background sky colour of 23/23/23 when sampled in PS (though I tend to leave mine at 30 during processing and almost the last step is to drop to 23-27 depending on the image - I'm probably guilty of leaving my backgrounds a bit too light...) You can do the work to sort out the filter offsets if you wish if you find that focus shifts materially between filters... but you might be surprised and not need to. Baaders are supposed to be parfocal (though I can't see to what f-ratio) so the difference might be small enough for you not to need to bother. I just focus for every 0.5 degree temperature change using the L filter and I'm good. If you do go down the Bhatinov route then I thoroughly recommend Bhatinov Grabber which, if you have not come across it already, takes the guesswork out of focussing and shows you the latitude you have across acceptable focus.

Just Six Numbers by Martin Rees is an excellent book...

+1 for a RACI finder... it's just so much easier when the image in the finder is the right way up! I must say that I find an illuminated finder invaluable too - even though my skies aren't jet black I do struggle to see the crosshairs sometimes. Having tried very many finders including Telrads, red dots, straight throughs of various kinds, a 9 x 50 illuminated crosshair RACI is the only finder I will ever need.

I must admit I felt the same way when I first saw this image Robin... for a second I asked myself why I bother! But actually images of this quality are inspirational for me, and give me something to aspire to... So I would urge you to post your images and enjoy what you've done. I actually like going back over my images and seeing how my processing has improved but there is always so much to learn and experiment with.

Thanks Robin. I've never used APP but it's literally just a case of bunging all of the registered subs together and stacking them so unless APP reads the fits header, sees the filters are different and refuses to stack them (which would be a bit odd) I'd have thought you'd be OK... and even then you could edit the fits header if you really had to. It's nice to see a consensus that the synthetic luminance has improved the image but as mentioned earlier I think capturing genuine luminance data is the way to go really. That's my plan for the future anyway...

Very many thanks for all the likes and comments, I really appreciate it. Thanks @tomato. I neglected to mention that I had to crop this more than I would have liked as I made a complete pigs ear of the framing. I made my own adapter to couple up the field flattener to my Atik OAG and removed the imaging gear to check some measurements... When I put it back I set the orientation of the camera by eye and didn't check a reference image - rookie error!

Great to see this project progress Jake - very envious!

There have been lots of really excellent M101's posted on SGL of late but for the record here is my attempt. This is the final data set of my spring galaxy imaging (although actually I captured the data for this before any of the others on a couple of clear nights in late March and I will be posting a reprocess of the Whale Galaxy at some point). This was first light for my new Esprit 150 and sadly I didn't nail the guiding (the scope is at the limit of what my CEM60-EC can handle) which has impacted on the blue channel in particular. I took 60, 180s subs in each channel and stacked the best 40 of each for a total integration time of 6hrs. 3 minute subs are just enough, I think. Only the very centre of HIP68503 (TYC3852-468-1) and a couple of the other bright stars were saturated and the signal in the galaxy was OK but I think 300 second subs would have produced a better result on the outer spiral arms so I think that I will use 5 minute subs for this sort of target with this set-up in the future. Kit was CEM60-EC; SW Esprit 150; Atik EFW2 with Astrodon filters; Atik OAG with a mono QHY5L II; Atik 460ex. Data acquired using SGPro, guided with PHD2 and processed in PI and PS. This version is the straight RGB, processed as best as I can: This is a version with all 120 of the R,G, and B subs stacked together to create a synthetic luminance. Whilst I didn't go crazy sharpening the resulting luminance layer, I think it does add a bit of detail and contrast. As an aside, in the second StarGaZine instalment, Nik Szymanek (for whom I have the greatest respect) was not convinced that a synthetic luminance adds anything, if I understood correctly because you obviously don't actually have any more data. However you are, of course, utilising that data in a different way. Stacking all of the subs using an algorithm like sigma clipping gives a different result to layering the 3 individual stacks (not least because you have 3 times the number of subs for the low-signal sky background which I'd argue is more or less the same through each filter unless you have bad light pollution). All that said, as I have been discussing with @MartinB, I think in future I will return to putting most of my effort in to getting a good luminance data set and bin the colour as I have done with previous images of M51 and M81. Finally here's the annotated version from PI. Apols for the long post! Clear skies, Ian

Many thanks Tom - as I've said, the lack of data meant that I had to do more noise reduction than I would normally... and all things considered it hasn't come out too bad. Ah thanks Martin! The synthetic luminance did help on this one due to the lack of data. The stacking algorithm took care of the dust buddies on the green filter in the luminance stack and it gave me something with enough signal to sharpen. I've just done the same with some data of M101 that I captured earlier in the year and that I'll post shortly, but it's had much less of an impact on that which has a couple of hours data per channel. So I have to be honest that the main reason why I didn't shoot luminance in this image is because I knew I had dust bunnies on my L filter and didn't have time to take everything apart and clean it properly. I haven't used flats on this image at all, but when I do use flats I tend to use them to correct vignetting only and keep my imaging train as spotless as possible to avoid dust bunnies. My images are better with flats but I don't want to capture a set of flats for every filter and every image so that's the compromise I'd usually take. Now all that said, I was keen to see how things work just capturing RGB with no luminance so have given it a go with a number of spring galaxy targets. The results are OK but of my own galaxy images (e.g. M51, M81) I prefer the ones where I have captured luminance and binned the colour - I recall that they were easier to process too... Cheers, Ian

Thanks for the info - very interesting. I like it just as much this morning as I did last night!

The first image looks really good to my eyes. If I were being really critical then perhaps the background sky is a tad too dark? Good work though!

I must admit that I am struggling with exactly this in my recent images. Now admittedly they are galaxy images with fairly modest star fields but I seem to be continually experimenting with how hard to stretch the stars in relation to the galaxy... I'm under less than pristine skies, so stars do tend to bloat a bit, but so much depends on the basics of how the image is stretched and I would love to know how you approached stretching this data and how long your subs were. This probably sounds stupid but I find a lot of the difficulty in this astrophotography lark is settling on exactly what you actually like. Achieving that in processing is then, of course, a whole other matter and it's easy to get blown off course. That's why this is such a great image; it was an instant 'like' and genuinely jaw dropping. Like you, Brendan, I much prefer slightly more subtle colours than some of the neon that one sees (and that I end up with myself if I get addicting to boosting saturation too much).

I think it's best to place the filter as close to the camera sensor as possible.

Sumptuous is the word that springs to my mind. That is a very, very fine image. To my eye and on my monitor you have the colours absolutely spot on, just the right amount of sharpening... As close to perfection as it gets for me so goodness knows what you could do if you had three times the data. Thanks for posting - love it.

Wowsers. Absolutely fantastic.

Very many thanks for taking the trouble to comment Olly - I've always really appreciated the encouragement you give to SGL members and your images remain an inspiration.

In honesty Robin, unless you can spend a couple of hours on it every day, I don't think so! I've gone down the route of doing pre-processing manually rather than using the script and I barely got through that in the trial period! Of course if you just follow a tutorial then you can at least get through a whole process. The software is really, really capable and there is lots about it I like but the documentation is, frankly, awful if you want to properly understand what you are doing. Some of it is so technical it's ridiculous and other processes are not really documented at all (even some pretty basic ones). Warren Keller's book is useful but far from comprehensive (that would be virtually impossible in fairness). From what I can work some of the online tutorials are plain wrong too, but there are lots of tutorials and the PI forum is helpful too with the developers often answering questions which is incredibly helpful. The user interface also takes some getting used to and feels like something written with little regard to usability (see for example icon naming conventions). Now all that said, when I got used to the basics and its idiosyncrasies I quite enjoyed learning it - still am learning it, of course, but still exporting to a layers-based photo editor for final processing... As it happens I couldn't get on with StarTools at all but I am pleased I purchased PI - my advice would be get the trial, follow some tutorials (starting with this one) and try to work out whether you like the interface and if you do, go ahead and buy it. As I say, it is very powerful and allows you to do a lot with the linear image...